Decarburization and refining method for chromium-containing molten steel

Abstract

A method for decarburizing and refining Cr-containing molten steel is provided, which reduces the amount of reducing material used and shortens the refining time. When decarburizing molten chromium steel containing 10.5 mass% or more of Cr, the ratio of CaO to Al2O3 contained in the slag is set to 15 or less, and the composition of the slag and the activity a of Cr2O3 in the slag are adjusted. Cr2O3 The activity a of Cr2O3 was calculated from the composition of the slag using a machine learning model that uses the relationship between Cr2O3 The measured or predicted values ​​of the slag composition are introduced into the trained model corresponding to the temperature T of the Cr-containing molten steel to obtain the activity a of Cr2O3. Cr2O3 The carbon activity a in the Cr-containing molten steel is estimated based on the carbon concentration [C] (mass%) of the Cr-containing molten steel. C Calculate ΔG using the following formula (1), and then adjust the CO partial pressure P CO Control sequentially. ΔG=ΔG 0 +RTln((a 2 Cr P 3 CO ) / (a Cr2O3 / a 3 C )) …(1)

Description

[Technical Field]

[0001] The present invention relates to a method for decarburizing and refining chromium-containing molten steel. [Background technology]

[0002] Conventionally, when refining chromium-containing steel, particularly molten chromium steel containing 10.5% or more chromium such as stainless steel, a widely used decarburization refining method is the AOD method, in which oxygen gas or a mixture of oxygen gas and an inert gas is blown into the molten steel contained in a refining vessel. As an improved version of the AOD method, for example, the AOD-VCR method is known, in which decarburization is performed by the conventional AOD method until the carbon content in the molten steel reaches 0.10 mass%, and then decarburization is performed using only inert gas in a reduced pressure atmosphere while stirring the molten steel.

[0003] Another known method for refining chromium-containing molten steel containing 10.5% or more chromium is the VOD process, in which oxygen is blown into the molten steel under reduced pressure to decarburize it. Another known method is the AOD-VOD process, in which the steel is roughly decarburized using the AOD process, then tapped with slag, and vacuum decarburized using oxides in the slag as the oxygen source in the VOD process.

[0004] When oxygen is blown into molten steel containing Cr, O2 reacts with the C in the molten steel to produce CO or CO2, directly promoting decarburization, and Cr is oxidized by the oxygen to produce Cr2O3, which then reacts with the C in the molten steel, promoting indirect decarburization. Furthermore, when molten steel is stirred in a reduced-pressure atmosphere, the partial pressure of CO gas in the atmosphere decreases, and the equilibrium reaction represented by the chemical equation 3C + Cr2O3 = 2Cr + CO advances to the right-hand side, accelerating the decarburization reaction.

[0005] In the equilibrium reaction shown in the chemical equation above, it would seem that the higher the degree of vacuum and the lower the CO partial pressure, the more rapidly decarburized the steel would be. However, in reality, a lower degree of vacuum requires a smaller amount of gas injection, which means it takes a longer time to reduce the carbon content in the molten steel to the target level. Furthermore, because chromium is abundant in molten steel, the injection of oxygen produces a large amount of Cr2O3. Therefore, after the decarburization process, the excess Cr2O3 must be reduced to Cr using reducing agents such as Al and Si. However, this Cr reduction process can increase costs.

[0006] Patent Document 1 describes a refining method in which a gas containing oxygen is injected into chromium-containing molten steel in a refining vessel, which includes a first step in which the vessel is set at a pressure in the range of 400 Torr to atmospheric pressure and the gas containing oxygen is injected; a second step in which the vessel is depressurized to 250 to 400 Torr and the gas containing oxygen is injected at a pressure of 0.4 Nm per ton of molten steel; 3 The method for refining chromium-containing molten steel is described, which comprises a second step in which gas is blown in at a rate of 1 / min or more, and a third step in which the pressure inside the vessel is reduced to 250 Torr or less and gas is blown in, and in which the second step is switched to the third step when the [C] concentration in the molten steel reaches 0.2 to 0.1 mass%.

