Laser processing apparatus, failure mode determination system, and failure mode determination method

The laser processing apparatus uses OCT to generate a three-dimensional image of keyhole depths and intensities, enabling accurate discrimination of defect modes through a machine learning model, addressing the inability of conventional systems to specify defects.

JP2026060795APending Publication Date: 2026-04-08PANASONIC INTELLECTUAL PROPERTY MANAGEMENT CO LTD
View PDF 1 Cites 0 Cited by

Patent Information

Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Filing Date
2024-09-27
Publication Date
2026-04-08

AI Technical Summary

Technical Problem

Conventional laser processing apparatuses cannot specify the types of defects in laser processing and discriminate between different defect modes, such as gap, penetration, and dissimilar material penetration defects.

Method used

A laser processing apparatus that uses optical coherence tomography (OCT) to generate a three-dimensional information image based on the reflected light of measurement light, determining the depth and intensity of keyholes formed during processing, and employs a discrimination unit to identify defect modes using a machine learning model.

Benefits of technology

Enables accurate discrimination of defect modes in laser processing, allowing for precise identification of defects and ensuring proper processing quality.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2026060795000001_ABST
    Figure 2026060795000001_ABST
Patent Text Reader

Abstract

The present invention provides a laser processing apparatus that can identify defective modes in laser processing. [Solution] A laser processing apparatus 1 that performs laser processing on a workpiece 2 by irradiating the workpiece 2 with processing laser light PL, comprising a first laser light source 21 that emits processing laser light PL, and a discrimination unit 53 that determines a defective mode of laser processing based on a three-dimensional information image generated from the reflected light of measurement light ML reflected from the workpiece 2, wherein the three-dimensional information image is an image showing the distribution of the depth of the keyhole 2b formed in the workpiece 2 by irradiation with processing laser light PL, the intensity of the reflected light of measurement light ML, and the frequency of occurrence of the depth of the keyhole 2b in relation to the intensity of the reflected light of measurement light ML.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present disclosure relates to a laser processing apparatus, a defect mode discrimination system, and a defect mode discrimination method.

Background Art

[0002] A laser processing apparatus that performs laser processing on a processing object by irradiating the processing object with laser light is known. As this type of laser processing apparatus, a laser welding machine that laser-welds a processing object by irradiating the processing object made of a metal material or the like with laser light is known (for example, Patent Document 1).

[0003] In addition, in the laser welding machine disclosed in Patent Document 1, the depth of a keyhole generated on the processing surface of the processing object during laser processing is measured by an optical coherence tomography (OCT) technique.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] However, in a conventional laser processing apparatus, the types of defects in laser processing cannot be specified, and the defect mode cannot be discriminated.

[0006] The present disclosure solves such problems, and an object thereof is to provide a laser processing apparatus, a defect mode discrimination system, and a defect mode discrimination system that can discriminate a defect mode in laser processing.

Means for Solving the Problems

[0007] To achieve the above objective, one embodiment of the laser processing apparatus according to the present disclosure is a laser processing apparatus that performs laser processing on a workpiece by irradiating the workpiece with processing laser light, comprising: a laser light source that emits processing laser light; and a discrimination unit that determines a defective mode of the laser processing based on a three-dimensional information image generated from the reflected light of a measurement light reflected from the workpiece, wherein the three-dimensional information image is an image showing the distribution of the depth of a keyhole formed in the workpiece by irradiation with processing laser light, the intensity of the reflected light, and the frequency of occurrence of the keyhole depth at the intensity of the reflected light.

[0008] Furthermore, one embodiment of the defect mode determination system according to the present disclosure is a defect mode determination system for determining a defect mode of laser processing when laser processing is performed on a workpiece by irradiating the workpiece with processing laser light, and comprises a determination unit that determines the defect mode of laser processing based on a three-dimensional information image generated from the reflected light of a measurement light reflected from the workpiece, the three-dimensional information image is an image showing the distribution of the depth of the keyhole formed in the workpiece by irradiation with processing laser light, the intensity of the reflected light, and the frequency of occurrence of the keyhole depth at the intensity of the reflected light.

