EP0137618B1

EP0137618B1 - Process and apparatus for adding calcium to a bath of molten ferrous material

Info

Publication number: EP0137618B1
Application number: EP84305326A
Authority: EP
Inventors: Joseph Gerard Kaiser; Emil Joseph Wirth
Original assignee: Pfizer Inc
Current assignee: Cessione minerals Technologies Inc
Priority date: 1983-08-12
Filing date: 1984-08-06
Publication date: 1988-06-22
Also published as: AU550957B2; BR8404033A; EP0137618A2; EP0137618A3; ES8607408A1; DK386284A; DK386284D0; AU3178384A; JPH0347909A; KR850001921A; ES8700330A1; DE3472274D1; ES545812A0; ES8607407A1; ES545813A0; KR880000468B1; JPH0369966B2; ES535098A0

Description

This invention relates to the field of processing of molten metals, in particular to a process and an apparatus for adding refining or alloying ingredients to improve properties of a metal being processed.
In the production of steel, a ferrous melt is typically produced in a suitable furnace and then tapped into a ladle where it is treated with one or more ingredients for refining or alloying purposes. Thus, it is well known to add calcium to the molten ferrous material at this point as a refining . agent for oxide inclusion flotation, oxide inclusion morphology modification, desulfurization, etc. Unfortunately, the low density (relative to steel), volatility and reactivity of calcium severely complicate the task of providing a satisfactory process for its addition to the molten material in the ladle.
In metal processing, and in particular steel processing, molten metal generally is separated from a quantity of slag which remains relatively solid and floats upon the surface of the molten metal. The slag is made up of various lower- density impurities, quantities of oxidized metals and the like. In order to feed an additive material into the molten metal, the additive must be placed below or caused to pass through the slag surface. Of course, additive materials intended to improve the properties of e.g., steel are typically relatively expensive and must be conserved. Any waste of a calcium-containing additive material, for example by the loss of material in the slag layer during addition, can have a major economic impact on the producer and the product. It is therefore highly desirable to feed the calcium well below the surface of the molten metal to the point where it will be most effective, and to mix the molten metal to evenly distribute the calcium additive therein.
A variety of techniques have been employed for the addition of calcium to the molten material in a steelmaking ladle. Bulk addition of calcium-containing particulate materials is unsatisfactory because these materials rapidly rise to the surface of the melt without spending a sufficient residence time therein. Efforts to increase residence time by pouring the particulate material directly into the tapping stream from the furnace give rise to excessive reaction of the calcium with atmospheric oxygen. Introductions of calcium-containing materials by plunging or the injection of clad projectiles into the melt generally provide adequate residence times but are complicated, expensive and time-consuming procedures. It has also been proposed to inject calcium-containing powders into a melt by inert gas injection through a refractory lance. Since sizable flows of gas are required to propel the powder into the molten ferrous material, a high level of turbulence is generated at the surface of the melt as the gas is released, thereby causing an excessive exposure of the molten ferrous material to oxygen and nitrogen in the atmosphere. Furthermore, after leaving the lance, the calcium tends to rise rapidly through the melt in the inert gas plume surrounding the lance or in upwelling molten material adjacent the plume. Thus, calcium residence time in the bath is unacceptably low.
In an attempt to overcome the above-mentioned problems, calcium has also been added to melts in steelmaking ladles in the form of a calcium metal-containing wire (clad or unclad) continuously fed through the upper surface of the melt. A major advantage of wire feeding is that large flows of gas are not needed, as in powder injection, to propel the calcium-containing material into the molten ferrous material. However, the high volatility of calcium hinders the attainment of an efficient utilization of the calcium added in surface wire feeding. If the wire does not penetrate to a sufficient depth below the surface before the calcium in the wire desolidifies, a low residence time and poor utilization of the calcium results along with a non-uniform treatment of the melt. It is particularly important that most or all of the input calcium remain unreacted until it descends below the depth at which the ferrostatic pressure is equal to the vapor pressure of calcium. This goal is difficult to achieve, even when a 'clad calcium metal-containing wire is employed. When calcium desolidifies at ferrostatic pressures lower than its vapor pressure, large calcium gas bubbles are formed that rise rapidly to the surface of the melt. The result is an inefficient, non-uniform treatment of the molten ferrous material and the generation of a large amount of turbulence at the surface of the melt.
U.S. Patent 4,154,604 discloses a method and apparatus for adding a wire to molten metal in a vessel through a refractory clad tube filled with pressurized inert gas. This patent does not, however, disclose the desirability of effecting the melting of wire constituents at a substantial dis- tancefrom the lowertip of the refractory clad tube in or directly below a region of downwelling of the molten metal. In fact, such a result is physically precluded in the preferred embodiment disclosed in said patent by the close proximity of the lower tip of the tube to the bottom wall of the vessel.
