Method for determining tapping condition of molten metal

Abstract

To provide a method for determining the tapping condition of a molten metal, capable of optimizing operation of a cooling roll device so as to obtain a metal strip having desired properties.SOLUTION: A metal melt flow S flowing out from a tapping port F1 is captured; a processing range 11 is set within the captured image; particle analysis processing is performed on an image 1a and / or its binary image 2a within the processing range 11 to calculate evaluation values for metal melt flow image parts 11a to 15a and 21a to 25a in the images; and based on the calculated evaluation values, the stability of the tapping condition is determined.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a method for determining the pouring state of molten metal, and more particularly to a method for determining the pouring state of molten metal that can be suitably used in a rapid cooling roll apparatus.

Background Art

[0002] A rapid cooling roll apparatus that efficiently produces thin strips of amorphous metal or magnetic metal by pouring molten metal from a nozzle onto the surface of a cooling roll and rapidly cooling it has attracted attention. For example, in Patent Document 1, in order to obtain a metal thin strip having a predetermined thickness, a molten metal supply apparatus is shown in which the tilt angle of a tilting melting furnace that supplies molten metal into a tundish is controlled so as to always maintain the molten metal surface height in the tundish constant.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, even if the molten metal surface height in the tundish is made constant, pouring of the molten metal onto the cooling roll may not be performed normally due to clogging of the pouring port or other reasons, and it may not be possible to obtain a metal thin strip having desired properties. Therefore, a method for reliably determining the pouring state has been desired.

[0005] Therefore, in view of such demands, an object of the present invention is to provide a method for determining the pouring state of molten metal that can optimize the operation of a cooling roll apparatus so as to obtain a metal thin strip having desired properties.

Means for Solving the Problems

[0006] In order to achieve the above object, in the present invention, the molten metal flow (S) flowing out from the tapping port (F1) is photographed, a processing range (11) is set in the photographed image, and the evaluation values of the molten metal flow image portions (11a to 15a, 21a to 25a) in these images are calculated by particle analysis processing for the image (1a) within the processing range (11) and / or its binarized image (2a), and the quality of the stable tapping state is determined based on the calculated evaluation values.

[0007] According to the present invention, compared with the conventional method of constantly controlling the molten metal surface height in the tundish, the quality of the tapping state is directly evaluated from the photographed image of the molten metal flow by particle analysis processing, so that the tapping state can be reliably determined. Based on this, the operation of the cooling roll device can be optimized to obtain a metal strip with desired properties.

[0008] Preferably, the evaluation value is at least one of the "height", "tapping width", "luminance change amount between the upper and lower parts", "average luminance", and "difference between the maximum or minimum luminance and the average luminance" of the tapping flow image portions (11a to 15a, 21a to 25a).

[0009] Preferably, when the determination of the stable tapping state is negative, at least one of the states of "discontinuous dripping" and "clogging" of the molten metal flow (S) is further determined.

[0010] Preferably, when there are a plurality of molten metal flows (S) and the determination of the stable tapping state is negative, the "confluence" state of the molten metal flows is further determined.

[0011] Preferably, for each of the corresponding photographed images (1a, 2a), the quality of the stable tapping state or the states of "discontinuous dripping", "clogging", and "confluence" are given as annotations to generate learning data for machine learning.

[0012] It should be noted that the reference numerals in the parentheses above are shown for reference in relation to the specific means described in the embodiments described later.

Effects of the Invention

[0013] As described above, according to the method for determining the molten metal pouring state of the present invention, the operation of the cooling roll device can be optimized to obtain a metal strip with desired properties.

Brief Description of the Drawings

[0014]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Embodiments for Carrying Out the Invention

[0015] In FIG. 1, the rapid cooling roll device includes an induction heating melting furnace F and a cooling roll R located below it. From the tapping port F1 provided at the bottom of the melting furnace F, a metal molten metal flow S flows down onto the outer peripheral surface of the rotating cooling roll R and hits the roll outer peripheral surface, where it is rapidly cooled to become a metal strip T and is sent out laterally. A camera 1 is provided facing the metal molten metal flow S, and the area including the metal molten metal flow S is photographed, and the photographed image 1a is sent from the camera 1 to a processing device 2 incorporating a computer.

