WO2011162417A1 - Defective electrode detecting device - Google Patents

Abstract

Welding defects in all types of products can be detected at an early stage. A defective electrode detecting device used for a workpiece made of a laminated aluminum alloy comprises: a reflected light collecting unit for collecting reflected light of light waves scattered from a welding region; an infrared light collecting unit for collecting infrared light; each sensor unit for extracting the reflected light and infrared light having predetermined wavelengths from the light waves collected in each collecting unit, converting the extracted lights to electric signals, and transmitting the electric signals to a welding state determination processing unit (11); and the welding state determination processing unit (11) for monitoring each of the signals during the time that elapses before the welding region has been solidified. The welding state determination processing unit includes a control and calculation means for monitoring the detected intensities of the reflected light and infrared light on a per time basis, an output means, and a storage means. When the peak value of the detected intensity of the reflected light after a predetermined time of 2 ms has elapsed is a predetermined threshold value of 20 or more, the welding state determination processing unit determines an abnormality as a "visible defect" if the peak value of the detected intensity of the infrared light is a predetermined threshold value of 0.6 or more and as a "invisible defect" if the peak value of the detected intensity of the infrared light is less than a threshold value of B.

Description

Defective electrode detector

The present invention relates to a battery, for example, a defective electrode detection device for detecting a welding defect of an aluminum alloy used for an electrode of a lithium ion battery.

For example, lithium ion batteries are widely used as electric vehicle batteries that are attracting attention as one of the measures to prevent global warming.

Lithium-ion batteries are made of aluminum alloy for electrodes, in order to reduce the size and weight, make it easier to release ions, and two foil-like aluminum alloys are stacked and welded at appropriate locations. To manufacture. Laser welding is generally applied to welding in order to minimize the thermal effect on the peripheral parts.

By the way, in-vehicle batteries are different from general-purpose batteries used in home appliances, mobile phones, personal computers, etc., and are incomparable to general-purpose batteries due to vibration during driving and the temperature of the environment where the car is placed. Used in.

For example, it must operate properly in places with irregular vibrations on unpaved roads, for example, in a cold area of minus 20 ° C or in a high temperature area exceeding 50 ° C. Cause serious consequences. Therefore, high-accuracy quality control is required for in-vehicle batteries.

In laser welding, multiple conditions, such as the surface condition of the material, the behavior of keyholes during welding, and the decrease in output due to deterioration (dirt) of optical components, can affect the welding result, resulting in weld defects. is there.

By the way, since the aluminum alloy has high thermal conductivity and linear expansion coefficient, the welded portion and its peripheral portion are likely to be deformed by irradiation with laser light. In addition, when used for high-precision joining in the micro-region, different welding quality conditions such as hole defects may occur due to the surface conditions of individual workpieces, for example, the occurrence of gaps and subtle differences in heat capacity at each joint location. It's easy to do. When distortion or a gap occurs in the workpiece, it causes a defect.

In addition, when the voltage or electrode material stability of the lithium ion battery is poor, there is a possibility that the electrode is placed in a very strong oxidation state or reduction state when the voltage rises during charging and becomes unstable.
For this reason, in charging, voltage control with extremely high accuracy is required, and welding during electrode manufacturing also requires welding with no defects.

Conventionally, the inspection of the welding quality of the electrodes has been confirmed by visual inspection after welding, a destructive inspection such as a tensile test and measurement of a weld cross section.

However, such inspection requires skill and is difficult to inspect all products. Furthermore, it is often not easy for skilled workers to elucidate complex defect factors in processing quality, and because it requires skills backed up by experience and intuition, it is indispensable to train skilled inspectors. Yes, a lot of people are devoted to this inspection process.

And, if there are welding defects, extra steps such as long-term production line stoppage and quality re-inspection occur, causing problems such as quality problems and cost increase.

Specifically, for example, there are the following problems.
A. Detection of weld defects Weld defects were detected by sampling after completion of the welding process. In addition, in order to grasp internal defects and external defects, a plurality of operations such as destruction and visual inspection are required, which is expensive.
I. Eliminating weld defects The investigation of the cause of weld defects was an off-line inspection, which required time. Moreover, since it discovered in the downstream process of a manufacturing process, there existed a fault that a production loss was large.
C. Line restoration The investigation of the cause of welding defects requires resetting the line because the welding conditions are reset based on the experience of the engineers.

JP 2000-042769 A

The present invention has been made in the above background, and in the manufacturing process of an electrode used as an automobile battery or the like, the molten state of a workpiece to be welded is monitored in real time, and welding defects for all products are detected at an early stage. With the goal.

In order to achieve the above object, a defect electrode detection apparatus according to the present invention is an apparatus for detecting the presence or absence of defects in a molten state at a welded part of a work to be welded by laser light, wherein the work is a foil-like aluminum alloy. A reflected light collecting unit that collects the reflected light scattered from the welding site, an infrared light collecting unit that collects the infrared light scattered from the welding site, and the reflected light collecting unit. The reflected light sensor unit extracts reflected light of a predetermined wavelength from the collected light wave, converts the extracted light into an electrical signal, and sends the signal to the welding state determination processing unit, and the infrared light collecting unit collects the reflected light. An infrared light sensor unit that extracts infrared light of a predetermined wavelength from the emitted light wave, converts the extracted light into an electrical signal, and sends the signal to a welding state determination processing unit; the reflected light and the infrared light Monitoring the time until the welded part is solidified with a signal based on A welding state determination processing unit, the welding state determination processing unit, a control / calculation unit that monitors detection intensity for each time of the reflected light and the infrared light, an output unit that outputs a calculation result, and a predetermined value Storage means for storing the threshold value, and first discriminating an abnormality in the peak value of the detection intensity for each reflected light time, and then discriminating an abnormality in the peak value of the detection intensity in each infrared light time It is characterized by that.
Further, according to the defect electrode detecting device according to item 1, wherein, when the peak value of the detected intensity after the predetermined time T ₁ elapses per the reflected light is the threshold value A than the predetermined, the detected intensity of infrared light When the peak value is greater than or equal to a predetermined threshold value B, it is determined as “a clear defect”, and when it is less than the threshold value B, it is determined as a “hidden defect”.
Further, in the defect electrode detecting device according to claim 1 or claim 2, wherein the threshold time T ₁ relating to the reflected light characterized in that it is a "2ms" later.
The defect electrode detection apparatus according to claim 3, wherein a threshold value A of detection intensity related to the reflected light is “20”.
The defect electrode detection apparatus according to claim 3 or 4, wherein a threshold value B of the detection intensity related to the infrared light is "0.6".
The defect electrode detection apparatus according to claim 3 or 4, wherein the melting state determination processing unit is configured to detect infrared light when a peak value of detection intensity for the infrared light is equal to or greater than a predetermined threshold value B. when the difference between the fall time and the reflected light rise time is a predetermined value (absolute value) T less than ₂ determines that the first abnormal pattern or the second failure pattern, the value (absolute value) T ₂ or more In some cases, it is distinguished from the second abnormal pattern or the third abnormal pattern.
The defect electrode detection apparatus according to claim 3 or 4, wherein the melting state determination processing unit is configured to detect infrared light when a peak value of detection intensity for the infrared light is equal to or greater than a predetermined threshold value B. fall time and a difference is less than a predetermined value (absolute value) T ₂ of the reflected light rise time, the value difference between the infrared light rising time and the reflected light rise time is predetermined (absolute value) T ₃ or more when determined that the first failure pattern, the value when the (absolute value) T less than _3, characterized in that to determine the second abnormal pattern.
The defect electrode detection apparatus according to claim 3 or 4, wherein the melting state determination processing unit is configured to detect infrared light when a peak value of detection intensity for the infrared light is equal to or greater than a predetermined threshold value B. a is the fall time and the difference of the reflected light rise time is a predetermined value (absolute value) T ₂ or more, the value difference between the infrared light rising time and the reflected light rise time is predetermined (absolute value) T ₃ or more when determined that the second failure pattern, the value when the (absolute value) T less than _3, characterized in that to determine the third abnormal pattern.
Further, in the defect electrode detection apparatus according to claim 1 or 2, each of the light collecting portions is provided to be inclined with respect to a welding site.
The defect electrode detection apparatus according to claim 9 is characterized in that an inclination angle of each of the light collecting portions is 50 degrees.