A novel process for adding calcium to a bath of molten ferrous material has now been discovered, which process comprises feeding a calcium metal-containing wire having a lower density than said ferrous material downwardly through a refractory lance inserted into said bath while providing a sufficient flow of inert gas through said lance to maintain the interior of the lance essentially free of said molten ferrous material and to induce substantial recirculatory stirring of said molten material, with the disposition of the lance in said bath and the composition, cross-sectional dimensions and feeding rate of said wire being such that (a) said wire bends substantially towards the horizontal direction after exiting from the wire outlet of the lance and before fully decomposing, and (b) at least a major part of the desolidification of the calcium in said wire occurs by melting in or directly below a region of downwelling of said molten ferrous material at a depth below the surface of said bath at which the ferrostatic pressure is greaterthan the vapor pressure of calcium at the temperature of said molten ferrous material. It is of course the buoyancy of the wire, resulting from its lower density than that of the melt, that causes itto bend. Preferably, while the wire is being fed through the lance, the wire outlet of the lance is positioned at a depth below the surface of said bath at which the ferrostatic pressure is greater than the vapor pressure of calcium at the melt temperature.
The desolidification of calcium at a ferrostatic pressure greater than its vapor pressure leads to the creation by melting of liquid calcium globules, which rise much more slowly through the melt (thus providing a much higher residence time) than do calcium gas bubbles. As these liquid globules slowly rise through the molten ferrous material in the bath, they eventually are transformed into a very large number of small gas bubbles that do not generate excessive turbulence when they reach the surface of the melt. Furthermore, according to the present invention, these liquid calcium globules rise through a region of downwelling in the circulatory motion of the melt in the bath. This countercurrent flow of the rising calcium and circulating molten ferrous phases greatly enhances the degree of contact between the calcium and the molten ferrous material and further increases the calcium residence time in the bath. As a result, the efficiency of utilization of the calcium refining additive is substantially improved.
Another advantage of the process of the present invention is that the inert gas flow rate in the lance can be varied independently of the wire feeding rate to optimize the internal melt circulatory stirring rate and the extent of slag/metal contact at the surface of the bath.
The present invention also includes a novel apparatus for efficiently adding a processing element in the form of a wire directly into a quantity of molten material, said apparatus comprising a heat-resistant nozzle having an outlet disposable beneath the surface of the molten material, means for feeding the wire into the nozzle, and means for concurrently injecting an inert gaseous medium into the nozzle together with the wire, thereby preventing closure of the nozzle by solidification of molten material therein while agitating the molten material by gas bubble agitation. A seal device having opposed, pressure- biased pistons engages the wire upstream (relative to wire feed) of the source of inert gas, which gas is fed together with the wire through a gas-tight conduit to the nozzle. A particular configuration of the bore of the nozzle maximizes the effect of the inert gas. A restriction in the flow path adjacent the outlet of the nozzle creates an area of increased gas velocity, whereby any irregularities which may occur in the feeding of wire do not give rise to the passage of molten metal into the interior of the nozzle.
The novel apparatus of the invention is inexpensive, convenient in use, effective in starting, stopping and during use, and requires use of the least amount of additive necessary to achieve a given concentration in the processed metal.
The invention will be described in detail with reference to various preferred embodiments thereof. Reference to these embodiments does not limit the scope of the invention, which is limited only by the scope of the claims. In the drawings:

Figure 1 is a schematic depiction of an apparatus suitable for use in the process of the present invention;
Figure 2 is a view taken along line 2-2 in Figure 1 showing the eccentric disposition of the refractory lance in the ladle;
Figure 3 is a chart that can be used to determine the critical depth of molten steel in a ladle, i.e. the depth below the surface of the molten steel at which the ferrostatic pressure equals the vapor pressure of calcium, as a function of temperature;
Figure 4 is a schematic perspective representation of an embodiment of the apparatus of the invention;
Figure 5 is a perspective view, partialy cut-away, of the nozzle of the invention shown in Figure 4;
Figure 6 is a cross-sectional view taken along line 3-3 in Figure 5;
Figure 7 is a detail view of the point of addition, that is, the outlet of the nozzle, also taken in cross-section along line 3-3 in Figure 5;
Figure 8 is a perspective view of the seal device of the invention shown in Figure 4;
Figure 9 is a section view taken along line 6-6 in Figure 8;
Figure 10 is a section view taken along line 7-7 in Figure 8; and
Figure 11 is an elevation view showing the preferred physical layout of parts shown schematically in Figure 4.

