MÉTODO PARA GERAR MODELO DE ESTIMATIVA DO TEOR DE CAL LIVRE PARA ESCÓRIA DE ACIARIA, MÉTODO PARA ESTIMAR TEOR DE CAL LIVRE DA ESCÓRIA DE ACIARIA, MÉTODO PARA GERENCIAR TRANSPORTE DA ESCÓRIA DE ACIARIA, E MÉTODO PARA FABRICAR ESCÓRIA DE FERRO E AÇO PARA CONSTRUÇÃO DE ESTRADAS
Patent Information
- Authority / Receiving Office
- BR · BR
- Patent Type
- Applications
- Current Assignee / Owner
- JFE STEEL CORP
- Filing Date
- 2024-01-26
- Publication Date
- 2026-08-04
Smart Images
Figure 00000000_0000_ABST
Abstract
Description
1 / 34 METHOD FOR GENERATING A MODEL TO ESTIMATE THE FREE LIME CONTENT OF STEEL SLAG, METHOD FOR ESTIMATING THE FREE LIME CONTENT OF STEEL SLAG, METHOD FOR MANAGING THE TRANSPORT OF STEEL SLAG, AND METHOD FOR MANUFACTURING IRON AND STEEL SLAG FOR ROAD CONSTRUCTION FIELD OF TECHNIQUE
[001] The present invention relates to a method for generating a model for estimating the free lime content of steelmaking slag, a method for estimating the free lime content of steelmaking slag, a method for managing the transportation of steelmaking slag, and a method for manufacturing iron and steel slag for road construction. BACKGROUND TECHNIQUES
[002] Among the iron and steel slags produced during the steelmaking process, steelmaking slags generated in the refining stage, specifically converter slag and electric arc furnace slag, are crystalline and harder than other types of slag. Steelmaking slags can be suitablely used as base materials for roads when cooled and then crushed to a predetermined particle size. However, some of the quicklime used during the refining process may remain in the steelmaking slags. The quicklime that remains in the steelmaking slags is called free lime. Alternatively, the quicklime that remains in the steelmaking slags may be called free calcium oxide, uncombined lime, or f-CaO or other terms. This free lime undergoes a hydration reaction when exposed to water, such as rainwater or seawater, resulting in volume expansion.Therefore, when steelmaking slag containing free lime is used as a base material, the volume expansion caused by the hydration reaction of the free lime in the steelmaking slag can result in the formation of... Petition 870250085420, dated 09 / 22 / 2025, page 7 / 52 2 / 34 of irregularities similar to ridges at the base. There is also the possibility of pop-outs, in which the base layer expands locally, forming a mound and breaking through the asphalt. The occurrence of such a condition can hinder the circulation of vehicles and pedestrians.
[003] The JIS A 5015:2018 standard, Iron and Steel Slag for Road Construction, specifies iron and steel slag (hereinafter referred to as iron and steel slag for road construction) used in roadbeds and hot asphalt mixes. Specifically, the JIS A 5015:2018 standard specifies the physical and chemical properties, as well as the particle size, of iron and steel slag for road construction according to the intended application of the iron and steel slag for road construction. For example, when steelmaking slag is used as iron and steel slag for road construction, the expansibility of the steelmaking slag must be 1.0% or less in terms of the expansion rate by immersion in water.
[004] To meet this requirement, after crushing, steel slag is subjected to an aging treatment, in which it reacts with air and water to stabilize the expansion stability of the steel slag.
[005] The steelmaking slag produced in a converter and its aging treatment will be described. In a converter, blowing is carried out in a discontinuous (batch) process for each charge. The operation of the converter under virtually identical conditions still easily generates variations in properties such as basicity, viscosity, and chemical composition, which are indicative of the reactivity of the steelmaking slag. Equipment-related errors can also occur. As a result, the free lime content of the steelmaking slag produced per operation inevitably varies. Petition 870250085420, dated 09 / 22 / 2025, page 8 / 52 3 / 34
[006] Currently, a mixture of steelmaking slag accumulated from multiple charges is stored in a slag yard for aging treatment. Aging treatment can generally be classified into atmospheric aging and steam aging. In atmospheric aging treatment, steelmaking slag is stored in a slag yard for a predetermined period, allowing the water present in the air to react with the free lime. In steam aging, steam is fed to the steelmaking slag stored in a slag yard for a predetermined period, allowing the steam (water) to react with the free lime. The expansion of the steelmaking slag is correlated with the free lime content. Regardless of which aging treatment is performed, it is appropriate to quantify the free lime content of the steelmaking slag and perform the aging treatment for a period of time according to the free lime content.Specifically, the storage period in the slag yard, i.e., the aging treatment time, is determined by sampling a portion of the steelmaking slag stored in the slag yard to assess its expansibility and free lime content and determine the aging treatment time based on these results. However, the steelmaking slag stored in the slag yard is a mixture of steelmaking slag from multiple operations, as described above. Therefore, the assessment of expansibility and free lime content described above only evaluates the expansibility and free lime content of steelmaking slag from one of the multiple feedstocks.
[007] Considering the safety of using steel slag as a base material for roads, the aging treatment is preferably carried out over a longer period than the time determined in the aging treatment based Petition 870250085420, dated 09 / 22 / 2025, page 9 / 52 4 / 34 in the expansibility and free lime content of the steelmaking slag obtained as described above. However, even if the aging treatment is carried out for a long period, as described above, there is a possibility that the expansibility of the steelmaking slag after treatment will not meet the requirements specified in the JIS A 5015:2018 standard. If the requirements are not met, the aging treatment is carried out again.
[008] The simple and quick assessment of the free lime content of steelmaking slag allows for the prediction of the expansibility of the steelmaking slag produced for each operation. Furthermore, the steelmaking slag for each operation can be subjected to aging treatment for an appropriate time. Even for a mixture of steelmaking slag from multiple operations, the total quantity, average value, and maximum value of the free lime content in the mixture are determined using the free lime content for each batch. Therefore, the steelmaking slag can also be subjected to aging treatment for an appropriate time according to the total quantity, average value, and maximum value described above.
[009] Several methods for quantifying the free lime content of steelmaking slag have been described. Non-Patent Literature 1 describes an ethylene glycol extraction method. In the ethylene glycol extraction method, steelmaking slag ground to 0.074 mm or less is stirred for one hour in ethylene glycol heated to 80 °C to dissolve the calcium oxide (CaO) contained in the steelmaking slag. Subsequently, the calcium (Ca) in the ethylene glycol is quantified to determine the free lime content of the steelmaking slag. The ethylene glycol extraction method is a practical standard method used in the steel industry.
[0010] Patent Literature 1 describes a method for quantifying the free lime content of steelmaking slag from the ratio between the integrated intensity of the nuclear magnetic resonance spectrum. Petition 870250085420, dated 09 / 22 / 2025, page 10 / 52 5 / 34 obtained from a mixture of steelmaking slag and a reagent using solid-state NMR.