[0007] The refining method described in Patent Document 1 aims to reduce the amount of inert gas and oxygen gas used and shorten the refining time, but there is room for further improvement in reducing the amount of reducing material used and shortening the refining time. [Prior art documents] [Patent documents]

[0008] [Patent Document 1] Patent No. 4895446 Summary of the Invention [Problem to be solved by the invention]

[0009] The present invention has been made in view of the above circumstances, and an object of the present invention is to provide a method for decarburizing and refining Cr-containing molten steel, which is capable of reducing the amount of reducing material used and shortening the refining time. [Means for solving the problem]

[0010] In order to solve the above problems, the present invention employs the following configuration. [1] When chromium-containing molten steel containing 10.5 mass% or more of Cr and slag containing at least Cr2O3 are contained in a smelting vessel, a mixed gas containing oxygen gas and an inert gas, oxygen gas, or an inert gas is blown into the chromium-containing molten steel while the inside of the smelting vessel is made into a reduced pressure atmosphere for decarburization, The ratio of CaO to Al2O3 contained in the slag (CaO / Al2O3 ratio) is set to 15 or less, The composition of the slag and the activity of the Cr2O3 contained in the slag Cr2O3 The activity a of the Cr2O3 was calculated from the composition of the slag using a machine learning model that uses the relationship between Cr2O3 A trained model for estimating the above is obtained in advance for each temperature T of the Cr-containing molten steel, The composition of the slag, the carbon concentration [C] (mass%) of the Cr-containing molten steel, the temperature of the Cr-containing molten steel, and the partial pressure P of CO gas in the smelting vessel CO are successively measured or successively predicted, The measured or predicted value of the composition of the slag is introduced into the trained model corresponding to the temperature T of the Cr-containing molten steel to calculate the activity a of the Cr2O3. Cr2O3 and successively estimating the carbon activity a in the Cr-containing molten steel based on the carbon concentration [C] (mass%) of the Cr-containing molten steel. C Seeking The activity of the Cr2O3 Cr2O3 and the estimated value of the carbon activity a C and the partial pressure P of the CO gas CO and are input into the following formula (1) to calculate ΔG, and the CO partial pressure P in the Cr-containing molten steel is adjusted so that this ΔG does not become 0 or more. CO This is a method for decarburizing and refining chromium-containing molten steel by sequentially controlling the above. ΔG=ΔG0 +RTln((a 2 Cr P 3 CO ) / (a Cr2O3 / a 3 C )) …(1) Cr2O3+3C=2Cr+3CO …(2) However, ΔG in equation (1) is the free energy change of the decarburization reaction shown in equation (2), and ΔG 0 is the standard free energy change of the decarburization reaction shown in the above formula (2) at 1 atmosphere and 25°C, T is the temperature of the Cr-containing molten steel, and R is the gas constant (R=8.314×10 -3 kJmol -1 K -1 ) and a Cr is the activity of Cr in the Cr-containing molten steel, and P CO is the partial pressure of CO in the refining vessel, and a Cr2O3 is the activity of Cr2O3, and a C is the activity of carbon. [2] The method for decarburizing and refining chromium-containing molten steel according to [1], wherein the composition of the slag is such that the contents (mass%) of FeO, SiO2, MnO, Cr2O3, CaO, Al2O3, and MgO contained in the slag are each determined based on 100 mass% of the total of FeO, SiO2, MnO, Cr2O3, CaO, Al2O3, and MgO. [Effects of the Invention]

[0011] According to the present invention, it is possible to provide a method for decarburizing and refining Cr-containing molten steel, which can reduce the amount of reducing material used and shorten the refining time. [Brief explanation of the drawings]

[0012] [Figure 1] 1 is a diagram showing a refining facility for carrying out a decarburization refining method for Cr-containing molten steel according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0013] The present inventors have conducted extensive research to reduce the amount of reducing material used and shorten the refining time in the decarburization refining of Cr-containing molten steel.

[0014] Molten chromium steel containing 10.5% or more by mass of Cr is decarburized using equipment such as AOD or VOD. Decarburization generally refers to the process of blowing oxygen-containing gas onto molten steel to convert the carbon (C) contained in the molten steel into CO or CO2, thereby removing it from the molten steel. In the case of molten chromium steel, most of the blown oxygen reacts with the chromium (Cr) present in large quantities in the molten steel to form Cr2O3, and it is thought that decarburization proceeds indirectly as the resulting Cr2O3 reacts with the C in the molten steel.