A suitable apparatus for use in.feeding a calcium metal-containing wire 1 into a bath 2 of molten ferrous material, e.g., steel, contained in a ladle 3 (which is open to the atmosphere) is schematically depicted in Figures 1 and 2. In the process of the present invention, wire 1 has a lower density than the molten ferrous material 2. As used herein, the term "calcium metal-containing wire" means that such a wire is comprised at least in part of unalloyed elemental metallic calcium as a distinct phase. The wire may also contain distinct phases of calcium alloys (e.g. a calcium-aluminum alloy) or calcium compounds (e.g. calcium silicide) or other ingredients being added to the molten ferrous material for refining or alloying purposes (e.g. aluminum, magnesium, rare earth elements). The calcium metal-containing wire may be clad (e.g. with a steel cladding) or unclad. In the former case, the calcium metal-containing core of the clad wire may itself be a wire or may exist in any other known form, e.g. a powder. Preferably, a surface layer4 of a basic synthetic slag containing e.g. lime and fluorspar is applied to the melt 2 prior to commencement of the wire feeding. As used herein, the terms "depth below the surface of the
bath", "depth below the surface of melt 2", etc., refer to the depth below the slag/molten metal interface.
As is best shown in Figure 1, wire 1 is fed into melt 2 downwardly through a refractory lance 5 inserted into the bath 2 of molten ferrous material. Simultaneously, a flow of gas inert to the molten ferrous material (e.g. argon) is provided through the lance. This inert gas exits from the wire outlet 6 of lance 5 and rises as a multiplicity of bubbles 7 surrounding lance 5 to the surface of the melt. The pressure and flow rate of the inert gas must be sufficient to maintain the inner bore of the lance free of molten ferrous material and thus prevent blockage of the bore by solidification of said material. Moreover, the inert gas pressure and flow rate should be sufficient to induce a substantial recirculatory stirring of the melt 2 in ladle 3 (note arrows in bath 2 in Figure 1). Preferably, however, the inert gas flow rate is not so high as to generate a large amount of turbulence on the surface of the melt as the bubbles 7 escape to the atmosphere. A preferred range for the flow rate of inert gas through lance 5 is from about 1.56-4.16x10-^e m3fs. kg (1.5x10-⁵ to about 4x10-⁵ standard ft³/ (min - Ib of melt)). Since the inert gas in lance 5 is not relied upon to propel the wire 1 into the melt, its flow rate through the lance can be adjusted independently of the wire feeding rate. The inert gas pressure in lance 5 must, of course, be greater than the ferrostatic pressure at the wire outlet.
As used herein, the term "refractory lance" means that at least those outermost longitudinal portions of lance 5 that come into contact with the molten ferrous material 2 are made of a refractory material (e.g. alumina) that is resistant to physical or chemical change while subjected to such contact. Preferably, lance 5 is straight and oriented in a vertical manner while wire 1 is being fed through it. However, lance 5 may also be tilted away from a vertical orientation during the wire feeding (but not horizontal). Also, the lance may have a "dog-legged" shape. The lance is provided with a wire inlet and a wire outlet, with the wire inlet at a higher elevation during use than the wire outlet. Usually, the wire outlet is at the lower tip of the lance. However, it is possible, e.g., to employ a lance having a side port wire outlet displaced from the lower tip of the lance.
In addition to lance 5, the apparatus shown in Figure 1 includes a wire spool 8, a mechanical wire feeder 9, an inert gas feeding and sealing assembly 10 and a gas-tight wire conduit 11 connecting assembly 10 to and supporting lance 5. Although not essential to the practice of the present invention, it is preferred to employ a mechanical wire feeder, an inert gas feeding and sealing assembly and a refractory lance of the types disclosed in Figures 4 to 11 herein. If wire 1 includes exposed elemental calcium metal at its outer surface, such as when it is an unclad calcium metal wire, conventional steps will have to be taken to protect the wire on spool 8 from atmospheric attack, such as maintaining spool 8 in a housing pressurized with calcium-inert gas.
In typical steelmaking operations, the temperature of the molten ferrous material 2 in ladle 3 ranges from 1538-1649°C (about 2800°F to about 3000°F). At these temperatures the vapor pressure of calcium is quite substantial. As discussed earlier, it is essential to the full success of the calcium addition operation that a major part (or all) of the desolidification of the elemental calcium metal in wire 1 occur by melting rather than by vaporization. Thus, this desolidification must occur below the critical depth in the melt, which is defined as that depth below the surface of the melt at which the ferrostatic pressure is equal to the vapor pressure of calcium (at the melt temperature). The critical depth may be readily determined as a function of temperature by using the chart provided in Figure 3. The rightmost curve in Figure 3 is a plot of calcium vapor pressure vs. temperature, while the leftmost curve is a plot of ferrostatic pressure vs. depth below the surface of the melt. At 1571°C (2860°F), for example, the vapor pressure of calcium is 1.57 atm. A ferrostatic pressure of 1.59x10⁵ Pascal (1.57 atm.) is experienced at a depth of 0,85 m (2.8 feet), which is thus the critical depth at 1571°C (2860°F).
At the heart of the present process invention is the concept of adjusting the disposition of lance 5 in melt 2 and the composition, cross-sectional dimensions and feeding rate of wire 1 so that