[0016] Figure 2 shows an example of the captured image 1a captured into the processing device 2. The captured image 1a shows a state where five metal hot water flow S1 to S5 flow down from a plurality (five locations in this embodiment) of hot water outlets F1 provided at the bottom of the melting furnace F located above, and collide with the lower cooling roll R respectively and change directions. Since the captured image 1a is captured from an oblique side rather than the front, accordingly, the rectangular line L defining the range (processing range) 11 of the captured image 1a to be processed gradually shortens in the height direction from the right end to the left end.

[0017] In the processing device 2, in addition to the image within the processing range 11 of the captured image 1a, a binarized image 2a (Figure 3) obtained by binarizing the inside of the processing range 11 by the discriminant analysis method is generated, and the determination process described below is performed by appropriately using the image 1a within the processing range 11 and the binarized image 2a. Note that the capture image 1a is captured into the processing device 2 at regular intervals with time stamps during the operation time.

[0018] A. Calculation by particle analysis Process an appropriate image among the captured image 1a or the binarized image 2a by the particle analysis method, and calculate the values of "height", "hot water width", "amount of change in luminance up and down" (luminance change amount between the upper and lower parts), "average luminance", and "difference from the average luminance" (difference between the maximum or minimum luminance and the average luminance) for each of the hot water flow image portions 11a to 15a, 21a to 25a. Table 1 shows an example of each of the above calculated values for the five hot water flow image portions 11a to 15a, 21a to 25a.

[0019]

Table 1

[0020] B. Determination of stable hot water outflow state In the processing device 2, the following determination is made based on the values of each of the hot water flow image portions 11a to 15a, 21a to 25a calculated by the particle analysis method, and when all of these determinations are "good", it is determined that it is in a "stable hot water outflow state" (steps 101 to 108 in Figure 4).

[0021] (The hot water is connected vertically: Step 101) As shown in FIG. 5, when the height of the hot water outflow image portion 22a is the same as the height of the processing range 11, it is determined as "good" because the hot water is vertically connected. When the height of the hot water outflow image portion 21a is lower (shorter) than the height of the processing range 11, it is determined as "bad".

[0022] (The hot water width is less than the threshold value: Step 102) As shown in FIG. 6, when the adjacent hot water outflow image portions 21a and 22a merge or water splashing 211 occurs, the hot water width exceeds the threshold value and becomes large. Therefore, when the hot water width is less than the predetermined threshold value, it is determined as "good", and when it is greater than or equal to the predetermined threshold value, it is determined as "bad".

[0023] (The amount of change in brightness up and down is less than the threshold value: Step 103) The amount of change in brightness up and down is the difference between the average brightness values of the upper part (20% of the total length) and the lower part (20% of the total length) of one hot water outflow image portion. As shown in FIG. 7, when the brightness value of the lower part of the hot water outflow image portion 11a is lower (darker) than the brightness value of the upper part by more than a predetermined threshold value, it is determined as "bad", and if it is less than the predetermined threshold value, it is determined as "good".

[0024] (The average brightness is greater than or equal to the threshold value: Step 104) As shown in FIG. 8, when the average brightness of the hot water outflow image portions 11a to 15a is less than the predetermined threshold value, it is considered that a clear captured image 1a has not been obtained due to the gas generated in front of the camera 1, so it is determined as "bad". When the average brightness of the hot water outflow image portions 11a to 15a is greater than or equal to the predetermined threshold value, it is determined as "good".

[0025] (The difference from the overall average brightness is less than the threshold value: Step 105) Calculate the overall average brightness of all the hot water outflow image portions 11a to 15a. When the difference between the average brightness of each hot water outflow image portion 11a to 15a and the overall average brightness is less than the predetermined threshold value (12a, 13a in FIG. 9), it is determined as "good". When it is lower than the threshold value (11a in FIG. 9), the flow of the hot water is considered bad and it is determined as "bad".

[0026] (Vertical division (hot water width less than the threshold value): Step 106) As shown in Fig. 10, a horizontal line L1 is further drawn at approximately the center of the rectangular line L that defines the processing range 11 to partition the processing range 11 vertically. If the difference between the upper water width and the lower water width of the hot water flow image portions 21a to 25a is less than the threshold value, it is considered [good]. If the above difference is greater than or equal to the threshold value, it is considered [bad] as there is a possibility that the hot water flows are merging.

[0027] (Vertical partitioning (same number): Step 107) Similarly, as shown in Fig. 10, if the number of upper hot water flow image portions 21a to 25a is the same as the number of lower hot water flow image portions 21a to 25a in the vertically partitioned processing range 11, it is considered "good". If they are different, it is considered "bad" as there is a possibility that the hot water flows are merging.