[0015] The reaction formula when Cr2O3 reacts with carbon (C) in molten steel to produce CO is expressed as formula (A). The free energy change ΔG of the reaction of formula (A) is expressed as formula (1).

[0016] Cr2O3+3C=2Cr+3CO …(A)

[0017] ΔG=ΔG 0 +RTln((a Cr 2 P CO 3 ) / (a Cr2O3 / a C 3 )) …(1)

[0018] However, ΔG in equation (1) is the free energy change of the decarburization reaction (reaction from the left side to the right side) shown in equation (A), and ΔG 0 is the standard free energy change of the decarburization reaction shown in the above formula (A) at 1 atmosphere and 25°C, T is the temperature of the Cr-containing molten steel, and R is the gas constant (R=8.314×10 -3 kJmol -1 K -1 ) and a Cr is the activity of Cr, and P CO is the partial pressure of CO, and a Cr2O3 is the activity of Cr2O3, and a Cis the activity of carbon.

[0019] The condition for decarburization to proceed, that is, for the reaction of equation (A) to proceed from the left side to the right side, is ΔG<0. Therefore, the Cr activity a Cr and CO partial pressure P CO is low, and Cr2O3 activity a Cr2O3 , C activity a C On the other hand, as decarburization progresses, that is, as C decreases in the molten steel, the C activity a C Therefore, in AOD and VOD, the C activity a C As the CO partial pressure P CO By lowering the pressure, specifically by increasing the degree of vacuum, we are operating in a way that maintains ΔG<0.

[0020] However, Cr2O3 does not exist as a pure substance in chromium-containing molten steel; most of it is contained in slag. The slag of chromium-containing molten steel contains CaO, Al2O3, SiO2, MgO, Cr2O3, etc., and it is known that the activity of the Cr2O3 contained in this slag varies depending on the slag composition and temperature. The higher the Cr2O3 concentration in the slag, the higher the activity tends to be, and in terms of the ease with which decarburization progresses, a higher Cr2O3 concentration is more favorable for the decarburization reaction.

[0021] The inventors further investigated the relationship between the composition of Cr2O3-containing slag and the activity of Cr2O3. As described above, chromium-containing molten steel slag contains CaO, Al2O3, SiO2, MgO, Cr2O3, etc., and it is known that the activity of Cr2O3 is not simply determined by the concentration of Cr2O3, etc. in the slag, but is also affected by the concentrations of other components besides Cr2O3. Therefore, using the relationship between the composition of each component in the slag and the Cr2O3 activity as training data, machine learning and neural network regression were performed. It was found that a certain correlation exists between the slag composition and the Cr2O3 activity for each temperature of the chromium-containing molten steel. Based on this, by measuring the temperature of the chromium-containing molten steel, calling up a trained model corresponding to that temperature, and using that trained model, it was possible to successfully calculate the Cr2O3 activity at a given time from the slag composition at that time. Note that the slag temperature is approximately the same as the temperature of the chromium-containing molten steel, so the temperature of the chromium-containing molten steel is used in this embodiment.

[0022] Conventionally, there are two known methods for estimating the activity of each component contained in slag. One method involves referencing existing data, but this method is not applicable to actual operations because it cannot handle complex component systems (systems containing a large number of components). The other method involves estimating the activity using general-purpose thermodynamic equilibrium calculation software. However, this method uses convergent calculations, which can take a long time. In addition, the results do not always converge and may not return a value, making it unsuitable for sequential calculations in actual operations.

[0023] Furthermore, if decarburization is completed while the Cr2O3 concentration is still high, a large amount of reducing agent will be required to reduce Cr2O3 to Cr, which is cost-intensive and the process of reducing Cr will also take time.

[0024] Therefore, it is important to have a slag composition that allows high Cr2O3 activity even when the Cr2O3 concentration in the slag is low. The inventors have found that Cr2O3 activity is affected by the CaO to Al2O3 concentration ratio (CaO / Al2O3 (hereinafter referred to as C / A)), and that the lower the C / A, the higher the Cr2O3 activity tends to be. In other words, to increase Cr2O3 activity, it is preferable to lower the C / A ratio.

[0025] That is, when the C / A ratio is higher than 15, the Cr2O3 activity remains low and changes little even when the C / A or Cr2O3 concentration changes, whereas when the C / A ratio is lower than 15, the change in Cr2O3 activity due to the change in Cr2O3 concentration increases when the C / A ratio is the same. Therefore, a smaller C / A ratio is advantageous for increasing the Cr2O3 activity while keeping the Cr2O3 concentration low.