[0009] Furthermore, one embodiment of the defect mode determination method relating to this disclosure is a defect mode determination method for determining a defect mode of laser processing when laser processing is performed on a workpiece by irradiating the workpiece with processing laser light, and includes the step of determining the defect mode of laser processing based on a three-dimensional information image generated based on the reflected light of a measurement light reflected from the workpiece, wherein the three-dimensional information image is an image showing the distribution of the depth of the keyhole formed in the workpiece by irradiation with processing laser light, the intensity of the reflected light, and the frequency of occurrence of the keyhole depth at the intensity of the reflected light. [Effects of the Invention]

[0010] According to this disclosure, it is possible to determine the defect mode in laser processing. [Brief explanation of the drawing]

[0011] [Figure 1] Figure 1 shows the process of measuring the depth of a keyhole formed in a workpiece. [Figure 2] Figure 2 shows the optical interference signal obtained according to the optical path difference between the reflected measurement light reflected by the keyhole and the reference light, and the data obtained by applying a Fast Fourier Transform to the optical interference signal. [Figure 3] Figure 3 is a schematic diagram showing data on keyhole depth according to the distance of the machined surface. [Figure 4] Figure 4 shows an example of a keyhole depth profile depending on the distance of the machined surface. [Figure 5] Figure 5 is a diagram illustrating the conditions when defects occur in the gap defect mode, penetration defect mode, and dissimilar material penetration defect mode. [Figure 6] Figure 6 is a schematic diagram showing data on keyhole depth according to the distance of the machined surface when a defect occurs. [Figure 7] Figure 7 shows an example of a keyhole depth profile corresponding to the distance of the machined surface when a defect occurs. [Figure 8] Figure 8 shows the configuration of a laser processing apparatus according to an embodiment. [Figure 9] Figure 9 shows the optical interference signal obtained according to the optical path difference between the reflected light of the measurement light reflected by the keyhole and the reference light in a laser processing apparatus according to the embodiment, and the data obtained by applying a Fast Fourier Transform to the optical interference signal. [Figure 10] Figure 10 shows an example of a keyhole depth profile (two-dimensional information image) corresponding to the distance of the processing surface in a laser processing apparatus according to an embodiment. [Figure 11] Figure 11 shows an example of a three-dimensional information image generated by the image generation unit in a laser processing apparatus according to an embodiment. [Figure 12]FIG. 12 is a diagram showing an example of a three-dimensional information image in a normal mode, a gap defect mode, a through defect mode, and a different material through defect mode. [Figure 13] FIG. 13 is a diagram for explaining a specific process of a discrimination unit in a laser processing apparatus according to an embodiment. [Figure 14] FIG. 14 is a diagram for explaining another specific process of a discrimination unit in a laser processing apparatus according to an embodiment.

MODE FOR CARRYING OUT THE INVENTION

[0012] (Background of the Invention) Prior to specifically describing embodiments of the present disclosure, the background of the invention of the present disclosure will be described.

[0013] [[ID=二十]]There is known a laser processing apparatus that can measure the depth of a keyhole generated on the processing surface of a processing object during laser processing using OCT technology. In this case, as shown in FIG. 1, by irradiating the processing object 2 with the measurement light ML together with the processing laser light (not shown), the optical path difference between the reflected light of the measurement light ML reflected from the bottom surface of the keyhole 2b formed on the processing surface 2a of the processing object 2 and the reference light is detected. Then, as shown in FIG. 2, by performing a fast Fourier transform (FFT) on the optical interference signal, the depth of the keyhole 2b based on the reflected light of the measurement light ML can be calculated. Further, as shown in FIG. 3, by continuously calculating the depth of the keyhole 2b, data on the depth of the keyhole 2b corresponding to the distance (position) on the processing surface 2a of the processing object 2 can be obtained. FIG. 4 shows an example of a profile of the actually obtained depth of the keyhole 2b.

[0014] Thus, in the conventional laser processing apparatus, the depth of the keyhole 2b can be calculated from the reflected light of the measurement light ML. However, in the conventional laser processing apparatus, it is not possible to determine whether the keyhole 2b is formed at the intended depth, and it is not possible to determine whether the laser processing is being performed normally on the workpiece 2. That is, in the conventional laser processing apparatus, even if a defect occurs during laser processing, the type of the defect in the laser processing cannot be specified, and the defect mode in the laser processing cannot be discriminated.

[0015] For example, as shown in FIG. 5, there are a plurality of defect modes in laser processing, such as a gap defect mode, a penetration defect mode, and a dissimilar material penetration defect mode. However, in the conventional laser processing apparatus, these defect modes cannot be discriminated.