Since the aluminum alloy used as the electrode of the lithium ion battery is made of a thin plate, if the workpiece is slightly distorted before or during welding when it is irradiated with laser light, the gap or the focus of the laser light will shift to the molten pool. Vigorous vibration (so-called “blow”) occurs.

Since this vibration energy does not affect the amount of heat, it is not detected by visible light that measures the amount of heat on the surface of the electrode work and infrared light that measures the amount of heat of the entire work. It is detected in the reflected light that measures the shape.

When the molten pool is exposed in this way, the appearance of electrode welding is the same as that of a good product, but since the penetration is shallow and an abnormality has occurred inside, for example, when used as an electrode of a lithium ion battery, due to vibration etc. It may cause leakage and cause an accident.

According to the present invention, it is possible to inspect all of the hidden weld defects of the welding electrode which is not apparent in appearance but has an internal abnormality in-line, and 100% mechanically detect the workpiece weld defects in real time. it can. Therefore, defective electrodes can be eliminated, and hence defective products of, for example, lithium ion batteries can be eliminated, so that the reliability of product quality can be ensured.

In addition, since welding defects can be detected upstream of the production line, loss can be suppressed, not only welding conditions but also the cause of defects up to the pretreatment stage can be analyzed instantaneously, and welding conditions can be quickly reset. Is possible.

(A) is a top view which shows embodiment of the defect electrode detection apparatus by this invention, (B) is the same front view. It is a block diagram which shows the Example of the reflected light condensing unit of FIG. It is a block diagram which shows the Example of the infrared-light condensing unit of FIG. It is a block diagram which shows the Example of the molten state discrimination | determination process part of FIG. (A) is a conceptual diagram showing “vibration state of melted part”, (B) is a conceptual diagram showing “sputtering”, (C) is a conceptual diagram showing “perforation of upper plate”, (D) is “upper and lower plate holes” It is a conceptual diagram which shows "aki". (E) is a conceptual diagram when it is welded well, and (F) is a conceptual diagram when it is not attached. It is a graph of the signal when the measurement signal of the welding workpiece | work obtained by the defect electrode detection apparatus by this invention is abnormal (molten part vibration state), (A) is reflected light, (B) is infrared light, (C) Indicates a signal of visible light measured as a comparative example. It is a graph of the signal when the measurement signal of the welding workpiece obtained by the defect electrode detection apparatus according to the present invention is abnormal (sputtering), (A) is reflected light, (B) is infrared light, and (C) is a comparative example. The signal of visible light measured as. It is a graph of the signal when the measurement signal of the welding workpiece | work obtained by the defect electrode detection apparatus by this invention is abnormal (upper-plate hole perforation), (A) is reflected light, (B) is infrared light, (C) Indicates a signal of visible light measured as a comparative example. It is a graph of the signal when the measurement signal of the welding workpiece obtained by the defective electrode detection apparatus according to the present invention is abnormal (upper and lower plate holes), (A) is reflected light, (B) is infrared light, (C) Indicates a signal of visible light measured as a comparative example. It is an image (drawing substitute photograph) when the measurement image of the welding workpiece obtained by the defective electrode detection device according to the present invention is abnormal (the melted portion vibration state). It is an image (drawing substitute photograph) when the measurement image of the welding workpiece obtained by the defect electrode detection apparatus according to the present invention is abnormal (occurrence of spatter). It is an image (drawing substitute photograph) when the measurement image of the welding workpiece obtained by the defective electrode detection device according to the present invention is abnormal (perforation of the upper plate). It is an image (drawing substitute photograph) when the measurement image of the welding workpiece obtained by the defective electrode detection apparatus according to the present invention is abnormal (upper and lower plate holes are provided). It is a graph of the signal in case the measurement signal of the welding workpiece | work obtained by the defect electrode detection apparatus by this invention is normal, (A) is reflected light, (B) is infrared light, (C) was measured as a comparative example. A visible light signal is shown. It is an image (drawing substitute photograph) when the measurement image of the welding workpiece obtained by the defective electrode detection apparatus according to the present invention is normal. It is a graph which shows the relationship between the visible light peak value as a comparative example, and a welding result. It is a graph which shows the relationship between a reflected light peak value and a welding result. It is a graph which shows the relationship between an infrared-light peak value and a welding result. It is a graph which shows the relationship between the acoustic signal peak value as a comparative example, and a welding result. It is a flowchart which shows the detection algorithm of the welding defect by this invention. (A) is an explanatory diagram of the rise time of the reflected light with respect to the sputter waveform example, (B) is an explanatory diagram of the rise time and the fall time of the infrared light, and (C) is a rise time of the reflected light with respect to the perforated waveform example. (D) is explanatory drawing of the rise time and fall time of the same infrared light. It is a graph showing the difference between the fall time of infrared light and the rise time of reflected light. It is a graph showing the difference between the rise time of infrared light and the rise time of reflected light.

Next, the defective electrode detection apparatus 1 according to the present invention will be described in more detail based on the drawings showing the embodiments. In addition, the same code | symbol is attached | subjected to the part which show | plays the same function, and the description is abbreviate | omitted.

FIG. 1 shows a schematic configuration diagram of a defective electrode detection apparatus 1. The defective electrode detection device 1 includes a reflected light collecting unit 3, an infrared light collecting unit 5, a reflected light sensor unit 7, an infrared light sensor unit 9, and a welding state determination processing unit 11. The infrared light sensor unit 9 includes an infrared light sensor unit 9a that detects infrared light of 1300 nm and an infrared light sensor unit 9b that detects infrared light of 1550 nm. Reference numeral 13 denotes a laser beam emitting unit, which is suspended from the work 15 and irradiates the work 15 placed on a table (not shown) with laser light. The work 15 is composed of an upper plate 15a and a lower plate 15b, and is placed on a table in a state where two foil-like aluminum alloys A3004 having a thickness of 0.1 mm are stacked. The reflected light condensing unit 3 and the infrared light condensing unit 5 are each provided with an inclination angle (α) inclined by 50 degrees with respect to the surface of the work 15, and the condensing point is aligned with the laser light irradiation unit. . The reflected light collecting unit 3 and the reflected light sensor unit 7 are connected by an optical fiber 17, and the infrared light collecting unit 5 and the infrared light sensor unit 9 are connected by optical fibers 19a and 19b. Connected.