[0028] C. Determination of the form of the hot water flow In addition to determining whether it is in a stable hot water output state, when it is not in a stable hot water output state, the following determination of the form of the hot water flow can also be performed.

[0029] (Discontinuous dripping) When the amount of change in the above brightness up and down exceeds the threshold value or the difference from the overall average brightness is greater than or equal to the threshold value, it is determined that the state of the hot water flow is in the "discontinuous dripping" state. Here, "discontinuous dripping" means a phenomenon where the hot water flow drops in drops.

[0030] (Clogging) When the number of hot water flows is less than a predetermined number (5 in this embodiment), it is determined that it is in the "clogging" state.

[0031] (Merging (upper)) In the above vertical partitioning, when the number of upper hot water flow image portions is less than the number of lower hot water flow image portions or the water width of the upper hot water flow image portion is greater than or equal to the threshold value, it is determined that it is in the "merging (upper)" state.

[0032] (Merging (lower)) In the above vertical partitioning, when the number of lower hot water flow image portions is less than the number of upper hot water flow image portions or the water width of the lower hot water flow image portion is greater than or equal to the threshold value, it is determined that it is in the "merging (lower)" state.

[0033] (Confluence (vertical and horizontal)) In the above vertical division, when the number of hot water flow image parts on the upper side and the lower side is less than a predetermined number (5 in this embodiment) and the hot water width of any of the hot water flow image parts on the upper side and the lower side is equal to or greater than a predetermined threshold, it is determined to be in the "Confluence (vertical and horizontal)" state.

[0034] (Number is okay (with wide width)) When the number of hot water flow image parts is a predetermined number but there is one with a hot water width equal to or greater than a predetermined threshold, it is determined to be in the "Number is okay (with wide width)" state.

[0035] (Unclear) When the average luminance of each hot water flow image part is less than a predetermined threshold, it is determined to be in the "Unclear" state. Here, "Unclear" refers to a phenomenon where the image is dark and the data has low reliability.

[0036] The determination of whether it is in the above "stable hot water state" is given as an annotation to the corresponding captured image 1a, or when it is not in the stable hot water state, the determination of the above "form of hot water flow" is further given as an annotation to the corresponding captured image 1a, and these are used as learning data to train a machine learning device, so that the operation of the quenching roll device can be optimized.

[0037] In the above embodiment, the case where there are multiple hot water flows has been described, but it goes without saying that it is also applicable when there is one hot water flow. In the above embodiment, "height", "hot water width", "luminance change amount between the upper and lower parts", "average luminance", and "difference between the maximum or minimum luminance and the average luminance" are calculated as evaluation values, but it is not always necessary to calculate all of these. In this case, a binarized image is not necessarily required. Also, it is not necessary to determine all of the states of "discontinuous dripping", "clogging", and "confluence".

Explanation of reference numerals

[0038] 1… Camera, 1a… Captured image, 11a, 12a, 13, 14a, 15a… Metal molten metal flow image part, 2… Processing device, 2a… Binary image, 21a, 22a, 23a, 24a, 25a… Metal molten metal flow image part, F… Induction heating melting furnace, F1… Tap hole, R… Cooling roll, S… Metal molten metal flow.

Claims

1. Photograph the molten metal flow discharging from the molten metal discharge port, set a processing range in the photographed image, calculate an evaluation value of the molten metal flow image part in the image by particle analysis processing for the image within the processing range and / or its binarized image, and determine the quality of the stable molten metal discharge state based on the calculated evaluation value. A method for determining the molten metal discharge state, characterized by the above.

2. The evaluation value is at least one of the "height", "width of the molten metal flow", "luminance change amount between the upper and lower parts", "average luminance", and "difference between the maximum or minimum luminance and the average luminance" of the molten metal flow image part. The method for determining the molten metal discharge state according to Claim 1.

3. When the determination of the stable molten metal discharge state is negative, further determine at least one of the states of "discontinuous dripping" and "clogging" of the molten metal flow. The method for determining the molten metal discharge state according to Claim 1.

4. When there are a plurality of the molten metal flows and the determination of the stable molten metal discharge state is negative, further determine the "confluence" state of the molten metal flows. The method for determining the molten metal discharge state according to Claim 1.

5. Assign the quality of the stable molten metal discharge state or the states of "discontinuous dripping", "clogging", and "confluence" as annotations to the corresponding photographed images to generate learning data for machine learning. The method for determining the molten metal discharge state according to any one of Claims 1 to 4.

Images