Induction heating flaw detection method and induction heating flaw detection device

Abstract

To provide an induction heating flaw detection method and an induction heating flaw detection device capable of easily discriminating a flaw even if blackbody treatment is incomplete.SOLUTION: An induction heating flaw detection method includes an application step of applying a black body paint, a heating step of heating a surface of a conductive material by changing a relative position between the conductive material and an induction heating unit, an extraction step of extracting a high temperature portion by acquiring a surface temperature of the conductive material after the heating is finished by a radiation thermometer, a tracking step of tracking the high temperature portion, and a determination step of determining whether the high temperature portion is a flaw or a noise based on a temporal temperature change of the high temperature portion. The induction heating flaw detector includes an application unit that applies black body paint, an induction heating unit that heats a surface of a conductive material by changing a relative position with the conductive material, an extraction unit that extracts a high temperature portion by acquiring a surface temperature of the conductive material after heating by a radiation thermometer, a tracking unit that tracks the high temperature portion, and a determination unit that determines whether the high temperature portion is a flaw or a noise based on a temporal temperature change of the high temperature portion.SELECTED DRAWING: Figure 6

Description

[Technical Field]

[0001] The present invention relates to an induction heating flaw detection method and an induction heating flaw detection device, and more specifically to an induction heating flaw detection method that can easily identify defects even if blackbody treatment is incomplete, and an induction heating flaw detection device that can easily identify defects even if blackbody treatment is incomplete. [Background technology]

[0002] BACKGROUND ART Conventionally, there has been known a flaw detection method for detecting flaws present in an object by heating the surface of the object and measuring the surface temperature with a radiation thermometer or the like. As such a flaw detection method, the following methods have been proposed.

[0003] For example, Patent Document 1 discloses a flaw detection method in which an induced current is generated in a conductive steel billet by electromagnetic induction, the surface of the billet is heated, the surface temperature of the billet is measured, and flaws are detected from non-uniformity in the surface temperature.

[0004] The same document states: (a) a heating step in which the surface of the billet is heated by an induced current having a penetration depth less than the depth of the flaw; (b) a temperature measurement step of measuring the surface temperature of the steel billet with a radiation thermometer; (c) a calculation and determination step of calculating a local temperature change amount on the billet surface based on the surface temperature and determining whether the local temperature change amount exceeds a negative threshold value; It is described that by providing the above, pinholes can be detected.

[0005] Furthermore, the same document states: (d) It is also described that in the calculation and determination step, cracks and the like can be detected by determining whether or not the amount of local temperature change exceeds a positive threshold value.

[0006] Furthermore, Patent Document 2 discloses a flaw detection method that uses pulse phase thermography to detect defects located at deep positions.

[0007] The same document states: (a) a heating step of pulse-heating a surface of an object; (b) a temperature detection step of detecting the surface temperature of the object at a set sampling frequency by a temperature detection means; (c) a data processing step of performing a Fourier transform on the data showing the relationship between the elapsed time from heating and the surface temperature to convert it into data showing the relationship between frequency and phase; (d) a display step of displaying the phase value at the set inspection frequency as an image; It is described that by providing the above, defects inside an object can be visually detected.

[0008] Furthermore, the document clarifies the relationship between the inspection frequency and the depth at which detectable defects exist, and states that by lowering the inspection frequency, defects that exist at deeper positions in the object can be detected.

[0009] The configuration of Patent Document 1 is premised on the use of a conductive material that has been subjected to a complete blackbody treatment. However, with actual conductive materials, even if black paint is applied, the emissivity will be below 1. In other words, with actual conductive materials, the reflectivity will be above 0, and the influence of reflection from surrounding objects will no longer be negligible.

[0010] Furthermore, in actual conductive materials, the amount of radiation directed toward a radiation thermometer or the like may change due to the presence of dirt or surface shape (waviness, unevenness, etc.). For this reason, the configuration of Patent Document 1 lacks redundancy in terms of blackbody detection, and there are cases where stains or the like are erroneously determined to be defects, so further improvement is required.

