Methods for evaluating the distribution of substances

Abstract

This invention provides a method for evaluating inclusion distribution that can investigate the correlation between the melting conditions of a metallic material and the distribution morphology of inclusions. [Solution] An inclusion distribution evaluation method comprising: an incorporation step of mixing simulated inclusions consisting of particles of a substance different from the metal material into the molten metal material during the metal material melting process; a solidification step of solidifying the metal material into which the simulated inclusions have been mixed in the incorporation step to obtain a melted material; and a detection step of performing ultrasonic testing on the melted material obtained in the solidification step to detect the distribution of simulated inclusions in the melted material.

Description

[Technical Field]

[0001] The present invention relates to an inclusion distribution evaluation method, and more particularly, to a method for evaluating the distribution of inclusions when smelting a metal material. [Background Art]

[0002] When smelting a metal material such as steel, it is desired to obtain a metal material with high cleanliness by suppressing the mixing of inclusions such as metal oxides, or by preventing inclusions from aggregating and localizing. By suppressing the mixing and localization of inclusions, the mechanical properties of the metal material, including fatigue properties, can be improved. Ultrasonic flaw detection can be suitably used to evaluate the amount and distribution of inclusions in metal materials such as steel. Even in Patent Documents 1 to 3, inclusions are detected by ultrasonic flaw detection for the purpose of evaluating the cleanliness of steel materials and the like. The use of ultrasonic flaw detection enables high-precision evaluation of the cleanliness of metal materials. [Prior Art Documents] [Patent Documents]

[0003] [Patent Document 1] Japanese Unexamined Patent Application Publication No. 2004-093227 [Patent Document 2] Japanese Unexamined Patent Application Publication No. 2006-349698 [Patent Document 3] Japanese Unexamined Patent Application Publication No. 2021-060373 [Summary of the Invention] [Problems to be Solved by the Invention]

[0004] As described above, ultrasonic testing allows for highly accurate evaluation of the distribution of inclusions in metallic materials. However, the distribution of inclusions, including their concentration and localization, is influenced by various factors such as raw materials and melting methods. Therefore, simply evaluating the distribution of inclusions in melted metallic materials is insufficient to verify these factors and aim for the production of metallic materials with minimal inclusion influence. For example, it is difficult to separate the influence of raw materials from the influence of the melting method on the distribution of inclusions. Furthermore, multiple parameters contribute to the melting process, making it even more difficult to individually verify the influence of these parameters. In addition, in metallic materials melted using conventional methods, the occurrence of inclusions is infrequent and probabilistic, making it difficult to systematically analyze how various factors affect the distribution of inclusions. Systematic research has not been sufficiently conducted on the relationship between the factors causing inclusion formation in raw materials and melting methods, the distribution morphology of the resulting inclusions, and the relationship between the distribution morphology of inclusions and the properties of metallic materials such as fatigue characteristics. Therefore, it is beneficial to melt metallic materials under various conditions and investigate how these conditions affect the distribution morphology of inclusions, as this will provide basic information for such research.

[0005] The problem that this invention aims to solve is to provide an inclusion distribution evaluation method that can investigate the correlation between the melting conditions of a metallic material and the distribution morphology of inclusions. [Means for solving the problem]

[0006] To solve the above problems, the inclusion distribution evaluation method according to the present invention has the following configuration. [1] The method for evaluating the distribution of inclusions according to the present invention includes: an incorporation step of mixing simulated inclusions consisting of particles of a substance different from the metal material into the molten metal material during the melting process of the metal material; a solidification step of solidifying the metal material into which the simulated inclusions have been mixed in the incorporation step to obtain a molten material; and a detection step of performing ultrasonic testing on the molten material obtained in the solidification step to detect the distribution of the simulated inclusions in the molten material.

[0007] [2] In the embodiment of [1] above, a plurality of molten materials may be produced by changing at least one of the method of carrying out the melting process and the method of mixing in the simulated inclusions, and then the detection process may be carried out for each of the plurality of molten materials, and the distribution of the simulated inclusions obtained in the detection process for the plurality of molten materials may be compared with each other.

[0008] [3] In the embodiment of [1] or [2] above, ultrasonic testing may be performed on a cross-sectional sample perpendicular to the interface where the metal material solidifies in contact with the gas phase during the solidification process.

[0009] [4] In the embodiment of [3] above, the inclusion distribution evaluation method further includes an analysis step of quantitatively analyzing the distribution of the simulated inclusions detected in the detection step, wherein the analysis step involves taking the position of the center of the interface in the cross-sectional sample as the origin, measuring the distance from the origin for each of the simulated inclusions, calculating the skewness of the distance distribution, and evaluating that the simulated inclusions are dispersed as the skewness takes a small positive value.

