JP7708639B2 - Refractory defect evaluation method - Google Patents

Description

The present invention relates to a method for evaluating defects in refractories, and in particular to a method that can easily grasp the relative positional relationship of defects present inside a refractory and their changes.

Refractories, including firebricks, are widely used in steelmaking facilities and other facilities that require high-temperature processes because they can withstand high temperatures. For example, in steelworks, refractories are used as linings for blast furnaces and converters.

However, refractories used in such high-temperature environments have the problem of being damaged by the impacts they receive during operation. A typical form of damage to refractories is thermal spalling. In other words, repeated stress caused by temperature changes causes cracks to form and grow, resulting in peeling and destruction of the refractory.

This type of damage to refractories does not occur instantly, but rather develops over a period of time. Therefore, in order to suppress damage caused by spalling over this long-term span and improve the furnace life of refractory equipment, it is necessary to understand the behavior of cracks that occur and grow within the refractory. To do this, it is necessary to measure and evaluate information about cracks that exist in the refractory and the voids (pores) that can be the origin of these cracks. In the following explanation, cracks and voids (pores) are collectively referred to as defects.

Conventional methods for evaluating defects present inside refractories and other materials include, for example, the following:

Non-Patent Document 1 specifies a method for evaluating the apparent porosity of refractories using the dry weight, submerged weight, and saturated weight.

Non-Patent Document 2 specifies a method for evaluating the pore size distribution of fine ceramics using mercury intrusion porosimetry.

Non-Patent Document 3 proposes a method for evaluating defects inside refractories using an X-ray CT scanner.

Patent Document 1 proposes a technology for detecting defects in a ceramic substrate by irradiating the ceramic substrate with linearly polarized light and observing the light reflected or transmitted through the ceramic substrate via a polarizer.

Patent Document 2 also proposes a technology that uses ultrasound to measure the thickness of the gaps between bricks in a lining for a heat-resistant container that is made up of multiple refractory bricks.

JP 2015-190957 A JP 2007-192546 A

JIS R 2205 (Method of measurement of apparent porosity, water absorption rate and specific gravity of firebricks) JIS R 1655 (Method of testing pore size distribution of fine ceramics molded bodies by mercury intrusion method) Keisuke Yamada et al., Refractories, Refractories Technology Association, 2019, Vol. 71, No. 5, pp. 206-210 Ippei Obayashi et al., Materia, Japan Institute of Metals, 2019, Vol. 58, No. 1, pp. 17-22

While these conventional techniques can evaluate defects present inside refractories and other materials, they have the following problems:

For example, the methods of Non-Patent Documents 1 and 2 can evaluate the apparent porosity and pore size distribution, but because defects of various shapes are distributed in a complex manner inside an actual refractory, the characteristics of the refractory cannot be accurately evaluated based on the simple apparent porosity and pore size distribution alone. In order to evaluate the characteristics of the refractory, particularly how cracks develop and progress when a load is applied to the refractory, more detailed information is required about the location of each defect and the relative positional relationship of multiple defects. In addition, the above-mentioned methods require filling the inside of the pores with water or mercury during measurement, so they can measure pores that are connected to the surface of the refractory (open pores), but cannot measure pores that are not connected to the surface (closed pores).

On the other hand, it is thought that information about closed pores can be obtained by using methods that use polarized light or ultrasound, such as those described in Patent Documents 1 and 2. However, even with these methods, it is still difficult to accurately evaluate the detailed distribution of individual defects or the relative positional relationships between multiple defects.

In addition, as proposed in Non-Patent Document 3, detailed position information of internal defects can be obtained by using X-ray CT scanning. However, the transmission image data obtained by X-ray CT scanning is extremely complex, making its analysis difficult. In particular, to understand the behavior of cracks that occur and grow within refractories, it is necessary to grasp the occurrence and growth behavior of defects when a load is applied to the refractory. To do this, it is necessary to conduct an analysis that includes the relative positional relationships of multiple adjacent defects. However, the reality is that such advanced analytical methods have not been established using transmission image data obtained by methods such as X-ray CT scanning.

The present invention was made in consideration of the above-mentioned circumstances, and aims to provide a method for evaluating defects in refractories that can easily grasp the relative positional relationship of defects present inside the refractory and their changes.