(a) the wire bends substantially towards the horizontal direction after exiting from the wire outlet of the lance and before fully decomposing, and
(b) at least a major part of the desolidification of the calcium in the wire occurs by melting in or directly below a region of downwelling of the molten ferrous material at a depth below the critical depth D (see Figure 1). As used herein, the term "disposition of the lance" or "lance disposition" contemplates both the depth of the lance in the bath and its position in horizontal planes through the bath (e.g. the plane of Figure 2), as well as the orientation of the lance with respect to the vertical (i.e. the degree and direction of its tilt, if any, away from the vertical). The four variables of lance disposition, wire composition, wire cross-sectional dimensions and wire feeding rate are interrelated, so that a change in one of said variables may require that one or more of the remaining variables be readjusted to continue obtaining the results (a) and (b) set forth above. Thus, for example, it is preferred that the lance be disposed so that is wire outlet 6 is positioned below the critical depth while the wire is being fed through the lance, as shown in Figure 1. However, it is also possible to operate with the wire outlet of the lance somewhat above the critical depth. In this case, it may be necessary to increase the wire feeding rate, increase the wire diameter or switch to a clad wire in order to continue the practice of the
present invention. It is also preferred that the lance 5 be non-centrally disposed in the ladle 3, as viewed in horizontal planes such as the plane of Figure 2. This eccentric disposition of lance 5 in ladle 3 serves to increase the volume of the target downwelling region in the recirculating melt 2 by concentrating downwelling on one side of the ladle (see Figure 1). Preferably, the distance between the longitudinal axis of lance 5 and the inner surface of the nearest ladle side wall (e.g. surface 12 in Figures 1 and 2) is from about 1/6 to about 1/3 of the longest linear dimension L of the bath, as viewed in horizontal planes. This longest linear dimension of the bath would be its major axis in the case of a ladle with elliptical or oval cross-section, its diameter in the case of a ladle with circular cross-section, its length in the case of a ladle with rectangular cross-section, etc.

Since the distance that a particular wire 1 will travel from the wire outlet 6 of lance 5 before fully decomposing will depend directly upon the wire feeding rate, this rate is a very important variable. In the practice of the present invention, decreasing the thickness of wire 1 or changing from a clad to unclad wire will tend towards requiring an increase in the wire feeding rate. Also, a higher melt temperature will tend to require a higher wire feeding rate. In the case in which wire 1 is an unclad calcium metal wire having a diameter of from about 8 mm to about 12 mm, lance 5 is straight and vertically-oriented in the bath, the wire outlet 6 of lance 5 is at the lower tip of the lance and is positioned below the critical depth D, the distance between the longitudinal axis of the lance and the inner surface of the nearest ladle side wall is from about 1/6 to about 1/3 of the longest linear dimension of the bath (in horizontal planes), and the temperature of the molten ferrous material 2 is from 1538-1649°C (about 2800°F to about 3000°F), a preferred range for the wire feeding rate in the practice of the present invention is from 2.54-5.08 m/s (about 500 ft/min to about 1000 ft/min).
The following examples illustrate the process of the invention but are not to be construed as limiting the same.