[0010] [5] Alternatively, in the embodiment of [3] above, the inclusion distribution evaluation method further includes an analysis step of quantitatively analyzing the distribution of the simulated inclusions detected in the detection step, wherein the analysis step involves taking the position of the center of the interface in the cross-sectional sample as the origin, determining a position vector with respect to the origin for each of the simulated inclusions, calculating a correlation coefficient between the position vectors, and evaluating that the simulated inclusions are dispersed as close to zero as possible. [Effects of the Invention]

[0011] In the inclusion distribution evaluation method according to the present invention having the configuration described in [1] above, a separately prepared simulated inclusion is mixed into the molten metal material during the metal material melting process. Since a predetermined amount of simulated inclusion can be reliably mixed into the metal material, the influence of the low frequency and uncertainty of natural inclusion occurrence in metal materials can be suppressed, and the distribution morphology of inclusions in the metal material can be evaluated by ultrasonic flaw detection in the detection process. After setting various conditions such as various parameters when melting the metal material, the raw materials used, and the method and timing of mixing in the simulated inclusion, the metal material melting process including the mixing in of the simulated inclusion is carried out, and the distribution morphology of the simulated inclusion is evaluated on the obtained molten material, information on the distribution morphology of inclusions under those conditions can be obtained, and information on the correlation between the conditions for melting the metal material and the distribution morphology of inclusions can be obtained. Furthermore, by changing parameters such as composition, size, and shape of the simulated inclusion to be mixed in, information on the correlation between those parameters and the distribution morphology of inclusions can also be obtained.

[0012] In the embodiment described in [2] above, multiple molten materials are produced by changing at least one of the method of carrying out the melting process and the method of mixing in simulated inclusions, and the distribution of simulated inclusions is compared with each other. This comparison allows for a direct comparison of the distribution patterns of simulated inclusions for different forms of inclusion formation caused by differences in the method of carrying out the melting process and the method of mixing in simulated inclusions. Based on the results of this comparison, it is possible to effectively investigate melting conditions for metal materials that can suppress the influence of inclusions, for example, by evaluating the conditions corresponding to the molten material with the highest dispersibility of simulated inclusions among the multiple molten materials as preferable.

[0013] In the embodiment described in [3] above, ultrasonic testing is performed on a cross-sectional sample perpendicular to the interface during solidification. During the solidification of metallic materials, inclusions tend to exhibit characteristic distribution patterns perpendicular to the interface, such as floating up and agglomerating at the interface. Therefore, by performing ultrasonic testing on a cross-section perpendicular to the interface of the molten material and evaluating the distribution of simulated inclusions, it becomes easier to clearly recognize the correlation between the various conditions for the melting of metallic materials and the distribution patterns of inclusions.

[0014] In the embodiment described in [4] above, the origin is set at the center of the interface in the cross-sectional sample, and the skewness of the distance distribution from the origin for each simulated inclusion is calculated. The smaller the positive value of the skewness, the more dispersed the simulated inclusions are considered to be. When simulated inclusions are localized near the interface, they are concentrated in a small range of distances from the origin to each simulated inclusion, and the skewness takes a large positive value. Conversely, when simulated inclusions are highly dispersed and not localized, the distances from the origin to each simulated inclusion are distributed over a wide range of values, and the skewness takes a small value close to zero. Therefore, a small positive value for the skewness accurately reflects the high dispersion of the simulated inclusions. By using skewness, the dispersion of simulated inclusions can be quantitatively evaluated, which makes it easier to correlate the melting conditions of the metallic material with the distribution morphology of the inclusions.

[0015] In the embodiment described in [5] above, the origin is set at the center of the interface in the cross-sectional sample, and a position vector relative to the origin is determined for each simulated inclusion. The correlation coefficient between the position vectors is calculated, and the closer the correlation coefficient is to zero, the more dispersed the simulated inclusions are considered to be. The more dispersed the simulated inclusions are and the less localized they are, the more diverse the distribution of the position vectors of each simulated inclusion relative to the origin becomes, and the smaller the absolute value of the correlation coefficient becomes. Therefore, a correlation coefficient close to zero accurately reflects the high degree of dispersion of the simulated inclusions. By using the correlation coefficient, the dispersion of the simulated inclusions can be quantitatively evaluated, which makes it easier to correlate the melting conditions of the metal material with the distribution morphology of the inclusions. [Brief explanation of the drawing]

[0016] [Figure 1] It is a flow diagram explaining the inclusion distribution evaluation method according to an embodiment of the present invention. [Figure 2] It is a schematic diagram showing three sample preparation methods, wherein (a) shows the mode of mixing simulated inclusions by charging into a furnace, (b) by placement in a tundish, and (c) by placement in a mold, respectively. [Figure 3] Regarding the detection step, (a) shows the cross-section sampling method, and (b) shows the ultrasonic flaw detection method for the cross-section sample. [Figure 4] It is a diagram explaining the distance method and the vector method as examples of the analysis step. (a) shows the case where the dispersibility of simulated inclusions is low, and (b) shows the case where the dispersibility of simulated inclusions is high. [Figure 5] It shows distribution images of simulated inclusions obtained by ultrasonic flaw detection, wherein (a) is the case of charging into a furnace, (b) is the case of placement in a tundish, and (c) is the case of placement in a mold. In each case, the left side shows the distribution at the R / 2 cross-section, and the right side shows the distribution at the central cross-section. [Figure 6] It shows the relationship between the distribution image of simulated inclusions on the central cross-section and the distance distribution of the simulated inclusions, as well as the skewness values, wherein (a) is the case of charging into a furnace, (b) is the case of placement in a tundish, and (c) is the case of placement in a mold. [Figure 7] It shows the relationship between the distribution image of simulated inclusions on the central cross-section and the position vector distribution of the simulated inclusions, as well as the correlation coefficient values, wherein (a) is the case of charging into a furnace, (b) is the case of placement in a tundish, and (c) is the case of placement in a mold. DETAILED DESCRIPTION OF THE INVENTION