[0011] Furthermore, the configuration of Patent Document 2 is based on static measurement. For this reason, it is necessary to temporarily stop the movement of the object to be inspected, which poses a problem of extremely poor throughput of the inspection itself.

[0012] Furthermore, the configuration of Patent Document 2 is designed to be applied to relatively small objects, such as those using lamp heating, and therefore the size of the object to be inspected is also limited. Furthermore, since the inspection frequency must be relatively high, there is also the problem that it takes a long time to detect defects and the like. [Prior art documents] [Patent documents]

[0013] [Patent Document 1] Patent No. 3391134 [Patent Document 2] Patent No. 5574261 Summary of the Invention [Problem to be solved by the invention]

[0014] The problem to be solved by the present invention is to provide an induction heating flaw detection method that can easily detect flaws even if the blackbody treatment is incomplete. Another problem to be solved by the present invention is to provide an induction heating flaw detector that can easily detect flaws even if the blackbody treatment is incomplete. [Means for solving the problem]

[0015] In order to solve the above problems, the induction heating flaw detection method according to the present invention comprises: a coating step of coating a surface of the conductive material with black paint; a heating step of heating the surface of the conductive material by changing the relative position of the conductive material after the black body paint is applied and the induction heating unit; an extraction step of obtaining the surface temperature of the conductive material after heating by the induction heating unit using a radiation thermometer and extracting a high-temperature portion; a tracking step of tracking the high temperature portion while acquiring the surface temperature; a determining step of determining whether the high temperature portion is a defect or noise based on the temperature change over time of the high temperature portion; Equipped with.

[0016] In order to solve the above problems, the induction heating flaw detector according to the present invention comprises: an application unit that applies black paint to the surface of the conductive material; an induction heating unit that heats the surface of the conductive material by changing the relative position of the conductive material after the black paint is applied; an extraction unit that acquires the surface temperature of the conductive material after heating by the induction heating unit using a radiation thermometer and extracts a high-temperature portion; a tracking unit that tracks the high temperature portion while acquiring the surface temperature; a determination unit that determines whether the high temperature portion is a defect or noise based on the temperature change over time of the high temperature portion; Equipped with. [Effects of the Invention]

[0017] [Induction heating flaw detection method] The induction heating flaw detection method according to the present invention includes a tracking step of tracking a high-temperature portion while acquiring a surface temperature. Therefore, by tracking the high temperature portion, it is possible to confirm the temperature change of the high temperature portion over time.

[0018] This makes it possible to easily determine whether the high temperature portion is a defect or noise based on the temperature change over time of the high temperature portion. In other words, even if a high temperature area is mistakenly recognized due to dirt, surface shape, etc., it can be determined to be noise by checking the temperature change over time, making it easy to distinguish between scratches and noise. As a result, even if the blackbody treatment is incomplete, defects can be easily identified.

[0019] [Induction heating flaw detection device] The induction heating flaw detector according to the present invention has a tracking unit that tracks the high-temperature portion while acquiring the surface temperature. Therefore, by tracking the high temperature portion, it is possible to confirm the temperature change of the high temperature portion over time.

[0020] This makes it possible to easily determine whether the high temperature portion is a defect or noise based on the temperature change over time of the high temperature portion. In other words, even if a high temperature area is mistakenly recognized due to dirt, surface shape, etc., it can be determined to be noise by checking the temperature change over time, making it easy to distinguish between scratches and noise. As a result, even if the blackbody treatment is incomplete, defects can be easily identified. [Brief explanation of the drawings]

[0021] [Figure 1] (a) is an explanatory diagram of induced current, (b) is an explanatory diagram of induced current concentration around the flaw, and (c) is a simulation result of temperature distribution around the flaw. [Figure 2] (a) is an explanatory diagram of external disturbance factors, etc., and (b) is a pixel map image diagram for each external disturbance, etc. [Figure 3] FIG. 10 is a diagram showing temperature changes over time in a high-temperature portion (flaw) and a high-temperature portion (noise). [Figure 4] FIG. 1 is an explanatory diagram of an induction heating flaw detector. [Figure 5] 10 is a flowchart of the induction heating flaw detector after heating is completed. [Figure 6] 10 is a flowchart of a determination unit. [Figure 7] FIG. 10 is a diagram showing temperature changes over time for magnetic steel material. DETAILED DESCRIPTION OF THE INVENTION

[0022] Hereinafter, embodiments of the present invention will be described in detail. The induction heating flaw detection method according to the present invention includes an application step, a heating step, an extraction step, a tracking step, and a determination step.

