US3754895A - Process for decarburization of steels - Google Patents

Abstract

Described herein is an improved process for decarburization of straight carbon as well as alloy-bearing steels without the loss through oxidation of valuable metallic alloying elements such as chromium, manganese, and iron, by careful control of the input of oxidizing material.

Description

Uited States Patent 1 I n11 Ramachandran et al. v

3,754,895 Aug. 28, 1973 PROCESS FOR DECARBURIZATION 0F [56] References Cited STEELS v UNITED STATES PATENTS Inventors: Sundaresan Ramachandran, 2.803.535 8/1957 Kosmider et al 75/60 Nat ona Heights; Basil U- lgwe, New 2,855,293 10/195? Savard CI 8].

' 3,003,865 10/196 Bridges Kensmgmn both of 3.046.107 7/1962 Nelson et a1 75/59 Assignee: Allegheny Ludlum Industries, Inc., 3.252.790 5/1966 Krivsky 75/60 Pittsburgh, Pa. 7 Filed: Jan 27, 1971 Primary Exammer -Henry W. Tamng, ll

Appl. No.: 110,324

Related US. Application Data Continuation of Ser. No. 801,017, Feb. 20, 1969,

Attorney-Vincent G. Gioia [57] ABSTRACT Described herein is an improved process for decarburization of straight carbon as well as alloy-bearing steels abandoned v without the loss through oxidation of valuable metallic U 8 Cl 75/60 75/59 alloying elements such as chromium, manganese. and

In. .CI czic 5/34 iron y careful comm] of the input of oxidizing mam rial Field of Search 75/59, 60

MELT BULK CARBON CONTENT (94) 4 Claims, 1 Drawing Figure TOTAL BLOW/N6 rms, umurss PATENTEDMIRZB W75 TOTAL BLOW/N6 TIME, MINUTES By V M lma.

Attorney PROCESS FOR DECARBURIZATION OF STEELS This is a continuation of application Ser. No. 801,017 filed Feb. 20, 1969, and now abandoned.

BACKGROUND OF THE INVENTION The availability of inexpensive gaseous oxygen in abundant quantities has led to its large-scale application in several phases of steel production. Present steel making practices involve a stage during which gaseous oxygen or an oxygen-bearing alloy, such as nickel oxide, is introduced into the molten or semi-molten charge for the purpose of reactingwith impurity elements, such as carbon, silicon, aluminum, etc., contained therein. Such oxygen-impurity reactions result in lowering the amounts of these elements to specified desirable levels by forming gaseous reaction products (oxides) which escape from the bath.

Whereas the use of oxygen or oxygen-bearing components results in rapid removal of the impurity elements, it also gives rise to simultaneous oxidation of useful elements such as iron and chromium. These elements are oxidized and transferred into the slag, thus making necessary a reducing stage during which ferroalloys, such as ferrosilicon, are charged into the bath for purposes of recovering the useful alloying elements from the slag.

The thermodynamic equilibrium existing during decarburization of chromium-bearing steel baths has been studied and the contents of carbon and chromium that exist in equilibrium at a given melt temperature 3 have been established. Of particular relevance to chromium steel decarburization are the works of I-Iilty (The Relation Between Chromium and Carbon in Chromium Steel Refining, Trans. AIME, Vol. I85, (1949), p. 9l and of I-Iilty, Rassbach and Crafts (Observations of Stainless Steel Melting Practices, Journal of the Iron and Steel Institutes (London), Vol. 180, (1955), p. l 16). The relationships developed in these studies show that under atmospheric pressure conditions around the molten bath, the minimum carbon level in equilibrium with a given chromium value decreases as the temperature of the melt increases. If decarburization in an oxygen-saturated bath proceeds below the equilibrium carbon level for given pressure and temperature conditions, carbon removal occurs simultaneously with chromium oxidation, and such oxidation continues until the chromium-carbon-temperature-pressure equilibrium is re-established. It is apparent therefore that the production of a high chromiumlow carbon steel would require relatively high temperatures in order to prevent chromium oxidation. Such high temperatures are impractical and lead to rapid deterioration of the refractory wall of the reactor furnace or ladle. A compromise alloy steel making practice thus involves the following steps: meltdown, decarburization by oxygen injection or by charging an oxygenbearing alloy at a tolerable temperature (such decarburization being accompanied by alloy oxidation into the slag), reduction of the alloying elements from the slag back into the metal, and finally finishing which entails the adjustment of the melt chemistry to conform with specifications. The recovery of the alloy values from the slag is never totally effective and usually between and percent of the charged oxidizable alloy elements are lost in the slag.

