JP5395545B2 - Manufacturing method of ultra high purity alloy ingot - Google Patents

Description

  The present invention relates to a method for producing an alloy ingot such as high-grade stainless steel or superalloy that requires ultra-high purity (very low impurity content), and in particular, a practical scale ingot having a product ingot weight of 10 kg or more. The present invention relates to a manufacturing method suitable for manufacturing.

  Known impurity elements that adversely affect the corrosion resistance of the alloy include carbon [C], nitrogen [N], oxygen [O], phosphorus [P], and sulfur [S]. It is also known that the corrosion resistance of the alloy is greatly improved by reducing the concentration of these impurity elements to the limit.

  The target concentration of these impurity elements is, for example, [C] + [N] + [O] + [P] + [S] <100 ppm. In the conventional mass production method of stainless steel, the total amount of these impurity elements is about 250 ppm even for highly purified stainless steel.

  On the other hand, in the case of using a high-purity alloy raw material such as electrolytic iron, electrolytic nickel, and metallic chromium, and performing melting by a vacuum induction melting apparatus, [P] and [S] are about 10 to 20 ppm, and [N ], [O] can be reduced to about 20 to 30 ppm, and [C] can be reduced to about 30 to 50 ppm, and an alloy ingot having a considerably high purity can be produced. However, it is necessary to use an expensive high-purity alloy raw material.

Further, it is known that a conventional refractory crucible vacuum induction melting method is difficult as a method for removing impurity elements such as [P] and [N] from a molten stainless steel with a high Cr content. In normal refining of [P] in steel materials, an oxidation refining method is used. [P] is converted to slag-like phosphorous oxide (P ₂ O ₅ ), and [P] is absorbed and removed by slag. To do. However, in the case of stainless steel with a high Cr content, [P], which is an alloy component, is also oxidized simultaneously with [P]. That is, in the normal refractory crucible vacuum induction melting method, there is a principle problem that oxidation refining is difficult in the case of stainless steel with a high Cr content.

On the other hand, as a technique for removing impurity elements such as phosphorus [P], a reduction refining technique as shown in Non-Patent Document 1 has been reported in the 1970s. In this report, an electroslag remelting (ESR) apparatus is used. In a water-cooled copper crucible with an inner diameter of 70 mm, CaF ₂ is used as molten slag, and a slag bath in which metallic Ca is dissolved is formed, and a melting and refining test of stainless steel (SUS304) is performed as a consumable electrode material. . As a result, [P], [Sn], [Pb], [As], [Sb], [Bi], [O], [S], [Se], [Te] which are impurity elements in molten stainless steel , [N] etc. can be removed and refined. This report is an early report of the reduction refining technique using metal Ca, and shows that impurity elements such as phosphorus [P] can be removed and refined in principle from the Cr-containing alloy by the reduction refining process. It is. However, in the ESR process used in this report, it is necessary to supply an alternating current to the slag bath itself and form the slag bath by the resistance heat generation. Therefore, when the amount of metallic Ca added to increase the refining effect is increased, the electrical resistance of the slag bath itself is remarkably lowered, and a sufficient calorific value cannot be obtained, making it difficult to form the slag bath itself. That is, there was a problem as a practical process.

In the 1980's, a reduction refining technique using a magnetic levitation type induction melting apparatus (cold crucible induction melting apparatus) using a water-cooled copper crucible as shown in Patent Documents 1 to 3 has been reported. In this refining technique, melting of stainless steel itself is induction-heated to form a molten alloy pool, to which metal Ca and calcium fluoride (CaF ₂ ) are added as a refining agent, and an impurity element such as phosphorus [P] Is to be removed. In these reports, stainless steel (SUS316L) is dissolved by 0.8 kg to 2.0 kg using a water-cooled copper crucible (cold crucible) having an inner diameter of φ60 mm or an inner diameter of φ84 mm. Thereafter, a rod-shaped refining agent in which calcium fluoride and metal Ca are melted is directly added to stainless steel (molten metal pool) to cause a reaction. Thereby, removal and refining of impurity elements such as phosphorus [P] can be performed. However, similar to the report described in Non-Patent Document 1, these reports were also the scale of a small-scale principle confirmation test. That is, whether or not the reduction refining technique is established for a molten alloy pool of 10 kg or more that seems to be practical, or the specific conditions when it is established, has not been grasped.

In the reductive refining method described above, calcium fluoride (CaF ₂ ) is used as a flux to form a molten calcium fluoride layer, in which a state in which metal Ca is dissolved is formed. Then, the metal Ca and [P] in the molten alloy pool are reacted to form a Ca ₃ P ₂ compound, and the Ca ₃ P ₂ compound is absorbed in the calcium fluoride bath. Thereby, dephosphorization [P] is performed. In this refining reaction, it is indispensable to use a molten flux such as CaF ₂ that can dissolve metal Ca. Therefore, it is necessary to use a water-cooled copper crucible that is a container that does not react with molten CaF ₂ or Ca. That is, this reductive refining technique cannot be applied to a normal refractory crucible type vacuum induction melting method. Therefore, for practical use, it is necessary to establish a reduction refining technique using a large-sized cold-crucible induction melting apparatus.

On the other hand, the present inventors have established a large-scale cold-crucible induction melting technique using a water-cooled copper crucible having an inner diameter of 400 mm or more, as shown in Patent Document 4, as a large-scale cold-crucible induction melting technique. . If the conditions for establishing the above-described reduction refining technique with such a large cold-crucible induction melting apparatus can be grasped, it is expected to be a refining technique on a practical scale. Other documents describing the cold-crucible induction melting technique include those described in Non-Patent Document 2 and Patent Documents 5 to 7, for example.

Y. Nakamura et.al: Refining of 18% Cr-8% Ni Steel with Ca-CaF2 Solution, Transaction ISIJ, Vol. 16, (1976) p.623 Iwasaki, Sugaya, Fukuzawa: Melting ultra-low phosphorus stainless steel by cold crucible levitation melting: Iron and steel Vol.88 (2002) No.7, p.413 Japanese Patent Laid-Open No. 11-246910 JP 2002-69589 A JP 2003-55744 A Japanese Patent Laid-Open No. 11-310833 JP 2003-342629 A JP 2007-154214 A JP 2007-155141 A

Therefore, in order to determine whether or not the above-described reduction refining technique is established in a molten alloy pool of 10 to 50 kg class, the present inventors have introduced a cold-crucible induction with a vacuum chamber having a water-cooled copper crucible having an inner diameter of 220 mm. Using a melting apparatus, 20 kg of stainless steel (SUS310 composition) was melted, and a test was attempted in which calcium fluoride and metallic Ca were added thereto. The atmosphere in the vacuum chamber was an inert gas atmosphere in which Ar gas was introduced up to 600 to 800 hPa after the inside of the vacuum chamber was evacuated in order to suppress oxidation loss of Ca. Using commercially available ferrochrome, low carbon steel, electrolytic nickel, etc. as alloy raw materials, 20 kg of stainless steel (SUS310 composition) was melted to form a molten alloy pool. Thereafter, an attempt was made to add a refining agent in which 400 g of powdered calcium fluoride (CaF ₂ ) and granular metal Ca: 100 g were added to the molten alloy pool. It became clear that black smoke (dust) was generated, and it was not possible to observe the radiant light from the molten alloy pool surface in just a few seconds.

  Here, in a normal melting operation, the molten alloy pool state is maintained while observing the molten state of the molten alloy pool surface. For this reason, since the surface of the molten alloy pool cannot be observed normally, the present inventors immediately solidify the molten alloy pool in a water-cooled copper crucible with the high-frequency heating power turned off (after 1 minute of addition). The operation was performed.

After the melting and refining test, the water-cooled copper crucible was allowed to stand overnight, and the vacuum chamber was opened the next day. The black dust almost settled and dropped and adhered to the floor and walls of the vacuum chamber. It was. In the water-cooled copper crucible, there were solidified stainless steel ingots and solidified slag (CaF ₂ -Ca). Since these were reduced in diameter due to solidification shrinkage, they could be removed from the water-cooled copper crucible. After removing the solidified stainless steel ingot and CaF ₂ -Ca slag, when observing the state of damage in the water-cooled copper crucible, there is no situation eroded by the CaF ₂ -Ca slag and the water-cooled copper crucible itself is healthy. Was confirmed. In addition, the entire amount of the metal Ca added as a refining agent was melted, but the calcium fluoride partially adhered to the upper side of the solidified slag (CaF ₂ -Ca) as a powder, and the entire amount was melted. It became clear that it was not possible.

  Next, an analytical sample was cut out from the solidified ingot taken out (solidified stainless steel ingot), and an analysis investigation was performed. As a result, it was confirmed that the [P] concentration, which was about 0.018 wt% in the molten alloy pool stage, was as small as 0.014 wt%, but was removed and refined. That is, it has been confirmed that there is a possibility of removal and refining in principle even with a molten metal pool weight scale (practical scale) of 10 kg or more. However, in order to obtain the removal and refining effect efficiently, it has been found that it is indispensable to investigate appropriate refining conditions.

  That is, the problem to be solved by the present invention is a reduction refining technique of an impurity element such as phosphorus [P] with metal Ca, which has been confirmed in principle in a laboratory scale test of a cold-crucible levitation melting apparatus using a water-cooled copper crucible Is to specify a specific method for developing a refining technology on a practical scale in which the product ingot weight is, for example, 10 kg or more.

  As a result of intensive studies to solve the above problems, the present inventors have found that the addition rate of metallic Ca is indispensable for removing impurity elements from a practical scale molten alloy pool having a product ingot weight of, for example, 10 kg or more. As a result, the present inventors have found a range such as the component composition of the flux and the addition ratio thereof and have completed the present invention based on this finding.

That is, the present invention is a molten pool forming step of forming an alloy molten pool having components adjusted to a predetermined alloy composition in an inert gas atmosphere by charging the alloy raw material into a water-cooled copper crucible of a cold crucible induction melting device, And a refining step of removing an impurity element containing at least phosphorus by adding a refining agent to the molten alloy pool. The refining agent is a mixture of metallic Ca and a Ca halide composition flux, and the Ca halide composition flux is CaF ₂ -CaO in which 5-30 wt% of calcium oxide is mixed in calcium fluoride, and chloride in calcium fluoride. CaF _{2-CaCl 2} calcium was blended 5-30 wt%, or, CaF ₂ was 5-30 wt% blending calcium oxide and calcium chloride in the calcium fluoride - a (CaO + CaCl _2), wherein relative to the weight of the molten alloy pool The addition rate of metal Ca is 0.5 wt% or more, and the addition rate of the Ca halide composition flux with respect to the weight of the molten alloy pool is set to be equal to or higher than the addition rate of the metal Ca.

  Further, in the present invention, after the refining agent is added to the molten alloy pool, 1/2 of the added metal Ca is equal to or longer than a time T1 corresponding to 1/2 of the time required for all of the added refining agent to melt. It is preferable to hold the molten alloy pool for a time T2 or less lost by evaporation.

  Furthermore, in the present invention, it is preferable to produce the ultra high purity alloy ingot having a phosphorus content of 2 ppm or less by performing the molten pool forming step and the refining step a plurality of times.

  Furthermore, in the present invention, the ingot produced by the method for producing the ultra-high purity alloy ingot described above is used as a primary ingot, and the primary ingot is introduced into the water-cooled copper crucible of the cold crucible induction melting apparatus, A second molten pool forming step for forming a molten alloy pool, and after adding a second refining agent to the molten alloy pool of the primary ingot, the inert gas in the chamber is exhausted and the exhaust state is maintained for 15 minutes or more. And a second refining step for removing impurity elements including at least carbon and calcium, wherein the second refining agent is an oxide that is an oxide of a predetermined main component element of an alloy composition such as iron oxide and the Ca. In order to oxidize the impurity element including at least carbon and calcium in the molten alloy pool of the primary ingot, the added weight of the oxidizing agent is a mixture with a halide composition flux. In the second refining step, the addition rate of the Ca halide composition flux with respect to the weight of the molten alloy pool of the primary ingot is set to 0.2. It is preferable to set it to 5 wt% or more and 5.0 wt% or less.