[0023] [1. Induction heating material flaw detection method] [1.1. Coating process] The coating step refers to a step of coating the surface of the conductive material with black paint.

[0024] 1.1.1. Conductive materials A conductive material is a material that has free electrons and can conduct electricity, and is a material whose surface can be easily heated by generating an induced current through electromagnetic induction, for example. The material, size, shape, etc. of the conductive material are not particularly limited, and the optimum material, size, shape, etc. can be selected appropriately depending on the purpose. For example, the conductive material may be magnetic steel, non-magnetic steel, copper alloy, nickel alloy, zinc alloy, aluminum alloy, magnesium alloy, pure aluminum, magnetic material, titanium, titanium alloy, or the like.

[0025] [1.1.2. Black paint] Black body paint is a material that, when applied to the surface of an object, reduces the object's reflectance and brings its emissivity closer to 1. The black body paint is not particularly limited, and an optimum one can be selected appropriately depending on the purpose.

[0026] When transmittance can be ignored, the emissivity is the value obtained by subtracting reflectance from 1. To reduce the influence of external disturbances, it is necessary to reduce the reflectance, so it is preferable that the emissivity of the conductive material after applying black paint is 0.9 or higher.

[0027] [1.1.2.1. Water] Here, when a radiation thermometer that uses wavelengths in the mid-infrared region (wavelength 3 to 5 μm) or far-infrared region (wavelength 8 to 14 μm) is used, water can be used as the black body paint. The water used as the black paint may contain a small amount of additives such as a surfactant to ensure wettability. Water is an ideal black paint because it is easy to apply and remove and has a low environmental impact.

[0028] [1.2. Heating process] The heating step refers to a step of heating the surface of the conductive material by changing the relative position between the conductive material and the induction heating unit after the black body paint has been applied.

[0029] [1.2.1. Induction heating section] The induction heating unit generates an induced current through electromagnetic induction in the conductive material, heating the surface of the conductive material. This unit is an induction coil connected to a high-frequency power supply (see Figure 1(a)). The shape, size, etc. of the induction heating part are not particularly limited, and the optimum one can be selected appropriately depending on the shape, size, etc. of the conductive material.

[0030] [1.2.2. Induced current] 1.2.2.1. Penetration Depth The penetration depth of the induced current from the induction heating unit (the depth at which the induced current becomes 1 / e of the surface) depends on the resistivity ρ of the conductive material, the frequency f of the high-frequency current from the high-frequency power supply, and the relative permeability μ of the conductive material. For example, if the conductive material is magnetic steel, the penetration depth will be shallow. If the flaw is a linear flaw (a defect that extends in the elongation direction due to rolling) and the penetration depth is shallower than the flaw depth, the induced current will concentrate around the flaw, as shown in Figure 1(b), causing the temperature of the flaw to be higher than the surrounding area.

[0031] On the other hand, if the penetration depth is deeper than the flaw depth, the induced current bypasses the flaw, making the flaw temperature lower than the surrounding temperature. However, the induced current concentrates at both ends of the flaw, making the temperature higher than the surrounding temperature. Therefore, flaws can be easily detected by optimizing the frequency of the high-frequency current depending on the conductive material and the flaw depth.

[0032] In addition, since the position of the flaw and the position of the high temperature area directly coincide, a shallow penetration depth is more convenient for analysis (see Figure 1(c)). Therefore, in the heating step, it is preferable to set the penetration depth of the induced current induced in the conductive material shallower than the flaw.