[0017] The following describes a method for evaluating the distribution of inclusions according to one embodiment of the present invention. As shown in Figure 1, the method for evaluating the distribution of inclusions according to this embodiment involves a melting process including a mixing process and a solidification process, and a detection process. Furthermore, although optional, an analysis process is performed after the detection process. In the method for evaluating inclusions according to this embodiment, a molten material mixed with simulated inclusions in the melting process is manufactured as a sample, and the distribution of simulated inclusions in the molten material is detected in the detection process. The detected distribution of simulated inclusions is then analyzed in detail in the analysis process. The obtained detection results of the distribution of simulated inclusions can be used as basic information regarding the distribution morphology of inclusions in the molten material. The following describes each process in order.

[0018] (1) Melting process including mixing and solidification steps In the melting process, a metal material is melted to produce a molten material as a sample. Any type of metal can be used as the metal material, but steel is preferably used. The melting process can be carried out in the same way as the normal melting of metal materials, but in the inclusion distribution evaluation method according to this embodiment, a mixing process is carried out in the middle of the melting process. In the mixing process, simulated inclusions are mixed into the molten metal material. Then, a solidification process is carried out to solidify the metal material into which the simulated inclusions were mixed in the mixing process, thereby obtaining the molten material.

[0019] The simulated inclusions added to the metal material during the mixing process consist of particles of a different substance from the metal material being melted. The simulated inclusions are composed of a substance that mimics the inclusions that occur in the melted material. Specifically, the simulated inclusions are preferably composed of particles of a substance having the same or similar composition and particle size as the inclusions that are expected to occur and be mixed into the metal material. When the metal material is steel, light metal oxides such as alumina (Al2O3), silica (SiO2), and magnesia (MgO) can be suitably used as constituent materials for the simulated inclusions. The particle size can be selected as appropriate, for example, a range of 1 μm to 500 μm can be exemplified. It is preferable that the simulated inclusion particles are classified to a predetermined particle size. The simulated inclusion particles may be added to the molten metal material in the form of the particles themselves, or, as shown in the examples, they may be added in a composite form with the main component metal of the metal material, such as a composite, from the viewpoint of improving handling. Even when a composite material is formed by combining simulated inclusions with a metal in this way, when the composite material is added to a molten metal material, the metal constituting the composite material melts, and the particles of the simulated inclusions are released into the molten metal. There is no particular requirement for the amount of simulated inclusions added to the metal material, but it is preferable that the concentration be higher than the concentration of inclusions naturally contained in a metal material in which no inclusions have been intentionally generated or added.

[0020] In the mixing process, there are no particular limitations on the specific method of mixing simulated inclusions into the molten metal material. As shown in Figure 2, when a metal material M is melted in an electric furnace 1, poured into a mold 3 via a tundish (TD) 2, and allowed to solidify in the mold 3, three mixing methods can be exemplified. In the furnace infusion method shown in Figure 2(a), simulated inclusions (composites) I are added to the metal material M molten in the electric furnace 1. In this case, the simulated inclusions I accurately simulate inclusions originating from the raw material or inclusions that are mixed in from the components of the electric furnace 1. In the TD infusion method shown in Figure 2(b), the simulated inclusions I are placed inside the tundish 2 in advance, and then the metal material M is poured into the tundish 2 from the electric furnace 1. In this case, the simulated inclusions I accurately simulate inclusions that are mixed in from the components of the tundish 2. In the mold arrangement shown in Figure 2(c), a simulated inclusion I is placed inside the mold 3 beforehand, and then the metal material M is poured into the mold 3 from the tundish 2. In this case, the simulated inclusion I accurately simulates an inclusion that would be mixed in from the constituent materials of the mold 3. In either case, as the metal material M cools within the mold 3, the metal material M containing the simulated inclusion I solidifies, and the solidification process is carried out.