  Further, in the present invention, it is preferable to add a deoxidizing element-based alloy component to the ingot produced by the above-described method for producing an ultra-high purity alloy ingot to form an alloy.

Further, in the present invention, the ingot produced by the above-described method for producing an ultra-high purity alloy ingot is used as a secondary ingot, and the secondary ingot is supplied to a water-cooled copper dish-like container of a cold hearth type electron beam melting apparatus. Forming a third molten metal pool in the water-cooled copper dish-shaped container and in the water-cooled copper mold adjacent to the water-cooled copper dish-shaped container at a pressure lower than 5 × 10 ⁻⁴ mbar. And a third refining step for removing carbon as an impurity element by adding a third refining agent to the molten alloy pool in the water-cooled copper dish-shaped vessel, wherein the third refining agent is iron oxide or the like In order to oxidize all the carbon as the impurity element in the molten alloy pool of the secondary ingot, the added weight of the third refining agent 1.0 of the calculated weight calculated in It is preferable to make it not less than twice and not more than 4.0 times.

  According to the present invention, a practical scale ingot of an ultra-high purity alloy having a product ingot weight of, for example, 10 kg or more can be manufactured by melting.

It is a mimetic diagram showing a cold crucible induction dissolution device. It is a graph which shows the relationship between the addition rate {Ca} M of metal Ca, a dephosphorization [P] rate, and a denitrification [N] rate. It is a graph which shows typically the change of the dephosphorization rate and the denitrification rate with respect to the retention time of a molten alloy pool. The figure which modeled metallic Ca reduction refining reaction (Figure 4 (a)), and the graph which shows the example which calculated two parameters (Kmelt (Ca + Flx), Kev (Ca)) in reaction model from test result (Figure 4 (b), (c)). It is a graph which shows the relationship between the internal diameter of a water-cooled copper crucible, and the optimal holding time after adding a refining agent. Correlation between the addition rate {Flx} M of the Ca halide composition flux and the decarburization [C] rate relative to the iron oxide addition ratio (WFe ₃ O ₄ / MFeO) and the molten alloy pool weight in the iron oxide addition / vacuum oxidation refining It is a graph. Correlation between iron oxide addition ratio (WFe ₃ O ₄ / MFeO) in iron oxide addition / vacuum oxidation refining and Ca halide composition flux addition rate {Flx} M to molten alloy pool weight and desiliconization [Si] rate It is a graph. It is a schematic diagram which shows a cold hearth type electron beam melting apparatus.

  Hereinafter, embodiments for carrying out the present invention will be described with reference to the drawings. In addition, this invention was made | formed using the result of the commissioned research from the Ministry of Education, Culture, Sports, Science and Technology.

(Ca reduction refining method using cold crucible induction melting apparatus)
As described above, according to the present invention, the alloy raw material is introduced into the water-cooled copper crucible 3 of the cold crucible induction melting apparatus 1 by the raw material feeder 2, for example, and the alloy raw material is induction melted by, for example, the coil 5 in an inert gas atmosphere. And a molten metal pool forming step for forming a molten alloy pool 6 whose components are adjusted to a predetermined alloy composition, and a refining step for removing an impurity element containing at least phosphorus by adding a refining agent to the formed molten alloy pool 6 And a method for producing an ultra-high purity alloy ingot. The refining agent is a mixture of metallic Ca and Ca halide composition flux. The Ca halide composition flux the calcium oxide to calcium fluoride CaF ₂ -CaO formulated 5-30 wt%, calcium chloride calcium fluoride CaF _{2-CaCl 2} was blended 5-30 wt%, or, calcium fluoride CaF ₂ that calcium oxide and calcium chloride were formulated 5-30 wt% - a (CaO + CaCl _2). In this manufacturing method, the addition rate of the metal Ca with respect to the weight of the molten alloy pool 6 is set to 0.5 wt% or more, and the addition rate of the Ca halide composition flux with respect to the weight of the molten alloy pool 6 is set to be higher than the addition rate of the metal Ca. (See attached FIG. 1).

  According to the method for producing an ultra high purity alloy ingot of the present invention, an ultra high purity alloy ingot having a product ingot weight of 10 kg or more can be produced. In order to perform a certain amount of hot working and form a part on a practical scale, a product ingot weight of at least about 10 kg is required.

  The component composition of the ultra-high purity alloy ingot can be varied depending on the purpose of use. That is, the manufacturing method of the present invention includes an alloy material mainly composed of Fe, an alloy material mainly composed of Ni, an alloy material mainly composed of Fe-Ni, or an alloy material mainly composed of Co, etc. It can be applied to alloy raw materials. Preferably, the production method of the present invention is applied to an Fe-based Ni-based alloy material mainly composed of Fe-Ni. Further, ultra-high purity means that the content of at least P, which is an impurity element, is extremely small as compared with ingots of the same type of alloy that are already widely used. Furthermore, ultra-high purity means that the contents of impurities such as P, S, N, Sn, and Pb, which are used as impurity elements, are generally very small compared to ingots of the same type of alloys that are already widely used. Means that.

In order to carry out the production method of the present invention, a magnetic levitation type cold crucible induction melting (CCIM) type melting apparatus (cold crucible type induction melting apparatus 1) having a water-cooled copper crucible 3 as shown in FIG. Is indispensable and cannot be applied to the induction melting method using a general refractory crucible. This is because it is necessary to use a Ca halide flux such as calcium fluoride (CaF ₂ ) as a refining flux. This is because, in the normal refractory crucible system, the refractory crucible is significantly melted by molten calcium fluoride, etc., and there is a risk of causing an accident such as erosion of the water-cooled copper coil for heating and a steam explosion.

  The inner diameter D (m) of the water-cooled copper crucible 3 of the cold crucible induction melting apparatus 1 is preferably 0.2 m or more in diameter in order to form a molten alloy pool 6 of 10 kg or more. When the inner diameter of the water-cooled copper crucible 3 is 0.2 m or less, the weight of the molten alloy pool 6 that can be formed is reduced, which is not practical for forming an molten alloy pool of 10 kg or more.

  For melting by the CCIM method, it is necessary to provide a vacuum chamber 4 for controlling the atmosphere. In order to prevent evaporation loss of alloy components from the alloy to be melted, it is effective to place the inside of the vacuum chamber 4 in an inert gas atmosphere into which Ar gas, He gas, or the like is introduced. In order to obtain an inert gas atmosphere, it is desirable to introduce Ar gas or the like after exhausting the inside of the vacuum chamber 4 with a vacuum pump in advance. This is because molten metal Ca used as a refining agent is very active, and therefore, if oxygen gas or the like is present in the vacuum chamber 4, Ca is oxidized and consumed before the refining reaction. In the reductive refining using metal Ca, it is desirable to reduce leakage from the vacuum chamber 4 as much as possible.

  In an actual refining operation, in the water-cooled copper crucible 3 of the cold crucible induction melting apparatus 1, an alloy molten pool 6 whose components are adjusted to a predetermined alloy composition is formed, and then the weight M (kg) of the molten alloy pool 6 is set. On the other hand, an impurity element such as phosphorus [P], sulfur [S], nitrogen [N], tin [Sn], lead [Pb], boron [B] is added by adding a refining agent that satisfies the following conditions: Remove and refine. An alloy solidified skull layer 8 is formed under the molten alloy pool 6. Such a state of Ca reduction refining is schematically represented as shown in FIG. Satisfying the refining conditions as described below is an indispensable condition for removing and refining the impurity elements and melting, for example, an Fe-based Ni-based alloy ingot of ultra-high purity (very low impurities).

  Using the cold-crucible induction melting apparatus 1 having a water-cooled copper crucible 3 having an inner diameter of φ220 mm, a number of tests and examinations were conducted, and the refining conditions revealed were as follows.

(1) The refining agent is a mixture of metallic Ca and Ca halide composition flux.
In the reduction refining method using metal Ca, it is necessary to hold metal Ca stably. The boiling point of the metallic Ca alone is 1484 ° C. The molten metal pool temperature of, for example, an Fe-based Ni-based alloy formed by the CCIM method is about 1520 ° C. (Fe-based) to about 1450 ° C. (Ni-based). Therefore, when metal Ca is added to the molten alloy pool 6 alone, the metal Ca having a boiling point of 1484 ° C. is almost evaporated and removed, so that there is no metal Ca for the refining reaction. Therefore, a means for reducing the vapor pressure of the metal Ca and causing the metal Ca to coexist with the molten alloy pool 6 is required. It is known that metal Ca dissolves in a molten Ca halide such as molten calcium fluoride (CaF ₂ ). The melting point of calcium fluoride is about 1410 ° C., and in the case of an Fe-based alloy, the temperature of the molten alloy pool 6 is relatively high. Therefore, the molten slag layer 7 is formed by heat transfer from the molten alloy pool 6. Is possible.

(2) Regarding the component composition of the Ca halide composition flux, calcium fluoride (CaF ₂ ) alone is used in Non-patent Document 1 and Patent Documents 1 to 3. However, in the case of stainless steel or Ni-based alloy with a high Ni content, the temperature of the molten alloy pool formed by the CCIM method is low, so that calcium fluoride (CaF ₂ ) alone is not melted and powdered. It was often observed that the refining reaction ended with the body intact. In such a case, it becomes difficult to effectively hold the metal Ca in the flux bath, and it becomes difficult to obtain the impurity removal refining effect. As a result of many tests, the present inventors use calcium flux (CaF ₂ ) as a main component, but use a flux added with a compound that lowers the melting point, and the added flux needs to be easily dissolved. I found. Calcium oxide (CaO) or calcium chloride (CaCl ₂ ) is selected as a compound that is added to calcium fluoride (CaF ₂ ) to lower the melting point and has little influence on the refining reaction, and these are selected as calcium fluoride (CaF ₂ ). It was found that the flux composition blended in ₂ ) is appropriate.

The specific composition of the Ca halide composition flux is CaF ₂ —CaO (5-30 wt%), CaF ₂ —CaCl ₂ (5-30 wt%), CaF ₂ — (CaO + CaCl ₂ ) (5-30 wt%), etc. It is. Here, for example, CaF ₂ —CaO (5 to 30 wt%) is obtained by blending calcium oxide with 5 to 30 wt% (ratio to calcium fluoride) in calcium fluoride. CaF ₂ − (CaO + CaCl ₂ ) (5 to 30 wt%) is a mixture of calcium fluoride and calcium chloride combined with calcium fluoride in a total of 5 to 30 wt%.

Ni relative content less Fe based alloys, among which a high melting point _CaF 2 -20CaO relatively (wt%) composition flux _(CaF 2 -CaO _(20wt%) Flux composition) is valid Met. The addition of small amounts of CaO and CaCl ₂ to CaF _2, melting point of the flux is lowered, it is easy to form a molten slag layer 7 by heat transfer from the molten alloy pool 6. However, when the CaO to be added is 30 wt% or more, it is presumed that the undissolved CaO is generated, but the flux becomes difficult to be dissolved and the fluidity of the slag bath is lowered. It was judged that it was not. Further, the addition of calcium chloride is effective for an alloy having a high Ni content and a low melting point because it has a large effect of lowering the melting point. However, since evaporation loss is severe, the addition of 30 wt% or more was also found to be 30 wt% or less because the refining operation became unstable.