[0026] Furthermore, in the above formulas (A) and (1), the higher the degree of vacuum is so increased as to maintain ΔG<0, the smaller ΔG in formula (1) becomes (more negative), which should make decarburization easier, but in reality, the higher the degree of vacuum is, the less gas can be injected, which increases the time required to reduce the C concentration in the Cr-containing molten steel to the specified concentration, and reduces operational efficiency. Ultimately, it is preferable to maintain ΔG at a value slightly lower than 0, and it has been found that to achieve this, it is necessary to increase the degree of vacuum in stages in accordance with the decrease in the C concentration (C activity).

[0027] The present invention has been made in light of the above findings. Hereinafter, embodiments of the present invention will be described.

[0028] In the method for decarburizing and refining chromium-containing molten steel of this embodiment, chromium-containing molten steel containing 10.5 mass % or more of Cr and slag containing at least Cr2O3 are placed in a smelting vessel, and one of a mixed gas containing oxygen gas and an inert gas, oxygen gas, and an inert gas is blown into the chromium-containing molten steel while creating a reduced pressure atmosphere in the smelting vessel for decarburization, and the ratio of CaO to Al2O3 contained in the slag (CaO / Al2O3 ratio) is set to 15 or less, and the composition of the slag and the activity a of the Cr2O3 contained in the slag are adjusted. Cr2O3 The activity a of the Cr2O3 was calculated from the composition of the slag using a machine learning model that uses the relationship between Cr2O3 A trained model for estimating the temperature of the Cr-containing molten steel is previously obtained for each temperature T of the Cr-containing molten steel, and the trained model for estimating the temperature of the Cr-containing molten steel is calculated based on the composition of the slag, the carbon concentration [C] (mass%) of the Cr-containing molten steel, the temperature of the Cr-containing molten steel, and the partial pressure P of the CO gas in the smelting vessel. CO The measured or predicted values ​​of the composition of the slag are introduced into the trained model corresponding to the temperature T of the Cr-containing molten steel to calculate the activity a of the Cr2O3. Cr2O3 and successively estimating the carbon activity a in the Cr-containing molten steel based on the carbon concentration [C] (mass%) of the Cr-containing molten steel. C The activity a of the Cr2O3 Cr2O3 and the estimated value of the carbon activity a C and the partial pressure P of the CO gas CO and are input into the following formula (1) to calculate ΔG, and the CO partial pressure P in the Cr-containing molten steel is adjusted so that this ΔG does not become 0 or more. CO Control sequentially.

[0029] ΔG=ΔG 0 +RTln((a 2 Cr P 3 CO ) / (a Cr2O3 / a 3 C )) …(1) Cr2O3+3C=2Cr+3CO …(2)

[0030] However, ΔG in equation (1) is the free energy change of the decarburization reaction shown in equation (2), and ΔG0 is the standard free energy change of the decarburization reaction shown in the above formula (2) at 1 atmosphere and 25°C, T is the temperature of the Cr-containing molten steel, and R is the gas constant (R=8.314×10 -3 kJmol -1 K -1 ) and a Cr is the activity of Cr in the Cr-containing molten steel, and P CO is the partial pressure of CO in the refining vessel, and a Cr2O3 is the activity of Cr2O3, and a C is the activity of carbon.

[0031] In this embodiment, decarburization refining is performed by AOD or VOD. For example, in this embodiment, when performing reduced-pressure refining, a refining vessel 1 shown in FIG. 1(a) is used. Furthermore, when performing atmospheric refining as preliminary decarburization before reduced-pressure refining, a refining vessel 1 shown in FIG. 1(b) is used. Refining gas is injected into the chromium-containing molten steel in the refining vessel through a bottom-blowing tuyere 2. The refining vessel 1 has a removable exhaust hood 3. During reduced-pressure refining, the exhaust hood 3 is attached to the refining vessel 1 as shown in FIG. 1(a), and the inside of the refining vessel is depressurized by gas suction. During atmospheric refining, the exhaust hood 3 is not attached as shown in FIG. 1(b), so gas can be injected not only through the bottom-blowing tuyere 2 but also through a top-blowing lance 12.