[0011] Patent Literature 2 describes a method for reducing the expansion of steelmaking slag by hydration by estimating the free lime content of the steelmaking slag from a four-component CaO-SiO2-MgO-Al2O3 phase diagram and controlling the composition of the steelmaking slag.
[0012] Patent Literature 3 describes a method for analyzing slag composition by calculating the amount of Ca contained in the slag in terms of CaO, assuming that all the Ca contained in the slag exists as CaO.
[0013] Non-Patent Literature 2 describes, as a related technique, a method for estimating an alternative indicator (L-value) correlated to free lime content using an estimation formula based on the composition and stoichiometry of a specific mineral phase of the slag containing Ca. LIST OF QUOTES Patent Literature
[0014] PTL 1: Publication of Examined No. 2012-42464
[0015] PTL 2: Publication of Examined No. 2001-64714
[0016] PTL 3: Publication of Examined No. 2016-48235 Non-Patented Literature Japanese Patent Application No. Japanese Patent Application No. Japanese Patent Application No.
[0017] NPL 1: Standardization of Techniques for Characterization of Free CaO in Iron and Steel Slag, Final Report of the Study Group on Standardization of Techniques for Characterization of Free CaO in Iron and Steel Slag, pages 88-95, March 28, 2013, edited by the Committee for Standardization of Techniques for Characterization of Petition 870250085420, dated 09 / 22 / 2025, page 11 / 52 6 / 34 Free CaO in Iron and Steel Slag, Analytical Technology Committee, Division of Process Technology, The Iron and Steel Institute of Japan
[0018] NPL 2: Wada, Kaname et al., Hardness Stabilization Treatment for Molten Converter Slag, Seitetsu Kenkyu (Iron and Steel Research in Japanese), N° 3011, 1980, páginas 59-70
[0019] NPL 3: Okazaki, Kodai et al., Examination of Measurement Method of the Hardness by Laser Induced Breakdown Spectroscopy (LIBS), The Atomic Energy Society of Japan, 2019 Spring Annual Meeting, Março de 2019
[0020] NPL 4: Murayama, Seiji et al., Plasma Emission Spectroscopy for Solid Sample Analysis: Measurement Methods Series 2, The Spectroscopical Society of Japan, Center for Academic Publications Japan, Outubro de 1982 SUMÁRIO DA INVENÇÃO Problema Técnico
[0021] The ethylene glycol extraction method described in Non-Patent Literature 1 involves the complexity characteristic of wet analysis and lacks speed. This method also requires operator proficiency to obtain highly accurate quantitative results. Since free lime (f-CaO) and calcium hydroxide (Ca(OH)2) contained in steelmaking slag are extracted simultaneously, this method requires quantification of the amount of Ca(OH)2 through thermogravimetric analysis and calculation of the free lime content, resulting in a highly laborious operation.
[0022] The method described in Patent Literature 1 involves complex sample preparation, requires specialized analytical equipment, and takes several days to complete the measurements, resulting in slow performance.
[0023] The method described in the Patent Literature2 estimates the free lime content assuming a thermodynamically equilibrium state Petition 870250085420, dated 09 / 22 / 2025, page 12 / 52 7 / 34 is ideal and lacks precision when applied to complex systems with multiple components, such as steelmaking slag.
[0024] The method of analyzing the composition of slag described in Patent Literature3 estimates the free lime content assuming that all the Ca contained in the slag exists as CaO, as described above. Therefore, the slag composition analysis method described in Patent Literature3 may not be able to determine the free lime content with high precision.
[0025] The present invention was developed to solve the above problems and provides a method for generating a model for estimating the free lime content of steelmaking slag by means of which the free lime content of steelmaking slag can be estimated in a simple, fast and highly accurate manner, a method for estimating the free lime content of steelmaking slag, a method for managing the transportation of steelmaking slag and a method for manufacturing iron and steel slag for road construction. Solution to the Problem
[0026] The inventors of the present invention focused on the emission spectrum obtained by subjecting steelmaking slag to laser-induced breakdown spectroscopy, thus concluding the present invention. The present invention was developed to achieve the above objective.
[0027] [1] A method for estimating the free lime content of steelmaking slag, the method comprising: a step of acquiring an emission spectrum by irradiating the steelmaking slag with a laser to convert a portion of a steelmaking slag surface into plasma and acquiring an emission spectrum from the plasma; and a step of estimating the free lime content by introducing the intensities of an emission spectrum, including peaks caused by emission lines of Ca, Fe and Si, within the emission spectrum acquired in the emission spectrum acquisition step into a content estimation model. Petition 870250085420, dated 09 / 22 / 2025, page 13 / 52 8 / 34 of free lime to generate the free lime content of laser-irradiated steelmaking slag.
[0028] [2] The method for estimating the free lime content in steelmaking slag, according to [1], wherein the estimation step includes introducing the intensities of an emission spectrum, including peaks caused by Al and P emission lines, into the free lime content estimation model.
[0029] [3] The method for estimating free lime content in steelmaking slag, according to [1], wherein the free lime content estimation model is constructed using partial least squares regression using the emission spectrum intensities, including peaks caused by the emission lines of Ca, Fe and Si, within the emission spectrum acquired in the emission spectrum acquisition step as explanatory variables and using the free lime content as the response variable.
[0030] [4] The method for estimating the free lime content in steelmaking slag, according to [3], in which the free lime content estimation model uses the intensities of an emission spectrum, including peaks caused by Al and P emission lines, as explanatory variables.
[0031] [5] The method for estimating free lime content in steelmaking slag, according to [1], wherein the free lime content estimation model is a trained machine learning model that uses the emission spectrum intensities, including peaks caused by the Ca, Fe and Si emission lines, within the emission spectrum acquired in the emission spectrum acquisition step as input data and the free lime content as output data.
[0032] [6] The method for estimating the free lime content in steelmaking slag, according to [5], wherein the free lime content estimation model uses the intensities of an emission spectrum, including Petition 870250085420, dated 09 / 22 / 2025, page 14 / 52 9 / 34 still the peaks caused by the Al and P emission lines, as input data.
[0033] [7] The method for estimating the free lime content in steelmaking slag, according to [1], wherein the emission spectrum acquired in the emission spectrum acquisition step has wavelengths of 270 nm or more and 410 nm or less.
[0034] [8] A method for producing iron and steel slag for road construction, the method comprising: estimating the free lime content of steelmaking slag using the method for estimating the free lime content of steelmaking slag according to [1]; and defining the aging treatment conditions for the steelmaking slag based on the estimated free lime content of the steelmaking slag.
[0035] [9] A method for managing the transport of steelmaking slag comprising: estimating the free lime content of the steelmaking slag using the method for estimating the free lime content of steelmaking slag according to [1]; and selecting a transport destination for the steelmaking slag based on the estimated free lime content of the steelmaking slag.