The present invention was made based on the above findings, and has the following gist:

1. A method for evaluating defects in a refractory material, comprising:
A transmission image data acquisition step of acquiring transmission image data of the refractory material;
an analysis step of performing a topological analysis on the transmission image data acquired in the transmission image data acquisition step;
A method for evaluating defects in a refractory material, comprising:

2. A method for evaluating defects in a refractory material as described in 1 above, in which the topological analysis in the analysis step is performed using persistent homology.

3. The refractory defect evaluation method described in 1 or 2 above, in which the transmission image data acquisition step is performed using an X-ray CT scanning device.

The method of the present invention allows for non-destructive evaluation of defects that exist three-dimensionally inside refractories. By utilizing topological analysis, the method of the present invention allows for easy understanding of information regarding the size and distribution of defects, which can be quantified from transmission image data obtained by techniques such as X-ray CT scanning. This allows for the identification of defects that cause destruction, and for an accurate evaluation of the stability (safety) of the refractory.

FIG. 1 is a schematic diagram showing an analysis method using persistent homology. FIG. 1 is a schematic diagram showing an analysis method using persistent homology. FIG. 1 is a schematic diagram showing a method of applying persistent homology to defect evaluation of refractory materials. FIG. 2 is a diagram showing transmission image data of an alumina refractory material obtained by X-ray CT scanning. 1 is an example of a persistence diagram obtained by analyzing transmission image data of an alumina refractory material obtained by X-ray CT scanning using persistent homology. 1 shows a load-crosshead displacement curve when a compressive load is applied to a refractory sample in an example. FIG. 1 is a persistence diagram at each time point in an embodiment.

The present invention will be described in detail below. Note that the present invention is not limited to this embodiment.

The refractory defect evaluation method according to one embodiment of the present invention is a refractory defect evaluation method for evaluating defects present inside the refractory, and includes a transmission image data acquisition process for acquiring a transmission image of the refractory, and an analysis process for performing a topological analysis on the transmission image data acquired in the transmission image data acquisition process.

[Refractories]
The refractory is not particularly limited and may be any refractory. The refractory may be either a shaped refractory or an unshaped refractory. Examples of the refractory include, but are not limited to, MgO-C bricks, Al ₂ O ₃ -SiC-C bricks, high alumina bricks, chamotte bricks, wax stone bricks, silica bricks, castable refractories, plastic refractories, mortar, and heat insulating refractories. The castable refractory is a refractory that can be poured (castable). The plastic refractory is literally a plastic refractory.

[defect]
The method of the present invention evaluates defects present inside the refractory. Here, the term "defect" refers to structural defects (also called structural defects) present inside the refractory, and includes cracks and voids (open pores and closed pores).

[Transmission image data acquisition process]
In the method of the present invention, first, transmission image data of the refractory is acquired (transmission image data acquisition step). The method of acquiring the transmission image data is not particularly limited, and any method can be used as long as it can acquire transmission image data including information about defects inside the refractory. From the viewpoint of versatility, it is preferable to acquire the transmission image data in the transmission image data acquisition step using an X-ray CT scanning device. Transmission image data can also be acquired using synchrotron X-rays, which have superior brightness and directivity compared to X-ray tubes, which are general X-ray sources.

[Analysis process]
Next, a topological analysis is performed on the transmission image data acquired in the transmission image data acquisition process (analysis process). As described above, many fine defects exist inside the actual refractory, and the size and position of the defects are not uniform, but are distributed in a three-dimensionally complex manner. Therefore, the transmission image data obtained by a method such as X-ray CT scanning is very complex, and it is not easy to understand the distribution of defects and their influence on material properties as it is. Therefore, by extracting relatively easy-to-understand parameters from the information on the huge number of defects contained in the transmission image data by the topological analysis, it becomes possible to easily grasp the relative positional relationship of defects existing inside the refractory and their changes.

The following describes the case where persistent homology is used as the topological analysis, but the topological analysis in the present invention is not limited to this and any method can be used.

Persistent homology is an analytical method as explained in Non-Patent Document 4, but we will provide an overview of it below.