[0016] As shown in FIG. 5(a), the gap defect mode is a defect mode in a state where a gap occurs between two workpieces 2 when the workpiece 2 is laser processed with two stacked. For example, when a metal member (such as a metal plate) is used as the workpiece 2, this is a case where a gap occurs between two stacked metal members when laser welding the two stacked metal members with a processing laser beam. In the gap defect mode, the measurement light ML penetrating the upper metal member has a branched reflection point in the gap, and the amount of light of the reflected light of the measurement light ML decreases.

[0017] As shown in FIG. 6, in the gap defect mode, as a result of a gap occurring between the two workpieces 2, the depth of the keyhole 2b will be obtained as a value increased from the intended depth. For example, in the gap defect mode, the result shown in FIG. 7(a) is obtained. As shown in FIG. 7(a), it can be seen that the depth of the keyhole 2b has increased compared to the result in the normal case of FIG. 4.

[0018] However, when the depth of the keyhole 2b is found to be greater than the intended depth, it is unclear whether the increase in keyhole 2b depth is due to a gap forming between the two workpieces 2, or due to an increase in the output of the processing laser beam. If the increase is due to an increase in the output of the processing laser beam, then the processing strength (welding strength, etc.) of the workpiece 2 is ensured, and therefore it is not a defect. However, if a gap forms between the two workpieces 2, then it is a defect. For this reason, when the depth of the keyhole 2b is found to be greater than the intended depth, it may be necessary to treat all cases as defects, resulting in a narrow likelihood threshold.

[0019] Furthermore, as shown in Figure 5(b), the penetration failure mode is a failure mode in which the measuring light ML penetrates the workpiece 2. For example, when a metal member (such as a metal plate) is used as the workpiece 2, laser welding is supposed to be performed without penetrating the metal member, but the processing laser light and the measuring light ML end up penetrating the metal member. Even if the measuring light ML penetrates the workpiece 2, a portion of the measuring light ML is reflected from the inner surface of the through hole in the workpiece 2 or from the bottom surface of the workpiece 2, and the depth of the keyhole 2b is calculated from the reflected light. However, in the penetration failure mode, the amount of reflected light from the measuring light ML is low.

[0020] In the through-hole failure mode, the depth of the keyhole 2b becomes approximately the same as the thickness of the workpiece 2. Therefore, in the through-hole failure mode, as shown in Figure 6, the depth of the keyhole 2b is approximately the same as in the normal state. For example, in the through-hole failure mode, the result shown in Figure 7(b) is obtained. As shown in Figure 7(b), it can be seen that the depth is similar to that of the normal state in Figure 4.

[0021] However, when the result obtained is that the depth of keyhole 2b is approximately the same as the thickness of workpiece 2, it is unclear whether the depth of keyhole 2b became the same as the thickness of workpiece 2 because the measuring light ML penetrated workpiece 2, or because exactly the same amount of material as the thickness of workpiece 2 was machined. If exactly the same amount of material as the thickness of workpiece 2 was machined, it is not a defect, but if the measuring light ML penetrated workpiece 2, it is a defect. For this reason, when the result obtained is that the depth of keyhole 2b is no different from the thickness of workpiece 2, it may be necessary to treat all of them as defects, resulting in a narrow likelihood of the threshold.

[0022] Furthermore, as shown in Figure 5(c), the dissimilar material penetration failure mode is a failure mode in which the measurement light ML penetrates from one of the two dissimilar materials to the other when the workpiece 2 contains two different dissimilar materials. For example, when laser welding two metal members (such as metal plates) made of different metal materials as the workpiece 2 by scanning the processing laser beam, if the processing area contains a part where only one of the two metal members is laser-welded (same material part) and a part where both of the two metal members are laser-welded (dissimilar material part), the processing laser beam penetrates from one of the two metal members to the other in the dissimilar material part. In the dissimilar material penetration failure mode, if the absorption characteristics of the measurement light ML differ for the two dissimilar materials, the amount of reflected light from the measurement light ML changes. For example, if the reflectance of the lower material of the two dissimilar materials is greater than the reflectance of the upper material of the two dissimilar materials, the amount of reflected light from the measurement light ML increases in the dissimilar material part.