[0036]
[10] A method for generating a free lime content estimation model for steelmaking slag for use in estimating the free lime content in steelmaking slag, the method comprising: a first acquisition step of irradiating the steelmaking slag with a laser to convert a portion of the steelmaking slag surface into plasma and acquiring an emission spectrum of the plasma; a second acquisition step to acquire the actual free lime content of the steelmaking slag whose emission spectrum was acquired in the first acquisition step; and a model generation step for acquiring multiple datasets, each consisting of pairs of the emission spectrum acquired in the first acquisition step and the actual free lime content acquired in the second acquisition step, and based on these sets Petition 870250085420, dated 09 / 22 / 2025, page 15 / 52 10 / 34 data points, generate a model to estimate the free lime content using the intensities of an emission spectrum, including peaks caused by emission lines in Ca, Fe, and Si, within the emission spectrum acquired in the first acquisition step as input and using the actual free lime content acquired in the second acquisition step as output.
[0037]
[11] The method for generating a model for estimating the free lime content for steelmaking slag according to
[10] , wherein the model generation step uses the intensities of an emission spectrum, including peaks caused by Al and P emission lines, as input. Advantageous Effects of the Invention
[0038] The present invention can provide a method for generating a free lime content estimation model for steelmaking slag whereby the free lime content of the steelmaking slag can be estimated in a simple, fast and highly accurate manner. According to the present invention, the use of the free lime content estimation model allows the free lime content of the steelmaking slag to be estimated in a simple, fast and highly accurate manner. As a result, the free lime content of the steelmaking slag for each operation can be estimated in a simple, fast and highly accurate manner and an appropriate aging treatment time can be defined based on the free lime content of the steelmaking slag for each operation. Even for a mixture of steelmaking slag from multiple feedstocks, an appropriate aging treatment time can be defined based on the maximum free lime content.As a result, it is possible to achieve expansion stability of iron and steel slag for road construction and to carry out efficient aging treatment, thereby reducing the manufacturing costs of steelmaking slag used as base material. Furthermore, a destination for... Petition 870250085420, dated 09 / 22 / 2025, page 16 / 52 11 / 34 Suitable transport for steelmaking slag can be selected based on the estimated free lime content, thereby improving the efficiency of steelmaking slag transport management at the facility. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] Figure 1 illustrates exemplary regression coefficients.
[0040] Figure 2 illustrates the relationship between the measured free lime content and the estimated free lime content by substituting the emission spectrum of the steelmaking slag samples shown in Table 1 into the model equation.
[0041] Figure 3 illustrates an emission spectrum assigned to Ca, Fe, Si, Al, and P in the wavelength range of 270 to 410 nm that is used as explanatory variables to generate the model equation and the emission spectrum intensity in Example of the Invention 4. DESCRIPTION OF THE MODALITIES
[0042] The inventors of the present invention investigated a method for estimating the free lime content of steelmaking slag in a simple, fast, and highly accurate manner. In this investigation, the inventors of the present invention focused on laser-induced breakdown spectroscopy (hereinafter referred to as LIBS), which is an elemental analysis method that allows direct analysis of samples in air without pretreatment and enables rapid and simultaneous multi-element analysis through simple operations. In LIBS, the steelmaking slag to be analyzed is irradiated with a high-energy pulsed laser to convert a portion of the steelmaking slag into plasma. In LIBS, the excited light, i.e., the plasma emission, obtained from the steelmaking slag plasma is collected and spectrally dispersed to perform quantitative analyses from the wavelengths and intensities of the emission lines that correspond to the elements. Petition 870250085420, dated 09 / 22 / 2025, page 17 / 52 12 / 34
[0043] The elements that constitute steelmaking slag exist primarily in the form of oxides. However, the elements that constitute steelmaking slag exist not only as oxides of individual elements, but also as composite oxides containing multiple elements. For example, calcium (Ca) exists as calcium silicates (e.g., CaSiO3), calcium aluminates (e.g., CaAl2O4), calcium ferrites (e.g., CaFeO2), and other forms. Therefore, it is difficult to calculate the quantities of individual element oxides and composite oxides from the quantity of each element contained in the steelmaking slag obtained by LIBS.
[0044] Non-Patent Literature 3 suggests that the emission spectrum obtained by LIBS may vary depending on the bonding state of the compounds. The inventors of the present invention focused on the emission spectrum obtained by LIBS and studied the estimation of free lime content based on the relationship between the emission spectrum obtained by LIBS and the free lime content.
[0045] The calcium-based compounds contained in steelmaking slag include calcium silicates (e.g., CaSiO3), calcium aluminates (e.g., CaAbO4), calcium ferrites (e.g., CaFeO2), and tricalcium phosphates (e.g., Ca3(PO4)2), as described above. The inventors of the present invention focused on the emission spectrum of the calcium-based compounds obtained using LIBS. Specifically, a model for estimating the free lime content is generated using the emission spectrum of the calcium-based compounds as explanatory variables and the actual free lime content as the response variable. The emission spectrum used as explanatory variables is an emission spectrum that includes at least one peak caused by the calcium emission line, at least one peak caused by the iron emission line, and at least one peak caused by the silicon emission line among the peaks caused by the calcium, iron, silicon, and silicon emission lines. Petition 870250085420, dated 09 / 22 / 2025, page 18 / 52 13 / 34 Al and P, which are elements that constitute compounds based on Ca. The inventors of the present invention have discovered that the free lime content can be adequately estimated.
[0046] The present invention was made based on the above findings. A preferred embodiment of the present invention will be described in detail below. Steelmaking slag
[0047] The steelmaking slag used in this embodiment may be converter slag or electric arc furnace slag. The steelmaking slag may be in the molten state or in the solid state. First Stage of Acquisition
[0048] Using a conventionally known laser-induced breakdown spectrometer, the surface of the steelmaking slag is irradiated with a pulsed laser to convert a portion of the slag surface into plasma, and the emission spectrum is acquired from the plasma. The steelmaking slag surface can be irradiated with a pulsed laser once or multiple times. The emission spectrum obtained by irradiating the steelmaking slag surface with a pulsed laser can be an emission spectrum obtained from a single laser irradiation or an emission spectrum obtained by the accumulation of multiple emission spectra obtained from multiple laser irradiations. To improve reliability, multiple emission spectra, each obtained from a single laser irradiation, or multiple emission spectra, each obtained by the accumulation of multiple emission spectra, can be used.The reliability of the emission spectrum is preferably ensured by the use of multiple emission spectra, each obtained by accumulating multiple emission spectra from multiple laser irradiations. The term a single pulsed laser irradiation, as described above, refers to the emission of a pulsed laser onto a single steelmaking slag. Petition 870250085420, dated 09 / 22 / 2025, page 19 / 52 14 / 34 only once. In the case of a double pulsed laser, where a pulsed laser is emitted twice in rapid succession, i.e., in a short period of time, the term a single pulsed laser irradiation refers to the emission of a pulsed laser onto steelmaking slag twice in a short period of time.