[0033] For example, when the flaw depth is 1 mm, it is preferable to adjust the frequency of the high frequency current according to the resistivity and relative permeability of the conductive material so that the penetration depth is less than 1 mm. The penetration depth is preferably 90% or less of the flaw depth, more preferably 50% or less, and even more preferably 10% or less of the flaw depth.

[0034] Relative Position Changing the relative positions of the conductive material and the induction heating unit may involve either displacing the conductive material relative to the stationary induction heating unit or displacing the induction heating unit relative to the stationary conductive material. For example, it may involve inserting a rod-shaped conductive material into a stationary induction coil or scanning the surface of a stationary plate-shaped conductive material with an induction coil. Furthermore, both the conductive material and the induction heating portion may be displaced relative to each other.

[0035] [1.3. Extraction process] The extraction step is a step of obtaining the surface temperature of the conductive material after heating by the induction heating unit using a radiation thermometer and extracting the high-temperature portion.

[0036] [1.3.1. Radiation thermometer] A radiation thermometer converts the radiant energy from the surface of an object into temperature. The radiation thermometer is not particularly limited, and an optimum one can be selected depending on the purpose. For example, the radiation thermometer may have an imaging element of a single element type, a one-dimensional array type, or a two-dimensional array type. To create a temperature map of the surface of an object, it is preferable to use a radiation thermometer that uses a two-dimensional array type image sensor, which can reduce the number of mechanical components.

[0037] The radiation thermometer may be one that uses wavelengths in the near-infrared region (0.7 to 2.5 μm), one that uses wavelengths in the mid-infrared region, or one that uses wavelengths in the far-infrared region.

[0038] [1.3.2. High temperature section] The extraction method for the high temperature part is not particularly limited, and the most suitable method can be selected appropriately depending on the purpose. For example, the temperature map of the conductive material surface obtained by a radiation thermometer may be subjected to image processing (for example, the pixel values ​​of high-temperature areas may be increased and the pixel values ​​of low-temperature areas may be decreased; hereinafter, an image subjected to such processing may be referred to as a "pixel map"), and an operator may visually extract the high-temperature areas from the background areas.

[0039] Furthermore, for example, AI inference can be used when extracting high-temperature areas from a pixel map. More specifically, by inputting pixel map data into a machine learning model that has undergone predetermined machine learning, it is possible to extract inferred high-temperature areas from the pixel map. Here, the machine learning model that can be used is YOLO, which separates and extracts the object from the background. Furthermore, by using AI inference, processing can be performed at high speed.

[0040] In a realistic environment, it is impossible to reduce the reflectance to 0, even for conductive materials coated with black paint. Therefore, as shown in Figure 2(a), the material is affected by external disturbances (reflections). A radiation thermometer converts the energy received from the surface of an object into a temperature without distinguishing between radiated and reflected energy. For this reason, as shown in Figure 2(b), the radiation thermometer may detect a part that is not actually hot as a high-temperature part.

[0041] As shown in Figure 1, defects are formed in high-temperature areas due to the concentration of induced current in the induction heating section, but as shown in Figure 2, high-temperature areas are not necessarily defects. In other words, defects always form high temperature areas, but conversely, high temperature areas do not always become defects.

[0042] [1.4. Tracking Process] The tracking step refers to a step of tracking the high temperature portion while acquiring the surface temperature. The method for tracking the high temperature portion is not particularly limited, and the most suitable method can be selected appropriately depending on the purpose.

[0043] For example, when a conductive material is moved relative to a stationary induction heating unit, the moving point of the high temperature portion can be easily predicted from the imaging frame rate of the radiation thermometer and the conveying speed of the conductive material. Therefore, even if the temperatures of the background and high temperature areas change, the high temperature areas can be easily tracked.

[0044] The method for extracting high temperature areas from each pixel map in the tracking step is not particularly limited, and the most suitable method can be selected appropriately depending on the purpose. For example, the extraction may be performed by visual inspection by an operator as described above, or may be performed using image processing, particularly AI inference.

[0045] [1.5. Judgment process] The determination step is a step of determining whether the high temperature portion is a defect or noise based on the temperature change of the high temperature portion over time.