Example 1

Clad calcium metal wire

1633 kg (3600 lbs) basic slag mix was added to the bottom of a ladle having an elliptical cross-section in horizontal planes, and 1.904x10⁵ Kg (210 tons) of molten steel was then tapped from a furnace into the ladle. The sulfur content of the steel, was reduced from 0.021 wt.% to 0.008 wt.% as a result of the tapping operation. A 2.44 m (8 ft) long straight refractory lance of the type described in Figures 5-7 was then disposed in the bath of molten steel, with the lance being vertically-oriented and positioned on the major axis of the elliptical ladle cross-section at a distance of about 1/3 of the length of said major axis from the inner surface of the nearest ladle side wall, and with its wire outlet at its lower tip being positioned 1.83 m (6 ft) below the surface of the molten steel bath. With pressurized 2.07 Pascal (30 psi) argon flowing through the lance at 5.7x10-³ m'/s (12 scfm), 914 m (3000 ft) of clad calcium metal wire (49 wt% calcium metal core-51 wt.% 0.025 cm (0.010 in) thick 1010 steel cladding) having a total diameter of 8 mm was then fed downwardly into the molten steel bath through the lance at a feed rate of 2.79 m/s (550 ft/ min). The temperature of the molten steel in the ladle was 1571°C (2860°F), which corresponds to a critical depth of 0.85 m (2.8 ft). After exiting from the lower tip of the lance, the wire bent substantially towards the horizontal direction.. Complete decomposition of the wire occurred at a distance of about 3.05 m (10 feet) from the lower tip of the lance. After completion of the wire feeding, the molten steel in the ladle was tapped and cast into appropriate molds. The cast steel product contained 0.22 wt.% carbon, 1.36 wt.% manganese, 0.03 wt.% aluminum, 0.12 wt.% vanadium, 0.005 wt.% sulfur and 45 ppm calcium. 100% inclusion modification was observed.