(3) Addition rate of metallic Ca to weight M (kg) of molten alloy pool 6 ({Ca} M (wt%)), and addition rate of Ca halide composition flux to weight M (kg) of molten alloy pool 6 ( As a result of a number of tests, {Flx} M (wt%)) is effective to satisfy the following range. Here, {Ca} M (wt%) and {Flx} M (wt%) are defined by the following equations.
{Ca} M = WCa / M × 100
{Flx} M = WFlx / M × 100
WCa: Weight of added metal Ca (kg)
WFlx: Weight of added Ca halide composition flux (kg)
M: Weight of molten alloy pool 6 (kg)
In addition, the weight of the metal Ca added as a refining agent and the weight of the flux are often arranged as a concentration in the flux in normal handling. However, since it is easier to directly grasp the required metal Ca amount and flux amount by organizing as a weight ratio with respect to the weight M (kg) of the molten alloy pool 6, here {Ca} M and {Flx} We decided to express those weights with a definition such as M. The weight M of the molten alloy pool 6 is equal to the weight of the alloy raw material before being put into the water-cooled copper crucible 3.

And it turned out that the specific requirement of {Ca} M and {Flx} M must satisfy the following equation.
0.5 ≦ {Ca} M and {Ca} M ≦ {Flx} M
When the mixture of the metal Ca and the Ca halide composition flux is added to the molten alloy pool 6, the evaporation of the metal Ca immediately starts and black smoke is generated. Under conditions where the metal Ca addition rate is low ({Ca} M <0.1), most of the added Ca is lost due to the initial addition Ca evaporation loss, resulting in almost no removal and refining effect. When the addition amount of metallic Ca is increased (0.4 <{Ca} M), the generated black smoke enters the vacuum chamber 4 within a few seconds to several tens of seconds after the addition, and from the surface of the molten alloy pool 6. The situation is such that the radiated light is also blocked, and it is difficult to observe the surface condition of the molten alloy pool 6.

At the initial stage of the present study, a test was performed under a condition in which the amount of metallic Ca to be added was reduced because of concern that the state of the surface of the molten alloy pool 6 could not be observed. For example, a 20 kg molten alloy pool (Fe-20Ni25Cr) is formed in a water-cooled copper crucible 3 having a diameter of 220 mm, and metal Ca: 30 g and CaF ₂ —CaO (8: 2): 270 g are added to the molten alloy pool. The test ({Ca} M = 0.15%, {Flx} M = 1.35%) was performed. Under such conditions, the field of view deteriorates due to Ca evaporation dust, but the surface itself of the molten alloy pool 6 can be observed.

However, under these conditions, the refining effect for removing impurities such as phosphorus [P] was only about 15 to 30% as the dephosphorization [P] rate, which was completely unsatisfactory. Here, the dephosphorization rate: ηp (%) and the denitrification [N] rate: ηN (%) are defined by the following equations.
ηp = ([P] 0− [P]) / [P] 0 × 100
ηN = ([N] 0− [N]) / [N] 0 × 100
[P] 0: [P] concentration before refining (wt%), [P]: [P] concentration after refining (wt%)
[N] 0: [N] concentration before refining (wt%), [N]: [N] concentration after refining (wt%)

  Further, the refining test was repeated under the condition that the amount of metallic Ca added is small ({Ca} M is small). As a result, even if the refining test for adding the metal Ca and the Ca halide composition flux is repeatedly performed, it is understood that the water-cooled copper crucible 3 is not damaged, and it is not necessary to visually observe the molten alloy pool 6 directly. I was able to grasp that. Then, the test which increased the addition rate {Ca} M of metal Ca and the addition rate {Flx} M of Ca halide composition flux was continued.

  In the refining operation (0.4 <{Ca} M) in which the addition rate of the metallic Ca is increased, the generated black smoke can enter the vacuum chamber 4 in several seconds to several tens of seconds after the addition of the metallic Ca. And even the radiated light from the surface of the molten alloy pool 6 is blocked, and it is impossible to visually observe the surface state of the molten alloy pool 6. However, it has been found that it is necessary to add a quantity of metal Ca so that even the radiated light from the molten alloy pool 6 cannot be observed, in order to obtain a refining effect.

  With respect to the addition rate {Ca} M of the metal Ca and the addition rate {Flx} M of the Ca halide composition flux, the weight M of the molten alloy pool 6 was varied in a range from 20 kg to 50 kg, and a number of conditions were varied. A refining test was conducted. When the results were arranged, a correlation as shown in FIG. 2 (a) was found between the dephosphorization rate (ηP) and {Ca} M. Although the variation of these correlations is large, the dephosphorization rate (ηP) increases as the Ca Ca addition rate {Ca} M to 6 wt M (kg) of the molten alloy pool increases, and {Ca} M is around 1.0. Thus, it has been found that the dephosphorization rate and the denitrification rate are close to the upper limit (80 to 100%). Melting ultra-high purity material ([P] <2ppm) with low impurity element content using inexpensive raw material with high impurity element content (eg [P] 0 = ferrochrome raw material containing about 200-300ppm) In order to do this, a plurality of such Ca reduction refining operations are required. Moreover, in order to minimize the number of times of refining, it is considered that at least 50% or more is necessary as the dephosphorization [P] rate in one refining operation. In order to obtain a dephosphorization [P] ratio of 50% or more, the addition ratio of metal Ca needs to satisfy at least 0.5 ≦ {Ca} M from the correlation shown in FIG.

  In addition, since the dephosphorization [P] rate almost reaches the upper limit when {Ca} M is in the vicinity of 1.0, a large refining effect cannot be expected even if the addition rate of metal Ca is excessively increased. Therefore, {Ca} M = 1.0 or so is desirable for the purpose of dephosphorization.

  On the other hand, as for denitrification [N], the correlation between the denitrification [N] rate and {Ca} M shown in FIG. 2 (b), which is almost the same as that of phosphorus, can be obtained. There is a tendency that the removal [N] rate is slightly lowered. This is presumably because nitrogen in the air easily flows into the vacuum chamber 4 and causes nitrogen contamination when a slight atmospheric leak occurs during the refining operation. Therefore, for the purpose of removing [N], {Ca} M = around 1.0 to 1.2 is desirable.

  In addition, regarding the addition rate of metal Ca, in principle, it is considered that as the amount of {Ca} M is increased, the effect of removing impurities is surely obtained. However, in the actual refining operation, if {Ca} M is made larger than 1.5, the Ca reduction refining operation itself is possible, but the amount of metallic Ca adhering to the ingot after refining operation and absorbed is large. Become. As a result, problems such as ignition of evaporated Ca dust easily occur in post-processing such as taking out the ingot. Furthermore, operation hindrance such as dust generation due to evaporation of [Ca] in the ingot in the next melting and refining process becomes remarkable. The inventors have experienced these problems. From the above, in order to perform a stable refining operation, it is preferable to set {Ca} M ≦ 1.5.

  Further, when {Flx} M is less than {Ca} M, the evaporation loss of the added metal Ca becomes remarkable, and stable refining operation cannot be performed. Therefore, the weight M (kg) of the molten alloy pool 6 is reduced. It was found that the condition that the addition rate {Flx} M (wt%) of the Ca halide composition flux was {Ca} M ≦ {Flx} M was necessary. Furthermore, in order to stably dissolve the metallic Ca in the Ca halide composition flux, it is desirable that {Flx} M is 1.5 times or more than {Ca} M.

  Regarding the addition of the Ca halide composition flux, it is preferable to add the Ca halide composition flux in advance before adding the mixed scouring agent of the metal Ca and the Ca halide composition flux. Thereby, a molten Ca halide composition slag layer can be formed in advance, and evaporation loss of the added metal Ca can be suppressed. The Ca halide composition flux added in advance is expressed as {Ca} M ≦ {Flx} M0 when the addition ratio {Flx} M0 (wt%) of the Ca halide composition flux with respect to 6 wt M (kg) of the molten alloy pool. It was effective to keep the amount satisfactory.

  On the other hand, in such a metal Ca—Ca halide reduction refining method, it is necessary to form the molten slag layer 7 by heat transfer from the molten alloy pool 6. Here, when the total amount of {Ca} M, {Flx} M, and {Flx} M0 is too large, it becomes difficult to form the molten slag layer 7. Numerous tests have shown that the total amount of {Ca} M, {Flx} M, and {Flx} M0 is preferably within 5% of 6 weight M of the molten alloy pool.

  The effect of removing and refining impurity elements other than [P] and [N], but [S] was easily removed. In addition, [Sn], [Pb], [Sb], and the like obtained a Ca reduction refining effect substantially the same as [P] and [N]. For boron [B], the de [B] rate at {Ca} M = 1.0 is about 20%, and separation and refining is possible.

  As described above, the addition rate {Ca} M of the metal Ca with respect to the weight of the molten alloy pool 6 is 0.5 wt% or more, and the addition rate {Flx} M of the Ca halide composition flux with respect to the weight of the molten alloy pool 6 is changed to the metal Ca. The addition rate of {Ca} M or more can secure 50% or more as the dephosphorization [P] rate. However, the removal and refining effect varies greatly depending on the retention time of the molten alloy pool 6 after the refining agent is added to the molten alloy pool 6. This is one of the factors that increase the variation in the correlation shown in FIG. 2. In order to obtain a high removal and refining effect, it is important to secure an appropriate molten metal pool 6 holding time. .

Therefore, the present inventors conducted the following test.
Using a cold crucible induction melting apparatus 1 having a water-cooled copper crucible 3 having an inner diameter of 220 mm, a stainless steel material (Fe-20Ni25Cr, Fe-35Ni25Cr) or the like is put into the water-cooled copper crucible 3 and weight M (20 kg, 40 kg, 50 kg) The molten alloy pool 6 is formed. Thereafter, Ca halide composition flux _{(CaF 2 -CaO (8: 2} ), CaF 2 -CaCl 2 -CaO (8: 1: 1) , etc.) added ({Flx} M0 = 1.5% ) and molten slag Layer 7 is pre-formed. Thereafter, a refining agent that is a mixture of metallic Ca and Ca halide composition flux is added under the conditions of {Ca} M = 1.0% and {Flx} M = 1.5%, and the molten alloy is melted for a certain period of time. The pool 6 and the molten slag layer 7 are held (2 to 60 minutes). Thereafter, the high-frequency heating power supply is immediately stopped, and the molten alloy pool 6 is rapidly solidified in the water-cooled copper crucible 3. Then, impurity elements such as phosphorus [P] and nitrogen [N] in the ingot that has been rapidly solidified are analyzed.

  The result of carrying out the above test is shown in FIG. FIG. 3 is a graph schematically showing changes in the dephosphorization rate and the denitrification rate with respect to the holding time of the molten alloy pool 6.

  The conditions of the test results schematically shown in FIG. 3 (a) are as follows. The weight of the molten alloy pool 6 was 20 kg. The weight of the added metal Ca is 200 g ({Ca} M = 1.0), the weight of the Ca halide composition flux is 300 g ({Flx} M = 1.5), and the weight of the Ca halide composition flux added in advance is 300 g ({Flx} M0 = 1.5).

  The test result conditions schematically shown in FIG. 3B are as follows. The weight of the molten alloy pool 6 was 50 kg. The weight of the metal Ca to be added is 500 g ({Ca} M = 1.0), the weight of the Ca halide composition flux is 750 g ({Flx} M = 1.5), and the weight of the Ca halide composition flux added in advance is It was set to 750 g ({Flx} M0 = 1.5).