[0032] In Figures 1(a) and 1(b), reference numeral 4 denotes molten steel, reference numeral 5 denotes bottom-blown gas, reference numeral 6 denotes slag, reference numeral 7 denotes an exhaust pipe, reference numeral 8 denotes a gas cooler, reference numeral 9 denotes a bag filter, reference numeral 10 denotes a steam ejector, and reference numeral 11 denotes a water pump.

[0033] In the decarburization refining method of this embodiment, chromium-containing molten steel containing 10.5 mass % or more of Cr and slag containing at least CrO are contained in a refining vessel, and decarburization is performed by creating a reduced pressure atmosphere in the refining vessel while blowing either a mixed gas containing oxygen gas and an inert gas, or oxygen gas or an inert gas into the chromium-containing molten steel.

[0034] The Cr-containing molten steel may contain alloy elements other than Cr.

[0035] The slag used in the decarburization refining method of this embodiment has a mass ratio of CaO to Al2O3 (CaO / Al2O3 ratio, hereinafter sometimes referred to as C / A) of 15 or less. As mentioned above, when the C / A is higher than 15, the Cr2O3 activity remains low and changes little even when the C / A or Cr2O3 concentration changes. However, when the C / A is lower than 15, the Cr2O3 activity changes significantly with changes in the Cr2O3 concentration. Therefore, a small C / A ratio is advantageous in order to increase the Cr2O3 activity while maintaining a low Cr2O3 concentration. Therefore, the C / A of the slag must be 15 or less.

[0036] Next, in this embodiment, the composition of the slag and the activity a of Cr2O3 contained in the slag are calculated in advance. Cr2O3 The activity a of Cr2O3 is calculated from the composition of the slag using a machine learning model that uses the relationship between Cr2O3 A trained model that estimates is obtained in advance for each temperature T of Cr-containing molten steel.

[0037] The training data is the composition of the slag and the activity of Cr2O3 contained in the slag. Cr2O3 A combination of these is used. Specifically, multiple types of slag are prepared to be used as training data. The composition of each slag is the content (mass%) of FeO, SiO2, MnO, Cr2O3, CaO, Al2O3, and MgO contained in the slag, where the total of FeO, SiO2, MnO, Cr2O3, CaO, Al2O3, and MgO is taken as 100 mass%. When entering the slag composition as training data, the content of each component is divided by 100 for standardization.

[0038] In addition, the activity of Cr2O3 contained in the slag a Cr2O3is measured for each slag by electromotive force measurement. For example, the target slag is melted in a metal Cr crucible, and the electromotive force measured with an oxygen probe is used to calculate the oxygen potential in the slag. CrO is calculated from the following formula: 1.5 The activity of (a CrO1.5 ) and square it to obtain the activity of Cr2O3 (a Cr2O3 ) can be calculated (Reference: "Oxidation state and activities of chromium oxides in CaO-SiO2-CrO x slag system”, Metallurgical and Materials Transactions B Process metallurgy and materials processing science, Volume 33, pages 595-603, 2002).

[0039]

number

[0040]

number

[0041] ΔG in formula (C) 0 is the standard free energy change of the reaction shown in the above formula (B) at 1 atmosphere and 25°C, and a CrO1.5 is CrO 1.5 is the activity of a Cr(s) is the activity of Cr(s), and P O2 is the oxygen potential.

[0042] The trained model may be generated using an information processing device such as a learning device. The learning device executes a learning process using training data stored in a storage device of the device itself. For example, training data is generated for each temperature. In this case, the learning device executes a learning process using the training data for each temperature, and generates a trained model for each temperature. As a specific example of such a learning process, for example, supervised learning for regression such as neural network regression (NNR) or support vector regression may be used. The learning device, for example, performs supervised learning to calculate the activity a of Cr2O3 based on input information indicating the composition of the slag. Cr2O3 The learning device generates a trained model for outputting the trained model. The learning device records the trained model in a storage device. The trained model obtained by the learning device may be transmitted to another device (e.g., an estimation device) and used therein.