[0049] The interval between pulses from a pulsed laser to a double pulsed laser, when used for LIBS, is preferably from several microseconds to tens of microseconds, more preferably, for example, 2 microseconds.
[0050] LIBS can be used for molten or solid steel slag. For molten steel slag, the emission spectrum is acquired, for example, by directly irradiating, with a pulsed laser, the molten steel slag received in a slag ladle after the converter blowdown. For solid steel slag, the steel slag is sampled from the slag ladle described above and cooled. A small piece of steel slag thus obtained is irradiated with a pulsed laser to acquire an emission spectrum. Solid steel slag tends to exhibit greater variations (differences) in its emission spectrum among the slag samples than molten steel slag due to compositional variations. Therefore, it is preferable to perform analyses at multiple points to reduce variations in the emission spectrum.Multipoint analysis refers to sampling multiple different locations in steelmaking slag from a predetermined batch, i.e., from a single batch of steelmaking slag, acquiring the emission spectrum from one location for each slag sample, and calculating their average value. Alternatively, multipoint analysis refers to sampling one location in steelmaking slag from a predetermined operation, i.e., from a single batch of steelmaking slag, acquiring the emission spectra from multiple locations. Petition 870250085420, dated 09 / 22 / 2025, page 20 / 52 15 / 34 in the sampled steelmaking slag and calculating its average value. Alternatively, multipoint analysis refers to combining these two multipoint analyses to acquire multiple emission spectra and calculate their average value. Second Stage of Acquisition
[0051] In this mode, the free lime content of the steelmaking slag whose emission spectrum was acquired in the first acquisition step is measured. The free lime content can be measured using steelmaking slag from the same batch as the steelmaking slag whose emission spectrum was acquired in the first acquisition step.
[0052] Specifically, the free lime content of steelmaking slag is measured by means of a conventionally known ethylene glycol extraction method. In this embodiment, the ethylene glycol extraction method is used as an example of a method for measuring the free lime content of steelmaking slag, however, the present invention is not limited to this method. The second acquisition step and the first acquisition step can be carried out in such a way that one of the two acquisition steps is carried out before the other acquisition step or the two acquisition steps are carried out almost simultaneously, i.e., in parallel. Model Generation Stage
[0053] A model for estimating the free lime content of steelmaking slag is generated using the emission spectrum acquired in the first acquisition step as explanatory variables and the measured free lime content acquired in the second acquisition step as a response variable.
[0054] Explanatory variables may include variables that have a certain degree of mutual correlation, and collinearity based on such correlation may compromise the accuracy of the estimate in simple multiple regression analyses. Since the constituent elements Petition 870250085420, dated 09 / 22 / 2025, page 21 / 52 16 / 34 of steelmaking slag may include elements that have a certain degree of mutual correlation; the explanatory variables may also exhibit a mutual correlation, even in the analysis of steelmaking slag. Therefore, it is preferable to select an analytical method that, in principle, is free from collinearity problems. Examples of such a method include partial least squares (PLS) regression.
[0055] Partial least squares regression involves transforming the explanatory variables into mutually uncorrelated principal component axes and then performing a regression analysis between a small number of principal components and a response variable. Therefore, partial least squares regression is suitable for estimating a response variable from multiple analysis values whose explanatory variables exhibit a correlation with each other and can achieve high estimation precision. In this embodiment, the free lime content of steelmaking slag is estimated using a model (partial least squares regression model) constructed through partial least squares regression as a model for estimating free lime content.
[0056] The emission spectrum to be included as explanatory variables in the free lime content estimation model will be specifically described as the data for generating the partial least squares regression model. The emission spectrum is an emission spectrum that includes at least one peak caused by the Ca emission line, at least one peak caused by the Fe emission line, and at least one peak from the Si emission line within the emission spectrum of the steelmaking slag acquired in the first acquisition step. This is because the peak caused by the Ca emission line, the peak caused by the Fe emission line, and the peak caused by the Si emission line can reflect information corresponding to the amount of compounds based on Ca. Preferably, the es Petition 870250085420, dated 09 / 22 / 2025, page 22 / 52 17 / 34 The emission spectrum, which also includes at least one peak caused by the Al emission line and at least one peak caused by the P emission line, is used as explanatory variables. For example, the emission spectrum described above could be the emission spectrum across the entire wavelength range measured in the first acquisition step. Alternatively, the emission spectrum could be an emission spectrum in the wavelength range of 400 to 540 nm, which is a partially continuous wavelength range within the entire wavelength range measured in the first acquisition step. Alternatively, the emission spectrum could be an emission spectrum in the wavelength range of 270 to 410 nm, which is a partially continuous wavelength range within the entire wavelength range measured in the first acquisition step.
[0057] The emission spectrum includes multiple emission lines resulting from complex transitions between excitation levels induced by a laser as an energy source. In other words, the emission spectrum is the data indicating the intensity of plasma emission at a specific plasma emission wavelength for each element. When multiple elements are present in steelmaking slag, the emission spectrum includes multiple emission lines specific to the respective excited elements. Each element has a representative or characteristic wavelength (hereinafter referred to as the central wavelength), which differs from those of other elements. In the emission spectrum, the data indicating the emission lines of the elements appear as waveforms (called peaks) centered on the central wavelengths of the respective elements.The central wavelengths for the elements Ca, Fe, Si, Al, and P described above could be, for example, the wavelengths described in Lite. Petition 870250085420, dated 09 / 22 / 2025, p. 23 / 52 18 / 34 Non-Patentable Ratura 4.
[0058] In this approach, it is preferable to acquire, as explanatory variables, intensity data at at least three points: one at the top of the peak, which corresponds to the central wavelength of each element in the emission spectrum, and at least one very close point on the peak waveform, both below and above the central wavelength. Alternatively, it is more preferable to acquire, as explanatory variables, intensity data at five or more points in total: one at the top of each element and at least two very close points on the peak waveform, below and above the central wavelength. The very close points can be defined experimentally in advance. Examples of very close points include two points on the peak waveform below and above the central wavelength, halfway up the top of the peak.Examples of very close points also include the intersection points between the baseline and the peak waveform or the trough points between adjacent peaks.
[0059] A partial least squares regression model that corresponds to the free lime content estimation model of this modality, which defines the relationship between the explanatory variables and the response variable, is generated.
[0060] As an embodiment of the present invention, a method for generating a model for estimating free lime content using ten samples of steelmaking slag will be described.