[0023] As shown in Figure 6, in the dissimilar material penetration failure mode, the same results are obtained as when the depth of the keyhole 2b is normal. In other words, the calculated depth value does not change between the same material portion and the dissimilar material portion, and the depth of the same material portion and the depth of the dissimilar material portion are the same. For example, in the dissimilar material penetration failure mode, the result shown in Figure 7(c) is obtained. As shown in Figure 7(c), the boundary between the same material portion and the dissimilar material portion is not discernible, and the depth of the same material portion and the depth of the dissimilar material portion are the same.

[0024] However, when the depth of the same material portion and the depth of the dissimilar material portion are found to be the same, it is unclear whether the measuring light ML penetrated from one of the two dissimilar materials to the other in the dissimilar material portion. In this case, the same material portion is not defective, but if the light penetrated from one of the two dissimilar materials to the other in the dissimilar material portion, it would be considered defective. Thus, until now, when the processing area of ​​workpiece 2 included both same material and dissimilar material portions, it was not possible to determine whether the light penetrated from one of the two dissimilar materials to the other in the dissimilar material portion, resulting in a narrow likelihood of the threshold.

[0025] Thus, with conventional laser processing equipment, even if a defect occurs during laser processing, it is not possible to identify the type of defect in the laser processing, and therefore it is not possible to determine the defect mode in the laser processing.

[0026] As a result of diligent research into these issues, the inventors of this invention have discovered a technique that can identify defect modes in laser processing by using the intensity of the reflected light of the measurement light ML.

[0027] The embodiments of this disclosure will be described below with reference to the drawings. The embodiments described below are all specific examples of this disclosure. Therefore, the numerical values, components, their arrangement and connection configurations, as well as the steps and their order, shown in the following embodiments are examples and are not intended to limit this disclosure. Accordingly, any components in the following embodiments that are not described in an independent claim will be described as optional components.

[0028] Furthermore, each figure is a schematic diagram and not necessarily a strictly accurate representation. In addition, the same reference numerals are used for substantially identical components in each figure, and redundant explanations are omitted or simplified. Moreover, in this specification, the terms "up" and "down" do not necessarily refer to the upward direction (vertically upward) and the downward direction (vertically downward) in absolute spatial perception.

[0029] Furthermore, in this specification and drawings, the X, Y, and Z axes represent the three axes of a three-dimensional Cartesian coordinate system. The X and Y axes are mutually orthogonal and both are orthogonal to the Z axis. In this embodiment, the Z axis direction is the vertical direction.

[0030] (Embodiment) First, the configuration of the laser processing apparatus 1 according to the embodiment will be explained using Figure 8. Figure 8 is a diagram showing the configuration of the laser processing apparatus 1 according to the embodiment.

[0031] The laser processing apparatus 1 is a device that performs laser processing on a workpiece 2 (object to be processed) by irradiating it with a processing laser beam PL. For example, laser processing is laser welding. In this case, the laser processing apparatus 1 is a laser welding machine, and for example, a metal member such as a metal plate is used as the workpiece 2, and laser welding is performed on the metal member by irradiating it with a processing laser beam PL.

[0032] Furthermore, the laser processing device 1 can not only perform laser processing on the workpiece 2, but can also measure the depth of the keyhole 2b that occurs on the processed surface 2a of the workpiece 2 during laser processing using OCT technology. Specifically, the laser processing device 1 irradiates the keyhole 2b with a measurement light ML and measures the depth of the keyhole 2b using the reflected light of the measurement light ML reflected from the keyhole 2b.

[0033] Furthermore, the laser processing apparatus 1 also has a function to determine the defective mode of the laser processing when irradiating the workpiece 2 with processing laser light PL and performing laser processing on the workpiece 2. Therefore, the laser processing apparatus 1 shown in Figure 8 may be configured as a defective mode determination system.

[0034] As shown in Figure 8, the laser processing apparatus 1 includes a processing head 10 that irradiates the workpiece 2 with processing laser light PL, a first laser light source 21 that emits processing laser light PL, a second laser light source 22 that emits measurement light ML, an optical interferometer 30, a detector 40, a processing unit 50, a control unit 60, a first drive unit 71, and a second drive unit 72.

[0035] The processing head 10 irradiates the workpiece 2 with processing laser light PL emitted from the first laser light source 21. The processing head 10 also irradiates the workpiece 2 with measurement light ML emitted from the second laser light source 22.

[0036] The machining head 10 includes a first mirror 11, a second mirror 12, a dichroic mirror 13, a collimating lens 14, and a lens 15.