[0043] The estimation device uses the trained model described above to estimate the activity a of Cr2O3. Cr2O3 The estimation device is configured using an information processing device. The estimation device performs processing using a trained model stored in the storage device of the estimation device itself. This trained model may be generated by the device itself, or may be generated by another device (for example, the learning device described above). When the trained model is generated by the device itself, the estimation device generates the trained model by executing processing equivalent to that of the learning device described above. The storage device of the estimation device stores trained models in association with each temperature. When the estimation device acquires the temperature of the slag to be estimated, for example, from an input device, it reads out the trained model stored in association with the temperature from the storage device. The estimation device performs estimation processing using the trained model and information indicating the composition of the slag input, for example, from the input device. The estimation processing determines the activity a of Cr2O3 in the slag to be estimated. Cr2O3 is estimated.

[0044] The processing of the above-described learning device and estimation device is executed by a control unit configured using a processor such as a CPU and a memory. The control unit functions when the processor executes a program. Note that all or part of the functions of the control unit may be realized using hardware such as an ASIC, PLD, or FPGA. The above-described program may be recorded on a computer-readable recording medium. Examples of computer-readable recording media include portable media such as flexible disks, magneto-optical disks, ROMs, CD-ROMs, and semiconductor storage devices (e.g., SSDs), as well as storage devices such as hard disks and semiconductor storage devices built into computer systems. The above-described program may be transmitted via a telecommunications line.

[0045] It is preferable to obtain a trained model in advance for each temperature T of Cr-containing molten steel. This is because when machine learning is performed by adding the temperature T of Cr-containing molten steel to the input data, the accuracy of the trained model tends to decrease.

[0046] In actual operation, during the progress of decarburization refining, the composition of the slag, the carbon concentration [C] (mass%) of the Cr-containing molten steel, and the temperature of the Cr-containing molten steel are successively measured. CO The frequency of these measurements may be, for example, every 1 to 60 seconds. Furthermore, the measurement intervals may be adjusted so that the frequency of these measurements and the frequency of adjusting the degree of vacuum are in the range of 20 to 7200 times per decarburization refining process.

[0047] In addition to being measured sequentially, the composition of the slag may also be predicted sequentially.

[0048] Furthermore, the carbon concentration [C] (mass%) of the Cr-containing molten steel may be successively predicted instead of being successively measured. For example, as described in

[0029] of Japanese Patent No. 3725312, the carbon concentration [C] (mass%) of the Cr-containing molten steel may be predicted by material balance calculation from the flow rate and composition of the exhaust gas discharged in the decarburization refining.

[0049] Furthermore, the Cr concentration in the Cr-containing molten steel can be estimated by material balance calculations based on the amount of Cr-containing molten steel charged, the Cr concentration, the amount of added alloys, and the Cr grade, and the Cr2O3 content in the slag can be predicted from the Cr concentration in the Cr-containing molten steel. Furthermore, since the amounts of other components in the slag remain almost unchanged, the concentrations of the other components can also be predicted automatically from material balance calculations.

[0050] Furthermore, the partial pressure P of CO in the refining vessel CO can be substituted by the degree of vacuum in the refining vessel.

[0051] The temperature of the Cr-containing molten steel can be estimated by either a method of directly measuring the temperature continuously or semi-continuously, or a method of calculating from an empirically known heat-up efficiency based on the temperature of the Cr-containing molten steel at the time of charging, or a combination of these.

[0052] Based on the measured or estimated temperature of the chromium-containing molten steel, the activity of Cr2O3 a Cr2O3 As described above, the trained model is calculated in advance for each temperature T of Cr-containing molten steel, so it is advisable to use a trained model that is suitable for the temperature T of Cr-containing molten steel.

[0053] Then, by introducing the measured or estimated values ​​of the slag composition into the trained model, the activity of Cr2O3, a Cr2O3 Ask for.

[0054] In addition, based on the measured or estimated carbon concentration [C] (mass%) of the Cr-containing molten steel, the carbon activity a C Calculate the activity of carbon a C The carbon activity coefficient can be calculated from the composition of the Cr-containing molten steel as a general method. The carbon activity coefficient is calculated from the interaction coefficient, which is publicly available thermodynamic data, and the carbon activity a is calculated from the product of the carbon activity coefficient and the carbon concentration of the Cr-containing molten steel. C can be obtained.

[0055] Furthermore, the activity of Cr a Cr Calculate the activity of Cr a Cr is the activity of carbon aC As in the case of (1), it can be generally determined from the composition of the Cr-containing molten steel. The activity coefficient of Cr is calculated from the interaction coefficient, which is publicly available thermodynamic data, and the activity of carbon a is calculated from the product of the activity coefficient of Cr and the Cr concentration in the Cr-containing molten steel. C can be obtained.