  As can be seen from FIGS. 3A and 3B, both the de [P] rate and the de [N] rate are low values immediately after the refining agent (mixture of metallic Ca and Ca halide composition flux) is added. . However, in the case of the 20 kg molten alloy pool 6 (FIG. 3A), the holding time is 4 to 7 minutes, and in the case of the 50 kg molten alloy pool 6 (FIG. 3B), the holding time is 10 to 17 minutes. The removal [P] rate and the removal [N] rate reach the maximum values (about 90 to 95%). However, when the molten metal state (dissolved state) is maintained for a longer time, on the contrary, both the [P] rate and the [N] rate tend to decrease again. Such a behavior showed a similar tendency even when the weight of the molten alloy pool 6 was 40 kg. However, the time when the de [P] rate and the de [N] rate are maximum is 40 kg for the molten alloy pool 6 (weight of added metal Ca: 400 g ({Ca} M = 1.0), Ca halide) The weight of the composition flux: 600 g ({Flx} M = 1.5) The weight of the Ca halide composition flux added in advance: 600 g ({Flx} M0 = 1.5)) was 8 to 11 minutes.

  From these test results, it was found that when the amount of the refining agent to be added is small (when the weight of the molten alloy pool 6 is small), the time until the maximum refining effect is obtained tends to be short.

  Further, in order to obtain a high [P] rate and a [N] rate, the retention time of the molten alloy pool 6 after the addition of a refining agent (mixture of metal Ca and Ca halide composition flux) is an appropriate time. The need to control the range is apparent from FIG. Therefore, as a refining technique, depending on conditions such as the inner diameter D of the water-cooled copper crucible 3, the weight M of the molten alloy pool 6, and the amount of refining agent to be added (the weight of the metal Ca and the weight of the Ca halide composition flux), It is essential to clarify the proper molten metal retention time after the addition of the refining agent (mixture of metallic Ca and Ca halide composition flux).

  Further, when the amount of the molten alloy pool 6 is different and the amount of the metallic Ca added and the amount of the Ca halide composition flux are different accordingly, the holding time of the molten alloy pool 6 at which the de [P] rate is maximized is different. From the test results, the following refining reaction mechanism was estimated.

Melting of the metallic Ca and Ca halide composition flux added to the molten alloy pool 6 is started by heat transfer from the molten alloy pool 6. As a result, the molten slag layer 7 (Ca + flux layer) is formed, and as the amount of metallic Ca dissolved in the slag increases, the following dephosphorization [P] reaction and denitrification [N ] The reaction is promoted.
2 [P] +3 (Ca) → (Ca ₃ P ₂ )
2 [N] +3 (Ca) → (Ca ₃ N ₂ )

On the other hand, (Ca) in the molten slag layer 7 continues to evaporate as Ca (g (gas)). Due to this evaporation loss, it is considered that the Ca concentration in the molten slag layer 7 gradually decreases once Ca is dissolved in the molten slag layer 7. Since the vapor pressure of the metallic Ca alone is very high and the evaporation rate of Ca is considerably increased, the decrease in the Ca concentration in the molten slag layer 7 also proceeds at a considerable rate. When the Ca concentration in the molten slag layer 7 is lowered, the (Ca ₃ P ₂ ) compound, the (Ca ₃ N ₂ ) compound, etc. once converted into a Ca compound and absorbed in the molten slag layer 7 are decomposed, [P], [N], etc. return to the molten alloy pool 6 again. That is, a so-called dephosphorization reaction such as the following formula occurs in the molten slag layer 7.
(Ca) → Ca (g) ↑
(Ca ₃ P ₂ ) → 2 [P] +3 (Ca)
(Ca ₃ N ₂ ) → 2 [N] +3 (Ca)

  For this reason, it is meaningless to take a molten metal refining time longer than necessary. When the molten alloy pool 6 is held for an extremely long time, all the [P] and [N] once absorbed in the molten slag layer 7 return to the molten alloy pool 6 side and no refining effect is obtained. It is also possible. Therefore, in the reductive refining method using metal Ca, time management after the refining agent is added is extremely important. Since the weight of the metal Ca to be added and the weight of the Ca halide composition flux differ depending on the weight of the molten alloy pool 6, in the management of the molten metal holding time, the weight of the added metal Ca and the weight of the Ca halide composition flux is used. Accordingly, the range of the appropriate molten metal holding time is different. Moreover, since the quantity of the molten slag layer 7 formed also changes with the internal diameters of the water-cooled copper crucible 3, this needs to be considered.

  Therefore, the present inventors modeled the melting and refining situation schematically shown in FIG. 1 as shown in FIG. FIG. 4 is a diagram modeling a reductive refining reaction by adding a refining agent (mixture of metallic Ca and Ca halide composition flux).

The inventors constructed the following reaction model.
Due to heat transfer from the molten alloy pool 6, the metal Ca added to the upper part of the molten alloy pool 6 and the Ca halide composition flux are sequentially dissolved to form a molten slag layer 7 (time steps 0 to 4 in FIG. 4). To the stage). When the molten slag layer 7 is formed, the [P] concentration and the [N] concentration in the molten alloy pool 6 and the (Ca ₃ P ₂ concentration in the molten slag layer 7 due to local equilibrium with Ca in the molten slag layer 7. ) Concentration and (Ca ₃ N ₂ ) concentration are determined, and the de [P] and de [N] reactions proceed. However, once the molten slag layer 7 is completely formed (time step 4 in FIG. 4), the Ca concentration in the molten slag layer 7 increases with time due to Ca evaporation loss from the molten slag layer 7. descend. This is a reaction model in which the [P] concentration and the [N] concentration in the molten alloy pool 6 again increase due to the local equilibrium reaction according to the decrease in the Ca concentration.

  In this reaction model, Kmelt (Ca + Flx) (kg / s) is defined as a melting rate constant at which the initial added metal Ca and added Ca halide composition flux are melted by heat transfer from the molten alloy pool 6 per unit area unit time. / m2). In addition, the Ca evaporation rate constant per unit time unit area in which the metallic Ca evaporates from the molten slag layer 7 was Kev (Ca) (kg / s / m 2). Then, Kmelt (Ca + Flx) and Kev (Ca) were used as parameters, and the values of these parameters were determined so as to be consistent with the test results when the weight of the molten alloy pool 6 was 20 kg, 40 kg, and 50 kg.

When 200 g of CaF ₂ CaO (8: 2) flux is added into the φ220 mm water-cooled copper crucible 3, the molten slag layer 7 is formed in about 1 to 2 minutes. From this, the dissolution rate constant (Kmelt (Ca + Flx)) of the scouring agent is estimated to be in the range of about 0.017 to 0.088 (kg / s / m 2). When the metal Ca is contained, the slag melting point is lowered, so that it is expected that the speed is slightly higher than this. On the other hand, the evaporation rate constant (Kev (Ca)) of Ca vapor from the molten slag layer 7 (for example, CaCaF ₂ CaO (25/60/15 wt%), etc.) is obtained by evacuation using the vapor pressure of pure Ca. When calculated from the equation for obtaining the evaporation rate constant, it is about 1-2 (kg / s / m 2). However, since Ca reduction refining is performed, for example, in the vicinity of 1 atm of Ar gas, the evaporation rate is estimated to be lower than this. Regarding this, the RGWard (JISI Vol201 (1963) p.11) shows that the evaporation rate of [Mn] in the molten steel pool is reduced to about 1/100 in an atmosphere of 1 atmosphere of Ar gas compared to that in a vacuum. ), It is expected that Kev (Ca) will be on the order of 0.01 to 0.02 (kg / s / m 2) if the same is assumed in the evaporation of Ca. Therefore, after adding the initial charge flux amount: {Flx} M0 = 1.5% (750 g) to the molten alloy pool 6 weight 50 kg to form the molten slag layer 7, the metal Ca: {Ca} M = 1.0% (500 g) + Ca halide composition flux: {Flx} M = 1.5% (750 g) was added, and the reaction was performed with respect to the test conditions in which the Ca refining test was carried out by changing the refining time. FIG. 4B shows the time required for dissolution of the added scouring agent when the dissolution rate constant of the scouring agent is changed using a model. In this test, the reaction time at which the highest de [P] rate was obtained was found to be about 10 to 17 minutes (600 to 1020 seconds). Therefore, it can be judged that the added scouring agent is completely dissolved in this time (600 to 1020 seconds) to form the molten slag layer 7. Thus, Kmelt (Ca + Flx) is in the range of about 0.03 to 0.06 (kg / s / m2). This value almost coincides with the rate constant estimated from the flux dissolution rate.

  Next, assuming that Kmelt (Ca + Flx) = 0.045 and changing the value of Kev (Ca), the obtained [P] rate is shown in FIG. 4C. Since the de [P] rate obtained in the actual refining test was about 88% to 94%, the corresponding Kev (Ca) value was 0.002 to 0.012 (kg / s / m2). This value is also generally consistent with the Ca evaporation rate constant value estimated from the vapor pressure of pure Ca, and a reasonable value is obtained.

Similarly, when 40 kg and 20 kg of molten metal pool conditions are used, values representing test results are obtained from the reaction model, and Kmelt (Ca + Flx) is 0.04 to 0.07 and 0.03 to 0.03, respectively. 06 (kg / s / m2), and Kev (Ca) becomes 0.002 to 0.012 and 0.002 to 0.012 (kg / s / m2), respectively. In Patent Document 2 and Non-Patent Document 2, a test result in which a 1.6 kg molten pool is created in a φ84 mm water-cooled copper crucible and a Ca + CaF ₂ scouring agent is added thereto and the refining time is changed is reported. Yes. As a result, it is disclosed that the highest de [P] rate (about 85 to 95%) is obtained 1 to 2 minutes after the addition. When this data was similarly analyzed using a reaction model, the Kmelt (Ca + Flx) in the test with a φ84 mm water-cooled copper crucible was about 0.04 to 0.09 (kg / s / m2), Kev. (Ca) is about 0.008 to 0.0027 (kg / s / m 2). Compared to the test with φ220mm water-cooled copper crucible 3, all the values are almost the same, but they are almost the same, and this reaction was applied to Ca reduction refining test using water-cooled copper crucible 3 of almost all diameters. A model can be applied.

  As a result of the above examination, the range of Kmelt (Ca + Flx) when completely melting a refining agent (mixture of metallic Ca and Ca halide composition flux) that satisfies a 6 to 20 kg molten alloy pool weight is In consideration of data variation, etc., 0.03 ≦ Kmelt (Ca + Flx) ≦ 0.1, while Kev (Ca) is about 0.002 ≦ Kev (Ca) ≦ 0.012. It has been found.

  If these values are used, the molten metal pool 6 is formed in the water-cooled copper crucible 3 of an arbitrary size, and an appropriate molten metal is maintained when a refining agent composed of a mixture of metallic Ca and Ca halide composition flux is added. You can set the time. Quantitatively, after adding a refining agent (mixture of metal Ca and Ca halide composition flux) to the molten alloy pool 6, an appropriate molten alloy pool 6 retention time T (min) is expressed as T1 ≦ T ≦ Satisfying the condition of T2 (min) is an effective condition for obtaining a high dephosphorization [P] rate, denitrification [N] rate, and the like.

Times T1 and T2 can be set as follows. (See Figure 3)
T1 (min) is ½ of the time required for the refining agent to completely melt after the refining agent (mixture of metal Ca and Ca halide composition flux) is added.
T1 = WCa + Flx / S / Kmelt (Ca + Flx) / 60/2
WCa + Flx = M × ({Ca} M + {Flx} M) / 100
WCa + Flx: Total weight of refining agent (kg)
S: πD ^2/4: horizontal cross-sectional area (m2) of the water-cooled copper crucible 3, D: inside diameter of the water-cooled copper crucible 3 (m)
Kmelt (Ca + Flx) (kg / s / m2) = twice the time T1, calculated as 0.1, for the added scouring agent to completely dissolve and form the molten slag layer 7 Time required. This time (T1 × 2) is the time at which the refining effect is maximized, but in reality, the variation is large. Therefore, the time T1 which is ½ of the time is set as the shortest molten alloy pool 6 holding time. It is necessary to hold the molten metal for a time longer than T1.