[0037] The first mirror 11 and the second mirror 12 each have a reflector with a reflective surface that specularly reflects incident light. The first mirror 11 and the second mirror 12 are movable mirrors that can be rotated on two or more axes. The first mirror 11 and the second mirror 12 are, for example, galvanometer mirrors.

[0038] The first mirror 11 is connected to the control unit 60 via the first drive unit 71. The first drive unit 71 operates the first mirror 11 based on instructions from the control unit 60. The second mirror 12 is connected to the control unit 60 via the second drive unit 72. The second drive unit 72 operates the second mirror 12 based on instructions from the control unit 60.

[0039] In this embodiment, the two axes on which the first mirror 11 and the second mirror 12 rotate are the X-axis and the Y-axis. Figure 8 shows only the rotational motion of the first mirror 11 and the second mirror 12 with the Y-axis as the axis of rotation. Note that the first mirror 11 and the second mirror 12 may each be configured to rotate on two or more axes.

[0040] The dichroic mirror 13 has the characteristic of transmitting light in the first wavelength range and reflecting light in the second wavelength range which is different from the first wavelength range. In this embodiment, the dichroic mirror 13 transmits the processing laser light PL and reflects the measurement light ML.

[0041] The collimating lens 14 focuses the light incident on it so that it becomes parallel light. In this embodiment, the measurement light ML emitted from the optical interferometer 30 is incident on the collimating lens 14. Therefore, the collimating lens 14 makes the measurement light ML emitted from the optical interferometer 30 into parallel light. The collimating lens 14 is also positioned in the optical path between the optical interferometer 30 and the second mirror 12. Therefore, the measurement light ML, which has become parallel light due to the collimating lens 14, is incident on the second mirror 12.

[0042] Lens 15 is a projection lens that projects the processing laser beam PL and the measurement beam ML onto the workpiece 2. Lens 15 focuses the light incident on it and emits it. Therefore, lens 15 focuses the processing laser beam PL and the measurement beam ML and irradiates the workpiece 2 with them. Specifically, lens 15 focuses the processing laser beam PL and the measurement beam ML onto the processing point 3 of the workpiece 2. As an example, lens 15 is an fθ lens.

[0043] The first laser light source 21 is a laser emission device that emits processing laser light PL for laser processing the workpiece 2. In this embodiment, the first laser light source 21 is a laser oscillator that generates and emits processing laser light PL. As an example, the first laser light source 21 oscillates laser light in single mode.

[0044] The processing laser beam PL emitted from the first laser light source 21 is input to the processing head 10 via the first inlet 16. The first inlet 16 is located in the processing head 10 at a position that allows the processing laser beam PL to be introduced into the dichroic mirror 13.

[0045] The second laser light source 22 is a laser emission device that emits measurement light ML for measuring the workpiece 2. In this embodiment, the measurement light ML is laser light, and the second laser light source 22 is a laser oscillator that generates and emits the measurement light ML, which is laser light. As an example, the second laser light source 22 oscillates laser light in single mode.

[0046] The measurement light ML emitted from the second laser light source 22 is input to the processing head 10 via the optical interferometer 30 and the second inlet 17. The second inlet 17 is positioned in the processing head 10 to allow the measurement light ML to be introduced into the collimating lens 14.

[0047] In this embodiment, both the processing laser light PL emitted from the first laser light source 21 and the measurement light ML emitted from the second laser light source 22 are infrared light. Furthermore, the peak wavelength of the processing laser light PL and the peak wavelength of the measurement light ML are different. For example, the peak wavelength of the processing laser light PL emitted from the first laser light source 21 is 1064 nm, and the peak wavelength of the measurement light ML emitted from the second laser light source 22 is 1300 nm.

[0048] Furthermore, the peak wavelengths of the processing laser light PL and the measurement light ML are not limited to those specified. The peak wavelength of the processing laser light PL may be any wavelength capable of processing the workpiece 2, and the peak wavelength of the measurement light ML may be any wavelength capable of measuring the keyhole 2b generated in the workpiece 2. In addition, the processing laser light PL and the measurement light ML are not limited to infrared light.