Example 2

Unclad calcium metal wire

The procedure of Example 1 may be repeated with the use of an unclad calcium metal wire. Operating equipment and conditions are substantially unchanged, except that an unclad 12 mm diameter calcium metal wire is fed to the bath of molten steel for one minute at a rate of 4.06 m/s (800 ft/min). After exiting from the wire outlet at the lower tip of the lance, the wire bends substantially towards the horizontal direction. Complete decomposition of the wire occurs at a distance of about 3.05 m (10 feet) from the lower tip of the lance.
A preferred embodiment of the apparatus of the invention is illustrated in Figures 4 to 11. One or more processing elements for treating a molten metal product are disposed in, or otherwise form a part of, a wire 20. Such elements are hereinafter sometimes referred to as being in wire-form. With reference to Figure 4 (a schematic view), the general objective is to convey the wire 20 from reel 22 to the quantity of molten metal 56 in receptacle 52. In order to accomplish such feeding, a feeding mechanism 24 draws the wire from the reel and advances the wire along a feed path. Adjacent the output portion, especially in the vicinity of nozzle 60, the wire 20 is carried in a gas-tight conduit 44. Inert gas is supplied to the gas-tight conduit, and a seal mechanism 30 located immediately upstream of the inert gas input prevents loss of inert gas around wire 20 in a direction backwards along the feed path.
Reference can be made to U.S. Patent 4,235,362 for a description of a suitable feed mechanism 24, including pinch rollers 26. A wide range of wire sizes and compositions are possible, including both sheathed and unsheathed wires. The invention will be described in detail, however, with reference to sheathed calcium-containing wire of approximately one cm diameter. Wires of this diameter, and wires of somewhat smaller diameter, are relatively rigid. Accordingly, the feed mechanism as well as the wire-carrying members must be capable of withstanding rough wear. Moreover, it should be expected that during feeding the relatively rigid wire will be prone to a certain amount of vibration and transverse displacement due to encountering discontinuities along the feed path, and due to bumps and bends in the wire.
The nozzle 60 of the invention, shown in detail in Figures 5 to 7, comprises a refractory ceramic casing 62, through which the calcium wire is conveyed in metallic conduit portions 66 and 70 to the ultimate outlet or discharge point 84. Refractory casing 62 may be made of alumina (A1₂0₃) or any other suitable refractory material such as those used to line kilns and the like.
The overall nozzle is made long enough to extend to a preselected depth in the reservoir of molten metal. It is usually preferred that the wire additive be discharged from the nozzle at least 0.91-1.52 m (3 to 5 feet) below the slag/metal interface. Accordingly, with, due regard to the high temperature and corrosive nature of the slag and metal, the refractory casing 62 should be on the order of 3.05 m (10 feet) long.
The nozzle 60 may be raised and lowered with respect to the metal receptacle 52, or vice versa, by means of appropriate mechanical linkages. As shown schematically in Figure 4, the metal receptacle 52 may be carried by a winch/conveying system, including yoke assembly 48. Alternatively, it may be preferable to raise and lower the entire feed mechanism as a unit, as shown in Figure 11. In any event, it is beneficial to avoid flexing the conduit 44.
The central wire-carrying portion of nozzle 60 includes a metallic conduit 66 leading to metallic conduit 70, through both of which the wire 20 is passed. The larger conduit 66 carries the wire to near the discharge opening 84 of nozzle 60. An enlarged bore 68 is formed at the end of large conduit 66, into which bore small conduit 70 is placed. Small conduit 70 and large conduit 66 are joined by threads, or by weld 72, or by other convenient means.
As shown in Figure 7, the discharge end of the smaller conduit 70, at the extreme end of nozzle 60, has an elongated, gradually tapered funnel-shaped section 80 of decreasing internal diameter in the direction of flow. Following the narrower end 82 of the funnel-shaped section, there is an abrupt increase in diameter, formed by a relatively short substantially cylindrical section 83 of substantially uniform diameter. The end of the uniform cylindrical section 83, opposite the narrower end 82 of the funnel-shaped section 80, forms the outlet 84 of the nozzle 60. As shown in Figure 7, this particular variation in diameter along the direction of wire travel has certain advantages. In particular, the cross-section is adapted to cooperatively prevent the molten metal 56 from running upwards into the nozzle. Otherwise, encroaching molten metal may solidify in the nozzle along the internal areas of conduits 66 and 70 and there bind the wire to the conduit. Simultaneously with exclusion of metal from the nozzle, the inert gas passing outwards through the nozzle together with wire 20 agitates the metal 56, mixing the additive and the molten metal, thus providing for a more even distribution of the additive material. The inert gas also functions to keep the nozzle cool.
In order to add the wire-form additive to the molten metal 56 at a point well below the surface of molten metal, it is necessary to overcome substantial fluid pressure in the molten metal. The fluid pressure is, of course, a function of the depth below the surface of molten metal. The particular pressure will depend upon the particular metal, but will usually be quite substantial at a depth of one or two meters. The pressure of inert gas supplied must overcome this fluid pressure in order to prevent molten metal 56 from rising in the nozzle. Should any molten metal be permitted to run into the nozzle, wire 20 can immediately be seized and welded to a conduit wall as the molten metal solidifies.
The additive material in the form of wire 20 melts after discharge into the reservoir of molten metal 56. Bubbles 88 of inert gas rise toward the surface of molten metal 56, agitating the molten metal and causing an overall flow therein, upwards adjacent the nozzle and downwards at other areas, namely around the periphery of the molten metal reservoir 52.