T2 (min) is a time during which 1/2 of the added metal Ca is lost from the molten slag layer 7 due to evaporation loss.
T2 = WCa / 2 / S / Kev (Ca) / 60
WCa = M × {Ca} M / 100
WCa: Weight of added metal Ca (kg)
The time T2 calculated as Kev (Ca)) (kg / s / m2) = 0.002 is the time when 1/2 of the added metal Ca evaporates and is lost, and the molten metal is retained within the time T2 or less. Need to be terminated.

  In actual refining operation, it is not necessary to hold until half of the added metal Ca is lost, and qualitatively, the added refining agent (mixture of metal Ca and Ca halide composition flux) is completely melted. Then, the refining effect of the time for forming the molten slag layer 7 (about twice as long as T1) is the maximum, and at that time, the molten alloy pool 6 is rapidly solidified as soon as possible, resulting in Ca evaporation loss. It is considered effective to suppress the reversion reaction. However, in the actual refining operation, after the addition of the refining agent, it is impossible to visually observe the surface of the molten alloy pool 6 by dust caused by Ca evaporation in a short time of several seconds to several tens of seconds. . Therefore, it is impossible to confirm by observation at which point the refining agent has completely melted, and it is necessary to manage only by time.

  In this case, as shown in FIG. 3, the impurity removal and refining effect has a steep gradient of time-dependent changes in the de [P] rate and de [N] rate at the stage where the refining agent melts. On the contrary, once the molten slag layer 7 is completely formed, when the Ca concentration proceeds and the Ca concentration in the molten slag layer 7 decreases, the de [P] rate and the de [N] rate are reduced. Time change is gradual. From the above, in order to ensure the removal [P] rate, it is more certain to set the holding time to be a little longer. Therefore, the upper limit is the time (T2) that ½ of the Ca content in the slag is lost by evaporation.

  With respect to this relationship, an example of a range of T1 and T2 of an appropriate molten metal holding time with respect to the inner diameter D (m) of the water-cooled copper crucible 3 is represented by the weight of the molten alloy pool 6 (L / D = 0.75: water-cooled copper The inner diameter of the crucible 3 × the weight of the cylindrical molten alloy pool 6 having a height (L) 6), the weight of the pre-added Ca halide composition flux ({Flx} M0 = 1.5), the Ca + flux refining conditions ({ The case of Ca} M = 1.0, {Flx} M = 1.5) is as shown in FIG.

FIG. 5 shows that the proper holding time varies depending on the inner diameter of the water-cooled copper crucible 3. Although there is some variation in test data and there are some problems in terms of accuracy, a more appropriate molten metal retention time range is Kmelt (Ca + Flx) (kg / s / m2) = 0.06 for T1. For the time T2, and the time T2, the time set as Kev (Ca)) (kg / s / m2) = 0.006 is within the range indicated by the dotted line and the two-dot chain line in FIG.
From the above, by using the relational expression shown here (T1 ≦ T ≦ T2), a refining operation by adding a metal Ca + Ca halide composition flux to be performed on the molten alloy pool 6 formed in a water-cooled copper crucible 3 of φ200 m or more. It is possible to set an appropriate time range.

  As described above, the addition rate {Ca} M of the metal Ca with respect to the weight of the molten alloy pool 6 is 0.5 wt% or more, and the addition rate {Flx} M of the Ca halide composition flux with respect to the weight of the molten alloy pool 6 is changed to the metal Ca. And the retention time T of the molten alloy pool 6 is T1 ≦ T ≦ T2, so that phosphorus [P], sulfur [S], tin [Sn], lead [Pb] It is possible to reduce the content of impurity elements such as] to 2 ppm or less.

For example, a ferrochrome raw material and a low carbon steel raw material (converter material) having a high impurity element content and a high-purity electrolytic Ni raw material are used as alloy raw materials. A cold crucible induction melting apparatus 1 having a water-cooled copper crucible 3 having a diameter of 220 mm is used. 50 kg of stainless steel (Fe-20Ni25Cr) is charged into the water-cooled copper crucible 3 to form the molten alloy pool 6. Thereafter, a metal Ca + flux refining agent (addition conditions: {Ca} M = 1.0, {Flx} M = 1.5) is added to the molten alloy pool 6 using CaF ₂ -20CaO flux. Then, a refining operation for holding the molten alloy pool 6 is performed for 11 to 15 minutes. In this case, the impurity element concentrations in the molten alloy pool 6 at the initial alloy raw material blending stage are about [P] 0 = 250 ppm and [N] 0 = 250 ppm. By reducing this with Ca, the de [P] rate = about 90% and the de [N] rate = about 85%, so that the impurity element concentrations are [P] 1 = 25 ppm, [N] 1 = 40 ppm. It will be about. When the second Ca reduction smelting is performed using this ingot as a raw material, [P] 2 = 3 ppm and [N] 2 = 6 ppm. Further, when the refining operation is performed for the third time, it is almost within the analytical limit of 2 ppm or less, that is, [P] 3 <2 ppm and [N] 3 <2 ppm.

  Actually, after performing CCIM-Ca reductive refining with φ220 mm water-cooled copper crucible 3 using a ferrochrome raw material having a high impurity element content and a low-carbon steel raw material (converter material) by such an operation, iron oxide GD-MS (glow glow), which can be used for chemical analysis and trace analysis of ordinary steel materials by adding vacuum to oxidative refining and further vacuum oxidative refining in the electron beam (EB) melting method to produce ingots. As a result of the impurity element analysis by the discharge mass) analysis method, the chemical analysis method has [P] <2 ppm (below the analysis limit), and the GD-MS analysis method has high purity up to about [P] = 0.5 ppm. Ingots (impurities removed) have been manufactured. Analytical values of other impurities measured by this GD-MS analysis method are as shown in Table 1, for example. Here, the GD-MS analysis method is an analysis method called glow discharge mass spectrometry, which is applied to a microanalysis of a semiconductor material or the like, and has a metal element / semiconductor element / insulator element up to about 0.01 ppm. This is an analytical method capable of microanalysis. Ingots produced using electrolytic iron, electrolytic nickel, and metallic chromium as melting raw materials by vacuum induction melting (VIM) -electron beam (EB) method, which is a conventional method for producing high-purity ingots, and commercially available stainless steel materials ( The analysis results of SUS304ULC are shown in Table 1 for comparison. This Ca reduction smelting method is very effective for removing and refining trace trump elements such as [P], [S], [Sn], [Sb], and [Pb]. It can be seen from Table 1 that it can be highly purified. As for nitrogen, since GD-MS analysis is impossible and the limit of chemical analysis is 5 ppm, it has been confirmed that impurity elements have been removed up to [N] <5 ppm.

  Table 1 shows a Ca reductive refining method, an oxidation refining method using a cold crucible induction melting apparatus 1 (details will be described later), and an oxidation refining method using a cold hearth type electron beam melting apparatus 11 (details will be described later). (Refer to Haas-type electron beam melting method)) (invention) Impurity element concentration in the ingot, and vacuum induction melting (VIM) -electron beam (EB) method (prior art) casting It is the table | surface which compared the impurity element density | concentration in a lump with the impurity element density | concentration in a commercial material (SUS304ULC).

  On the other hand, when using electrolytic iron, electrolytic Ni, metallic Cr, etc., which are commercially available high-purity raw materials that are relatively easily available, as the melting raw material (alloy raw material), the impurity element concentration in the molten alloy pool 6 at the initial blending stage [P] 0 = 10 ppm, [S] 0 = 10 ppm, and [N] 0 = 20 ppm. For this, it has been confirmed that [P] 1 <1 ppm and [S] 1 <1 ppm can be achieved only by performing Ca refining once. In addition, trace trump element impurities such as [Sn], [Pb], and [Sb] have been confirmed to be 1 ppm or less as analyzed by the GD-MS measurement method.

(Oxidative refining method using cold crucible induction melting equipment)
As described above, a practical-scale alloy ingot having a content of impurity elements such as phosphorus [P] of 2 ppm or less is produced by a Ca reductive refining method using a mixture of metal Ca and Ca halide composition flux as a refining agent. be able to.

  However, in an alloy ingot refined using metal Ca, a considerable amount of metal Ca remains in the ingot, particularly in an alloy ingot containing Ni as an alloy component. For example, in the case of an alloy having a Ni content of 0 (zero), the residual [Ca] concentration = about 0.02 wt%, and in the case of an alloy of about 20 wt%, the residual [Ca] concentration = about 0.05 wt%. It becomes. In an alloy having a Ni content of about 35 wt%, the residual [Ca] concentration is about 0.09 wt%. In an alloy having a Ni content of about 45 wt%, the residual [Ca] concentration = about 0.12 wt%. Further, in a Ni-based alloy having a Ni content of about 60 wt%, the residual [Ca] concentration may be about 0.5 wt%. Thus, a considerable amount of [Ca] remains in the alloy ingot refined using metal Ca. It is known that alkaline earth elements such as [Ca] and alkali metal elements in the alloy ingot deteriorate the corrosion resistance. Therefore, in order to produce an ultra-high purity alloy ingot, the content of these impurity elements is preferably reduced to 0.001 wt% or less, desirably 1 ppm or less.

  Further, during reduction refining using metal Ca, elements such as carbon and aluminum that are relatively easily reduced by Ca are likely to increase as impurities. [C] may cause an increase in concentration of about 30 ppm, and [Al] and [Si] may increase in concentration by about 0.005 wt%. It is considered that these are because organic substances and ceramics adhering in the chamber (vacuum chamber 4) are reduced by the metal Ca and absorbed into the molten alloy pool 6. This has been found to be a phenomenon that often occurs when reductive refining using metallic Ca.

  On the other hand, in an alloy that requires ultra-high purity, carbon [C] <10 ppm, silicon [Si] <0.01 wt%, or the like may be required. Therefore, a technique for removing [C], [Si] and the like together with [Ca] is required. Further, boron [B] may significantly deteriorate the corrosion resistance, and [B] <1 ppm may be required.

In order to remove the [C] impurity element in the molten stainless steel pool, it is necessary to promote the decarburization reaction by the reaction of the following formula.
[C] + [O] → CO (g) ↑
In order to proceed with this reaction, it is necessary to supply oxygen [O] into the molten alloy pool 6 and remove generated CO (g (gas)) to reduce the CO partial pressure and promote the reaction. . As a specific means, oxygen gas can be used for supplying oxygen to the molten alloy pool 6. However, for removing CO (g), it is effective to evacuate and continue to discharge the generated CO gas. Therefore, rather than supplying oxygen in a gaseous state, iron oxide serving as a solid oxygen source, etc. It is more effective to add an oxidant which is an oxide of a main component element of a predetermined alloy composition. In particular, residual oxygen [O] is often 5 ppm or less in an alloy ingot reductively refined with metal Ca, and is in a state where there is no oxygen source. Source supply is essential. As an oxidizing agent, iron oxide (Fe ₃ O ₄ , Fe ₂ O ₃ etc.) is used in the case of Fe-based alloys, iron oxide or nickel oxide is used in the case of Fe-Ni based alloys, nickel oxide is used in Ni-based alloys, Co In the case of a base alloy, cobalt oxide or the like can be applied. These oxidizing agents are solid oxidizing agents, in other words, oxidizing agents made of metal oxides.