[0049] The optical interferometer 30 generates an optical interference signal based on the measurement light ML. In this embodiment, the optical interferometer 30 uses optical coherence tomography (OCT) to generate an optical interference signal based on the optical path difference between the reflected light of the measurement light ML reflected from the workpiece 2 and the reference light. Specifically, the optical interferometer 30 generates the optical interference signal by wavelength-swept optical coherence tomography (SS-OCT). Therefore, the second laser light source 22 that emits the measurement light ML is a wavelength-swept light source, and emits the measurement light ML by sweeping the wavelength in time.

[0050] The optical interferometer 30 includes a beam splitter 31 and a reference mirror 32. The beam splitter 31 reflects a portion of the measurement light ML emitted from the second laser light source 22 and transmits the other portion of the measurement light ML emitted from the second laser light source 22. The measurement light ML reflected by the beam splitter 31 is reflected by the reference mirror 32 and travels toward the detector 40 as reference light. Meanwhile, the measurement light ML that has passed through the beam splitter 31 is irradiated onto the workpiece 2 via the processing head 10 and reflected by the workpiece 2. The reflected light of the measurement light ML reflected by the workpiece 2 returns to the optical interferometer 30 via the processing head 10, is reflected by the beam splitter 31 and travels toward the detector 40. As a result, an optical interference signal is generated based on the optical path difference between the reflected light of the measurement light ML that was reflected by the workpiece 2 and the reference light of the measurement light ML that was not reflected by the workpiece 2.

[0051] The optical interference signal generated by the optical interferometer 30 is detected by the detector 40. In other words, the detector 40 detects the optical interference signal generated based on the optical path difference between the reflected light of the measurement light ML and the reference light of the measurement light ML. The detector 40 is, for example, a photodetector such as a photodiode.

[0052] The processing unit 50 processes the optical interference signal detected by the detector 40. The processing unit 50 includes a measurement unit 51, an image generation unit 52, and a discrimination unit 53.

[0053] The measurement unit 51 measures the depth of the keyhole 2b formed in the workpiece 2 during laser processing based on the optical interference signal detected by the detector 40. Specifically, as shown in Figure 9, the measurement unit 51 calculates the depth of the keyhole 2b by applying a Fast Fourier Transform (FFT) to the optical interference signal detected by the detector 40. For example, the horizontal axis of the signal waveform shown in Figure 9 corresponds to the depth of the keyhole 2b.

[0054] In this embodiment, the measurement light ML is emitted after being swept over time. Therefore, the measurement unit 51 applies a fast Fourier transform to the optical interference signal generated by the measurement light ML that is emitted after being swept over time, thereby obtaining point cloud data of the depths of multiple keyholes 2b corresponding to the distance (position) on the machined surface 2a of the workpiece 2, as shown in Figure 10.

[0055] Figure 10 shows an example of a keyhole 2b depth profile obtained in this manner. As shown in Figure 10, the keyhole 2b depth profile is image data, and the depth of the keyhole 2b, calculated from the optical interference signal generated by the reflected light of the measurement light ML that is swept over time and emitted, is plotted as a cloud of countless points. In other words, the depths of multiple keyholes 2b are obtained as a two-dimensional information image distributed in a two-dimensional Cartesian coordinate system with the depth of the keyhole 2b as the first axis and the distance (position) on the machined surface 2a as the second axis. The two-dimensional information image shown in Figure 10 is generated by the optical interference signal detected by the detector 40. This two-dimensional information image may be generated by the measurement unit 51, the image generation unit 52, or another image generation unit in the processing unit 50.

[0056] The image generation unit 52 generates a three-dimensional information image based on the reflected light of the measurement light ML reflected from the workpiece 2. Specifically, the image generation unit 52 generates a three-dimensional information image based on the two-dimensional information image obtained by the measurement unit 51. In other words, the three-dimensional information image is generated from the optical interference signal used when calculating the two-dimensional information image.