[0021] The sample produced as a molten material through a melting process that includes an incorporation process and a solidification process may be of one type or multiple types. When multiple types are produced, it is preferable to produce multiple samples by changing at least one of the methods of carrying out the melting process and the method of incorporating the simulated inclusions. Changes in the method of carrying out the melting process include the selection of various parameters such as temperature, required time, and operating speed at each stage of melting, the shape and size of the components used such as the mold, and the raw materials used. Changes in the method of incorporating the simulated inclusions include the selection of elements such as which method to use—furnace loading, TD placement, or mold placement—as described above, and how and at what stage of the melting process the simulated inclusions are incorporated into the metal material. Furthermore, the composition, particle size, shape, and other compositional and quantifiable aspects of the simulated inclusions can also be changed. In this way, when multiple samples are formed, the subsequent detection and analysis processes can be carried out for each of these samples, and the obtained results can be compared with each other. When preparing multiple samples for comparison, and comparing the effects of the melting process and the method of introducing simulated inclusions, it is preferable to keep the composition and quantity of the metal material, as well as the composition and quantity of the simulated inclusions, consistent.

[0022] (2) Detection process In the detection step, ultrasonic testing is performed on the sample obtained as molten material in the melting step to detect the distribution of simulated inclusions in the sample. The shape of the sample to be subjected to ultrasonic testing is not particularly limited, but in order to analyze the distribution of simulated inclusions inside the sample in detail, it is preferable to perform ultrasonic testing on a cross-sectional sample obtained by cutting the sample. In particular, it is preferable to prepare a cross-sectional sample by cutting the molten material of the sample on a plane perpendicular to the interface. Here, the interface of the molten material refers to the interface where the metal material comes into contact with the gas phase and solidifies in the solidification step, that is, the surface of the molten material that constituted the gas-liquid interface. As shown in Figure 2, when molten metal is poured into a mold from above and solidified, the upper surface of the molten metal becomes the interface, and when continuous casting is performed from above downwards, the lower end face becomes the interface. In this specification, the direction perpendicular to these interfaces is referred to as the up and down direction, and the direction toward the interface is referred to as the up direction. Furthermore, the aggregation of (simulated) inclusions in the up direction is referred to as "floating". However, these vertical directions do not necessarily correspond to the direction of gravity, and as mentioned above, they are based on the interface.

[0023] Figure 3(a) shows an example of a method for preparing a cross-sectional sample when using a frustoconical mold as shown in Figure 2. The molten material of the sample should be cut along the direction perpendicular to the interface Sa (up and down direction in the figure), which corresponds to the central axis direction of the frustoconical mold. Here, two types of cross-sectional samples are shown: a central cross-section, S1, which passes through the center of the circular interface Sa, and an R / 2 cross-section, S2, which passes through a position half the radius of the circle. During the melting process, inclusions tend to float and aggregate at the interface of the metal material. Therefore, by inspecting cross-sections obtained by cutting the sample perpendicular to the interface, such as the central cross-section and the R / 2 cross-section, a wealth of information can be obtained regarding the distribution of simulated inclusions.

[0024] Figure 3(b) schematically shows an example of an ultrasonic flaw detection method in the detection process. Here, an ultrasonic probe P that transmits and receives ultrasonic waves is placed perpendicular to the surface of the cross-sectional sample S. Then, ultrasonic waves are transmitted and received to the sample surface (indicated by dashed arrows) while scanning the ultrasonic probe P along the surface of the cross-sectional sample S, and flaw detection is performed using the C-scan method. In the illustrated configuration, the ultrasonic probe P is scanned in a zigzag pattern (indicated by solid arrows). It is preferable to perform ultrasonic flaw detection with the cross-sectional sample S and ultrasonic probe P placed in water.

[0025] In a sample, at locations where simulated inclusions are present, the intensity of the reflected ultrasound detected by the ultrasonic probe increases due to reflection from these inclusions. Therefore, by mapping the intensity of the reflected signal at each measurement point while scanning with the ultrasonic probe, simulated inclusions can be detected as points with strong reflected signal intensity, and their distribution can be detected. This yields a two-dimensional image showing the spatial distribution of simulated inclusions. Furthermore, since the reflected intensity increases when the simulated inclusions are aggregated, information regarding the presence and degree of aggregation of the simulated inclusions can be obtained based on the detected reflected signal intensity. Alternatively, by adjusting the detection sensitivity of the ultrasonic probe, simulated inclusions aggregated to a size above a predetermined level can be selectively detected. In cases like these, where information on reflected intensity is used, it is preferable to calibrate the detection sensitivity separately from performing ultrasonic testing on the actual sample. For calibration, for example, a metal plate can be prepared as a calibration sample, with a hole having a bottom surface formed on the back surface at a position midway along its thickness. Ultrasonic testing can then be performed on the calibration sample from the front surface, and the reflection intensity from the bottom surface of the hole can be recorded. Since the amplitude of the ultrasonic reflection wave is proportional to the area, the proportional relationship between the bottom area of ​​the hole in the calibration sample and the intensity of the reflection wave detected in the actual sample can be correlated with the size of the simulated inclusion.

[0026] It is preferable to binarize the flaw detection image obtained by ultrasonic testing. Binarization clarifies the spatial distribution of simulated inclusions. The threshold for binarization should be set so as to distinguish between locations where simulated inclusions to be detected are present and locations where they are not. However, as mentioned above, by utilizing the fact that the amplitude of the reflected wave is proportional to the area of ​​the simulated inclusion, the threshold can also be set to select simulated inclusions that are larger than a certain size, such as those that have a significant impact on the properties of the molten material due to aggregation, etc.