The decreasing internal diameter of conduit 70 is intended to maximize the gas velocity immediately adjacent the ultimate outlet 84 of the nozzle. Along the area 80 of decreasing cross-section, the gas, at constant pressure, increases in velocity up to the restriction 82. Immediately beyond the restriction, an open cavity or chamber formed by the uniformly cylindrical section 83 of the bore serves to space the restriction 82 from the molten metal 56, further guarding against the entry of molten metal into the restricted orifice 82.
By virtue of the construction described above, the wire is maintained well clear of the lowermost edges of the conduit 70 which are unavoidably exposed to the molten metal, and cannot be welded to these edges by solidifying metal cooled by contact with the nozzle. As the wire 20 is fed, it can be expected to vibrate and rattle around the allowed space in restricted orifice 82. However, the wire remains centrally positioned in the discharge opening 84 even if resting against an edge of the restricted orifice wall. The space which is left open between the wire and the wall of restricted orifice 82 is small enough that the gas pressure overcomes the fluid pressure of displaced molten metal, othewise tending to flow up the nozzle. Interactive movement of the wire and the inert gas enhance the ability of the nozzle to resist clogging.
Should no provision be made for feeding from some form of sealed inert gas reservoir, a sub- stantiat quantity of the inert gas will be vented into the atmosphere and will not function to prevent the entry of molten metal 56 into the nozzle 60. Accordingly, seal mechanism 30 is provided to prevent a backwash of inert gas. Seal mechanism 30 comprises a housing having at least one pair of opposed pistons 32 having contoured sealing surfaces for slidably engaging the wire moving therebetween, which clasp the advancing additive wire 20 in a gas-tight fashion. Downstream of the opposed pistons 32, the inert gas is fed from inert gas source 31 via conduit 33 to the area of wire 20, the wire now being enclosed in a gas-tight conduit 44 leading from seal 30 to the nozzle 60. The particulars of the seal mechanism are shown schematically in Figure 4 and in more detail in Figures 8 to 10. A compressed air source 34 is preferably used to drive opposed pistons 32 against wire 20. Spring biasing, hydraulic pressure or the like are also possible. A manifold 36 may be used to equally distribute the air pressure of compressor 34 or other source. Opposed pistons 32 are slidably mounted in gas-tight cylinders, and sealed therein by means of resilient "O"-rings, for example two per piston. The equalization of gas pressure by means of manifold 36 results in equal pressure on opposed pairs of axially aligned pistons 32, at each stage thereof. Two stages or pairs of opposed pistons are shown, disposed in parallel relationship. It will be appreciated that the opposed pistons may likewise be mounted at right angles, or as otherwise desired. The pairs may also be operated independently such that one pair provides an atmosphere seal and the other pair provides an inert gaseous medium seal.
The housing of seal unit 30 is preferably made of steel. The pistons 32, mounted in the cylinders of the housing, are made of a durable plastic material. The pistons may, for example, be made from or coated with teflon@, nylon@, or the like.
The housing of seal 30 is provided with an enlarged, funnel-shaped input orifice 35, adapted to "capture" the advancing end of wire 20. It may be necessary to additionally spring-bias opposed pistons 32, or provide for a manual adjustment, in order to ensure their central alignment during the initial loading of wire 20. Once loaded, however, the seal mechanism 30 will compensate for variation in the transverse position of wire 20 with respect to the seal 30, while maintaining the gas-tight seal thereof. Inasmuch as the sheathed wire is quite stiff, it is necessary to allow some variation in alignment in order to prevent undue friction and to maintain the seal.
A suitable control mechanism may be connected simultaneously to the pinch roller wire feed device 24 and to the inert gas pressure control 42. To avoid waste, the gas control 42 should be left closed until the wire becomes engaged by opposed pistons 32 of seal 30. In any event, no particular gas pressure is required until the injector nozzle 60 is brought into proximity with the molten metal 56, or the slag 54 thereupon. At this point, the feeder and inert gas pressure control may be simultaneously activated, and the nozzle plunged into the molten metal. Melting additive and inert gas are discharged at the nozzle orifice, well below the slag/ metal interface.
A preferred physical arrangement of the system is shown in Figure 11. Virtually the entire system is disposed upon a pivotally-mounted table 120, which pivots on hinge 122. A hydraulic or pneumatic lifting device 124 is operable to lift and lower the table 120 about its pivot, thereby raising and lowering nozzle 60 with respect to the molten metal 56 in container 52. The lifting mechanism may likewise be incorporated under the common inert gas/wire feed control.
The nozzle 60 is formed with a bore having, with respect to the direction of additive feed and inert gas flow, a substantially cylindrical section of substantially uniform diameter, followed by a tapered section of decreasing diameter terminating at an aperture having a radius only slightly larger than that of the wire and a second substantially cylindrical section of substantially uniform diameter larger than that of said aperture, whereby the wire remains spaced from the internal edges of the nozzle conduit adjacent the outlet. An abrupt transition between the tapered and second cylindrical sections creates a restricted diameter orifice with increased gas velocity therein, past which orifice the molten metal does not backflow.
The essential features of the invention having been disclosed, further variations will now become apparent to persons skilled in the art. Reference should be made to the appended claims, rather than the foregoing specification, as indicating the true scope of the subject invention.