[0056] Then, the activity of each substance obtained (a Cr , a Cr2O3 , a C ) and the partial pressure of CO, P CO and are introduced into the above formula (1) to calculate ΔG (change in free energy). 0 is the standard free energy change of the decarburization reaction shown in the above formula (A) at 1 atmosphere and 25°C, T is the temperature of the Cr-containing molten steel, and R is the gas constant (R=8.314×10 -3 kJmol -1 K -1 )

[0057] If ΔG is in a range less than 0, it is determined that the reaction rate of the decarburization reaction is appropriate, and operation is continued as is.

[0058] However, when the decarburization reaction progresses to a certain extent and the carbon concentration in the Cr-containing molten steel decreases, ΔG becomes 0 or more, and the reaction rate of the decarburization reaction slows down. In this case, the degree of vacuum in the smelting vessel is increased to increase the partial pressure P CO This reduces ΔG to less than 0, increasing the reaction rate of the decarburization reaction and shortening the time required for decarburization refining.

[0059] On the other hand, as the degree of vacuum increases, the partial pressure of CO CO This lowers the ΔG, which in turn reduces the reaction rate of the decarburization reaction, which appears to increase. However, if the degree of vacuum becomes too high, the amount of oxygen-containing gas injected into the Cr-containing molten steel must be reduced to maintain the vacuum, which in turn reduces the reaction rate of the decarburization reaction and lengthens the refining time. Therefore, it is preferable to adjust the degree of vacuum so that ΔG is -300 kJ / mol or more.

[0060] As described above, according to the decarburization refining method for Cr-containing molten steel of this embodiment, the activity a Cr2O3 By successively estimating the free energy change ΔG of the decarburization reaction proceeding in the refining vessel, the degree of vacuum in the refining vessel is estimated, and the estimated ΔG is dynamically controlled so that it does not become 0 or more, and preferably falls within a desired range less than 0. This makes it possible to appropriately control the reaction rate of the decarburization reaction, thereby shortening the refining time. Furthermore, by setting the CaO to Al2O3 concentration ratio (CaO / Al2O3) in the slag to 15 or less, decarburization refining can be performed under conditions that increase the Cr2O3 activity even if the Cr2O3 concentration in the slag is low, and the Cr2O3 concentration in the slag can be reduced. Decarburization can be completed when the Cr2O3 concentration is low, and the amount of reducing agent used to reduce Cr2O3 to Cr can be reduced. [Example]

[0061] The present invention was applied to the production of SUS304 stainless steel (8% Ni-18% by mass Cr) in an AOD furnace with a 60-ton molten steel capacity, as shown in Figure 1(a) and Figure 1(b). Atmospheric pressure refining was performed in the manner shown in Figure 1(b), and reduced pressure refining was performed by spraying oxygen-containing gas in the manner shown in Figure 1(a).

[0062] The [C] concentration in the molten steel at the start of melting was approximately 1.7%, and decarburization refining was carried out to reduce it to approximately 0.01%. Decarburization refining was carried out in two ways: one where the machine learning model of the present invention was used to successively estimate the Cr2O3 activity in the slag to calculate the ΔG of the decarburization reaction, and the degree of vacuum was controlled so that ΔG would not exceed 0; and one where the machine learning model of the present invention was not used and the Cr2O3 activity in the slag was kept constant to calculate the ΔG of the decarburization reaction, and the degree of vacuum was controlled so that ΔG would not exceed 0.

[0063] After that, the pressure inside the vessel was returned to atmospheric pressure, and an Al alloy was added as a reducing agent to reduce the chromium that had become concentrated during decarburization. The reduction treatment was carried out by blowing in only Ar gas, and the steel was tapped into a ladle.

[0064] Table 1 shows the amount of reducing material used and the time from charging molten steel into the AOD to tapping it, expressed as an index, when the Cr2O3 activity in the slag was estimated using the machine learning model of the present invention (Invention Examples 1 and 2) and when it was not used (Reference Examples 3 and 4).