In an effort to reduce metallic alloy oxidation, several practices have been proposed and some of these are described in patents. U.S. Pat. No. 3,003,865 (issued to John B. Bridges, Oct. 10, 1961) discloses a process whereby dry gaseous material selected from the group consisting of air, mixtures of air and oxygen, and mixtures of oxygen and an inert gas, is blown through molten steel for purposes of decarburization. Oxygen content of the gas mixture is specified in the range 15 to percent, and sizable reductions in alloy oxidation of chromium-containing melts are achieved, but complete prevention of metallic loss is not realized.

An improvement in the process of mixed. gas decarburization is disclosed in US. Pat. No. 3,046,107 (issued July 24, 1962, to Edward C. Nelson and Neal R. Griffing). A quantitative relationship is presented based on the previously mentioned thermodynamic equilibrium between chromium content, carbon content, temperature and ambient pressure of the bath. The oxygen content by volume of the decarburizing gas is given by the expression:

where %Cr and %C are respectively the chromium and carbon contents, T is the temperature of the melt in degrees Kelvin and Z is an empirical constant whose value 0 depends on the adopted decarburization practice. A

practice based on the above relationship was proposed and calls for one or more reduction steps in the proportion of oxygen in the decarburizing gas as the bath carbon content decreases.

A further purported improvement of the process is described in U.S. Pat. No. 3,252,790 (issued May 24, 1966, to William A. Krivsky). By considering the thermodynamic equilibrium conditions during decarburization, a relationship is developed giving the realizable bath carbon content in terms of the chromium content, %Cr, in equilibrium with that level of carbon, the ambient pressure around the melt, P, and K the thermodynamic equilibrium constant for the carbon-oxygen reaction derived from the thermodynamic activities of carbon and chromium at a given temperature:

The above decarburization processes which are based on quantitative thermodynamic considerations result in substantial improvements in the amount of chromium and other metallics oxidized to slag during decarburization. In chromium-bearing heats, actual losses of chromium could amount toas high as 10-15 percent of the charged chromium, the exact quantity being a function'of the particular decarburization and gas injection techniques employed as well as the initial and desired final carbon and chromium contents of the melt.

PRESENT INVENTION The process of the present invention provides improved decarburization by accounting for the kinetics of the carbon-oxygen reaction, and oxygen consumption efficiency (herein defined as the degree to which the supplied oxygen preferentially oxidizes melt carbon). The systems kinetics are taken into consideration in the development of a quantitative relationship for the tolerable oxygen content of the decarburizing gas required to attain given decarburization without excessive chromium loss. Furthermore, the equilibriumcarbon'concentration is recognized to exist at the gasmetal reaction interface only and as a consequence the bulk carbon concentration in the melt differs from the equilibrium value previously used.

The decarburization reaction in chromium-bearing steels occurs'at the gas-metal interface. A thin liquid film separates the bulk metal from the gas phase, and the diffusion rate of either oxygen or carbon across this film determines the rate of carbon-oxygen reaction. At normal steel making temperatures, the reaction occurs instantaneously as soon as carbon and oxygen are in contact. When the melt carbon content is relatively high, the rate of decarburization is determined largely by the rate of oxygen supply, and since carbon is available in relatively large amounts at the gasmetal interface, it is preferentially oxidized and nosignificant metallic alloy oxidation occurs. However, below a certain carbon content, determined by the melt and tempera ture pressure as well as chromium content, the decarburization rate is determined by the rate of carbon diffusion across the liquid film. If oxygen is supplied in amounts exceeding that required to combine with the carbon arriving gas-metal interface, only a fraction of the oxygen reacts with carbon, the balance combining with valuable alloy elements resulting in undesirable metallic oxidation.