[0027] The simulated inclusions introduced into the metal material during sample preparation mimic the inclusions that naturally occur and become mixed into that metal material, and there is a high probability that both will exhibit similar behavior within the metal material. Therefore, by examining the distribution pattern of the simulated inclusions in the sample, it is possible to gain insights into the distribution pattern of inclusions that actually occur and become mixed into the metal material. Thus, if the distribution of simulated inclusions in the sample can be detected by performing a detection process, this distribution can be correlated with the conditions for melting the metal material, and the correlation can be evaluated to investigate the behavior of inclusions in the actual melted material.

[0028] For example, if, in a sample prepared under certain conditions, the simulated inclusions are dispersed over a wide area within the sample, as shown in Figure 4(b1), then those conditions can be judged as excellent, as they are less likely to cause aggregation or localization of inclusions and produce a molten material with minimal inclusion influence. On the other hand, if, in a sample prepared under certain conditions, the simulated inclusions are concentrated in a specific area within the sample, typically near the center of the interface as shown in Figure (a1), then those conditions can be judged as more likely to cause aggregation or localization of inclusions and should be improved. Furthermore, as illustrated in the three forms in Figure 2, when multiple samples are prepared by changing conditions such as the method of the molten process or the method of mixing in the simulated inclusions, the detection process can be performed on each of these samples, and the resulting distribution of simulated inclusions can be compared with each other. Based on this comparison, the various conditions for manufacturing the molten material and the distribution of inclusions can be correlated, and insights into the correlation between the manufacturing conditions of the molten material and the distribution of inclusions can be obtained in an easily understandable form. For example, by adopting manufacturing conditions corresponding to the sample with the least aggregation and localization of simulated inclusions among several samples being compared, or manufacturing conditions close to those, it can be determined that the aggregation and localization of inclusions can be reduced in metallic materials.

[0029] (3) Analysis process As described above, if the distribution of simulated inclusions in the sample is obtained during the detection process, evaluation and consideration of the correlation between the melting conditions of the metal material and the distribution morphology of the inclusions may be performed based on the distribution image itself. However, by performing an analysis process to quantitatively analyze and quantify the distribution, it becomes easier to perform evaluation and consideration in a clear and systematic manner.

[0030] In the analysis process, the distribution of simulated inclusions detected in the detection process is quantitatively analyzed. Here, the degree of dispersion of the simulated inclusions is quantified. Two specific methods for quantification are described below: the distance method, which utilizes the skewness of the distance distribution of the simulated inclusions, and the vector method, which utilizes the correlation coefficient of the position vectors of the simulated inclusions.

[0031] Here, we consider two cases for the distribution of simulated inclusions in the cross-sectional sample obtained in the detection process: low dispersion of simulated inclusions (a1) and high dispersion of simulated inclusions (b1), as schematically shown in Figure 4. For both the distance method and the vector method, the distribution of simulated inclusions is analyzed with respect to the origin, as shown in Figures 4(a2) and (b2), respectively. The origin is set at the center of the upper end of the cross-sectional sample, i.e., at the center of the interface. In molten materials, inclusions tend to float to the interface and aggregate easily, so by setting the origin at the upper end of the cross-sectional sample in this way, the degree of localization accompanied by the floating of simulated inclusions can be evaluated with high accuracy.

[0032] When using the distance method, the distance from the origin is measured for each simulated inclusion particle in the distribution image. More specifically, the distance from the origin to the centroid of the simulated inclusion particle is measured. In the configuration shown in Figure 4, the length of the arrows extending from the origin to the simulated inclusion particle in Figures 4(a2) and (b2) indicates the distance. Next, the distance values ​​obtained for all particles are statistically processed. Specifically, as shown in Figures 4(a3) and (b3), the number of particles that take each distance is tallied. As shown in Figure 4(a3), when the dispersion of the simulated inclusion is low, a large number of data points are distributed in the range where the distance from the origin is small, and the statistical data shows a relatively sharp peak-like distribution in the region of small distance. In contrast, as shown in Figure 4(b3), when the dispersion of the simulated inclusion is high, many data points are distributed in the region where the distance from the origin is large, and the statistical data does not take a sharp peak-like distribution, but shows a smooth distribution.

[0033] Skewness is used to quantitatively analyze the differences in the distribution of these distances. Skewness (Sk) is defined by the following equation (1), and a large positive value indicates that the data points are concentrated in a region of small distances. On the other hand, a skewness value close to 0 indicates that the data points are widely distributed across various distances.

number

[0034] As shown in Figure 4(a), when the simulated inclusions float to the interface and are localized near the origin, the skewness of the distance distribution takes a large positive value. On the other hand, as shown in Figure 4(b), when the simulated inclusions do not float and are dispersed over a wide area, the skewness takes a value close to zero. Therefore, the smaller the positive value of the skewness, the more the simulated inclusions are dispersed and not localized at the interface, meaning that the dispersion of the simulated inclusions is high. Thus, the skewness of the distance distribution of simulated inclusions serves as a quantitative indicator of the degree of dispersion of the simulated inclusions. Note that when the skewness takes a small negative value (a large negative absolute value), it indicates that the simulated inclusions have settled and are localized downwards, but downward localization of simulated inclusions does not practically occur in actual molten materials.