Claims

1. A process for adding calcium to a bath (2) of molten ferrous material which comprises feeding a calcium metal-containing wire (1) having a lower density than said ferrous material down- wardlythrough a refractory lance (5) inserted into said bath (2), said lance (5) containing inert gas surrounding said wire (1) at a pressure sufficient to maintain the lance (5) essentially free of said molten ferrous material, at least a major part of the desolidification of the calcium in said wire (1) occurring at a depth below the surface of said bath (2) at which the ferrostatic pressure is greater than the vapor pressure of calcium at the temperature of said molten ferrous material, characterized in that a sufficient flow of said inert gas is provided through said lance (5) to induce substantial recirculatory stirring of said molten material, and the disposition of the lance (5) in said bath (2) and the composition, cross-sectional dimensions and feeding rate of said wire (1) are such that (a) said wire (1) bends substantially towards the horizontal direction after exiting from the wire outlet (6) of said lance (5) and before fully decomposing, and (b) at least a major part of the desolidifcation of the calcium in said wire (1) occurs by melting in or directly below a region of down-welling of said molten ferrous material.

2. A process of Claim 1, wherein the wire outlet (6) of said lance (5) is positioned, while said wire (1) is being fed through said lance (5), at a depth below the surface of said bath (2) at which the ferrostatic pressure is greater than the vapor pressure of calcium at the temperature of said molten ferrous material.

3. A process of Claim 2, wherein said lance (5) is straight, said wire outlet (6) is at the lower tip of the lance (5), said lance (5) is vertically-oriented while said wire (1) is being fed through it, and said lance (5) is eccentrically-disposed in said bath (2), as viewed in horizontal planes, while said wire (1) is being fed through said lance (5).

4. A process of Claim 3, wherein said bath (2) is held in a vessel (3) having bottom and generally vertical side walls and the distance between the longitudinal axis of said lance (5) and the inner surface (12) of the vessel side wall nearest thereto is from about 1/6 to about 1/3 of the longest linear dimension (L) of said bath (2) in horizontal cross-sectional planes.

5. An apparatus for injecting a processing element in the form of a wire (20) below the surface of a molten material, comprising: a heat resistant nozzle (60), movable into an operative position wherein an inlet is disposed above said surface and an outlet (84) is disposed below said surface, with said nozzle (60) being provided with a bore having a funnel-shaped section (80) terminating in an aperture (82) having a radius only slightly larger than that of said wire (20) and a substantially cylindrical section (83) of substantially uniform diameter into which said aperture (82) opens, the diameter of said uniform diameter section (83) being sufficiently larger than the diameter of said aperture (82) to form an abrupt transition between said two sections (80, 83), and with the opening at the end of said substantially uniform diameter section (83) opposite said aperture (82) forming the outlet (84) of the nozzle;

a gas-tight conduit (44) connected at one end to the nozzle inlet;

means (30) disposed at the other end of the conduit (44) for sealably receiving the wire (20);

means (31, 33, 42) for injecting an inert gas into the conduit (44); and

means (24) for feeding the wire (20) into and through the conduit (44) and nozzle (60) and directly into the interior of the molten material.

6. The apparatus of Claim 5 wherein said means (30) for sealably receiving the wire (20) comprises at least one pair of pressure activated pistons (32) disposed coaxially and movable toward one another, with the pistons (32) having contoured sealing surfaces for slidably engaging the wire (20) moving therebetween; and wherein said means for injecting an inert gas into the conduit (44) comprises a pressurized source (31) of inert gas and means (33,42) for connecting said source (31) to the means (30) for sealably receiving the wire (20) downstream of the at least one pair of pistons (32) and upstream of the conduit (44).

7. The apparatus of Claim 6 wherein said means (30) for sealably receiving the wire (20) further comprises a gas manifold (36) for driving the pistons (32) under equal pressure and a pressurized gas source (34) for the manifold (36).

8. The apparatus of claim 6 wherein said means (30) for sealably receiving the wire (20) comprises two pairs of pistons (32) separated from one another along the direction of wire feeding, with one of said pairs forming an atmosphere seal and the other forming an inert gas seal.

9. A nozzle (60) for injecting a processing element in the form of a wire (20) directly into the interior of molten material and the like, comprising an elongated casing (62) formed from highly temperature resistant material and having an external bore through which the wire (20) may be fed directly into the molten material, said bore having a termination opening (84) through which the wire (20) exits the nozzle (60), characterized in that said bore is characterized by a restriction in diameter (82) upstream of the termination opening (84), a funnel-shaped section (80) upstream of said restriction (82) and an abrupt transition from said funnel-shaped section (80) to a section (83) of substantially uniform diameter downstream of said restriction (82) and adjacent the termination opening (84), said funnel-shaped section (80) being relatively longer axially than said uniform diameter section (83).