As shown in FIG. 2, the reflected light condensing unit 3 is provided with a condensing lens 4 a at the tip of a cylindrical cylinder 4 and condenses the reflected light scattered from the melting part 16 of the work 15. The other end of the cylinder 4 is provided with another condenser lens 4b. Thereby, the reflected light obtained from the condensing point of the condensing lens 4a is sent to the condensing lens 4b as parallel light.

The light is sent to the reflected light sensor unit 7 through the optical fiber 17. As shown in FIG. 2, the reflected light sensor unit 7 includes a YAG laser light transmission filter 21, a visible light to infrared light photodiode 23, an amplifier 25, and an analog filter 27 connected in series. Extract wavelengths longer than infrared light.

That is, the YAG laser light transmission filter 21 has a transmittance of 0% for light in the wavelength band of 400 nm to 900 nm and a transmittance for light of 900 nm or more is 90% or more, thereby cutting the visible light region.

The light generated with high intensity during laser light irradiation is mainly plasma light (visible light region) and laser scattered light (YAG light wavelength: 1064 nm), so if visible light is removed, only the laser light is detected. Can do.

The visible light to infrared light photodiode 23 has a sensitivity wavelength range of 320 nm to 1100 nm and is set to a maximum sensitivity wavelength of 960 nm, thereby extracting a wavelength of 1070 nm to 1090 nm, and the detected light is converted into an electric signal. The The measurement signal converted into the electrical signal is amplified by the amplifier 25, the noise component is cut by the analog filter 27, and sent to the molten state determination processing unit 11.

FIG. 3 shows the infrared light collecting unit 5. The infrared light condensing unit 5 includes housings 29 and 30 each having a condensing lens. A rectangular parallelepiped casing 31 is accommodated in the housing 29, and a condenser lens 33 and a condenser lens 35 are provided at both ends of the casing 31, respectively. A rectangular parallelepiped housing 39 is provided in the housing 30 connected to the housing 29 via a filter 37. A condensing lens 41 is provided at the front end of the casing 39, and a dichroic mirror 43 is provided at the other end with an inclination of 45 degrees. The dichroic mirror 43 transmits visible light and reflects only light of 1300 nm to 1550 nm.

The infrared sensor unit 9 is connected to the housing 30. That is, as shown in FIG. 3, a housing 49 is provided orthogonal to the housing 30, and a rectangular parallelepiped casing 50 is provided in the housing 49. At both ends of the casing 50, an infrared light transmission filter 51 that transmits 1200 nm or more and an interference filter 55 that transmits only a wavelength of 1550 nm with a half-value width of 30 nm are provided, and a dichroic mirror 53 is inclined at 45 degrees in the middle. Provided. A condensing lens 57 condenses the light transmitted through the interference filter 55. An infrared light photodiode 59, an amplifier 61, and an analog filter 63 are connected to each of the condensing lenses 57 in series, thereby forming an infrared light sensor unit 9b that detects infrared light of 1550 nm. .

The infrared light photodiode 59 has a sensitivity wavelength range of 900 nm to 1700 nm and is set to a maximum sensitivity wavelength of 1550 nm, whereby the detected light is converted into an electric signal. The measurement signal converted into the electric signal is amplified by the amplifier 61, and then the noise component is cut by the analog filter 63 and sent to the molten state determination processing unit 11.

An interference filter 65 that transmits only a wavelength of 1300 nm with a half-value width of 30 nm is separately provided at a position orthogonal to the dichroic mirror 53. A condensing lens 67 condenses the light transmitted through the interference filter 65. An infrared light photodiode 69, an amplifier 71, and an analog filter 73 are connected in series with the condenser lens 67, thereby constituting an infrared light sensor unit 9 a that detects 1300 nm infrared light. .

The infrared light photodiode 69 has a sensitivity wavelength range of 900 nm to 1700 nm and is set to a maximum sensitivity wavelength of 1300 nm, whereby the detected light is converted into an electric signal. The measurement signal converted into the electric signal is amplified by the amplifier 71, and then the noise component is cut by the analog filter 73 and sent to the molten state determination processing unit 11.

The focal length (f) of the condenser lens 4a of the condenser unit 3 is 70 mm, and the focal length (f) of the condenser lens 4b is 20 mm. The focal length (f) of the condensing lens 33 of the infrared light condensing unit 5 is 300 mm, the focal length (f) of the condensing lens 35 is 150 mm, and the focal length (f) of the condensing lens 41 is 60 mm. The focal length (f) of the lens 57 was 30 mm, and the focal length (f) of the condenser lens 67 was 30 mm.

FIG. 4 shows the welding state determination processing unit 11, which processes the detection intensity for each time for the signal based on the reflected light and infrared light transmitted. The welding state determination processing unit 11 is connected to an A / D converter 75 that converts a signal based on the reflected light and infrared light transmitted to a digital signal, and the A / D converter 75 and a high-speed signal processor 79. A field programmable gate array (FPGA) 77 that communicates with each other, a storage means 81 that is connected to the high-speed signal processor 79 and communicates with each other, and an alarm signal that is connected to the high-speed signal processor 79 and that is defective. And output means 83 for outputting. The FPGA 77 is coupled with the A / D converter 75 to monitor the detected intensity of reflected light and infrared light for each time, and is coupled with the high-speed signal processor 79 to control and calculate data. The storage means 81, the threshold value A threshold time for reflected light T ₁ and the detection intensity is previously set and stored, also the threshold B of the detected intensity for the infrared light is previously set and stored. The high-speed signal processor 79 calculates the difference T ₂ between the infrared light fall time and the reflected light rise time, and the difference T ₃ between the infrared light rise time and the reflected light rise time, which will be described in detail later. The output means 83 emits an alarm sound when the threshold value A is exceeded or related to the threshold value B.

The laser beam is irradiated from the laser beam emitting unit 13 onto the required part of the workpiece 15, and the welding operation starts. The reflected light scattered from the melting part 16 is converted into parallel light from the condensing point of the condensing lens 4 a and sent to the condensing lens 4 b and transmitted to the reflected light sensor unit 7 through the optical fiber 17. The transmitted light is cut in the visible light region by the YAG laser light transmission filter 21, and a wavelength longer than the infrared light is extracted. The extracted light is converted into an electrical signal by the visible light to infrared light photodiode 23, amplified, and then the signal from which the noise component has been cut by the analog filter 27 is sent to the welding state determination processing unit 11. It is done.

On the other hand, the infrared light scattered from the melting portion 16 becomes parallel light from the condensing point of the condensing lens 33 and is sent to the condensing lens 35 and is condensed again by the condensing lens 35, and then the filter 37. After that, the condenser lens 41 becomes a confocal point. By using the confocal point, the influence of unnecessary scattered light other than the focal part of the condenser lens 33 is reduced. The light that has become parallel light by the condenser lens 41 is transmitted and removed by the dichroic mirror 43, and only the infrared light component of 1300 nm to 1550 nm is reflected in the direction of the infrared light sensor unit 9. The reflected light is further attenuated by the infrared light transmission filter 51, the 1550 nm light component is transmitted by the dichroic mirror 53, and the 1300 nm light component is reflected.

In the infrared light sensor unit 9b, the transmitted light of the former is transmitted by the interference filter 55 only at 1550 nm with a half-value width of 30 nm, and is condensed on the condensing lens 57, and is transmitted to the infrared light photodiode 59 as an electric signal. Is converted to This signal is further amplified by the amplifier 61, the noise component is cut by the analog filter 63, and transmitted to the welding state determination processing unit 11.