[0035] When using the vector method, a position vector is determined for each simulated inclusion particle in the distribution image, relative to the origin. More specifically, the position vector of the centroid of the simulated inclusion particle is determined relative to the origin. In the configuration shown in Figure 4, the arrows extending from the origin to the simulated inclusion particles in Figures 4(a2) and (b2) correspond to these position vectors. Next, the position vectors obtained for all particles are statistically processed. Figures 4(a4) and (b4) show plots of the x-component (horizontal component of the distribution image) and y-component (vertical component of the distribution image) of the position vector of each particle. As shown in Figure 4(a4), when the dispersion of the simulated inclusion is low, the data points are densely clustered, and there are many data points where the x-component and / or y-component take close values ​​to each other. In contrast, as shown in Figure 4(b4), when the dispersion of the simulated inclusion is high, the data points are sparsely dispersed, and the x-component and y-component take values ​​over a wide range.

[0036] To quantitatively analyze the differences in the distribution of these position vectors, we use the correlation coefficient. The correlation coefficient of position vectors can be calculated as the correlation coefficient of data points in plots of x and y components, such as in Figures 4(a4) and (b4). The correlation coefficient (r) can be calculated using the following equation (2), and a larger absolute value indicates that the variability of the data points is small and the correlation between position vectors is high. On the other hand, a smaller absolute value of the correlation coefficient indicates that the data points are scattered and the correlation between position vectors is low.

number

[0037] As shown in Figure 4(a), when the simulated inclusions are localized in certain areas, such as near the interface of the sample cross-section, the correlation coefficient of the position vectors takes a large absolute value far from zero. On the other hand, as shown in Figure 4(b), when the simulated inclusions are not localized due to leachation or the like, and are dispersed over a wide area, the correlation coefficient takes a small absolute value close to zero. Therefore, the closer the correlation coefficient is to zero, the more the simulated inclusions are dispersed over a wide area without being localized by leachation at the interface, meaning that the dispersion of the simulated inclusions is high. Thus, the correlation coefficient of the position vectors of the simulated inclusions can also be used as a quantitative indicator of the degree of dispersion of the simulated inclusions.

[0038] (4) Use of information on the distribution of simulated inclusions As described above, in the inclusion distribution evaluation method according to this embodiment, the relationship between the metal material melting process and the inclusion dispersion can be evaluated based on the distribution image of simulated inclusions obtained in the detection process, and further based on quantitative information regarding the degree of dispersion of the simulated inclusions obtained by performing an analysis process. Simulated inclusions are modeled after inclusions that naturally occur and are mixed into metal materials, and insights into how natural inclusions are distributed in metal materials can be obtained based on the distribution of simulated inclusions. Since natural inclusions occur infrequently, and their presence and degree are affected by probability, it is difficult to systematically investigate the distribution of natural inclusions themselves. However, by adding a predetermined amount of simulated inclusions and performing a melting process, and investigating the distribution of the simulated inclusions by ultrasonic testing, the behavior of the inclusions can be reliably analyzed, and systematic information can be accumulated.

[0039] Furthermore, by preparing samples by specifying various parameters such as temperature, time, and operating speed at each stage of the melting process, as well as various conditions involved in the melting process, such as the shape and size of the components used (including molds) and the raw materials used, and by specifying the timing and method of adding simulated inclusions, and then investigating the distribution of the simulated inclusions, information on the correlation between these conditions and the distribution of the simulated inclusions can be obtained. In particular, by changing some of these conditions and comparing the distributions of the simulated inclusions, it is possible to understand how the changes in conditions affect the distribution of the inclusions. In addition, by changing the composition, size, and shape of the added simulated inclusions and comparing their distributions, insights into the differences in the behavior of inclusions with different compositions can be obtained.

[0040] By analyzing the distribution of simulated inclusions, information about the distribution of inclusions can be obtained, which can then be used to consider the causes of inclusion formation. For example, it is possible to examine whether the main cause of inclusion formation lies in the raw material or in the melting process. Furthermore, the above information can be used to consider melting conditions that minimize the influence of inclusions. Inclusions in metallic materials can degrade the mechanical properties of the metallic material, including fatigue properties. However, if the amount of inclusions is small and they are not localized in a specific area, the impact on mechanical properties will not be significant. On the other hand, if inclusions are localized in a specific area, the impact on mechanical properties may become significant. Therefore, in order to minimize the influence of inclusions, the melting conditions should be set so that the simulated inclusions are dispersed over a wide area, as shown in Figure 4(b), rather than being localized near the interface, as shown in Figure 4(a). [Examples]

[0041] Examples of the present invention are shown below. In these examples, the inclusion distribution evaluation method described above was actually carried out, and the distribution of simulated inclusions was compared for samples with different methods of simulated inclusion contamination. However, the present invention is not limited to these examples.