[0057] Figure 11 shows an example of a three-dimensional information image generated by the image generation unit 52. As shown in Figure 11, the three-dimensional information image is an image that shows the distribution of the depth of the keyhole 2b formed in the workpiece when the processing laser light PL is irradiated, the intensity of the reflected light of the measurement light ML reflected from the workpiece 2, and the frequency of occurrence of the depth of the keyhole 2b in relation to the intensity of the reflected light of the measurement light ML. In other words, the three-dimensional information image generated by the image generation unit 52 is data that shows the distribution of the depth of the keyhole 2b, the intensity of the reflected light of the measurement light ML, and the frequency of occurrence of the depth of the keyhole 2b in relation to the intensity of the reflected light of the measurement light ML in a three-dimensional orthogonal coordinate system, in which the depth of the keyhole 2b is the first axis and the intensity of the reflected light of the measurement light ML is the second axis, with the frequency of occurrence of the depth of the keyhole 2b in relation to the intensity of the reflected light of the measurement light ML added as the third axis to a two-dimensional orthogonal coordinate system in which the depth of the keyhole 2b is the first axis and the intensity of the reflected light of the measurement light ML is the second axis. In other words, the depth of the keyhole 2b and the intensity of the reflected light of the measurement light ML are expressed in a planar coordinate system, and the frequency of occurrence of the keyhole 2b depth in relation to the intensity of the reflected light of the measurement light ML is expressed as information in the height direction of that planar coordinate system. Note that the frequency of occurrence of the keyhole 2b depth in relation to the intensity of the reflected light of the measurement light ML represents the ratio of the keyhole 2b depth that occurs for a given intensity of reflected light of the measurement light ML.

[0058] In Figure 11, the intensity of the reflected light of the measurement light ML can be calculated from the signal waveform obtained by applying a fast Fourier transform to the optical interference signal detected by the detector 40. Specifically, the vertical axis of the signal waveform shown in Figure 9 corresponds to the intensity of the reflected light of the measurement light ML.

[0059] The discrimination unit 53 determines whether laser processing is being performed in normal mode or in defective mode based on the three-dimensional information image generated by the image generation unit 52. In other words, the discrimination unit 53 determines whether laser processing is being performed normally or not based on the three-dimensional information image generated by the image generation unit 52.

[0060] Furthermore, the discrimination unit 53 determines the defect mode of the laser processing based on the three-dimensional information image generated by the image generation unit 52. Specifically, the discrimination unit 53 identifies one defect mode from among multiple defect modes based on the three-dimensional information image generated by the image generation unit 52. In this embodiment, the multiple defect modes include a gap defect mode, a penetration defect mode, and a dissimilar material penetration defect mode, as shown in Figure 5. Therefore, the discrimination unit 53 can determine which of the gap defect mode, penetration defect mode, and dissimilar material penetration defect mode occurred during laser processing based on the three-dimensional information image generated by the image generation unit 52.

[0061] In this case, as shown in Figure 12, the three-dimensional information images for the normal mode, the gap defect mode, the penetration defect mode, and the dissimilar material penetration defect mode have different characteristics from each other.

[0062] Figure 12 shows examples of three-dimensional information images in normal mode, gap defect mode, penetration defect mode, and dissimilar material penetration defect mode. In Figure 12, in normal mode, gap defect mode, and penetration defect mode, laser welding is performed using an aluminum metal plate as the workpiece 2, while in dissimilar material penetration defect mode, laser welding is performed with a copper metal plate placed underneath an aluminum metal plate.

[0063] Figure 12(a) shows a three-dimensional information image in normal mode. In other words, it is a three-dimensional information image when laser welding is performed normally. As shown in Figure 12(a), in the three-dimensional information image in normal mode, the peaks in the frequency of occurrence of keyhole 2b depth appear concentrated in one place. For example, in the three-dimensional information image in normal mode, there is one peak in the frequency of occurrence of keyhole 2b depth. Note that the three-dimensional information image in normal mode in Figure 12(a) is the same as the three-dimensional information image shown in Figure 11 and was generated based on the two-dimensional information image shown in Figure 10.

[0064] Figure 12(b) shows a three-dimensional information image of the gap defect mode. As can be seen by comparing Figures 12(a) and (b), in the three-dimensional information image of the gap defect mode, the peak of the occurrence frequency of the keyhole 2b depth is divided into two locations, compared to the three-dimensional information image of the normal mode. For example, in the three-dimensional information image of the gap defect mode, the occurrence frequency of the keyhole 2b depth appears as two separate peaks. Note that the three-dimensional information image in Figure 12(b) was generated based on the two-dimensional information image shown in Figure 7(a).

[0065] Figure 12(c) shows a three-dimensional information image of the penetration failure mode. As can be seen by comparing Figures 12(a) and (c), the peak position of the occurrence frequency of keyhole 2b depth has shifted in the three-dimensional information image of the penetration failure mode compared to the three-dimensional information image of the normal mode. Specifically, in Figure 12(c), the peak position of the occurrence frequency of keyhole 2b depth has shifted to the lower left compared to the three-dimensional information image of the normal mode. Note that the three-dimensional information image in Figure 12(c) was generated based on the two-dimensional information image shown in Figure 7(b).