As an example, we will explain the case where a point cloud arrangement on a two-dimensional plane as shown in Figure 1 is analyzed using persistent homology. Here, each point represents the position of the object to be analyzed (in this case, a defect). Then, a circle is virtually set with each point as its center, and the radius of the circle is increased.

If we focus on the four points 1, 2, 3, and 4 surrounded by dashed lines among the points shown in Figure 1, the circles centered on each point behave as shown in Figure 2 as the radius is increased to r1, r2, r3, r4, and r5. In persistent homology, we focus on the creation and disappearance of holes that are formed between adjacent circles when the radius of the circle is increased in this way.

That is, when the radius is r1, the circles are not in contact (Fig. 2(a)), but when the radius is increased to r2, the four circles come into contact, forming a hole _1-2-3-4 between the circles on the point arrangement 1-2-3-4 (Fig. 2(b)). When the radius is further increased to r3, the circumferences of the upper and lower circles come into contact, causing the hole _1-2-3-4 to disappear and split into two holes _1-2-4 and _2-3-4 (Fig. 2(c)). The corresponding relationship between the creation and disappearance of this hole _1-2-3-4 is the pair of radii (r2, r3) of the virtual circles when the hole _1-2-3-4 is created and disappeared, and is called the birth-death-pear that represents the point arrangement 1-2-3-4. Among the elements that make up the birth-death-pear, the one that corresponds to the creation of the hole is called the birth value, and the one that corresponds to the disappearance of the hole is called the death value.

When the radius of the virtual circle is further increased and becomes r4, hole _2-3-4 of the two holes disappears (Fig. (2d)), and when the radius becomes r5, the remaining hole _1-2-4 also disappears (Fig. (2e)). (r3, r4) and (r3, r5), which correspond to point arrangements 2-3-4 and 1-2-4, respectively, are obtained as birth-death-pears representing each point cloud arrangement.

The birth value is half the maximum distance between adjacent points surrounding a hole, and the death value is the radius of a circle inscribed in at least two points in the point cloud. From this, it can be seen that the birth value and death value are quantified to represent the proximity distance between the point clouds surrounding a hole and the size of the hole surrounded by the point cloud, respectively. Furthermore, the larger the difference between the birth value and the death value, the more easily it can be identified as an element that indicates a characteristic point cloud arrangement that spreads over a wider space.

The birth-death-pear obtained in this way consolidates the structural information of the original point cloud data. A diagram in which the obtained birth-death-pear is plotted with the birth value on the horizontal axis and the death value on the vertical axis is called a persistence diagram (PD), and by using a persistence diagram, the structural information of multiple point cloud arrangements can be visually represented and quantitatively analyzed. Note that persistence diagrams have orders, with a zeroth order persistence diagram containing connected components of point clouds, a first order persistence diagram containing information on the formation of two-dimensional holes by point clouds, and a second order persistence diagram containing information on the point cloud arrangement related to the formation of three-dimensional cavities.

When persistent homology is actually applied to the analysis of defects in refractory materials, a first-order persistence diagram that takes into account the volume of the defect is used. First, for each defect present in the transmission image data acquired in the transmission image data acquisition process, a sphere having the same volume as the defect (defect volume equivalent sphere) is placed so that the center of the defect volume equivalent sphere coincides with the center of gravity of the defect (Figure 3(a)). Then, the radius of each defect volume equivalent sphere is increased as shown in Figures 3(b) to (d). The sphere with the increased radius of the defect volume equivalent sphere is hereinafter referred to as a virtual sphere.

As the radius is gradually increased, at a certain point, a two-dimensional hole _1-2-3 is generated that is surrounded by a virtual sphere with points 1, 2, and 3 as its center, as shown in Figure 3(c). Consider a plane that includes the triangle 1-2-3 at this point. On the plane, draw a tangent from the tangent or intersection of the virtual sphere cross section to the sphere cross section circle equivalent to the defect volume, and the length of that tangent is the birth value.

If the radius is further increased, at a certain point, the hole _1-2-3 disappears as shown in Fig. 3(d). Therefore, at this point, the length of the tangent drawn from the intersection of the imaginary spherical cross section to the spherical cross section circle equivalent to the defect volume is set as the death value.