[0061] In this embodiment, the free lime content (hereinafter referred to as the measured f-CaO value) for ten samples of steelmaking slag was measured using the conventionally known ethylene glycol extraction method. The measured f-CaO values (% by mass, hereinafter simply referred to as %) are summarized Petition 870250085420, dated 09 / 22 / 2025, page 24 / 52 19 / 34 of those in Table 1. The LIBS emission spectrum for ten steelmaking slag samples was acquired as described above. The model equation (free lime content estimation model) represented by Equation (1) below, where the emission spectrum was used as the explanatory variable and the measured f-CaO (%) value of each steelmaking slag sample shown in Table 1 was used as the response variable, was generated. The wavelength range of the emission spectrum of each steelmaking slag sample used as an explanatory variable was 270 to 410 nm, and the emission spectrum intensity data were acquired at 464 points (n = 464) within this wavelength range. Table 1 Sample of steelmaking slag. Measured f-CaO value (% by mass). A 2.53 B 1.81 C 1.46 D 7.64 E 0.94 F 6.49 G 4.60 H 6.13 I 3.51 J 1.85 Y - ko + klxXl + k2xX2+k3xX3 + knxXn (1)
[0062] In Equation (1), Y represents the free lime content (f-CaO) and Xi to Xn represent the emission spectrum intensity data corresponding to the explanatory variables, ko to kn represent the regression coefficients obtained by expanding the latent variables and Figure 1 illustrates exemplary regression coefficients. The regression coefficients from ko to kn may be associated with variables Petition 870250085420, dated 09 / 22 / 2025, p. 25 / 52 20 / 34 latent variables and can incorporate their weights. The relationship between the latent variables and the regression equation is described below. Y = bo + bixTi + b2 x T2 +......+ br x Tr ··· (2)
[0063] In Equation (2), Ti and Tr represent latent variables and bo and br represent regression coefficients for the latent variables. Specifically, Ti represents a first latent variable, T2 represents a second latent variable, and Tr represents the r-th latent variable. The latent variables can be represented by the following equations. Ti = W11 X Xi + W12 X X2 + ··· + Win X Xn T2 = W2i X Xi + W22 X X2 + ··· + W2n X Xn Tr = Wri X Xi + Wr2 X X2 + ··· + Wrn X Xn
[0064] Wii a Wmsão os pesos das varias entrada para as varias latentes.
[0065] In addition to the ten steelmaking slag samples described above, a steelmaking slag with an unknown free lime content (referred to as the steelmaking slag of interest) was used to acquire the LIBS emission spectrum in the same manner described above. The steelmaking slag of interest is the steelmaking slag to be subjected to aging treatment and may be, for example, converter slag or electric arc furnace slag. The steelmaking slag of interest may be in the molten or solid state. The LIBS emission spectrum acquisition step of the steelmaking slag of interest corresponds to the emission spectrum acquisition step in this mode. Estimation Stage
[0066] The free lime content (% by mass, hereinafter simply referred to as %) of the steelmaking slag of interest is calculated using the free lime content estimation model represented by Equation (i). Specifically, the LIBS emission spectrum of the slag Petition 870250085420, dated 09 / 22 / 2025, page 26 / 52 21 / 34 steelmaking slag of interest is replaced as the explanatory variable in the free lime content estimation model represented by Equation (1). The free lime content (%) of the steelmaking slag of interest is calculated as the response variable.
[0067] Figure 2 illustrates the relationship between the measured free lime content (hereinafter referred to as the measured f-CaO value) (%) of the steel slag samples presented in Table 1 and the estimated free lime content (hereinafter referred to as the estimated f-CaO value) (%) obtained by substituting the emission spectrum of the steel slag of interest into the equation of the model above. The steel slag samples presented in Table 1 are considered steel slags of interest with unknown free lime contents. The horizontal axis of Figure 2 represents the measured f-CaO value (% by mass, hereinafter simply referred to as %) and the vertical axis represents the estimated f-CaO value (% by mass, hereinafter simply referred to as %). The coefficient of determination (R2) is 0.9718, indicating a very high correlation between the measured f-CaO value (%) and the estimated f-CaO value (%). According to this method, the free lime content can be predicted with high accuracy.
[0068] In this embodiment, the free lime content of the steelmaking slag is estimated for each operation, for example, in a converter or electric arc furnace, using the free lime content estimation model described above. Subsequently, the appropriate aging treatment time for the steelmaking slag for each charge is defined based on the estimated f-CaO (%) value to perform the aging treatment. For a mixture of steelmaking slag from multiple charges, the aging treatment time is defined based on the maximum or average value of the estimated f-CaO (%) value for each charge or on the total amount of f-CaO to perform the aging treatment. Petition 870250085420, dated 09 / 22 / 2025, page 27 / 52 22 / 34
[0069] Alternatively, the storage location for steel slag is selected for each load based on the estimated f-CaO (%) value. Specifically, steel slags with similar estimated f-CaO (%) values are stored in the same location. This allows for efficient aging treatment. The aging treatment time defined based on the estimated f-CaO (%) value and the steel slag storage location selected based on the estimated f-CaO (%) value correspond to the aging treatment conditions for steel slag in the method of manufacturing iron and steel slag for road construction according to this embodiment.
[0070] According to this method, the emission spectrum of each steelmaking slag can be easily obtained using LIBS. The free lime content (%) can be estimated simply, quickly, and with high accuracy based on the emission spectrum and the free lime content estimation model. Even though the free lime content (%) of the steelmaking slag varies from batch to batch, the free lime content (%) of the steelmaking slag for each batch can be estimated simply, quickly, and with high accuracy. This allows the definition of the ideal aging treatment time for the steelmaking slag for each batch. Furthermore, it is also possible to optimize the storage location of the steelmaking slag for each batch depending on its free lime content (%).Therefore, even though the free lime content (%) of steelmaking slag varies from batch to batch, the excess or deficiency of the aging treatment time of the steelmaking slag for each batch can be reduced to improve the efficiency of the aging treatment, reducing the costs of the aging treatment of the steelmaking slag.
[0071] The free lime content of steelmaking slag can be sufficiently reduced by means of the above aging treatment. Petition 870250085420, dated 09 / 22 / 2025, page 28 / 52 23 / 34 Therefore, steel slag for road construction, composed of steel slag, i.e., steel slag for base materials, will hardly expand even when exposed to water, thus improving the expansion stability and quality stability of the steel slag for base materials. According to this method, the free lime content (%) can be estimated simply, quickly, and with high accuracy. It is also possible to manage the transport of steel slag properly and quickly based on the estimated free lime content (%) and improve the efficiency of slag transport management at the plant. For example, even if the estimated free lime content (%) of the steel slag deviates from the permitted range, it is possible to quickly select a transport destination in applications such as calcium-enhanced materials, where the expansibility of the steel slag is not a problem or is less likely to cause problems.The free lime content (%) in steelmaking slag can be estimated after aging treatment, instead of estimating the free lime content (%) in the steelmaking slag before aging treatment. This allows for a simple and quick determination of whether the free lime content of the steelmaking slag has been sufficiently reduced by the aging treatment. The allowable range described above refers to the free lime content of the steelmaking slag that meets the requirements specified in JIS A 5015:2018 Iron and steel slag for road construction.