On the other hand, in the infrared light sensor unit 9a, the light reflected by the dichroic mirror 53 is transmitted by the interference filter 65 only at 1300 nm with a half-value width of 30 nm, and is condensed on the condensing lens 67, and the infrared light photodiode 69 is obtained. Is converted into an electrical signal. This signal is further amplified by the amplifier 71, the noise component is cut by the analog filter 73, and transmitted to the welding state determination processing unit 11.

FIG. 5 is a conceptual diagram of a welding defect. In (A), 16a is a “melting part vibration state” in which the welding defect is visually invisible, but in FIG. 5B, 16b is a solution during welding. "Spatter" in which a part of the spatter is scattered outside the melted portion 16, in (C) 16c is "upper plate perforated" where only the upper plate 15a is overlaid, and in (D) 16d is overlaid. Each of the upper plate 15a and the lower plate 15b indicates “upper and lower plate holes” in which holes are generated. In the figure, 2 indicates a laser beam. As a comparative example, “16 g” indicates the case where the welding is successfully performed in (E), and “16 e” indicates the case where the upper plate 15 a and the lower plate 15 b which are not welded are shown in (F). Comparing (A) and (E), the melted part vibration state 16a of (A) is characterized in that the surface is waved by the vibration of the melted part 16. Further, comparing (A) and (F), the melted part vibration state 16a in (A) is characterized in that the upper and lower plates 15a and 15b are welded.

Such weld defects in the melted part 16 are as shown in FIGS. 6 to 9 when viewed from waveform data, and as shown in FIGS. 10 to 13 as viewed from image data. FIG. 6 shows the waveform data of the “molten portion vibration state”, and FIG. 10 shows the image data. FIG. 7 shows the waveform data of “spatter”, and FIG. 11 shows the image data. FIG. 8 shows the waveform data of “perforated upper plate”, and FIG. 12 shows the image data. FIG. 9 shows the waveform data of “upper and lower plate holes”, and FIG. 13 shows the image data. As a comparative example, FIG. 14 shows waveform data in the case of good welding without welding defects, and FIG. 15 shows the image data. In each of the above drawings, the X axis represents time (ms), and the Y axis represents detection intensity (voltage × predetermined coefficient).

The image data was taken with a high-speed video camera (manufactured by nc Corporation, maximum shooting speed: 200,000 f / s, shortest exposure time: 1 / 2,000,000 seconds) installed obliquely to the laser beam irradiation. . Photographing is performed by using a high-power semiconductor laser with a wavelength of 940 nm for illumination through an interference filter that transmits a 962.5 nm band (half-value width 79 nm) in front of the lens, thereby suppressing halation due to laser reflected light and plume light, and a laser irradiation unit Was carried out so that it could be observed.

First, in FIGS. 15 and 14 showing normal welding, in the case of the image shown in FIG. 15, it is confirmed that the molten portion 16 is gradually formed immediately after the start of irradiation, and the diameter of the molten pool increases with time. Is done.

When the reflected light obtained at this time is confirmed, the detected intensity immediately after the start of irradiation is the highest at about 20, and after that, it changes at about 10 (FIG. 14 (A)). This is because immediately after the start of irradiation, since the reflectivity of the surface of the work 15 is high, strong laser reflected light can be obtained, but after that, the laser is absorbed inside the work 15 and the reflected laser light is reduced. As in the case of visible light, infrared light has a small detection intensity immediately after the start of irradiation, and gradually increases from around 3.8 ms, and a maximum detection intensity of about 0.25 is obtained at the end of irradiation (FIG. 14). (B)). This is because infrared light captures the heat input on the surface of the work 15, and the intensity increases as the surface temperature and the area of the melted part 16 increase due to laser absorption. Unlike the case of reflected light, visible light has a low detection intensity immediately after the start of irradiation, and the waveform increases from around 4.2 ms, and saturates at a detection intensity of about 10 from around 8 ms to the end of irradiation (FIG. 14 ( C)). This is because the intensity tends to increase as the laser is absorbed by the work 15 in order to capture the plume accompanied by the metal vapor generated from the work 15 by the laser absorption of the visible light. When the workpiece is welded well, almost the same image and waveform characteristics as described above can be obtained.

6 and 10 showing the vibration state of the melted part, in the case of the image shown in FIG. 10, the behavior was almost the same as that during good welding from the start of laser irradiation to around 3.2 ms, but the melted from around 3.2 ms. Part 16 begins to vibrate. This vibration converges once at about 5.8 ms after about 5.3 ms, and starts again at about 5.8 ms to 8.4 ms. Thereafter, no abnormal signal occurs until the end of laser irradiation.

See the reflected light waveform at this time based on Fig. 6. The reflected light waveform at this time is approximately the same as that during good welding from 0 ms to around 3.2 ms (see FIG. 14), but a detection intensity of 20 or more is detected around 3.2 ms to 5.3 ms. Attenuation occurs in the vicinity of about 5.3 ms to 5.8 ms, and a waveform having a detection intensity of about 20 is obtained again in the vicinity of about 5.8 ms to 8.4 ms. After that, it attenuates until the end of laser irradiation. This waveform behavior corresponds to the vibration generation timing of the melting part 16 of the high-speed camera in FIG. Therefore, it can be said that the waveform of the measurement signal shown in FIG. 6, particularly the reflected light waveform, clearly captures the vibration phenomenon.

Next, when comparing the image data of FIG. 10 showing the melted portion vibration state and FIG. 15 showing normal welding, there is no difference in appearance in the melted portion, but the waveform data of FIG. 6 and FIG. When comparing, there are significant differences. That is, in the reflected light shown in (A), two peak wave groups appear in a mountain shape in FIG. 6 (A), whereas in FIG. 14 (A), no peak wave group appears until the end. The remarkable difference is obvious at first glance. Therefore, it is possible to accurately and easily detect the “hidden defect” “molten state vibration state” in real time.

Here, there is no difference between the infrared light waveform data shown in FIGS. 6B and 14B and the visible light waveform data shown in FIGS. 6C and 14C. You have to be careful.

That is, the infrared light waveform has almost the same shape as that during good welding, and there is almost no fluctuation in the waveform (FIGS. 6B and 14B).
In addition, the visible light waveform can be confirmed to increase or decrease in waveform shape almost in synchronization with the timing of the reflected light that captures the molten state vibration state, but the abnormal signal is small with a detection intensity of 10 or less, and it is difficult to distinguish from good welding ( FIG. 6 (C) and FIG. 14 (C)).

This is thought to be due to the following reasons. Since the reflected light captures the laser diffused light, the waveform change accompanying the increase in scattered light due to the fluctuation of the molten pool surface due to the vibration of the molten portion 16 can be clearly confirmed, whereas the visible light is This is because the increase and decrease of the plume is captured, so that it is not substantially affected by the minute vibration of the melting part 16, and the infrared light intensity is not detected because of the temperature and area change of the melting part 16. Further, since the work 15 is a thin plate having a thickness of 0.1 mm, it is considered that the work 15 is easily distorted even under the same conditions, and vibration is caused by a slight shift of the gap and the laser focus.