It is apparent from the above that the decarburization rate is a function of the gas-metal surface area available at any stage of decarburization. When decarburization is achieved by blowing the decarburizing gas onto the surface of the melt, the reaction rate depends on the amount of surface and the rate'at which fresh surface is exposed to the reactant gas. A preferred means of optimizing the gas-metal contact area is by injecting the gas by sub-surface means into the volume of melt and thus generating therein a large number of fine bubbles. The size of these bubbles relates to the dimension of the injecting orifice. The mean bubble size formed at an orifice is a function of the orifice Reynolds number, N where:

and where d circular orifice throat diameter,

V exit velocity of gas jet through the orifice,

p gas density,

p. gas viscosity,

W gas mass flow rate through the orifice.

in practicing the invention, oxidizing gas is introduced at a mean equivalent bubble diameter D in accordance with the following equations:

D 0.18 d N, for laminar gas flow, and (l) D 0.28 N for turbulent gas flow 5 2 When the Reynolds number at an orifice exceeds 10,000, the generated spherical bubbles are small and of a virtually constant diameter less than about 0.2

v. (SD/2) U2 where g is the gravitational constant.-

It has been previously mentioned that the transfer of carbon across the liquid film separating'the decarburizing gas phase from the bulk liquid metal phase is a factor that determines the rate and efficiency of carbonoxygen interaction especially at relatively low carbon contents in the melt. Various models have been proposed for the mechanism of this carbon transfer, and corresponding to each model is a mass transfer coefficient. For instance, by analogy to the two film model for gas absorption, a carbon concentration gradient exists across the liquid film separating liquid metal from the decarburizing gas. The bulk metal is assumed to have a uniform carbon concentration. Additionally, the carbon concentration gradient is established immediately upon gas-metal contact and persists throughout the contact period. in the case of gas bubble decarburization, this contact time is the bubble residence time within the melt. The carbon mass transfer coefficient corresponding to this model, K, is then given by:

I (3) where D diffusion coefficient of carbon in the liquid film at the temperature of the melt.

8 thickness of the film.

An alternative and perhaps preferred model has been advanced by R. liigbie (Transactions of the American institute of Chemical Engineers, Vol. 31, 1934-35, p. 365) wherein he proposes that the said liquid film is continually replenished during the contact period or bubble residence time. Consequently, the equilibrium carbon concentration gradient is never attained across a film before it flows around and past the gas. The carbon mass transfer coefficient in such cases is given by: K: 2 c/ r] "2 v where t, h/V gas-metal contact period or bubble residence time, and h bubble rise height.

The process of the present invention incorporates the provisions of the above discussion into a decarburization rate equation of the following form:

dn/dr AK (C C AK m/lOOM C C (5) where dn/dt melt decarburization rate (moles per sec.)

A total reaction interfacial area (cm).

K carbon mass transfer coefficient.

C bulk carbon concentration (in moles/cm).

C, interfacial carbon concentration.

p melt density. 0'92 1 M molecular weight of carbon.

%C,, bulk percentage carbon concentration.

%C, interfacial percentage carbon concentration.

The above relationship (5) is the basis for the process of the present invention. Equivalent expressions can be substituted for the various terms contained in the ex- 5 pression. The substituted values would depend on the particular gas injection technique and carbon transfer model adopted.

For instance, by assuming that the decarburizing gas is injected into the liquid metal via one or more submerged orifices offering a bubble rise height of H inches, and by adopting the carbon transfer model of Higbie discussed above, an expression of the following form is obtained by substituting for the terms of the relation given above:

1 I: m T

where P is the pressure of the bubbles (atmospheres).

T is bath temperature in R. %O is the volume percentage of oxygen in the bubbles.- As indicated previously, the equilibrium carbon concentration is- 2C 0 (3) 2C0 (g) C CO,(g) 2CO(g) it should be noted that a given volume of steam or-carbon dioxide reacts with only half as much melt carbon as the same volume of oxygen. Consequently, in a mixture of oxygen, steam, and carbon dioxide, %Ox is represented by:

%Ox %o, 1/2 %co %H,o)

A revised form of equation (6) is therefore:

If p is taken as approximately equal to 450 lb/ft and D as 7.532 X 10 ft lsec, and standard values are substituted for M, g, and R, equation (7) becomes:

The bubble pressure P is the sum of the ambient pressure in the vessel, the ferrostatic head, and the surface tension pressure of the bubble.