In addition, as an alloy element other than carbon, oxides such as [Si], [Al], [Ti], [Zr], [Hf], [B], and [Ca] are heated compared to iron and nickel. In order to remove the active metal element that is mechanically stable, it is necessary to proceed with the following oxidative refining reaction and separate and remove each oxide in the slack.
[Si] +2 [O] → (SiO ₂ )
2 [Al] +3 [O] → (Al ₂ O ₃ )
[Ti] +2 [O] → (TiO ₂ )
[Zr] +2 [O] → (ZrO ₂ )
2 [B] +3 [O] → (B ₂ O ₃ )
[Ca] + [O] → (CaO)

[Ca] in the alloy ingot is removed by evaporation only by melting, but a slight amount may remain in the alloy ingot. In order to remove this completely, it is also important to carry out oxidative refining. When these elements are converted into oxides and absorbed as slag, they are reacted with CaO in the molten slag layer 7 to reduce the activity of these components in the molten slag layer 7 as a more stable compound. However, it is effective for the removal and refining reaction. For example, SiO ₂ formed by oxidizing [Si] reacts with CaO to form a stable compound such as Ca ₂ SiO ₄ , thereby reducing the SiO ₂ activity in the molten slag layer 7. It is effective to facilitate the progress of the oxidation reaction of [Si]. Similarly, with regard to Al ₂ O ₃ , TiO ₂ , B ₂ O _3, etc., it is effective for oxidative removal refining of these elements to react with CaO to form a compound and reduce the activity in the molten slag layer 7. . Therefore, it is effective to match the amount of CaO in the molten slag layer 7 with the amount of various oxides generated by the oxidation reaction.

  Also in the present oxidation refining, a cold crucible induction melting apparatus 1 having a water-cooled copper crucible 3 having an inner diameter of 0.2 m or more is used as in the above-described Ca refining refining (see FIG. 1). In this oxidation refining, the ingot (alloy ingot) manufactured by the above-described Ca reduction refining method using a mixture of metal Ca and Ca halide composition flux as a refining agent is further subjected to refining treatment (oxidative refining treatment). After the ingot (alloy ingot) manufactured by the above-mentioned Ca reduction refining method is put into the water-cooled copper crucible 3 and the molten alloy pool 6 whose components are adjusted to a predetermined alloy composition is formed in an inert gas atmosphere. By performing a refining operation that satisfies the following conditions, carbon [C], silicon [Si], aluminum [Al], titanium [Ti], zirconium [Zr], hafnium [Hf], boron [B], etc. Remove and refine impurity elements of active elements. Ar gas or He gas is used as the inert gas.

(1) As a refining agent, a refining agent (corresponding to a second refining agent) obtained by mixing an oxidizing agent which is an oxide of a predetermined main component element of an alloy composition such as iron oxide and a Ca halide composition flux is used as a molten alloy pool 6. To add to.
Fe ₂ O ₃ , Fe ₃ O ₄ , or the like is used as iron oxide serving as a typical oxygen source. As a result of the reaction, it is necessary to stably absorb the oxide generated in the molten slag layer 7. Therefore, as a flux having a high oxide absorption capacity, a flux obtained by adding CaO to a Ca halide flux such as calcium fluoride (CaF ₂ ) or calcium chloride (CaCl ₂ ) is used.

(2) The Ca halide-based flux is a component system having a lower melting point than the basic component CaF ₂ , and is CaF ₂ —CaO (5 to 30 wt%), CaF ₂ —CaCl ₂ (5 to 30 wt%), or CaF. _{Use a 2-} (CaO + CaCl ₂ ) (5-30 wt%) system (Ca halide composition flux).
These components of the Ca halide composition flux have the same composition as the Ca halide composition flux used in the Ca reduction refining described above. This is because it is effective for reaction promotion to lower the melting point of the flux and easily melt the added flux by heat transfer from the molten alloy pool 6 to form the molten slag layer 7. In order to absorb the oxide generated by the oxidation reaction, a flux containing CaO is more effective. A component system in which the amount of CaO necessary for making the generated oxide a stable compound is previously contained in the flux becomes more effective.

(3) Regarding the addition rate of the Ca halide composition flux, the flux addition weight ratio {Flx} M with respect to the weight of the molten alloy pool 6 satisfies the range of 0.5 ≦ {Flx} M ≦ 5.0 (wt%). about.
If the addition ratio of the Ca halide composition flux is too small, the effect of absorbing the generated oxide cannot be obtained. Therefore, a flux amount of at least 0.5 wt% of the weight of the molten alloy pool 6 is necessary. On the other hand, when {Flx} M is 5.0 wt% or more, as in the case of metallic Ca reduction refining, the heat transfer from the molten alloy pool 6 is insufficient and the molten slag layer 7 is difficult to be formed. % Is the upper limit.

(4) The added weight (WFexOy (kg)) of iron oxide (FexOy), which is a representative oxidizing agent in the above mixed refining agent, satisfies the following conditions.
0.2 × MFeO ≦ WFexOy ≦ 4.0 × MFeO
here,
MFeO = M / 100 × ([C / 12.01 + 2 × [Si] /28.09+1.5× [Al] /26.98+2× [Ti] /47.9+2× [Zr] /91.22+2× [Hf ] /178.49+1.5× [B] /10.811+ [Ca] /40.08- [O] /15.9994) / y × (55.85 × x + 16.0 × y)
M: Weight of molten alloy pool 6 (kg)
[C]: C concentration (wt%) in the molten alloy pool 6
[Si]: Si concentration (wt%) in the molten alloy pool 6
[Al] Al concentration in molten alloy pool 6 (wt%)
[Ti]: Ti concentration (wt%) in the molten alloy pool 6
[Zr]: Zr concentration (wt%) in the molten alloy pool 6
[Hf]: Hf concentration (wt%) in the molten alloy pool 6
[B]: B concentration (wt%) in the molten alloy pool 6
[Ca]: Ca concentration (wt%) in the molten alloy pool 6
[O]: O concentration in molten alloy pool 6 (wt%)

[C], [Si], [Al], and the like are slightly mixed from the fireproof sheet for receiving the scattered molten metal provided on the outer periphery of the water-cooled copper crucible 3 during Ca reduction refining. Therefore, in the calculation of the amount of oxygen to be added, it is necessary to consider the amount of pickup during the Ca reduction refining. Specifically, based on the test results, [C] is about 30 ppm, and [Si] and [Al] are about 50 ppm each. There is a possibility of pickup due to contamination. As the concentration of each element in the above calculation formula, it is necessary to use a blended calculated value using an analytical value of the charged raw material. [C] is CO gas, [Si] is SiO ₂ slag, [Al] is Al ₂ O ₃ slag, [Ti], [Zr], and [Hf] are TiO ₂ slag, In order to convert ZrO ₂ slag, HfO ₂ slag, [B] into B ₂ O ₃ slag, and [Ca] into CaO slag, divide by the atomic weight of each element to obtain the required number of moles of oxygen. This is an equation for calculating the weight required for supplying oxygen with FexOy iron oxide.

Although not included in the above formula, there are [alkali metal: EI], [alkaline earth metal: EII], [Y], [lanthanum-based / actinium-based metal: ER] and the like as active metals. In this case, it is necessary to add the amount of oxygen necessary to convert these into slag such as EI ₂ O, EIIO, Y ₂ O ₃ , ER ₂ O _3, etc. In addition, manganese [Mn] is not so active as compared to Fe, and is relatively difficult to remove by oxidative refining, and moreover, has a characteristic that it is easily removed by evaporation in high vacuum melting. Therefore, it is not a target for removal by oxidation refining here. When nickel oxide or cobalt oxide is used as the oxidizing agent, Fe (atomic weight: 55.85) in the above formula is replaced with Ni (atomic weight: 58.71) or Co (atomic weight: 58.93). Calculate it.

As a result of some smelting and refining tests, when decarburization [C] is the main purpose, the iron oxide weight that needs to be added: WFexOy (kg) is calculated as shown in FIG. : MFeO (kg) is required to be 0.2 times or more, and the addition rate of the flux to be added simultaneously: {Flx} M was found to be 0.5 wt% or more. When WFexOy is WFexOy <0.2 × MFeO and {Flx} M is {Flx} M <0.5, the de [C] rate is also greatly reduced, and it is difficult to obtain a refining effect. In addition, the oxidation removal refining of [Si] and [B], whose thermodynamic stability of the oxide is less than that of elements such as [Al] and [Ti], is shown in WFe ₃ O ₄ / FIG. As shown in the relationship between MFeO and {Flx} M and the de [Si] rate, if it is desired to obtain 50% or more as the de [Si] rate, the weight of iron oxide added is 1.0 × It is clear that it is effective that MFeO <WFexOy is satisfied and {Flx} M satisfies 3.0 <{Flx} M. Furthermore, in order to increase the removal [Si] rate more reliably, WFexOy has to be about 1.5 to 2 times that of MFeO. This is because not all the added iron oxide is used in the reaction. On the other hand, when the weight of iron oxide is added more than 4 times the required amount, oxidation loss such as Cr of the alloy component increases excessively, which is not desirable.

When the weight of iron oxide to be added is about 0.2 to 1.0 times that of MFeO and vacuum oxidation refining is performed, [C], [Al], [Ti], [Zr], [Zr] Although removal and refining of Hf], [Ca], etc. proceeds, removal of [Si], [B], etc. often does not proceed so significantly. However, [Si], [B], etc. can be removed and refined by setting the addition amount of iron oxide to 1.0 times or more of the MFeO amount and the Ca halide composition flux addition rate {Flx} M to 3 wt% or more. It becomes like this. For example, the [B] concentration is 50 ppm before refining → 1 ppm after refining. The concentration of [Si] is 0.22 wt% before refining → <0.01 wt% after refining. Furthermore, the amount of iron oxide added is 2.0 times or more the amount of MFeO, a CaF ₂ -CaO-based or CaF _2- (CaO + CaCl ₂ ) -based flux is used, and the Ca halide composition with respect to the weight of the molten alloy pool 6 When the flux addition rate {Flx} M is 3.0 wt% or more, the removal and refining of [Si] and [B] becomes easier.

  However, if the weight of iron oxide to be added is more than four times the amount of MFeO, the oxidation reaction becomes too intense, and oxidation loss such as Cr, which is an alloy element, will be generated significantly. That is not preferable.

Specifically, for example, a stainless steel having a composition of Fe-20Ni25Cr, a 50 kg molten metal using a commercially available low-cost raw material such as a ferrochrome raw material, a low carbon steel raw material, and an electrolytic Ni raw material that is a high-purity raw material. When the pool is formed, the impurity element concentration in the molten pool is [C] = 0.02, [Si] = 0.21, [Al] = 0.015, [Ti] = 0.0, [B ] = 0.005, [Ca] = 0.001, [O] = 0.02 wt%. The molar amount of oxygen necessary to oxidize them is calculated as 8.9 mol. If this oxygen is supplied by iron oxide Fe ₃ O ₄ , Fe ₃ O ₄ becomes MFeO = 515 g. Therefore, when about 800 g (about 1.5 times) of Fe ₃ O ₄ is added, [C] decreases to about 0.005 wt% and [Al] decreases to about 0.003 wt%. However, [Si] is about 0.19 wt%, and [B] is about 0.004 wt%. However, when 1300 g (2.5 times) of Fe ₃ O ₄ is added, [Si] decreases to 0.01 wt% or less, and [B] decreases to approximately 0.0001 wt%. The Ca halide composition flux used at this time is a CaF ₂ —CaO (8: 2) composition, and the addition rate is {Flx} M = 4.0% (2000 g).

  By this refining, even if commercially available low-priced raw materials such as ferrochrome and low carbon steel are used as melting raw materials, the impurity concentration ([Si ] <0.01 wt%) has been confirmed to be able to be removed and refined.