Next, comparing FIG. 7 to FIG. 9 showing the welding abnormality and FIG. 14 showing the normal welding, the presence or absence of the peak wave in any of reflected light, infrared light and visible light is remarkable on the waveform data. Differences are observed. By comparing this point with the image data of FIGS. 11 to 13 showing the welding abnormality and FIG. 15 showing the normal welding, it is possible to confirm the presence of the welding defect from the appearance. Therefore, “sputter”, “upper plate hole”, and “upper and lower plate holes” that are “obvious defects” can be detected accurately and easily in real time.

This point will be described in detail. In the case of the occurrence of spatter, in the case of the image shown in FIG. 11, the behavior is almost the same as that during good welding from the start of laser irradiation to around 7.5 ms, but the spatter 16 b scatters from the melted part 16 around 7.5 ms. At this time, there is almost no change in the melted part 16 until just before the spatter 16b is scattered. After the spatter 16b is scattered, the fluctuation of the melting part 16 is converged, and the fluctuation is completely stopped in the vicinity of 10.4 ms.

Compare this with the waveforms shown in Figure 7. The reflected light waveform is almost the same as that during good welding from the start of irradiation to around 7.5 ms, but a detected intensity of 60 or more is detected around 7.5 ms and attenuates around 8.8 ms (FIG. 7 ( A)). The infrared light waveform is substantially the same as that during good welding from the start of irradiation to around 7.5 ms, but a detected intensity of 1.2 or more is detected around 7.5 ms and attenuates around 11.0 ms ( FIG. 7 (B)). The visible light waveform is almost the same as that during good welding from the start of irradiation to around 7.5 ms, but a detected intensity of 40 or more is detected around 7.5 ms and attenuates around 9.5 ms (FIG. 7 ( C)).

As described above, the waveform behavior of reflected light, infrared light, and visible light corresponds to the timing of spatter generation by the high-speed camera, and therefore, welding defects (spatter) are discriminated based on the waveform of the measurement signal shown in FIG. Can do.

In the case of the perforated upper plate, in the case of the image shown in FIG. 12, a hole is generated in the upper plate 15 a of the workpiece 15 in the vicinity of 10.6 ms.

Referring to the reflected light obtained at this time, a waveform with a detection intensity of around 20 is obtained around 10.6 ms. This is because the laser 2 is reflected from the surface on the lower plate 15b side due to the occurrence of perforation, and the scattered object is detected (FIG. 8A). The waveform of infrared light increases from around 3.2 ms, starts to increase more rapidly around 7.8 ms, and the detection intensity becomes 0.8 or more around 8.5 ms. After that, the detected intensity is attenuated to about 0.15 in the vicinity of 11.1 ms where the perforation of the upper plate 15a occurs. This is probably because the melted part 16 observing the heat disappears when the hole is generated, so that the waveform is attenuated in the vicinity of 11.1 ms. Thereafter, the detection intensity increases, and reaches a detection intensity of about 0.3 at the end of laser irradiation. This is presumably because the surface of the lower plate 15b captures the temperature rise due to laser absorption (FIG. 8B). Visible light has a waveform with a detected intensity of 20 or more from around 7.5 ms, and is greatly attenuated around 11.1 ms where holes in the upper plate 15a are generated. When perforation occurs, the plume disappears, so the waveform is thought to have attenuated around 11.1 ms. In addition, the detected intensity is higher than that during good welding. This is because a gap is generated between the upper plate 15a and the lower plate 15b, so that only the upper plate 15a is heated, and heat tends to accumulate, and the plume emits light. It is thought that the strength has increased.

In the case of the upper and lower plate holes, it is difficult to distinguish from the image shown in FIG. 13 that good welding is performed until the vicinity of 9.1 ms, but it appears that a hole is generated in the upper plate 15 a near 9.1 ms. It can be seen. Further, the melting portion 16 vibrates instantaneously in the vicinity of 10.8 ms, and a hole is also generated in the lower plate 15b in the vicinity of 13.1 ms.

When the reflected light obtained at this time is confirmed, a waveform with a detection intensity of around 60 is obtained around 9.1 ms. Thereafter, the reflected light attenuates due to the occurrence of perforations, but a peak with a detection intensity of about 60 occurs around 10.8 ms. Further, a waveform with a detection intensity of about 40 is obtained again at around 13.1 ms where a hole is generated in the lower plate 15b (FIG. 9A). Infrared light has a waveform increasing from around 3.5 ms, and a detection intensity of 0.4 or more is obtained around 9.1 ms where perforation of the upper plate 15a occurs. Immediately after that, a peak with a detection intensity of about 0.5 is obtained in the vicinity of 13.1 ms where the holes in the lower plate 15b are generated (FIG. 9B). Visible light has a waveform with a detected intensity of 20 or more from around 9.1 ms, is attenuated immediately afterwards, and a waveform with the same behavior is seen again around 13.1 ms where a hole is generated in the lower plate 15b. (FIG. 9C).
From these, it can be seen that the waveform of reflected light tends to rise in the vicinity of the occurrence of perforation, and the waveform of visible light and infrared light tends to fall when perforation occurs.

Next, based on FIG. 16 to FIG. 19, the effectiveness of detection of welding defects by the above three light waves, that is, reflected light, infrared light and visible light will be examined. The test conditions are the same as the implementation conditions described in the description of FIGS. The number of test workpieces is 60 pieces. For condensing and detecting visible light, devices having the same configuration as the reflected light collecting unit 3 and the reflected light sensor unit 7 used for collecting and detecting reflected light were used. In the figure, “◇” indicates good welding, “■” indicates the molten state vibration state, “×” indicates spatter, “Δ” indicates the upper plate hole, and “▲” indicates the upper and lower plate holes.

FIG. 16 is a graph in which the peak value of visible light is plotted for each welding defect classified in FIG. 5 in order to evaluate the effectiveness of detection of the welding defect by visible light. In FIG. 16, the peak values of “good welding” indicated by “◇” and “vibration state of melted portion” indicated by “■” are distributed with a detected intensity of 30 or less from the middle stage to the latter stage of laser light irradiation. . On the other hand, “spatter” indicated by “x”, “perforated upper and lower plates” indicated by “▲”, and “perforated upper plate” indicated by “Δ” are distributed with a detection intensity of about 20 to 100. There is a tendency to divide into the group of “good welding” and “vibration state of melted part” and the group of “sputtering”, “perforation of upper plate”, and “perforation of upper and lower plates” at a detection intensity of about 20-30. Since there are mixed data, it is difficult to clearly identify welding defects from only visible light or from visible light and other light wave components.

FIG. 17 is a graph in which the peak value of the reflected light is plotted for each welding defect classified in FIG. 5 in order to evaluate the detection effectiveness of the weld defect by the reflected light. In the case of the reflected light shown in FIG. 17, there is a tendency that peak values are generated on all time axes regardless of the presence or absence of welding defects, and that a certain amount of time is required for melting the melted portion by laser light. The earliest initial peak value can be adopted, and the initial peak value in this embodiment is “2.0” ms. In the case of “good welding”, the peak value is detected at a detection intensity of less than “20”, “melting part vibration state” indicated by “■”, “spatter” indicated by “×”, “▲”. Since the peaks of “upper and lower plate perforations” indicated by “Δ” and “upper plate perforation” indicated by “Δ” are all distributed at a detection intensity of about 20 to 100, the reflected light data in FIG. It can be used for discriminant evaluation.