[0046] 1.5.1. High temperature section Here, the temperature of the high temperature section is not particularly limited, and may be the maximum temperature of the high temperature section, the average temperature of the high temperature section, or the temperature of any one section of the temperature map of the high temperature section. When the temperature of the high temperature section is taken as the temperature of any one section in the temperature map of the high temperature section, it is preferable that the temperature change over time of the high temperature section is tracked and taken as the change over time in the temperature of this one section.

[0047] [1.5.2. Background part] The background is the area of ​​the pixel map other than the high temperature areas. The temperature of the background portion is not particularly limited, and may be the average temperature of a predetermined region near the high temperature portion, or the temperature of any one section of a temperature map of the predetermined region. When the temperature of one section of a temperature map of a predetermined region is used as the temperature of the background portion, it is preferable to track this one section and use the change in temperature of this one section as the change in temperature over time as the temperature of the background portion. The predetermined region is an experimentally determined region.

[0048] [1.5.3. Judgment method] The method for determining whether a high temperature portion is a defect or noise based on the temperature change over time of the high temperature portion is not particularly limited, and the most suitable method can be selected appropriately depending on the purpose.

[0049] [1.5.3.1. Example 1] In the case of a flaw, an induced current is concentrated during the heating process, resulting in a portion where the temperature actually rises compared to other portions. Therefore, for example, it is possible to determine whether the high temperature portion is a defect or noise from the temperature of the high temperature portion after heating is completed. More specifically, if there is a period in which the temperature of the high-temperature portion after heating is completed is above a predetermined temperature, the high-temperature portion can be determined to be a defect, and if there is no period in which the temperature is above the predetermined temperature, the high-temperature portion can be determined to be noise. The predetermined temperature and the predetermined period for which heat can be maintained are values ​​determined experimentally, and may vary depending on the depth (harmfulness) of the flaw, etc. For example, the predetermined temperature and the predetermined period can be set according to the harmfulness of the flaw to be detected, such as when the flaw is heated to room temperature +10°C for 0.1 seconds.

[0050] [1.5.3.2. Example 2] In scratches, depending on the penetration depth, heat pools may form at relatively deep positions during the heating process. It takes time for the heat from these pools to affect the surface, and the temperature may rise after heating is completed (see Figure 3). On the other hand, this does not happen with noise, so the temperature reaches its maximum at the end of heating and then decreases monotonically (see Figure 3).

[0051] Therefore, for example, if a period in which the temperature of the high-temperature part rises is confirmed after heating is completed, the high-temperature part can be determined to be a defect, and if a period in which the temperature of the high-temperature part rises is not confirmed, the high-temperature part can be determined to be noise.

[0052] [1.5.3.3. Example 3] If the cooling rate (temperature drop rate of the background portion) is high, the temperature drop rate will be greater than the temperature rise rate even in the case of a defect, and the temperature change over time may decrease monotonically. Therefore, for example, it is possible to determine whether a high temperature portion is a defect or noise from the change in the temperature difference between the high temperature portion and the background portion after heating is completed. More specifically, if a period of increasing temperature difference between the high temperature area and the background area is confirmed after heating is completed, the high temperature area can be determined to be a defect, and if no period of increasing temperature difference is confirmed, the high temperature area can be determined to be noise. In this case, the temperature of the high temperature portion also increases relative to the background portion.

[0053] The determination process may involve an operator determining defects and noise based on temperature changes over time in each high-temperature portion, or may involve a program determining defects and noise in accordance with a specific algorithm.

[0054] In addition, in order to eliminate the influence of sudden fluctuations in the temperature changes of the high temperature portion and the background portion, the determination step may be performed using the temperature changes of the high temperature portion and the background portion that have been subjected to a smoothing process. The smoothing process can be performed using a known method such as moving average processing within a predetermined interval.

[0055] [2. Effect] The induction heating flaw detection method according to the present invention includes a tracking step of tracking a high-temperature portion while acquiring a surface temperature. Therefore, by tracking the high temperature portion, it is possible to confirm the temperature change of the high temperature portion over time.