(5) After adding a refining agent (second refining agent) that is a mixture of Ca halide composition flux and iron oxide, the inert gas in the chamber (vacuum chamber 4) is evacuated (evacuated), Hold in the state (vacuum state (exhaust state)) for 15 minutes or more.
When a refining agent, which is a mixture of Ca halide composition flux and iron oxide, is added to the molten alloy pool 6 and the molten slag layer 7 is formed, the chamber is evacuated with an oil rotary pump, for example. If necessary, evacuation is performed with a mechanical booster pump, diffusion pump, etc., and oxidative refining is performed in a vacuum atmosphere.

  When the degree of vacuum is 10 hPa or less, fine droplets (splash) are scattered from the molten alloy pool 6. It is presumed that when [C] reacts with [O] and CO (g) gas is released, the surrounding molten metal is blown away. In this phenomenon, when the [C] concentration in the molten alloy pool 6 decreases, the reaction ends. For example, even in the case of the initial concentration [C] 0 = 250 ppm, intense splashing is completed in about 10 minutes, and the [C] concentration is about 20 to 60 ppm. In addition, it is observed that the surface temperature of the molten alloy pool 6 is increased by holding it under vacuum, the added flux is easily melted, and the molten slag layer 7 is stably formed.

  After such de [C] reaction, it is observed that molten slag pieces considered to be oxides float on the surface of the molten alloy pool 6. These are gradually absorbed by the molten slag layer 7 of Ca halide composition flux and disappear. In about 15 minutes, it is observed that almost no oxide slag is spilled out and floating oxide on the surface of the molten alloy pool 6 is reduced. Therefore, it is necessary to hold the molten metal for 15 minutes or more.

The operation of adding iron oxide to the molten alloy pool is described in Patent Document 5. Specifically, iron oxide and CaF ₂ are added to a ₂ kg molten pool in a φ84 mm water-cooled copper crucible under an inert gas atmosphere. This condition is the same in that iron oxide is used. However, in the present invention, a molten alloy pool 6 of 10 kg or more is formed in a water-cooled copper crucible 3 of φ200 mm or more, and the conditions for adding iron oxide and low melting point Ca halide composition flux on a practical scale are clarified. This is different from the prior art. Furthermore, in the present invention, the de [C] reaction is promoted by the vacuum evacuation operation, thereby promoting the formation of the molten slag layer 7 and absorbing the formed oxide. This is different from the prior art. That is, it is a great feature of the present invention that the conditions for obtaining a specific refining effect are clearly indicated.

(Combination of Ca reduction refining method and oxidation refining method)
It is also preferable to perform the refining process by appropriately combining the above-described Ca reduction refining method and the above-described oxidation refining method. From the molten alloy pool 6 whose components are adjusted to a predetermined alloy composition, phosphorus [P], sulfur [S], nitrogen [N], trace trump elements ([Sn], [Pb], [As], [As] Sb], [Bi], [Se], etc.), boron [B], etc. are removed by Ca reduction refining. Furthermore, boron [B], carbon [C], silicon [Si], aluminum [Al], titanium [Ti], Zr [Zr], calcium [Ca], alkali metal elements, and the like are subjected to iron oxide-added vacuum oxidation refining. Remove. Thus, impurity elements such as phosphorus [P], sulfur [S], tin [Sn], and lead [Pb] can be reduced to 2 ppm or less. Moreover, nitrogen [N] can be reduced to 5 ppm or less, and silicon [Si], aluminum [Al], titanium [Ti], zirconium [Zr], etc. can be reduced to 100 ppm or less. Furthermore, carbon [C] can be reduced to 50 ppm or less, and silicon [B], calcium [Ca], or the like can be reduced to 1 ppm or less.

  Since the pick-up of elements such as [C], [Si], [Al], and [Ca] is generated by the Ca reduction refining method, the oxidation refining method (vacuum oxidation refining) is performed as a subsequent process of the Ca reduction refining. It is desirable to do. However, if the content of impurity elements [C], [Si], [Al], [Ti], etc., in the melting raw material is extremely large, oxidation refining is performed before Ca reduction refining, and [C] , [Si], [Al], [Ti] and the like are effective in reducing the concentration.

  For example, when stainless steel scrap or the like is used as a melting raw material (alloy raw material), depending on the alloy composition, the content of [C] is high, and alloy components such as [Si], [Al], [Ti], etc. May be included in an amount of about 1 wt%, or may be an alloy to which [Zr], [B] or the like is added by several tens of ppm. In such a case, it is necessary to significantly increase the amount of added iron oxide during oxidative refining. When such oxidative refining with an increased amount of added iron oxide is performed, the oxygen [O] concentration in the molten alloy pool 6 may be significantly increased. When the active metal alloy element is added to the molten alloy pool 6 having a high oxygen content in order to adjust the composition to the final alloy composition, the component specification is not satisfied due to a significant decrease in yield, and the amount of non-metallic inclusions generated May cause a significant increase in In such a case, it is desirable to carry out the oxidation refining in two steps. First, oxidation refining (vacuum oxidation refining) is performed to remove [C], [Si], [Al], [Ti], [B] and the like. Thereafter, Ca reduction refining is performed to remove [P], [S], [Sn], [Pb], [N] and the like. Then, several tens to several hundred ppm of [C], [Si], [Al], [Ca], etc. picked up in the Ca reduction refining process are removed by vacuum oxidation refining (second time). Thus, oxygen [O] is not excessively left in the molten alloy pool 6.

  As described above, by carrying out refining by appropriately combining the Ca reduction refining method and the oxidation refining method (vacuum oxidation refining method), the ultra-high purity ultra-high purity (very low impurity) Fe-based Ni-based alloy ingot Melting and refining becomes possible, and an ultra-high purity alloy ingot can be manufactured.

(Addition of deoxidized elemental alloy components)
An ingot produced by the above-described oxidation refining method, or an ingot produced by appropriately combining the above-described Ca reduction refining method and the above-described oxidation refining method as a melting raw material, and a deoxidizing element-based alloy in the melting raw material It is preferable to alloy by adding components. Deoxidizing element-based alloy components include [Si], [Al], [Ti], [Zr], [B], and the like. Thereby, the ultra high purity (very low impurity) Fe-base Ni-base alloy ingot of a predetermined alloy composition can be melted.

  In the production of an ultra-high purity alloy ingot, when the vacuum oxidation refining is performed as a final process, the molten alloy pool 6 may contain a large amount of oxygen [O]. When these (deoxidizing element type) alloy components are added to the molten alloy pool 6, their deoxidized oxides are generated. In order to separate and remove such deoxidized oxide from the molten alloy pool 6, it is effective to add the Ca halide composition flux together with the addition of the alloy components. In this case, the addition rate of the Ca halide composition flux is suitably about 0.5 to 2 wt% for {Flx} M.

(Cold hearth electron beam melting method)
An ingot produced by appropriately combining the above-described Ca reduction refining method and the above-described oxidation refining method, or an ingot formed by adding a deoxidizing element-based alloy component to the ingot and forming a secondary ingot As (alloy raw material), it is preferable to perform further decarburization [C] and deoxidation [O] by the cold hearth type electron beam melting method described below.

  Using a cold crucible induction melting apparatus 1 having a water-cooled copper crucible 3 having an inner diameter of 200 mm or more, the weight of the molten alloy pool is 10 kg or more, and metal Ca reduction refining is carried out to obtain [P], [S], [Sn ], [Pb], [N], [B] and the like are removed and then oxidation refining (vacuum oxidation refining) is performed, and [C], [Al], [Ti], [Zr], An alloy ingot is manufactured by removing and refining impurity elements such as [Ca], [Si], and [B]. Or the ingot which added the deoxidation element type alloy component to the said alloy ingot as needed is manufactured. After that, by the cold hearth type electron beam melting method described below, de [C] and de [O] can be applied to remove and refine to [C] <10 ppm and [O] <10 ppm. By this method, it is possible to produce the Fe-base Ni-base alloy ingot with the highest purity (very low impurities).

  The cold hearth type electron beam melting method will be described below. FIG. 8 is a schematic diagram showing the cold hearth type electron beam melting apparatus 11.

  The oxidation refining method described above uses an induction melting method that uses a water-cooled copper crucible 3 with an inner diameter of 200 mm or more to produce a practical ingot of an ultra-high purity Fe-based Ni-based alloy having a product ingot weight of 10 kg or more by a melting refining method. Is the method. However, due to the characteristics of the cold crucible induction melting apparatus 1, the above-described oxidation refining method may be difficult to implement under a high vacuum. Therefore, removal of the generated CO gas may be insufficient in the decarbonization [C] refining. That is, the de [C] reaction cannot be promoted, and it may be difficult to remove carbon by [C] <10 ppm.

  On the other hand, the electron beam melting method that performs melting-vacuum refining under high vacuum can be completely removed by removal refining to the limit of carbon as an impurity element, the above-described Ca reduction refining method, and the above-described oxidizing refining method. Vacuum removal refining of nitrogen [N], manganese [Mn], etc. that has not been possible becomes possible.

  Also in the electron beam melting method using an electron beam as a heat source, the water-cooled copper dish-shaped container 9 (water-cooled copper hearth) having an inner dimension of the melting container of 0.2 × 0.2 m or more and the hot water from the water-cooled copper dish-shaped container 9 By using the cold hearth type electron beam melting apparatus 11 comprising the water-cooled copper mold 10 that receives the molten metal to form an ingot, the effect of removing the impurity elements can be further enhanced.

  In the actual melting operation, a rod-shaped and lump melting raw material 12 (corresponding to a secondary ingot) is supplied from the facing side of the outlet of the water-cooled copper dish-shaped container 9, and on the water-cooled copper dish-shaped container 9. Dissolve. An ingot formed by pouring molten metal overflowing from the hearth outlet of the water-cooled copper dish-shaped container 9 into the water-cooled mold 10 provided adjacent to the water-cooled copper dish-shaped container 9 and solidifying in the water-cooled mold 10 is formed. A long ingot can be melted by successively drawing downward. In such a melting method, by performing a refining operation that satisfies the following conditions, carbon [C], which is an impurity element, can be reliably removed and refined, and there are few carbon impurities that could not be produced by conventional methods. An ingot can be produced.

(1) Perform melting in a high vacuum atmosphere (<5 × 10 ⁻⁴ mbar).
Since the decarburization reaction is more easily promoted as the degree of vacuum is higher, it is desirable that the decarburization reaction be performed in a high vacuum atmosphere in order to perform decarbonization to the limit. The reason why “<5 × 10 ⁻⁴ mbar” is used is that a small amount of Ar gas may be introduced into the vacuum chamber 4. In the case where an inert gas such as Ar gas is not introduced into the vacuum chamber 4, it is desirable to perform melting at a pressure lower than 1 × 10 ⁻⁴ mbar.

(2) After the molten alloy pool 13 having a predetermined alloy composition is formed in the water-cooled copper dish-shaped container 9 and the water-cooled copper mold 10 for producing the solidified ingot, the molten raw material 12 (corresponding to the secondary ingot); Sending out an oxidizing agent (FexOy etc.) on the water-cooled copper dish-like container 9 (in this embodiment, iron oxide is put on the melted raw material 12 and sent out).
When the oxygen concentration in the melted raw material 12 is insufficient, it is not desorbed [C] even under high vacuum conditions. Therefore, it is necessary to supply [O] necessary for the oxidation of [C]. However, since the electron beam melting method is performed under a high vacuum, it is difficult to supply oxygen gas. Therefore, a method of supplying an oxidant, which is an oxide of a predetermined main component element of an alloy composition such as high-purity iron oxide, together with a dissolved raw material as a solid oxygen source is effective. In this case, the finely divided iron oxide is caught in the gas flow and scattered at the initial evacuation stage of the electron beam melting, reaches the vacuum pump, and damages the vacuum pump. Therefore, it is desirable to perform iron oxide sintering in advance and add the iron oxide in the form of granules. As an oxidizing agent, iron oxide (Fe ₃ O ₄ , Fe ₂ O ₃ etc.) is used in the case of Fe-based alloys, iron oxide or nickel oxide is used in the case of Fe-Ni based alloys, nickel oxide is used in Ni-based alloys, Co In the case of a base alloy, cobalt oxide or the like can be applied. These oxidizing agents are solid oxidizing agents, in other words, oxidizing agents made of metal oxides.