FIG. 18 is a graph in which the peak value of infrared light is plotted for each welding defect classified in FIG. 5 in order to evaluate the effectiveness of detection of welding defects by infrared light. In the case of the infrared light shown in FIG. 18, “good welding” is dense with a detected intensity of “0.3” or less in the vicinity of about 12 ms to 15 ms, while the “molten vibration state” which is a welding abnormality is The detection intensity is concentrated below “0.6”, and “spatter”, “upper plate hole”, and “upper and lower plate holes” are concentrated above the detection intensity “0.8”. From these, the infrared light data in FIG. 18 can be used for discriminating evaluation of welding defects.

FIG. 19 is a graph in which the peak value of the acoustic signal is plotted for each welding defect classified in FIG. 5 in order to evaluate the detection effectiveness of the welding defect by the acoustic signal as a comparative example. This is because the acoustic change at the time of laser light irradiation is measured at an angle of 50 ° with respect to the irradiation surface and 100 mm from the laser irradiation part by using a super-directional microphone (AT815b type, manufactured by audio-technica, frequency characteristics: 50 to 20,000 Hz). The result was compared with the results of detection of welding defects by reflected light, visible light, and infrared light. In the case of an acoustic signal, more than half of the melted portion vibration states in which defect detection on the appearance is impossible are mixed with “good welding” in the vicinity of the detection intensity “0.0”. Therefore, it is difficult to use an index for identifying a welding defect from an acoustic signal.

From FIG. 16 to FIG. 19, the light wave that can be used for discriminating a welding defect of a workpiece spot-welded with an aluminum alloy among the three light waves is two components of reflected light and infrared light, and visible light is used. You can't do it. Next, the order of reflected light and infrared light for discrimination of welding defects was examined. It is important to first determine whether or not there is an abnormality with respect to the peak value of the detection intensity of the reflected light according to FIGS. 6 to 9 described above, and then the abnormality is detected with respect to the peak value of the detection intensity of the infrared light. It is determined whether or not there is. In other words, whether or not there is an order of reflected light and infrared light for determining weld defects is “Yes”, and reflected light and infrared light cannot be used simultaneously, and infrared light is used. After that, the reflected light cannot be used. Furthermore, it can be seen that by determining the initial peak value of the reflected light, it is possible to detect only the presence or absence of welding defects from the reflected light alone.

FIG. 20 is a flowchart of the algorithm employed under this assumption. First, in step 1, a peak value after 2 milliseconds (2 ms) from the start of irradiation of laser light with respect to reflected light is set as a detection target (S1, see FIG. 17). In this embodiment, the threshold time T ₁ is best to two milliseconds (2 ms).

Next, the detected intensity peak value of the reflected light is discriminated. When the detected intensity is less than “20” after “2 ms”, it is discriminated as “good welding” (S2a). That is, as shown in FIG. 17, in the case of “good welding”, the peak value is detected with a detection intensity of less than 20, and in this embodiment, it is optimal to set the detection intensity threshold A to “20”.

On the other hand, if the detected intensity is “20” or more after “2 ms”, there is some welding defect (S2b).

Next, the detected intensity peak value of infrared light is discriminated, and when it is less than “0.6”, it is discriminated as a hidden defect (melted portion vibration state) (S3a). On the other hand, when it is “0.6” or more, the probability of appearance of “visible defects” increases, and this is determined at the initial stage of welding (S3b).

Next, when “obvious defect” in step 3b is determined, the value (absolute value) of “infrared light fall time (ms) −reflected light rise time (ms)” is “0.8”. Whether it is less than or greater than is determined.

When it is less than “0.8” in step 3b, the value (absolute value) of “infrared light rise time (ms) −reflected light rise time (ms)” is determined (S4a). When the value of step 4a is “0.5” or more, it is determined as a “perforated defect” as the first abnormal pattern (S5a), and when it is less than “0.5”, “sputtering” as the second abnormal pattern. Or “perforated defect” (S5b).

Further, when “0.8” or more in step 3b, the value (absolute value) of “infrared light rise time (ms) −reflected light rise time (ms)” is determined (S4b). When the value of step 4b is "0.5" or more, it is determined as either "spatter" or "perforated defect" as the second abnormal pattern (S5b), and when it is less than "0.5" It is determined as “spatter” as an abnormal pattern (S5c).

The basis for enabling discrimination of the hidden defect (melted portion vibration state) in step 3a (S3a) is shown in FIG. That is, as shown in FIG. 18, “good welding” is dense at a detection intensity of 0.3 or less in the vicinity of about 12 ms to 15 ms with respect to the detection intensity peak value of infrared light. It is based on the fact that the concentration is less than 0.6, and “spatter”, “upper plate hole”, and “upper and lower plate holes” are concentrated at a detection intensity of 0.8 or more. Therefore, in this embodiment, the infrared light detection intensity threshold B is optimally set to “0.6”.

The basis for enabling discrimination by S5a to S5c is based on FIG. 21 to FIG. As shown in FIG. 21A and FIG. 21B, the behavior of the reflected light waveform and the infrared light waveform are similar when sputtering occurs, and the waveforms rise almost simultaneously. On the other hand, as shown in FIGS. 21C and 21D, when perforation occurs, the reflected light waveform tends to rise at the timing when the infrared light waveform falls.

21A and 21B, “20” is used as the threshold value A for reflected light detection intensity and “0.6” is used as the threshold value B for infrared light detection intensity, and “2 ms from the start of laser light irradiation. ”And subsequent times, the time of reflected light detection intensity“ 20 ”or more is“ reflected light rise time a ”, the time of 1300 nm infrared light detection intensity“ 0.6 ”or more is“ infrared light rise time b ”, The time less than “0.6” after the rise of infrared light is defined as “infrared light fall time c”. The difference T _{3 (absolute} value) of the "infrared light rise time b" and "reflected light rise time a", or the difference T _{2 (absolute} "infrared light fall times c" and "reflected light rise time a" Value).

In FIGS. 21C and 21D, “20” is used as the threshold value A for reflected light detection intensity and “0.6” is used as the threshold value B for infrared light detection intensity. After 2 ms, the time of reflected light detection intensity “20” or more is “reflected light rise time d”, and the time of 1550 nm infrared light detection intensity “0.6” or more is “infrared light rise time e”. The time less than “0.6” after the rise of infrared light is defined as “infrared light fall time f”. The difference T _{3 (absolute} value) of the "infrared light rise time e" and "reflected light rise time d", or the difference T _{2 (absolute} and "infrared light fall time f""reflected light rise time d ' Value).
By these, welding defects are classified.

From FIG. 22, when a threshold value is set at “0.8 ms”, in the case of spatter occurrence indicated by “×”, all of the seven samples are 0.8 ms or more, and in the case of a hole indicated by “Δ” (here “ 11 pieces out of 12 samples were less than 0.8 ms, “upper and lower plate holes” and “upper plate holes”. This is because reflected light rises at the timing when infrared light falls when perforation occurs.

The reason why reflected light rises at the timing of infrared light fall will be described in detail.
The infrared light waveform represents the amount of heat of the melting part 16 and the fluctuation of the area of the melting part 16, and when the perforated 16c is generated, the melting part 16 is naturally lost, and the waveform is greatly lowered. That is, the infrared light waveform falls at the timing when the perforated 16c is generated (see FIG. 21D).
On the other hand, the reflected light waveform represents the fluctuation of the laser scattered light, and the laser beam 2 is absorbed by the melting part 16 when the melting part 16 is stable until immediately before the perforation 16c is generated. However, there is no fluctuation in the waveform, but a hole 16c is generated in the upper plate 15a, and the surface of the lower plate 15b that is not melted is irradiated with the laser beam 2, so that the laser scattered light from the lower plate 15b side is strong. Thus, a large reflected light waveform is detected. That is, the reflected light waveform rises at the timing when the hole 16c is generated. (See FIG. 21C)
From the above, the infrared light falls and the reflected light tends to rise at the moment when the perforated 16c is generated.