Figures 4 and 5 show the results of applying the refractory defect evaluation method of the present invention to an alumina refractory. Specifically, a sample with a diameter of φ10 mm and a height of 10 mm was cut out from an alumina refractory with an apparent porosity of 14.8%. Next, an X-ray CT scanner was used to measure the internal structure of the sample, and three-dimensional transmission image data (Figure 4) was obtained. Next, a topological analysis was performed on the transmission image data using the persistent homology described above to analyze the void distribution. Figure 5 is an example of a persistence diagram obtained by the analysis.

In this way, persistent homology analysis is possible even for complex porous materials such as refractories.

In addition, even if the location of the refractory sample has changed, the identity of the material can be assessed by performing topological analyses such as persistent homology.

Furthermore, by analyzing the changes in the purge stent diagram when conditions change, such as when a mechanical load (bending load, compressive load, etc.) is applied to the refractory material, it is possible to track the changes that occur within the material.

For example, when a bending or compressive load is applied and cracks appear inside the refractory, new combinations of birth and death values appear in certain areas on the persistence diagram. Furthermore, when these cracks grow and merge with adjacent cracks, the combinations of birth and death values in certain areas on the persistence diagram decrease. In this way, changes in refractory materials can be interpreted from changes in the persistence diagram.

In addition, by performing reverse analysis and correlation analysis from the persistence diagram, it is possible to identify initial cracks and voids that are prone to crack initiation or sensitive to fracture.

In this way, the refractory defect evaluation method of the present invention makes it possible to non-destructively and visually evaluate how cracks develop when a mechanical load is applied to a refractory.

The following describes the effects and advantages of the present invention using examples. Note that the present invention is not limited to the following examples.

Using the defect evaluation method for refractories of the present invention, we evaluated how voids (cracks) are generated inside the refractory when a mechanical load is applied to the refractory sample. As the refractory sample, an alumina refractory with an apparent porosity of 14%, a diameter of 10 mm, and a height of 10 mm was used.

A compressive load was applied to the refractory sample in the following pattern.
(1) No load (initial state)
(2) A load of 4.3 kN (corresponding to 70% of the compressive strength) was applied. (3) Loading was removed. (4) A load of 5.0 kN (corresponding to 80% of the compressive strength) was applied. (5) Loading was removed.

Figure 6 shows the load-crosshead displacement curve when a compressive load is applied in the above pattern.

At each of the above points (1) to (5), an X-ray CT scanner was used to photograph the void distribution in the refractory sample to obtain transmission image data, and the void and crack distribution in the obtained transmission image data was analyzed using two-dimensional persistent homology.

Figure 7 shows two-dimensional persistence diagrams for each of the above (1) to (5). The above (1) corresponds to Figure 7(a), (2) to Figure 7(b), (3) to Figure 7(c), (4) to Figure 7(d), and (5) to Figure 7(e). In Figures 7(b) and (c), the number of points has increased in the range of death values of 400 to 600 (area indicated by dashed lines in the figure) compared to Figure 7(a). This shows that applying a load equivalent to 70% of the compressive strength caused new cracks of 400 μm to 600 μm in size, and that these cracks remained even after the load was removed.

However, looking at Figures 7(d) and (e), applying a load equivalent to 80% of the compressive strength and then unloading it reduces the number of points in the death value range of 400 to 600. This shows that applying a load equivalent to 80% of the compressive strength caused some of the newly generated cracks of 400 μm to 600 μm in size to coalesce.

In this way, according to the refractory defect evaluation method of the present invention, it is possible to easily grasp how defects occur and coalesce when a load is applied to the refractory.

Claims (3)

A method for evaluating defects in a refractory material, comprising:
The refractory material is a shaped refractory material or an unshaped refractory material,
a transmission image data acquisition step of acquiring transmission image data of the refractory both in a state where a mechanical load is applied to the refractory and in a state where a mechanical load is not applied to the refractory;
an analysis step of performing a topological analysis on the transmission image data acquired in the transmission image data acquisition step;
Including,
The method for evaluating defects in a refractory material, wherein the mechanical load is a bending load or a compressive load . The refractory defect evaluation method according to claim 1, in which the topological analysis in the analysis step is performed using persistent homology. The refractory defect evaluation method according to claim 1 or 2, wherein the transmission image data acquisition step is performed using an X-ray CT scanning device.