As described, the present invention assumes that ear bon alone is removed by the injected oxidizing gases. Usually, however, other elements such as silicon are present in the melt and are preferentially oxidized to low equilibrium level prior to the commencement of carbon removal. The gas injection schedule, when silicon is present in amounts greater thanthe equilibrium level, should be based on the consideration that at the beginning of the blow, only silicon in part is oxidized into the slag and in part escapes as the volatile oxid'e until equilibrium is attained. The amount of oxygen or oxygen equivalent required for this phase of the process can be computed, assuming chemical stoichiometry, from the reaction:

EXAMPLES OF THE PRACTICE OF THE INVENTION throughout the melt, and also effects adequate mixing of the bath.

The measurements of total gas and oxygen flow rates are made with the usual volume flow measuring devices such as flow meters and orifice plates. The gas composition is determined with conventional gas analyzing techniques, for example mass spectrometers, etc.

The practice of the invention involves some or all of the following steps:

a. Determination of the initial bath carbon content and temperature. The carbon concentration can be obtained using the rapid carbon analysis techniques familiar to those skilled in the art of steel making. Bath temperature is determined with the aid of devices such as immersion thermocouples or pyrometers. Knowledge of the concentrations of other alloying elements in the melt is also desirable, but these concentrations are generally taken to be equivalent to those in the original charge material.

b. Determine the final carbon content desired in the melt at the end of the decarburization step. If decarburization is to be achieved in one step (that is, by injecting a constant gas flow rate and constant gas composition throughout the process), the desired carbon should be that specified for the heat. On the other hand, for a process employing more than one step, the desired carbon would be some intermediate value that is less than the initial carbon concentration. To ap proach continuous operation, decreasing carbon concentrations may be programmed.

c. The partial pressure of CO in the evolved gases is given by P where:

P (Volume flow rate of CO in evolved gases/Total volume flow rate of evolved gases) P (9 From a knowledge of the injected gas flow rate and composition, as well as the fraction of the injected gas that reacts to form CO Pco is easily computed when P is given. Since the oxidizing gases react with carbon according to the equations:

C 2%) 2CO(g) equation (9) can be rewritten as: I P00 F (Input OzHyunpui 20) was... so.)

(1+X) (Input 02) +y(In ut H O +e(Input Equation (9a) neglects CO to CO transformation.

Where this occurs, it must be taken into account in determining Pco.

d. The decarburization reaction for a melt containing, in addition to carbon and chromium, the alloying elements Mn, Cu, Mo, and Ni, can be represented as:

The equlibrium carbon concentration for such a melt at a temperature of T K is given by: v,

%Ceq Poo Exp (24,880/T,, 16.221 0.0552 X %Cr 0.0368 X Cu 0.0207 X Mo 0.0276 X Ni) V (%Cr %Mn) If it is assumed that the desired end-point carbon is I identical to the equilibrium carbon concentration, the

corresponding value of P can then be computed by substituting appropriate values in equation (10). The gas composition required to achieve this P is then determined from equation (9a). By maintaining such a gas composition, preferential C oxidation can be accomplished from the initial C level to the said equilibrium level.

e. Determine the time desired to achieve carbon removal to the stated level. Such a determination is based primarily on economic considerations. This in turn fixed the oxygen flow rate and the total gas flow rate in accordance with the desired gas composition. The oxygen rate is deduced from the relations:

24 lb ofC react with 359 scf O to yield 718 cu. ft. CO (b) f. Determine (in the case of submerged gas injection) the number of orifices necessary to effect uniform gas dispersion within the melt volume. Determine for each orifice the diameter required to achieve a Reynolds number of at least 10 according to equation (1). The mean bubble diameter will thus be assumed constant at about 0.18 inch. By using standard gas flow equations, calculate the pressure necessary to deliver the desired total gas flow rates through the calculated orifice sizes. g. From economic as well as practical considerations, determine the depth of bubble injection and average height of rise. The latter quantity is determined partly by the injector design employed and partly by the pressure and diameter of the orifice.

h. By substituting in equation (8) the bulk carbon attainable with the given system is calculated. If this value is greater than the desired endpoint carbon, the

injected gas composition is varied to yield a lower Pco value.

i. Steps (b) (h) are repeated. The bulk carbon determined in (h) now becomes the initial carbon.