Therefore, by numerically detecting the above, it becomes possible to find a welding defect (perforated 16c), and to quickly reset the welding conditions for eliminating the welding defect (perforated 16c). it can.

From FIG. 23, when a threshold value is set at “0.5 ms”, in the case of spatter occurrence indicated by “x”, 6 out of 7 samples are less than 0.5 ms, and in the case of perforation indicated by “Δ” 11 out of 12 were 0.5 ms or longer. This is because, as described above, the waveform rise times of infrared light and reflected light are similar when sputtering occurs, and the reflected light rises at the timing when the infrared light rises.

The reason why reflected light rises at the timing of rising of infrared light will be described in detail.
The infrared light waveform represents the amount of heat of the melting part 16 and the fluctuation of the area of the melting part 16, and when the spatter 16b is scattered, there are a lot of scattered objects within the observation area of the infrared light sensor unit 9a instantaneously. As a result, the scattered matter has substantially the same heat as that of the melted portion 16, and as a result, the amount of infrared light increases and the waveform greatly increases. That is, the infrared light waveform rises at the timing when the sputter 16b is generated (FIG. 21B).
On the other hand, in the reflected light waveform, a lot of scattered light is generated due to the abrupt behavior of the melting portion 16 when the sputter 16b is generated. Therefore, the reflected light also rises at the timing when the sputter 16b is generated (FIG. 21A).
As described above, the infrared light rises and the reflected light tends to rise at the moment when the sputter 16b is generated.

Therefore, by numerically detecting the above, it becomes possible to find a welding defect (spatter 16b), and to quickly reset the welding conditions for eliminating the welding defect (spatter 16b).

Here, the reason why the sensor sensitivity is adjusted to the initial peak value will be explained. In the present invention, the initial peak value is used for setting the detection intensity threshold value of the reflected light in order to easily catch a subtle intensity change due to the melted portion vibration state. In the present invention, the light collecting portions 3 and 5 are obliquely arranged with respect to the work 15 and the reflected light and infrared light are measured from the oblique direction. Since the later scattered light can be caught more clearly, it is possible to discriminate subtle changes in intensity due to the melted portion vibration state.

In addition, we will supplement the “molten part vibration state” here. In the present invention, since the workpiece 15 is a workpiece obtained by spot welding a foil-like aluminum alloy, when irradiated with laser light, compared to a workpiece made of another material, even under irradiation under the same conditions, or before welding. There is a possibility that slight distortion occurs in the workpiece. As a result, gaps and laser beam defocusing occur between the upper and lower workpieces, and vigorous vibrations (“molten part vibration state”) occur in the molten pool. This intense vibration causes the reflected pool to scatter because the molten pool surface undulates, so that the intensity change at this time can be caught in the reflected beam. On the other hand, since the amount of heat of the molten part 16 is detected in the infrared light, unless the area of the molten pool changes significantly, it cannot be caught because it does not appear as a large waveform change.

The above results are summarized in Tables 1 and 2.

According to the above-described embodiment, it is possible to determine the “molten portion vibration state”, which is a “hidden defect” that is visually invisible, at the initial stage of welding. Since the workpiece 15 according to the present invention is a workpiece obtained by spot welding a foil-like aluminum alloy, even a small gap greatly affects the welding strength. Therefore, although gap management is important, distortion or a gap during welding may easily occur in the work 15 due to laser light irradiation. If the workpiece 15 welded in such a state is used as an electrode, there is a risk of battery leakage due to insufficient welding, defects in electrical conductivity or operating voltage. In particular, if there is vibration or the like, the probability of generating the above-described defect increases.

Since the defect electrode detection device according to the present invention is digitized including not only welding conditions but also welding phenomena and constructed an algorithm, it detects such welding defects such as minute distortions and gaps in real time as follows. can do. Therefore, it is effective for quality control of battery electrodes that require precision, and there is an effect that repair can be handled by early detection of welding defects. Specifically, it is as follows.

A. Detection of Welding Defects Welding defects can be 100% mechanically detected during welding by performing 100% inspection during the welding process. This eliminates the need for an inspection process downstream of the manufacturing process and contributes to cost reduction.
I. Eliminating weld defects Because the investigation of the cause of weld defects is in-line inspection, it can be solved instantly. Moreover, since this elucidation is performed in an upstream process during laser welding, production loss is small.
C. Line restoration Because it is possible to immediately investigate the cause of a welding defect, it is possible to quickly reset the conditions and remove the cause of the defect, thereby strengthening the quality control system.

The present invention is not limited to the embodiment described above. For example, the threshold value A regarding the detection intensity of the reflected light is not limited to “20”, and an appropriate threshold value can be adopted depending on the thickness, components, and the like of the aluminum alloy. This is shown in Table 3.

Table 3 shows data measured by changing only the threshold value of Step 1 (S1) in FIG. 20 and fixing other threshold values. The numerical values of the sections where the vertical axis and the horizontal axis intersect are the evaluation items for the set threshold value. The detection probability (%) is shown.

From Table 3, in the case of this example, when the threshold value of the detection intensity of the reflected light is set to “23”, the detection probability of “good welding” becomes “96%”, and when the threshold value is set to “17”, “melting part” The detection probability of “vibration state” is “88%”, which increases the probability of occurrence of an error, which is inappropriate.

Further, the threshold value B relating to the detection intensity of infrared light is not limited to “0.6”, and an appropriate threshold value can be adopted depending on the thickness, components, etc. of the aluminum alloy. This is shown in Table 4.

Table 4 shows data measured by changing only the threshold value of step 2 (S2b) in FIG. 20 and fixing other threshold values. The numerical values of the sections intersected by the vertical axis and the horizontal axis are the evaluation items for the set threshold value. The detection probability (%) is shown.

From Table 4, in the case of this example, when the threshold value of the detection intensity of infrared light is set to “0.9”, the detection probability of “molten portion vibration state” becomes “93%”, and the threshold value is set to “0.5”. ”Is not suitable because the detection probability of“ either spatter or perforated ”becomes“ 95% ”and the probability of occurrence of an error increases.

Further, not limited to "2.0" ms applies threshold time T ₁ of the reflected light, the aluminum alloy thickness can be appropriately adopted threshold by components like. This is shown in Tables 5 (1) and (2).

Table 5 (1), (2) is varied only thresholds T ₁ in step 1 (S1) in FIG. 20, a data measured by fixing the other threshold, each section crossing the longitudinal axis and the horizontal axis The numerical value of indicates the detection probability (%) of each evaluation item with respect to the set threshold.

From Tables 5 (1) and 5 (2), in the case of this example, the generation time of the reflected light peak value is “good welding”, “melted portion vibration state”, “spatter” when the threshold is set to “2.0” ms or more. , “Perforated” can be detected with a probability of “100%”, but if the threshold is set to “1.6” ms or less, the detection probability of “molten vibration state” will be “93%”, resulting in an error. This is not suitable because the probability of occurrence increases.