[0056] This makes it possible to easily determine whether the high temperature portion is a defect or noise based on the temperature change over time of the high temperature portion. In other words, even if a high temperature area is mistakenly recognized due to dirt, surface shape, etc., it can be determined to be noise by checking the temperature change over time, making it easy to distinguish between scratches and noise.

[0057] Furthermore, the induction heating flaw detection method according to the present invention can detect flaws regardless of the type of conductive material. That is, flaws can be detected even in non-magnetic metal materials such as non-magnetic steel, aluminum alloys, and copper alloys, which cannot be subjected to magnetic particle flaw detection.

[0058] Furthermore, in the tracking step, tracking of the high temperature portion can be performed without stopping the relative displacement between the conductive material and the induction heating portion. This allows for dynamic flaw detection, making it possible to detect flaws over a wide area in a short period of time.

[0059] [3. Induction heating flaw detection device] [3.1. Application part] FIG. 4 shows an induction heating flaw detector 1 according to the present invention. The induction heating flaw detector 1 includes an application unit 3 that applies black paint to the surface of the conductive material 2 .

[0060] [3.2. Induction heating section] The induction heating flaw detection device 1 includes an induction heating unit 6 whose relative position with respect to the conductive material 2 is changed by a transport unit 5 that transports the conductive material 2 after the black body paint has been applied. The conductive material 2 has its surface heated as it passes through the induction heating unit 6.

[0061] The induction heating unit 6 is stationary and connected to a high-frequency power supply (not shown). The penetration depth of the induced current can be easily changed by changing the frequency of the high-frequency current from the high-frequency power supply. Furthermore, the transport unit 5 can easily change the transport speed of the conductive material 2 by changing the rotation speed of the roller 5a.

[0062] [3.3. Extraction part] The induction heating flaw detector 1 is provided with an extraction unit 9 that acquires the surface temperature of the conductive material 2 after heating by the induction heating unit 6 is completed using a radiation thermometer 8 and extracts high-temperature portions. Here, the extraction unit 9 extracts the high temperature portion based on AI inference.

[0063] The extraction unit 9 includes a known computer having a CPU, a GPU, and memories such as a ROM and a RAM. The radiation thermometer 8 has a two-dimensional array type imaging element, and transmits temperature map data and pixel map data according to the resolution of the imaging element to the extraction unit 9 at a predetermined imaging frame rate.

[0064] The memory of the computer of the extraction unit 9 stores a machine learning model that has undergone predetermined machine learning. Then, pixel map data from the radiation thermometer 8 is input into this machine learning model, and the inferred high temperature areas are extracted from the pixel map. By comparing the pixel map with the temperature map, the temperature of the extracted high temperature area can be easily determined.

[0065] [3.4. Tracking section] The induction heating flaw detector 1 includes a tracking unit 11 that tracks the high-temperature portion while acquiring the surface temperature. The tracking unit 11 includes a known computer having a CPU, a GPU, and memories such as ROM and RAM. The tracking unit 11 and the extraction unit 9 share a common computer, and the tracking unit 11 can also use the temperature map data and pixel map data.

[0066] Here, since the transport speed of the conductive material 2 by the transport unit 5 and the imaging frame rate of the radiation thermometer 8 are known in advance, the high temperature area can be tracked according to a program pre-stored in the memory of the computer of the tracking unit 11.

[0067] [3.5. Judgment part] The induction heating flaw detector 1 includes a determination unit 12 that determines whether a high-temperature portion is a flaw or noise based on the temperature change over time of the high-temperature portion. The determination unit 12 includes a known computer having a CPU, a GPU, and memories such as ROM and RAM. The determination unit 12, the tracking unit 11, and the extraction unit 9 share a common computer, and the determination unit 12 can also use the temperature map data and pixel map data.

[0068] In addition, in the judgment unit 12, according to a program stored in advance in the memory of the computer of the judgment unit 12, if a temperature rise in the high-temperature part is confirmed, the high-temperature part is judged to be a defect, and if a temperature rise in the high-temperature part is not confirmed, the high-temperature part is judged to be noise.