[0065] In Table 1, the index was calculated using Reference Example No. 4 as the standard. Table 1 also shows that cases where both the amount of reducing material used and the time from charging molten steel into the AOD to tapping were superior to those of Reference Example No. 4 and therefore highly economical were evaluated as S, while cases where only one of them was superior and therefore economical were evaluated as A. "Excellent" refers to cases where the amount of reducing material used was small or the time from charging molten steel into the AOD to tapping was short. Reference Example No. 3 was calculated based on a lower Cr2O3 activity in the slag than Reference Example No. 4.

[0066] As is clear from Table 1, Examples 1 and 2 are cases in which the degree of vacuum was controlled by estimating the Cr2O3 activity in the slag using the machine learning model of the present invention, and although the operating temperatures were different, Examples 1 and 2 were superior in both the amount of reducing material used and the time from charging the slag into the AOD to tapping the steel.

[0067] Reference Example No. 3 is superior to Reference Example No. 4 in terms of the amount of reducing material used, but this is thought to be because Reference Example No. 3 underestimated the Cr2O3 activity in the slag, which kept ΔG low and reduced the oxidation loss of Cr in the molten steel.On the other hand, Reference Example No. 3 took longer than No. 4 from charging the molten steel into the AOD to tapping, but this is thought to be because the timing of switching the vacuum level was earlier, which forced the amount of oxygen-containing gas injected to be reduced, increasing the time required for decarburization.

[0068] [Table 1] [Explanation of symbols]

[0069] 1...refining vessel, 4...molten steel, 6...slag.

Claims

【Request Item 1】 Cr-containing molten steel containing 10.5 mass % or more of Cr, ２ O ３ and slag containing at least CaO and Al contained in the slag ２ O ３ The ratio (CaO / Al ２ O ３ Ratio) is 15 or less, The composition of the slag and the Cr contained in the slag ２ O ３ Activity of a Ｃｒ２Ｏ３ The Cr content is calculated from the composition of the slag using a machine learning model that uses the relationship between the ２ O ３ Activity of a Ｃｒ２Ｏ３ A trained model for estimating the above is obtained in advance for each temperature T of the Cr-containing molten steel, The composition of the slag, the carbon concentration [C] (mass%) of the Cr-containing molten steel, the temperature of the Cr-containing molten steel, and the partial pressure P of CO gas in the smelting vessel ＣＯ are successively measured or successively predicted, The measured or predicted value of the composition of the slag is introduced into the trained model corresponding to the temperature T of the Cr-containing molten steel to obtain the Cr content. ２ O ３ Activity of a Ｃｒ２Ｏ３ is successively estimated, and the activity a of carbon in the Cr-containing molten steel is calculated based on the carbon concentration [C] (mass%) of the Cr-containing molten steel. Ｃ Seeking The Cr ２ O ３ Activity of a Ｃｒ２Ｏ３ and the estimated value of the carbon activity a Ｃ and the partial pressure P of the CO gas ＣＯ and are input into the following formula (1) to calculate ΔG, and the CO partial pressure P in the Cr-containing molten steel is adjusted so that ΔG does not become 0 or more. ＣＯ A method for decarburizing and refining Cr-containing molten steel, in which the above steps are sequentially controlled. ΔG=ΔG ０ +RTln((a ２ Ｃｒ ・P ３ ＣＯ ) / (a Ｃｒ２Ｏ３ / a ３ Ｃ )) …(1) Cr ２ O ３ +3C=2Cr+3CO …(2) However, ΔG in formula (1) is the free energy change of the decarburization reaction shown in formula (2), and ΔG ０ is the standard free energy change of the decarburization reaction shown in the above formula (2) at 1 atmosphere and 25°C, T is the temperature of the Cr-containing molten steel, and R is the gas constant (R = 8.314 × 10 －３ kJmol －１ K －１ ) and a Ｃｒ is the activity of Cr in the Cr-containing molten steel, and P ＣＯ is the partial pressure of CO in the refining vessel, and a Ｃｒ２Ｏ３ is Cr ２ O ３ is the activity of a Ｃ is the activity of carbon. 【Request Item 2】 The composition of the slag is FeO, SiO ２ , MnO, Cr ２ O ３ , CaO, Al ２ O ３ and MgO is 100 mass %. ２ , MnO, Cr ２ O ３ , CaO, Al ２ O ３ The method for decarburizing and refining Cr-containing molten steel according to claim 1, wherein the respective contents (mass%) of Sn and MgO are Sn and MgO.

Images