The threshold time T ₂ of the infrared light, as shown in Table 6, "good welding" Other than "0.8ms""melt unit vibration state", "sputter, either perforated" each "100%" Can be detected with a probability of. Thus for a threshold time T ₂ of the infrared light, the aluminum alloy thickness can be appropriately adopted threshold by components like.

As shown in Table 7, the infrared light threshold time T ₃ is “100%” for “good welding”, “melted portion vibration state”, and “sputtered or perforated” other than “0.5 ms”. Can be detected with a probability of. Thus for a threshold time T ₃ of the infrared light, the aluminum alloy thickness can be appropriately adopted threshold by components like.

The band of reflected light extracted as detection light in the wavelength range of 1070 nm to 1090 nm, the band of infrared light in the range of about 1300 nm (± 15 nm), and the band of about 1550 nm (± 15 nm) should be understood as examples. Yes, some width in front and back in the wavelength range is allowed.

The inclination of the light collecting portions 3 and 5 with respect to the work 15 is not limited to 50 degrees. In this case, if the tilt angle is larger than 50 degrees and stands up too much, the influence of the initial specular reflection of the reflected light becomes undesirably large, and if the tilt angle is smaller than 50 degrees, it is not desirable. There is a disadvantage that the detection intensity is weakened.

As long as the welding state determination processing unit 11 has a function of processing the detection intensity for each time with respect to the signal based on the reflected light and infrared light transmitted, the apparatus and means constituting this can be selected as appropriate. It is. In addition, the output unit may output any abnormal state. For example, the output unit may output an abnormal state on a screen, an input / output (IO) port, or other external output device.

Regarding the reflected light condensing unit 3, the infrared light condensing unit 5, the reflected light sensor unit 7, the infrared light sensor unit 9, and the welding state discrimination processing unit 11, the distance proximity or physical integrity of each of the above parts is not questioned. Absent.

The present invention can be used in the battery industry such as automobile batteries and other batteries.

DESCRIPTION OF SYMBOLS 1 Defect electrode detection apparatus 2 Laser beam 3 Reflected light condensing unit 4 Cylinder 4a Condensing lens 4b Condensing lens 5 Infrared light condensing unit 7 Reflected light sensor unit 9 Infrared light sensor unit 9a Infrared light sensor unit 9b Red External light sensor unit 11 Welding state determination processing unit 13 Laser beam emitting unit 15 Work 15a Upper plate 15b Lower plate 16 Melting portion 16a Melting portion vibration state 16b Sputter 16c Upper plate perforation 16d Upper and lower plate perforations 16e Non-adhesive welding 16g Normal welding 17 Optical fiber 19a Optical fiber 19b Optical fiber 21 YAG laser light transmission filter 23 Photodiode for visible light to infrared light 25 Amplifier 27 Analog filter 29 Housing 30 Housing 31 Housing 33 Condensing lens 35 Condensing lens 37 Filter 39 Housing 41 Condensing Lens 43 Dichroic Mirror 49 Housing 50 Housing 51 Infrared Light Transmission Filter 53 Dichroic Mirror 55 Interference Filter 57 Condensing Lens 59 Infrared Light Photodiode 61 Amplifier 63 Analog Filter 65 Interference Filter 67 Condensing Lens 69 Photodiode for infrared light 71 Amplifier 73 Analog filter 75 A / D converter 77 FPGA
79 High-speed signal processor 81 Storage means 83 Output means

Claims (10)

1. An apparatus for detecting the presence or absence of a defect in a molten state with respect to a welded portion of a workpiece to be welded by laser light,
   The workpiece is made of a foil-like aluminum alloy,
   A reflected light condensing part for condensing the reflected light scattered from the welding site;
   An infrared light condensing part that condenses the infrared light scattered from the welding site;
   A reflected light sensor unit that extracts reflected light of a predetermined wavelength from the light wave collected by the reflected light collecting unit, converts the extracted light into an electrical signal, and sends the signal to a welding state determination processing unit;
   An infrared light sensor unit that extracts infrared light of a predetermined wavelength from the light wave collected by the infrared light condensing unit, converts the extracted light into an electrical signal, and sends a nuclear signal to the welding state determination processing unit; ,
   It consists of a welding state determination processing unit that monitors the time until the welded part is solidified with the signal based on the reflected light and the infrared light,
   The welding state determination processing unit is a control / calculation unit that monitors detection intensity for each time of the reflected light and the infrared light, an output unit that outputs a calculation result, and a storage unit that stores a predetermined threshold value. A defect electrode detection apparatus comprising: firstly determining an abnormality with respect to a peak value of detection intensity for each time of reflected light; and then determining an abnormality with respect to a peak value of detection intensity for each time of infrared light. .
2. 2. The defect electrode detection apparatus according to claim 1, wherein the peak value of the detection intensity of the infrared light when the peak value of the detection intensity after the elapse of a predetermined time T _{1 for the} reflected light is equal to or greater than a predetermined threshold A. 3. A defective electrode detection device characterized in that when it is equal to or greater than a predetermined threshold B, it is determined as a “clear defect”, and when it is less than the threshold B, it is determined as a “hidden defect”.
3. In the defect electrode detecting device according to claim 1 or claim 2, wherein the defect electrode detecting apparatus characterized by threshold time T ₁ relating to the reflected light is "2ms" later.
4. 4. The defect electrode detection apparatus according to claim 3, wherein a threshold value A of detection intensity related to the reflected light is “20”.
5. 5. The defect electrode detection apparatus according to claim 3, wherein a threshold value B of detection intensity related to the infrared light is “0.6”.
6. 5. The defect electrode detection device according to claim 3, wherein the melting state determination processing unit detects the falling edge of the infrared light when the peak value of the detection intensity for the infrared light is equal to or greater than a predetermined threshold value B. 6. time and when the difference of the reflected light rise time is a predetermined value (absolute value) T less than ₂ determines that the first abnormal pattern or the second failure pattern, the value (absolute value) when it is T ₂ or more Is determined as a second abnormal pattern or a third abnormal pattern.
7. 5. The defect electrode detection device according to claim 3, wherein the melting state determination processing unit detects the falling edge of the infrared light when the peak value of the detection intensity for the infrared light is equal to or greater than a predetermined threshold value B. 6. a time and the reflected light difference rise time the predetermined value (absolute value) T less than _2, the value difference between the infrared light rising time and the reflected light rise time is predetermined (absolute value) T ₃ or more first abnormality pattern and to determine, the value (absolute value) when less than T ₃ defect electrodes detecting apparatus characterized by determining a second abnormal pattern when the.
8. 5. The defect electrode detection device according to claim 3, wherein the melting state determination processing unit detects the falling edge of the infrared light when the peak value of the detection intensity for the infrared light is equal to or greater than a predetermined threshold value B. 6. a is time difference of the reflected light rise time is a predetermined value (absolute value) T ₂ or more, the value difference between the infrared light rising time and the reflected light rise time is predetermined (absolute value) T ₃ or more defect electrode detector and the second failure pattern and to determine, the value (absolute value) when less than T _3, characterized in that to determine the third abnormal pattern when the.
9. 3. The defect electrode detection device according to claim 1, wherein each of the light condensing portions is provided to be inclined with respect to a welding site.
10. 10. The defect electrode detection apparatus according to claim 9, wherein an inclination angle of each of the light collecting portions is 50 degrees.

Images