The above sequence illustrates a single step decarburization process. By adopting infinitesimally small reduction steps in desired end-point carbon, a smoothly continuous carbon reduction and gas blow schedule can be generated. The gas ratio reductions can be achieved automatically in this case by a properly designed fiow rate control device.

The followingmore specific example will further i1- lustrate practice of the invention.

Decarburization of 10 percent Chromium Steel By applying the procedure outlined above, a six-step decarburization program is computed for a 150 pound melt containing initially 10 percent chromium and 0.5 percent carbon and employing only oxygen and argon, a constant melt temperature of 3000F and a constant gas flow rate of 3 scfm. An oxygen-carbon utilization efficiency (oxygen-carbon utilization efiiciency fraction of injected oxygen which reacts with melt carbon) of 90 percent was employed; the balance of the oxygen was taken to leave the bath unreacted.

The melt was made in a 500 lb. Basic Oxygen Furv nace. The decarburization gas was supplied via a topsubmerged water-cooled lance whose two orifices (of diameter one-sixteenth inch each) were machined from Type 304 stainless steel. Each orifice was inclined at 10 to the lance axis in order to promote bubble dispersion in the bath. A refractory insulation surrounded the orifices, whose tips were immersed about 2 inches below the melt surface.

The analysis of the melt at the start of gas injection was:

C 0.52% Si 0.080%

Mn 0.018% AI 0.030%

The gas supply program developed was as follows:

Duration of Blow with Oxygen in Gas Mixture Particular Gas Mixture Cumulative Blow Time 12 minutes 20 seconds Samples for chemical analysis were obtained at times of 1% minutes, 4% minutes, 5 minutes, 7 minutes and 12% minutes from the start of blow.

The attached figure shows the theoretical and experimental decarburization paths for this heat. The bath temperature during the blow ranged from 2990 F at the start of blow to 3100 F after 4 minutes, and had fallen to 2960 at the end of the process. As can be seen, the melt was decarburized to 0.073 percent carbon.

A variation of the practice of the present invention is applied to a process where it is desired to intentionally lower the levels of certain elements present in amounts higher than those specified for the desired heat. In such instances oxygen alone or oxygen contents in excess of those recommended by equation (8) would be employed. Since effectively oxygen is thus supplied at rates higher than those necessary to react exclusively with the gas-metal interfacial carbon, the

excess oxygen would oxidize those metallics, such as Cr, whose levels it is desired to lower to specifications.

From equation (8) it is seen that in addition to equivalent oxidizer content, %Ox, other quantitites in the expression can be varied in an attempt to control the decarburization process. Such quantities as bubble pressure, bubble diameter (i.e., effective gas-metal contact area), height of bubble rise, and bath temperature can be varied in a manner similar to those outlined above for effecting preferential oxidation of carbon in alloy steels. In fact, the optimum practice of the invention would involve adjustment of more than one of these variables.

We claim:

1. A method of decarburizing molten steel which comprises measuring the carbon content of said steel,

calculating the necessary decarburization rate to %Ox 0.081875 c, Ceq) [T VTf/PD 1 where Ox is the equivalent oxygen content,

C is the desired carbon content,

Ceq is the equilibrium carbon concentration,

T is the temperature of the melt in degrees Rankin,

H is the bubble rise height in inches,

P is the pressure of the bubbles in atmospheres, and

D is the meanequivalent bubble diameter; controlling the size of said bubbles of oxidizing gas and the point of introduction thereof into said steel melt in accordance with the foregoing equation to provide said necessary decarburization rate.

2. A method according to claim 1 wherein said oxidizing gas contains steam.

3. A method according to claim 1 wherein said oxidizing gas contains carbon dioxide.

4. A method according to claim 1 wherein said oxidizing gas includes a diluent gas from the group consisting of argon, helium, nitrogen, and carbon monoxide. a: a:

Claims (3)

1. 2. A method according to claim 1 wherein said oxidizing gas contains steam.
2. 3. A method according to claim 1 wherein said oxidizing gas contains carbon dioxide.
3. 4. A method according to claim 1 wherein said oxidizing gas includes a diluent gas from the group consisting of argon, helium, nitrogen, and carbon monoxide.

Images