  In addition, when the raw material feeder is attached to the cold hearth type electron beam melting apparatus 11, after the molten alloy pool 13 is formed, the molten raw material 12 and the oxidizing agent (FexOy, etc.) are combined with a water-cooled copper dish-shaped container. 9 There is no need to send it up. In this case, an oxidizing agent (such as FexOy or the like corresponding to the third refining agent) may be added to the molten alloy pool 13 in the water-cooled copper dish-like container 9 in accordance with the melting state of the melting raw material 12 by the raw material feeder.

(3) The added weight (WFexOy (kg)) of the iron oxide (FexOy) should be within a range satisfying 1.0 × MFeO ≦ WFexOy ≦ 4.0 × MFeO.
MFeO = WM / 100 × ([C] /12.01) / y × (55.85 × x + 16.0 × y)
WM: weight of molten raw material (kg), [C]: C concentration in molten raw material (wt%)
As a result of the test, it has been proved that the addition weight of iron oxide is the same as the calculation amount necessary for CO gasification of the target carbon or four times as much. When the addition amount is too small, the desorption [C] becomes insufficient. When the addition amount is too large, the oxygen [O] concentration in the ingot tends to increase, and empirically, it is about 2 to 3 times the calculated amount. In many cases, the amount added is appropriate.

  By applying this cold hearth type electron beam melting method, floating separation of non-metallic inclusions in the ingot progresses, and it has been confirmed that it is very effective as an oxygen removal method. By this refining, carbon [C] in the ingot can be reduced to 10 ppm or less and oxygen [O] can also be reduced to 10 ppm or less. When the conditions are appropriate, [C] <5 ppm, [O] <5 ppm. Satisfactory analysis results may be obtained. [Mn] of the alloy component is evaporated and removed in the electron beam melting process, and the [Mn] concentration is often 0.01 wt% or less.

(Example)
The structural schematic diagram of the test apparatus used for confirming the refining effect is as shown in FIGS. 1 and 8, and the general specifications of the equipment are as follows.
(1) Cold Crucible Induction Melting (CCIM) device 1
High frequency power supply Maximum output: 400kW, Frequency: 3000Hz
Water-cooled copper crucible 3 Inner diameter: φ220, Number of segments: 24
Ultimate vacuum 10 ^-2 mbar level Vacuum evacuation device Rotary pump, mechanical booster pump (2) Cold hearth electron beam melting (EBCHR) device 11
High voltage power supply Acceleration voltage: 40kV, Maximum output: 300kW
Two electron beam guns 14 Ultimate vacuum 10 ^-6 mbar level Vacuum evacuation device Rotary pump, mechanical booster pump, diffusion pump Raw material supply mechanism Maximum φ210 × 1000Lmm
Ingot drawing mechanism Maximum φ200 × 1000Lmm

(Results of Ca reduction refining test using cold-crucible induction melting device)
Table 2 shows the effect of the Ca reduction refining test using a refining agent that is a mixture of metallic Ca and Ca halide composition flux. In this Ca reduction refining test, a cold crucible induction melting apparatus 1 having a water-cooled copper crucible 3 having an inner diameter of 220 mm was used. An alloy molten pool 6 having a Fe-20Ni25Cr composition or a Fe-35Ni25Cr composition was formed using a ferrochrome material or a low carbon steel material, which is a low-priced material having a high impurity element content, as an alloy material. Metal Ca and Ca halide composition flux are added to the molten alloy pool 6 and a Ca reductive refining test is performed. The phosphorus removal [P] effects are summarized as shown in Table 2.

  As can be seen from Table 2, when the Ca addition ratio {Ca} M to the weight of the molten alloy pool 6 is {Ca} M = 0.2, only a low de- [P] effect was obtained. The refining effect is not achieved until the Ca addition rate is increased to 0.5 <{Ca} M or more. Further, in Comparative Example 3, Examples 3 and 4, when the Ca halide composition flux addition rate {Flx} M is too small, the de- [P] rate is lowered even if the amount of metallic Ca addition is sufficient. That is, it can be seen that {Ca} M ≦ {Flx} M is necessary to obtain the refining effect.

  Further, in the Fe-20Ni25Cr alloy and the Fe-35Ni25Cr alloy, the influence of the retention time of the molten alloy pool 6 after addition of metallic Ca on the de- [P] effect when the weight of the molten alloy pool 6 is 20 kg and 50 kg. From the results of examining the above, it can be seen that there is an optimum molten metal holding time depending on the addition amount of metallic Ca and the addition amount of Ca halide composition flux.

(Results of oxidation refining test using cold crucible induction melting device)
In addition, for ingots subjected to Ca reduction refining (test results are shown in Table 2), an oxidation refining test using a refining agent (second refining agent) that is a mixture of iron oxide and Ca halide composition flux. The effects are shown in Table 3. This oxidation refining test uses a CCIM apparatus. In this oxidation refining test, ferrochrome (FeCr), ultra-low carbon steel, and electrolytic Ni are used as melting raw materials.

As can be seen from Table 3, if the amount of iron oxide is too small (WFe ₃ O ₄ /MFeO<0.2), the de [C] effect cannot be obtained. For the purpose of removing [C], the effect can be obtained by setting WFe ₃ O ₄ /MFeO=0.2 or more and {Flx} M = 0.5 or more. In order to obtain the de- [Si] and de- [B] effects, the effect can be obtained by setting WFe ₃ O ₄ /MFeO=1.0 or more and {Flx} M = 3.0 or more. become. However, when WFe ₃ O ₄ /MFeO=5.0, violent droplet splashing (splash) accompanying decarburization reaction occurred during vacuum refining, and stable refining operation could not be performed. That is, there is an upper limit to the amount of iron oxide added.

(Test result by cold hearth electron beam melting method)
Furthermore, a CCIM device (cold-crucible induction melting device 1) having a water-cooled copper crucible 3 with a diameter of 220 mm is used as a melting raw material with an ingot subjected to refining treatment in the order of Ca reduction refining and oxidation refining (vacuum oxidation refining). Table 4 shows the effect of removing [C] when iron oxide is added to the melting raw material and melting is performed by the hearth-type electron beam melting apparatus 11.

  The impurity element concentration (wt%) before the test was [C] = 0.005 and [O] = 0.003. The amount of MFeO calculated from the content of these impurity elements was 0.011 kg.

As can be seen from Table 4, by adding an appropriate amount of iron oxide, de- [C] is promoted and an ingot having an impurity element [C] concentration of 10 ppm or less can be produced. That is, the WFexOy / MFeO ratio is preferably 1.0 or more and 4.0 or less, and more preferably 2.0 or more and 3.0 or less. The iron oxide agent is obtained by sintering fine Fe ₂ O ₃ powder at 1250 ° C. into granular Fe ₂ O ₃ .

Regarding the addition amount of granular Fe ₂ O ₃ , if the addition amount is too small compared to the expected oxygen amount for the decarburization reaction, the decarburization reaction becomes insufficient. On the other hand, if excessively added, decarburization proceeds sufficiently, but there is a problem that the oxygen content becomes too high. That is, there is an appropriate range when adding iron oxide in EBCHR.

  Although the embodiments of the present invention have been described above, the present invention is not limited to the above-described embodiments, and various modifications can be made as long as they are described in the claims. .

1: Cold crucible induction melting device 2: Raw material feeder 3: Water-cooled copper crucible 4: Vacuum chamber 6: Molten alloy pool 7: Molten slag layer

Claims (6)

A molten pool formation step of forming an alloy molten pool with components adjusted to a predetermined alloy composition in an inert gas atmosphere by charging the alloy raw material into a water-cooled copper crucible of a cold crucible induction melting device;
A refining step of adding a refining agent to the molten alloy pool to remove impurity elements including at least phosphorus;
With
The refining agent is a mixture of metallic Ca and Ca halide composition flux,
The Ca halide composition flux the calcium oxide to calcium fluoride CaF ₂ -CaO formulated 5-30 wt%, calcium chloride calcium fluoride CaF _{2-CaCl 2} was blended 5-30 wt%, or, calcium fluoride CaF ₂ − (CaO + CaCl ₂ ) containing 5 to 30 wt% of calcium oxide and calcium chloride,
The addition ratio of the metal Ca with respect to the weight of the molten alloy pool is 0.5 wt% or more,
A method for producing an ultra-high purity alloy ingot, wherein an addition rate of the Ca halide composition flux with respect to a weight of the molten alloy pool is equal to or higher than an addition rate of the metal Ca. In the manufacturing method of the ultra-high purity alloy ingot according to claim 1,
After the refining agent is added to the molten alloy pool, a time T1 or more that is 1/2 of the time required for all of the added refining agent to melt, and a time that 1/2 of the added metal Ca is lost by evaporation The method for producing an ultra-high purity alloy ingot, wherein the molten alloy pool is held for T2 or less. In the manufacturing method of the ultra-high purity alloy ingot according to claim 2,
A method for producing an ultra high purity alloy ingot, wherein the molten pool forming step and the refining step are performed a plurality of times to produce an ultra high purity alloy ingot having a phosphorus content of 2 ppm or less. The ingot produced by the method for producing an ultra-high purity alloy ingot according to any one of claims 1 to 3 is a primary ingot,
A second molten metal pool forming step of forming the molten alloy pool by charging the primary ingot into a water-cooled copper crucible of the cold crucible induction melting device;
After the second refining agent is added to the molten alloy pool of the primary ingot, the inert gas in the chamber is exhausted and the exhaust state is maintained for 15 minutes or more to remove impurity elements including at least carbon and calcium. 2 refining process,
With
The second refining agent is a mixture of an oxidizing agent that is an oxide of a predetermined alloy composition main component element such as iron oxide and the Ca halide composition flux,
The addition weight of the oxidizing agent is 0.2 times or more and 4.0 times the calculated weight calculated for oxidizing all the impurity elements including at least carbon and calcium in the molten alloy pool of the primary ingot. And
In the second refining step, the addition ratio of the Ca halide composition flux with respect to the weight of the molten alloy pool of the primary ingot is 0.5 wt% or more and 5.0 wt% or less, A method for producing a pure alloy ingot.   An ingot produced by the method for producing an ultra high purity alloy ingot according to claim 4 is added with a deoxidizing element-based alloy component to form an alloy. Production method. The ingot produced by the method for producing an ultra-high purity alloy ingot according to any one of claims 1 to 5 is a secondary ingot,
The secondary ingot is supplied to a water-cooled copper dish-shaped container of a cold hearth type electron beam melting apparatus, and the water-cooled copper dish-shaped container and the water-cooled copper dish-shaped container are used at a pressure lower than 5 × 10 ⁻⁴ mbar. A third molten pool forming step for forming a molten alloy pool in a water-cooled copper mold adjacent to the container;
A third refining step of adding a third refining agent to the molten alloy pool in the water-cooled copper dish-like container to remove carbon as an impurity element;
With
The third refining agent is an oxidant that is an oxide of a main component element of a predetermined alloy composition such as iron oxide,
The added weight of the third refining agent is 1.0 times or more and 4.0 times or less of the calculated weight calculated to oxidize all the carbon as the impurity element in the molten alloy pool of the secondary ingot. A method for producing an ultra-high purity alloy ingot, wherein

Images