[0072] As described above, in this embodiment, the free lime content estimation model using the LIBS emission spectrum of steelmaking slag as explanatory variables, i.e., input data, and the free lime content as a response variable, i.e., output data, is generated in advance. The LIBS emission spectrum of the steelmaking slag whose free lime content (%) is to be estimated is substituted into the free lime content estimation model. This allows Petition 870250085420, dated 09 / 22 / 2025, page 29 / 52 24 / 34 that the free lime content (%) of steelmaking slag is estimated in a simple, fast and highly accurate manner. The expansibility of steelmaking slag can also be estimated based on the estimated free lime content (%). Estimating the free lime content (%) of steelmaking slag using the free lime content estimation model for steelmaking slag corresponds to the method for estimating the free lime content of steelmaking slag of the present invention. Since the method for estimating the free lime content of steelmaking slag of the present invention is a simple method, as described above, the free lime content (%) can be estimated in approximately 40 minutes. Therefore, the free lime content (%) of steelmaking slag for each operation can be estimated quickly.
[0073] In this approach, a partial least squares regression model is used as an estimation model for free lime content; however, a trained machine learning model can be used instead. In other words, multiple datasets, each composed of the explanatory variables and the response variable, are acquired, and machine learning is performed using these datasets as training data to generate a machine learning model trained as an estimation model for free lime content. The LIBS emission spectrum of steelmaking slag with unknown free lime content (%) can then be fed into the trained machine learning model to generate the estimated free lime content.
[0074] In this modality, the wavelength range of the emission spectrum of each steelmaking slag sample was from 270 to 410 nm, and the intensity data of the emission spectrum at 464 points (n = 464) within this wavelength range were used as explanatory variables. However, instead of this emission spectrum, any emission spectrum that includes a peak caused Petition 870250085420, dated 09 / 22 / 2025, page 30 / 52 25 / 34 saturates caused by the Ca emission line, a peak caused by the Fe emission line, and a peak caused by the Si emission line can be used as emission spectra for the explanatory variables. In this embodiment, ten samples of steelmaking slag were used to generate the model equation. However, 10 or more samples of steelmaking slag can be used, or 20 or more samples of steelmaking slag are more preferably used. EXAMPLES
[0075] Examples in which the free lime content (%) of steelmaking slag is estimated using the free lime content estimation model for steelmaking slag according to this embodiment will be described below. The present invention is not limited to the Examples below.
[0076] The emission spectrum obtained by irradiating each of the steelmaking slag samples presented in Table 1 with a pulsed laser (100 mJ, 15 Hz) for one second and the accumulated emission spectra were used as explanatory variables. The measured f-CaO value (%) of each steelmaking slag sample presented in Table 1 was used as the response variable. A model equation (free lime content estimation model) was generated based on the explanatory variables, the response variable, and Equation (1). The emission spectrum intensity data corresponding to the wavelengths selected in each of the Examples of the Invention 1 to 3 were used as those of the emission spectrum used as explanatory variables for the generation of the model equation. OriginPro2017 (registered trademark) was used to perform the partial least squares regression analysis to generate the model equation.
[0077] Three types of steelmaking slag (X, Y, and Z) were prepared as steelmaking slag samples to estimate the free lime content (%). Each type of steelmaking slag (X, Y, and Z) was a slag block. Petition 870250085420, dated 09 / 22 / 2025, page 31 / 52 26 / 34 of steelmaking slag of approximately 3 cm collected from steelmaking slag discharged after decarburization blowing. The surface of each type of steelmaking slag (X, Y, Z) was irradiated with a pulsed laser (100 mJ, 15 Hz) for one second, as per the model equation generation. The emission spectra acquired over one second were accumulated and the resulting emission spectrum was used as explanatory variables for each type of steelmaking slag.
[0078] Subsequently, the emission spectrum obtained from each type of steelmaking slag (X, Y, Z) was substituted as explanatory variables in the free lime content estimation model to estimate the free lime content (%) of each type of steelmaking slag (X, Y, Z). However, the emission spectrum intensity data corresponding to the same wavelengths as the explanatory variables used to generate the model equation were used as explanatory variables.For comparison, the free lime content (%) of each type of steelmaking slag (X, Y, Z) was measured using the ethylene glycol extraction method. The estimated f-CaO (%) values are compared with the measured f-CaO (%) values.
[0079] In Example of Invention 1, the emission spectrum of each type of steelmaking slag (X, Y, Z) used as explanatory variables was an emission spectrum in the wavelength range of 270 to 410 nm, which is a wavelength range that includes all emission lines of Ca, Fe, Si, Al, and P. Specifically, intensity data at 464 points of the emission spectrum in the wavelength range of 270 to 410 nm were used as explanatory variables. The free lime content estimation model was generated from the slag emission spectrum described above in Table 1, and the free lime content (%) was estimated from the emission spectrum of each type of steelmaking slag (X, Y, Z) based on the model equation. Petition 870250085420, dated 09 / 22 / 2025, page 32 / 52 27 / 34
[0080] In Example of Invention 2, the emission spectrum of each type of steelmaking slag (X, Y, Z) used as explanatory variables was an emission spectrum in the wavelength range of 400 to 540 nm, which is a wavelength range that includes all emission lines of Ca, Fe, Si, Al, and P. Specifically, intensity data at 464 points of the emission spectrum in the wavelength range of 400 to 540 nm were used as explanatory variables. In Example of Invention 2, the free lime content (%) was estimated in the same manner as in Example of Invention 1, except for the explanatory variables used to generate the model and estimate the free lime content.
[0081] In Example of Invention 3, the emission spectrum of each type of steelmaking slag (X, Y, Z) used as explanatory variables includes intensity data at the tops of peaks assigned to Ca, Fe, Si, Al, and P in the wavelength range of 270 to 410 nm, and at two points immediately before and after each peak top. Specifically, intensity data for Ca include the emission spectrum intensities at the top of the Ca peak (393.4 nm) and at points (393.1 nm, 393.7 nm) before and after the top of the Ca peak. Intensity data for Fe include the emission spectrum intensities at the top of the Fe peak (374.9 nm) and at points (374.6 nm, 375.2 nm) before and after the top of the Fe peak. Intensity data for Si include the emission spectrum intensities at the top of the Si peak (288.1 nm) and at points (287.8 nm, 288.4 nm) before and after the top of the Si peak.The intensity data for Al include the emission spectrum intensities at the top of the Al peak (309.2 nm) and at the points (308.9 nm, 309.5 nm) before and after the top of the Al peak. The intensity data for P include the emission spectrum intensities at the top of the... Petition 870250085420, dated 09 / 22 / 2025, page 33 / 52 28 / 34 peak of P (405.9 nm) and at points (405.6 nm, 406.2 nm) before and after the top of the P peak. A total of 15 intensity data points from each emission spectrum described above were used as explanatory variables. In Example of Invention 3, the free lime content (%) was estimated in the same manner as in Example of Invention 1, except for the explanatory variables used to generate the model and estimate the free lime content (%).