[0042] [Method for preparing the sample] (1) Preparation of simulated inclusions As a simulated inclusion, alumina (Al2O3) powder classified to a particle size of 100 μm or less was prepared. This alumina powder and iron powder were mixed in a volume ratio of 10:30, and then cold-rolled into a cylindrical shape (φ25 mm × H40 mm) at a molding pressure of 10 tons to produce a composite. The obtained composite was crushed and classified to a particle size of 100 μm or less to obtain composite powder.

[0043] (2) Melting and addition of simulated inclusions As a sample, a molten material equivalent to SAE9254 was prepared by casting. As shown in Figures 1 and 2, the sample was prepared by melting the metal material in an electric furnace, pouring it into a mold via a tundish, and allowing it to solidify. At this time, as shown in Figures 2(a) to (c), the composite powder of the simulated inclusions prepared above was mixed into the metal material using three different methods. Specifically, the simulated inclusions were mixed into the sample using three methods: (a) inlet injection, (b) placement in the tundish, and (c) placement in the mold. In all cases, the amount of composite powder mixed in was 180 g per 150 kg of metal material.

[0044] (3) Preparation of cross-sectional samples For each of the three obtained samples, two types of cross-sectional samples were prepared by cutting perpendicular to the interface, as shown in Figure 3(a): an R / 2 cross-section and a central cross-section. All cross-sectional samples were flat plates with a thickness of 10 mm.

[0045] [Method for evaluating samples] As part of the inspection process, ultrasonic testing was performed on each cross-sectional sample prepared above. Specifically, an ultrasonic probe (center frequency 15 MHz, focal length 20 mm) was placed 15 mm from the surface of the cross-sectional sample in water, and flaw detection measurements were performed by scanning along the sample surface in a zigzag pattern at 0.1 mm intervals, and the reflected intensity was mapped. At this time, the flaw detection sensitivity was calibrated so that simulated inclusions with a particle size of 150 μm or more could be selectively mapped, and the obtained detection images were binarized. The flaw detection sensitivity was calibrated by performing ultrasonic testing from the surface on a calibration sample, which was a flat-bottomed hole with a diameter of φ0.5 mm formed on the back side of a metal plate, and recording the intensity of the reflected wave from the bottom of the hole. This was done by utilizing the fact that the reflected wave amplitude is proportional to the area. Simulated inclusions with a particle size of 150 μm or more correspond to aggregates of multiple simulated inclusion particles.

[0046] Once a distribution image of simulated inclusions was obtained by ultrasonic testing, an analysis process was performed to quantitatively evaluate the degree of dispersion of the simulated inclusions in each cross-sectional sample. The evaluation was carried out using two methods: the distance method and the vector method, as described above.

[0047] [Evaluation Results] Figure 5 shows the distribution of simulated inclusions after binarization for three different methods of introducing simulated inclusions: (a) in-furnace injection, (b) placement within the TD (transfer tray), and (c) placement within the mold. In each figure, the left side shows the distribution in the R / 2 cross-section, and the right side shows the distribution in the central cross-section. The upper part of the distribution image corresponds to the interface during casting, i.e., the upper part of the mold.

[0048] As shown in Figure 5, in all of (a) to (c), the simulated inclusions observed as white dots are distributed more densely in the central cross-section than in the R / 2 cross-section, and even within the central cross-section, they are distributed more densely in a relatively upper position. From these findings, it can be confirmed that the simulated inclusions tend to float and aggregate at the interface during solidification, especially in the central part of the interface.

[0049] However, a detailed comparison of the distribution of simulated inclusions in the distribution images (a) to (c) reveals differences in the morphology of aggregation. First, in the case of (b) where the material is placed inside the tundish, compared to (a) and (c), it can be seen that the localization of simulated inclusions near the center of the interface is significantly greater in the central cross-section. In other words, the floating of simulated inclusions is remarkably pronounced. This is thought to be due to the fact that the floating of simulated inclusions has already occurred within the tundish. In both the case of (a) where the material is introduced into the furnace and the case of (c) where the material is placed inside the mold, the localization of simulated inclusions near the center of the interface is suppressed. However, in (a), the simulated inclusions are distributed in the central cross-section in a shape resembling a multi-stage open umbrella or a fishbone-like structure. This is thought to correspond to the mixing and stirring that occurs when the molten metal containing simulated inclusions is introduced into the mold from the electric furnace via the tundish. On the other hand, (c) is characterized by a greater distribution of simulated inclusions on the bottom side of the R / 2 cross-section compared to (a) and (b). This is thought to be because the simulated inclusions were already present at the bottom of the mold before the molten metal was poured in.