[0069] Here, the processing performed after heating by the induction heating unit 6 of the induction heating flaw detector 1 is completed will be described with reference to the flowchart shown in FIG. First, in step S1, temperature map data based on the imaging data acquired by the radiation thermometer 8 and pixel map data are sent to the extraction unit 9. Next, in step S2, the pixel map data is input to the machine learning model stored in the extraction unit 9.

[0070] Next, in step S3, the extraction unit 9 extracts the inferred high temperature areas from the pixel map, and determines whether or not there is data for the previous high temperature area for each extracted high temperature area. If there is data on the previous high temperature area (Y), the process moves to step S4. If there is no data of the previous high temperature part (N), the extracted high temperature part is stored and the process proceeds to step S6.

[0071] Next, in step S4, the tracking unit 11 tracks the high temperature area and checks the temperature of the high temperature area by comparing the pixel map with the temperature map. Next, in step S5, the determining unit 12 determines whether the high temperature portion is a flaw or noise based on the temperature change over time of the high temperature portion.

[0072] Next, in step S6, it is determined whether or not the process should be continued. If the processing is to be continued (Y), the process returns to step S1, and new temperature map data and pixel map data based on new image data captured by the radiation thermometer 8 are acquired. If the process is not to be continued, the process is terminated. The determination of whether or not to continue the processing may be made based on whether or not the processing has continued for a predetermined period of time, or based on whether or not there is an external stop signal.

[0073] Here, the processing in the determination unit 12 will be described with reference to the flowchart shown in FIG. First, in step S51, the temperatures of the respective high temperature portions after heating is completed are acquired and stored.

[0074] Next, in step S52, it is determined whether or not there is previous temperature data for each high temperature portion. If there is previous temperature data (Y), the process proceeds to step S53. If there is no previous temperature data (N), the process ends.

[0075] Next, in step S53, it is determined whether or not the temperature has risen in each high temperature portion. If the temperature has risen (Y), the counter is incremented by one, and the process proceeds to step S54. If the temperature has not risen (N), the process proceeds to step S56, where the high temperature portion is determined to be noise, and the process ends.

[0076] Next, in step S54, it is determined whether the temperature rise has continued for a predetermined period of time. If the temperature rise continues for a predetermined period (Y), the process moves to step S55, where the high temperature portion is determined to be a flaw. If the temperature rise has not continued for the predetermined period (N), the process ends. This process is performed for each image data of the radiation thermometer 8 and for each high temperature portion. Once a high temperature area is determined to be a defect or noise, it is excluded from the tracking target, thereby saving computational resources and speeding up processing.

[0077] Here, the predetermined period calculated from the counter number is a period determined experimentally. By setting a predetermined period, redundancy is ensured against sudden temperature increases caused by fluctuations or the like.

[0078] [4. Effect] The induction heating flaw detector 1 according to the present invention has a tracking unit 11 that tracks the high-temperature portion while acquiring the surface temperature. Therefore, by tracking the high temperature portion, it is possible to confirm the temperature change of the high temperature portion over time.

[0079] This makes it possible to easily determine whether the high temperature portion is a defect or noise based on the temperature change over time of the high temperature portion. In other words, even if a high temperature area is mistakenly recognized due to dirt, surface shape, etc., it can be determined to be noise by checking the temperature change over time, making it easy to distinguish between scratches and noise.

[0080] Furthermore, the induction heating flaw detector 1 uses a computer for the processing of the extraction unit 9, the tracking unit 11, and the determination unit 12, so that the processing can be performed at high speed. Therefore, flaws can be determined in a short time. [Example]

[0081] [1. Test Method] The induction heating flaw detector 1 was used to detect flaws. A piece of steel made of SUS304 steel was used as the conductive material 2, and an aqueous solution of an activator was used as the black paint to improve wettability. The radiation thermometer 8 used was one that utilizes wavelengths in the far infrared region. The frequency of the high frequency current of the high frequency power supply was adjusted to about 100 kHz so that the penetration depth was shallower than the depth of the flaw to be detected. In addition, YOLO was used as the machine learning model in the extraction section 9. Furthermore, the rotation speed of the roller 5a of the conveying section 5 was adjusted so that the temperature at the flaw after heating was increased for a predetermined period (about 3 frames).