[0082] In Comparative Example 1, the peak attributed to Ca in the emission spectrum of each type of steelmaking slag (X, Y, Z) was used as an explanatory variable. Specifically, 15 intensity data points corresponding to the upper part of the Ca peak (393.4 nm) and the Ca peak were used as explanatory variables. In Comparative Example 1, the free lime content (%) was estimated in the same manner as in Example of the Invention 1, except for the explanatory variables used to generate the model and estimate the free lime content (%).
[0083] In Comparative Example 2, the estimated free lime content (%) for each type of steelmaking slag (X, Y, Z) was calculated using the alternative indicator (L value) described in Non-Patent Literature 2. Specifically, the pre-calculated values of the slag components (% by weight) of each type of steelmaking slag (X, Y, Z) were substituted into the equation for calculating L described in Non-Patent Literature 2 to obtain the L value («estimated free lime content»).
[0084] Results of Examples of the Invention 1 to 3, Example Comparative Example 1 and Comparative Example 2 are summarized in Table 2. Table 2 suggests that the measured f-CaO values (%) for all types of steelmaking slag (X, Y, Z) generally show better agreement with the estimated f-CaO values (%) in Examples of the Invention 1 to 3 than with the estimated f-CaO values (%) in Comparative Example 1 and Comparative Example 2. Petition 870250085420, dated 09 / 22 / 2025, page 34 / 52 29 / 34 In other words, the deviation between the measured values of f-CaO (%) and the estimated values of f-CaO (%) in Examples of the Invention 1 to 3 is smaller than the deviation between the measured values of f-CaO (%) and the estimated values of f-CaO (%) in Comparative Example 1 and Comparative Example 2. Regarding the standard deviation σd of the error that indicates precision, the standard deviations od of the error in Comparative Example 1 and Comparative Example 2 were, respectively: 0.92 and 0.62 for three types of steelmaking slag (X, Y, Z). The standard deviation od of the error in Example of Invention 1 was 0.38, the standard deviation od of the error in Example of Invention 2 was 0.45 and the standard deviation od of the error in Example of Invention 3 was 0.52. These results indicate that the free lime content (%) can be predicted with high accuracy according to this method. Table 2 Samples Measured fCaO Value (% by mass) Example of the Invention 1 Example of the Invention 2 Example of the Invention 3 Comparative Example 1 Comparative Example 2 Estimated fCaO Value (PLS) (% by mass) Estimated fCaO Value (PLS) (% by mass) Estimated f-CaO Value (PLS) (% by mass) Estimated fCaO Value (PLS) (% by mass) Estimated fCaO Value (% by mass) Wavelength range (270 to 410 nm) Wavelength range (400 to 540 nm) 15 intensity data points, peaks assigned to Si, Al, Ca, Fe and P 15 intensity data points, peaks assigned to Ca according to the literature X 6.60 5.76 5.69 6.02 5.56 4.23 Y 2.95 3.02 3.08 3.59 3.91 1.99 Z 2.50 2.28 1.79 2.81 3.34 1.35
[0085] According to this approach, by preparing the free lime content estimation model in advance, the free lime content (%) can be easily estimated based on the emission spectrum of the steelmaking slag without the need for complex analytical methods, such as the ethylene glycol extraction method. The lime content Petition 870250085420, dated 09 / 22 / 2025, page 35 / 52 30 / 34 free (%) can be estimated with high precision using partial least squares regression.
[0086] While the ethylene glycol extraction method took seven hours to determine the free lime content (%) of steelmaking slag, this method takes approximately six minutes to estimate the free lime content (%) of steelmaking slag. Therefore, the method for estimating the free lime content (%) of steelmaking slag according to the present invention significantly contributes to reducing the time required to estimate the free lime content (%). The free lime content (%) can be estimated quickly.
[0087] The Example (Example of Invention 4) corresponding to a method for managing the transport of steelmaking slag according to the present embodiment will be described below. In Example of Invention 4, the free lime content (%) of the steelmaking slag is estimated using the method for estimating the free lime content of steelmaking slag according to this embodiment, and the transport destination of the steelmaking slag is selected based on the estimated free lime content (%) of the steelmaking slag. The present invention is not limited to the Example below.
[0088] In Example of Invention 4, the emission spectrum obtained by irradiating 20 steelmaking slag operations in a slag vessel with a pulsed laser (100 mJ, 15 Hz) for one second and the accumulation of emission spectra were used as explanatory variables. The measured f-CaO value (%) of each steelmaking slag obtained by the ethylene glycol extraction method was used as the response variable. A model equation (free lime content estimation model) was generated based on the explanatory variables, the response variable, and Equation (1). Figure 3 illustrates an emission spectrum attributed to Ca, Fe, Si, Al, and P in the wavelength range of 270 to 410 nm that is used as variables Petition 870250085420, dated 09 / 22 / 2025, page 36 / 52 31 / 34 explanatory variables to generate the model equation and the emission spectrum intensity in Example of Invention 4. In Example of Invention 4, a total of 60 intensity data points, including the top of the peak of each peak shown in Figure 3 and two points immediately before and after each top of the peak (not shown in Figure 3), were used as explanatory variables to generate the model equation. Five measurements were taken for each steelmaking slag. OriginPro2017 (registered trademark) was used to perform the partial least squares regression analysis to generate the model equation.
[0089] Next, the emission spectrum obtained from the steelmaking slag (Chapters Nos. 1 to 7) subjected to the transport management method, i.e., the steelmaking slag whose free lime content (%) was to be estimated, was substituted as an explanatory variable in the free lime content estimation model described above. The free lime content (%) of each steelmaking slag (Chapters Nos. 1 to 7) was adequately estimated (estimated value of f-CaO). However, the emission spectrum corresponding to the same wavelengths as the explanatory variables used to generate the model equation was used as explanatory variables. The estimated value of f-CaO (%) obtained as described above was compared with the measured free lime content (measured value of f-CaO) (%). To measure the free lime content (%), samples for use in measuring the free lime content (%) were first collected from steelmaking slag (Chapters Nos. 1 to 7) before the steelmaking slag (Chapters Nos.Samples from Chapters 1 to 7 were spread on a slag yard. The free lime content (%) in each of the steelmaking slag samples (Chapters Nos. 1 to 7) was measured using the ethylene glycol extraction method (measured value of f-CaO). The estimated value of f-CaO (%) was then compared with the measured value of f-CaO (%). OriginPro2017 (registered trademark) was used to perform the analysis. Petition 870250085420, dated 09 / 22 / 2025, page 37 / 52 32 / 34 partial least squares regression.