[0050] Figure 6 shows the results of distance-based analysis for each sample: (a) in-furnace placement, (b) TD placement, and (c) mold placement. For each case, the distribution image of simulated inclusions in the central cross-section (the same image as in Figure 5), a graph showing the distribution of distances from the origin set at the center of the interface to each simulated inclusion, and the skewness value in that distance distribution are shown. Looking at the distance distribution graph, in (b), there is a tendency for the distribution to be more heavily skewed towards the short-distance side (left side of the graph) than in (a) and (c). The skewness value also reflects this, with small values ​​of 0.2 or less in (a) and (c), while only in (b) it is a large value exceeding 0.4. This distance distribution and skewness value indicate that in (b), the simulated inclusions tend to float and localize near the center of the interface more strongly than in (a) and (c). In other words, the dispersion of simulated inclusions is higher in (a) and (c) compared to (b). This is consistent with the trend obtained from the visual comparison of the distribution patterns in Figure 5. Furthermore, it more clearly shows the differences in the degree of dispersion than the visual comparison.

[0051] Figure 7 shows the results of vector analysis for each sample: (a) placed in the furnace, (b) placed in the TD, and (c) placed in the mold. For each case, along with the distribution image of the simulated inclusions in the central cross-section (the same image as in Figure 5), a graph showing the position vectors of each simulated inclusion in terms of x and y components, relative to an origin set at the center of the interface, and the correlation coefficient values ​​of these position vectors are shown. Comparing the correlation coefficient values, (a) and (c) are below 0.1 in absolute value, while (b) is above 0.1 in absolute value. In other words, (b) is further from zero than (a) and (c). In the case of (b), it is thought that the distribution of many data points near the bottom center of the position vector distribution graph strongly contributes to the magnitude of the absolute value of the correlation coefficient. The data points near the bottom center of the graph are those of simulated inclusions distributed near the center of the interface. The correlation coefficient values ​​indicate that in (b), the simulated inclusions tend to float and localize near the center of the interface more strongly than in (a) and (c). In other words, (a) and (c) have higher dispersion of simulated inclusions compared to (b). This is consistent with the trend obtained from the visual comparison of the distribution patterns in Figure 5, and the results of the distance method evaluation described above. Furthermore, it more clearly shows the difference in the degree of dispersion than the visual comparison.

[0052] As described above, the high dispersibility of simulated inclusions, as indicated by the small positive value of skewness in the distance method and the proximity of the correlation coefficient to zero in the vector method, is better in (a) furnace incorporation and (c) mold incorporation than in (b) tundish incorporation. In particular, the dispersibility indicated by these values ​​is higher in (c) mold incorporation. From this comparison, it can be concluded that inclusions mixed into the metal material in the tundish during the melting process have a significant impact on the metal material after solidification, but the impact of inclusions mixed in the electric furnace or mold, especially in the mold, can be kept relatively small.

[0053] The embodiments of the present invention have been described above. The present invention is not particularly limited to these embodiments, and various modifications are possible. [Explanation of symbols]

[0054] 1 Electric furnace 2 Tan Dish (TD) 3. Mold I. Simulated Inclusions (Composites) M Metal material P Ultrasound probe S cross-sectional sample Sa interface S1 is a cross-sectional sample corresponding to the central cross-section. Cross-sectional sample corresponding to the S2 R / 2 section

Claims

1. During the melting process of a metal material, an incorporation step is taken in which simulated inclusions consisting of particles of a substance different from the metal material are mixed into the molten metal material. A solidification step is performed to solidify the metal material into which the simulated inclusions have been mixed in the mixing step in order to obtain a melted material, A method for evaluating the distribution of inclusions, comprising: a detection step of performing ultrasonic testing on the molten material obtained in the solidification step to detect the distribution of the simulated inclusions in the molten material.

2. Multiple molten materials are produced by changing at least one of the method of carrying out the melting process and the method of mixing in the simulated inclusions, and then the detection process is carried out for each of the multiple molten materials. The method for evaluating the distribution of inclusions according to claim 1, comprising comparing the distribution of the simulated inclusions obtained in the detection step with respect to the plurality of melted materials.

3. The method for evaluating the distribution of inclusions according to claim 1 or 2, wherein in the detection step, ultrasonic testing is performed on a cross-sectional sample perpendicular to the interface where the metal material solidified in contact with the gas phase during the solidification step.

4. The process further includes an analysis step for quantitatively analyzing the distribution of the simulated inclusions detected in the detection step, In the aforementioned analysis process, Using the central position of the interface in the cross-sectional sample as the origin, the distance from the origin to each of the simulated inclusions is measured, The method for evaluating the distribution of inclusions according to claim 3, comprising calculating the skewness of the distance distribution and evaluating that the simulated inclusions are more dispersed in distribution the smaller the positive value of the skewness.

5. The process further includes an analysis step for quantitatively analyzing the distribution of the simulated inclusions detected in the detection step, In the aforementioned analysis process, Using the central position of the interface in the cross-sectional sample as the origin, a position vector is determined for each of the simulated inclusions relative to the origin, The method for evaluating the distribution of inclusions according to claim 3, comprising calculating a correlation coefficient between the position vectors, and evaluating that the simulated inclusions are dispersed and distributed as the correlation coefficient approaches zero.

Images