[0082] [2. Results] The results are shown in Figure 7. From Figure 7, two points are extracted as high-temperature areas, and it was confirmed that the temperature changes over time at each high-temperature area were different. In the high temperature area A, the temperature rose once after heating was completed, and then decreased monotonically. On the other hand, in the high temperature area B, the temperature decreased monotonically after heating was completed. The temperature curve over time of the high temperature portion B and the temperature curve over time of the background portion (not shown) were almost parallel to each other.

[0083] The high temperature area A was determined to be a defect because the temperature increase continued for a predetermined period (3 frames). On the other hand, in the high temperature area B, no temperature rise was observed, and it was therefore determined to be noise. It was confirmed that the high temperature part A was actually a defect, and the high temperature part B was actually noise.

[0084] Although the embodiments of the present invention have been described in detail above, the present invention is not limited to the above embodiments, and various modifications are possible within the scope of the gist of the present invention. [Industrial Applicability]

[0085] The induction heating flaw detection method according to the present invention can be used as a method for detecting flaws in conductive materials. The induction heating flaw detector according to the present invention can be used as a device for detecting flaws in conductive materials. [Explanation of symbols]

[0086] 1 induction heating flaw detector 2 conductive material 3 coating section 6 induction heating section 8 radiation thermometer 9 Extraction unit 11 Tracking unit 12 Judgment unit

Claims

1. a coating step of coating a surface of the conductive material with black paint; a heating step of heating a surface of the conductive material by changing a relative position between the conductive material and an induction heating unit after the black body paint is applied; an extraction step of obtaining a surface temperature of the conductive material after heating by the induction heating unit using a radiation thermometer and extracting a high-temperature portion; a tracking step of tracking the high temperature portion while acquiring the surface temperature; a determining step of determining whether the high-temperature portion is a defect or noise based on a temperature change over time of the high-temperature portion; An induction heating flaw detection method comprising:

2. The heating step comprises:

2. The induction heating flaw detection method according to claim 1, wherein the penetration depth of the induced current induced in the conductive material is set shallower than the flaw.

3. The determination step includes: When a temperature rise of the high temperature portion is confirmed, the high temperature portion is determined to be a flaw; 2. The induction heating flaw detection method according to claim 1, wherein the high temperature portion is determined to be noise when no temperature rise is confirmed in the high temperature portion.

4. The extraction step comprises: The induction heating flaw detection method according to claim 1, wherein the high temperature portion is extracted based on AI inference.

5. 2. The induction heating flaw detection method according to claim 1, wherein the black body paint is water.

6. an application unit that applies black paint to the surface of the conductive material; an induction heating unit that heats a surface of the conductive material by changing a relative position with respect to the conductive material after the black body paint is applied; an extraction unit that acquires the surface temperature of the conductive material after heating by the induction heating unit using a radiation thermometer and extracts a high-temperature portion; a tracking unit that tracks the high-temperature portion while acquiring the surface temperature; a determination unit that determines whether the high-temperature portion is a defect or noise based on a temperature change over time of the high-temperature portion; An induction heating flaw detection device.

7. The heating unit is 7. The induction heating flaw detection method according to claim 6, wherein the penetration depth of the induced current induced in the conductive material is set shallower than the flaw.

8. The determination unit When a temperature rise of the high temperature portion is confirmed, the high temperature portion is determined to be a flaw; 7. The induction heating flaw detector according to claim 6, wherein the high temperature portion is determined to be noise when no temperature rise in the high temperature portion is confirmed.

9. The extraction unit The induction heating flaw detection device according to claim 6, wherein the high temperature portion is extracted based on AI inference.

10. 7. The induction heating flaw detector according to claim 6, wherein the black body paint is water.

Images