[0090] Steelmaking slag (Chapters Nos. 1 to 7) is steelmaking slag received in a slag container for transport within the steel mill after decarburization treatment in a 300-ton capacity converter-type refining furnace. The slag container is integrated into a wagon and can be towed by a diesel-powered vehicle to move along the rails between the steel mill and the slag yard. During the stop before the slag container was coupled to the diesel-powered towing vehicle, the emission spectrum of the steelmaking slag (Chapters Nos. 1 to 7) received in the slag container was obtained. Specifically, the surface of the steelmaking slag discharged into the slag bath after decarburization blasting was irradiated with a pulsed laser (100 mJ, 15 Hz). The emission spectra acquired over one second were accumulated and the resulting emission spectrum was used as the emission spectrum of the steelmaking slag (Chapters Nos. 1 to 7).
[0091] Subsequently, if the estimated f-CaO (%) value of the steelmaking slag was equal to or less than a predetermined limit based on the estimated free lime content (estimated f-CaO) (%), the steelmaking slag was determined to be low-free lime slag. The low-free lime slag was then discharged into a slag yard 1 constructed along the tracks. If the estimated f-CaO (%) value of the steelmaking slag exceeded a predetermined limit, the steelmaking slag was determined to be high-free lime slag. The high-free lime slag was then discharged into a slag yard 2 constructed along the tracks. When the estimated f-CaO (%) value of the steelmaking slag was 5.20, the final product, iron and steel slag for road construction, showed deterioration of expansion properties. Therefore, the predetermined limit was set at 5.20. Petition 870250085420, dated 09 / 22 / 2025, page 38 / 52 33 / 34
[0092] The results of Example of Invention 4 are summarized in Table 3. The estimated f-CaO (%) value of each steelmaking slag (Chapters Nos. 1 to 7) in Example of Invention 4 generally exhibits good agreement with the measured f-CaO (%) value. The standard deviation od of the precision-indicating error was 0.96 for seven steelmaking slag charges (Chapters Nos. 1 to 7). Table 3 Charge No. Example of the Invention 4 Classification Measured f-CaO Value (% by mass) Estimated f-CaO Value (% by mass) Slag Yard 1 7.88 8.38 2 2 5.60 7.10 2 3 3.81 4.91 1 4 5.37 5.42 2 5 1.90 1.20 1 6 0.83 0.63 1 7 4.02 2.52 1
[0093] In Example of Invention 4, the estimated f-CaO (%) value was obtained on-site in a simple, quick and highly accurate manner, as described above. Steelmaking slag can be managed by transporting steelmaking slag with an estimated f-CaO (%) value equal to or less than the limit to slag yard 1 and steelmaking slag with an estimated f-CaO (%) value greater than the limit to slag yard 2.
[0094] These results indicate that, according to this method, the free lime content (%) of steelmaking slag can be estimated online in a simple, fast and highly accurate way, and steelmaking slag can be managed according to its free lime content (%).
[0095] The simple, quick and highly accurate estimation of the free lime content (%) in steelmaking slag also allows the establishment of an appropriate aging treatment time for the slag. Petition 870250085420, dated 09 / 22 / 2025, page 39 / 52 34 / 34 steel slag for each load. This can prevent or reduce insufficient aging of the steel slag and can efficiently supply iron and steel slag for road construction with stable quality (expansibility). The ideal destination for transport can also be selected according to the free lime content (%) of the steel slag, allowing for efficient transport management. Petition 870250085420, dated 09 / 22 / 2025, pages 40 / 52
Claims
1 / 3 CLAIMS 1. A method for estimating the free lime content of steelmaking slag, characterized in that it comprises: a step of acquiring an emission spectrum by irradiating the steelmaking slag with a laser to convert a portion of the steelmaking slag surface into plasma and acquiring an emission spectrum from the plasma; and a step of estimating the free lime content by inserting the intensities of an emission spectrum, including peaks caused by Ca, Fe, and Si emission lines, within the emission spectrum acquired in the emission spectrum acquisition step into a free lime content estimation model to generate a free lime content of the laser-irradiated steelmaking slag.
2. A method for estimating the free lime content in steelmaking slag, according to claim 1, characterized in that the estimation step includes introducing the intensities of an emission spectrum, including peaks caused by Al and P emission lines, into the free lime content estimation model.
3. Method for estimating free lime content in steelmaking slag, according to claim 1, characterized in that the free lime content estimation model is constructed using partial least squares regression using the emission spectrum intensities, including peaks caused by the emission lines of Ca, Fe and Si, within the emission spectrum acquired in the emission spectrum acquisition step as explanatory variables and using the free lime content as a response variable.
4. Method for estimating free lime content in steelmaking slag, according to claim 3, characterized in that the free lime content estimation model uses the intensities of an emission spectrum, including peaks caused by emission lines of Al and P, as explanatory variables. Petition 870250085420, dated 09 / 22 / 2025, page 41 / 52 2 / 3 5. Method for estimating free lime content in steelmaking slag, according to claim 1, characterized in that the free lime content estimation model is a trained machine learning model that uses the emission spectrum intensities, including peaks caused by Ca, Fe, and Si emission lines, within the emission spectrum acquired in the emission spectrum acquisition step as input data and the free lime content as output data.
6. Method for estimating the free lime content of steelmaking slag, according to claim 5, characterized in that the free lime content estimation model uses the intensities of an emission spectrum, including peaks caused by Al and P emission lines, as input data.
7. Method for estimating the free lime content of steelmaking slag, according to claim 1, characterized in that the emission spectrum acquired in the emission spectrum acquisition step has wavelengths of 270 nm or more and 410 nm or less.
8. Method for manufacturing iron and steel slag for road construction, characterized in that it comprises: estimating the free lime content of steelmaking slag using the method for estimating a free lime content of steelmaking slag as defined in claim 1; and defining the aging treatment conditions for the steelmaking slag based on the estimated free lime content of the steelmaking slag.
9. Method for managing the transport of steelmaking slag, characterized in that it comprises: estimating a free lime content of the steelmaking slag using the method for estimating a free lime content of the steelmaking slag as defined in claim 1; and selecting a transport destination for the steelmaking slag based on the estimated free lime content of the steelmaking slag.
10. Method for generating a model for estimating the free lime content of steelmaking slag for use in estimating the free lime content of steelmaking slag, characterized in that it comprises: a first acquisition step of irradiating the steelmaking slag with a laser to convert a portion of the slag surface into plasma and acquiring an emission spectrum of the plasma; a second acquisition step to acquire an actual free lime content of the steelmaking slag whose emission spectrum was acquired in the first acquisition step;and a model generation step to acquire multiple datasets, each consisting of pairs of the emission spectrum acquired in the first acquisition step and the actual free lime content acquired in the second acquisition step, and, based on these datasets, generate a model to estimate the free lime content using the intensities of an emission spectrum, including peaks caused by emission lines in Ca, Fe, and Si, within the emission spectrum acquired in the first acquisition step as input and using the actual free lime content acquired in the second acquisition step as output.
11. Method for generating a model for estimating the free lime content of steelmaking slag, according to claim 10, characterized in that the model generation step uses the intensities of an emission spectrum, including peaks caused by Al and P emission lines, as input. Petition 870250085420, dated 09 / 22 / 2025, pp. 43 / 52