[0066] Figure 12(d) shows a three-dimensional information image of the dissimilar material penetration failure mode. As can be seen by comparing Figures 12(a) and (d), in the three-dimensional information image of the dissimilar material penetration failure mode, the peak of the occurrence frequency of keyhole 2b depth appears as a linear extension in the horizontal direction, compared to the three-dimensional information image of the normal mode. Specifically, since copper has a higher reflectivity than aluminum, in Figure 12(d), the processing laser beam PL penetrates the metal plate made of aluminum and reaches the metal plate made of copper. As a result, compared to the three-dimensional information image of the normal mode, the peak of the occurrence frequency of keyhole 2b depth appears as a linear extension to the right. Note that the three-dimensional information image in Figure 12(d) is generated based on the two-dimensional information image shown in Figure 7(c).

[0067] As described above, the characteristics of the three-dimensional information image in normal mode and the three-dimensional information image in defective mode are different. Therefore, the discrimination unit 53 can easily determine whether the laser processing is being performed normally or not by identifying the three-dimensional information image generated by the image generation unit 52.

[0068] Furthermore, since the characteristics of the three-dimensional information image differ for the gap defect mode, the penetration defect mode, and the dissimilar material penetration defect mode, the discrimination unit 53 can easily identify whether the defect occurring during laser processing is a gap defect mode, a penetration defect mode, or a dissimilar material penetration defect mode by identifying the three-dimensional information image generated by the image generation unit 52.

[0069] The control unit 60 controls the on / off switching of the first laser light source 21. In other words, it emits and stops the processing laser beam PL. This allows the processing laser beam PL to laser process any position on the processing surface 2a of the workpiece 2 in any pattern. The control unit 60 also controls the on / off switching of the second laser light source 22. In other words, it emits and stops the measurement light ML. For example, the on / off switching of the first laser light source 21 and the on / off switching of the second laser light source 22 are the same. In this case, when the processing laser beam PL is irradiated onto the workpiece 2, the measurement light ML is also irradiated onto the workpiece 2.

[0070] Next, the operation of the laser processing apparatus 1 will be explained with reference to Figure 8.

[0071] As shown in Figure 8, the processing laser beam PL emitted from the first laser light source 21 is input to the processing head 10. In the processing head 10, the processing laser beam PL passes through the dichroic mirror 13 and is reflected by the first mirror 11. The processing laser beam PL reflected by the first mirror 11 passes through the lens 15 and is focused onto the processing surface 2a, which is the surface of the workpiece 2. As a result, the processing point 3 of the workpiece 2 is laser processed by the processing laser beam PL. At this time, the processing point 3 irradiated by the processing laser beam PL melts, forming a molten pool 3a in the workpiece 2. In addition, molten metal evaporates from the molten pool 3a, and the pressure of the vapor generated during evaporation forms a keyhole 2b in the workpiece 2.

[0072] In this embodiment, the first mirror 11 is a galvanometer mirror and the lens 15 is an fθ lens, so the first mirror 11 and the lens 15 constitute an optical scanning system. Therefore, by rotating the first mirror 11 by a predetermined angle from its origin position, the arrival position of the processing laser beam PL on the processing surface 2a can be controlled. This allows the processing laser beam PL to be scanned and irradiated onto any processing point 3 on the processing surface 2a. The amount of movement of the first mirror 11 to irradiate the processing laser beam PL onto the desired processing point 3 (i.e., the angle by which the first mirror 11 is rotated from its origin position) can be uniquely set if the positional relationship of each optical component constituting the processing head 10 and the distance from the lens 15 to the processing surface 2a are determined. This allows the processing laser beam PL to be irradiated onto the desired processing point 3.

[0073] In this case, the distance from the lens 15 to the processing surface 2a should be set to a distance where the focal point where the processing laser beam PL is most focused coincides with the processing surface 2a. This allows for the most efficient processing of the workpiece 2 using the processing laser beam PL. Note that the distance from the lens 15 to the processing surface 2a is not limited to this, and can be determined to any appropriate distance depending on the application of the processing.

[0074] Meanwhile, the measurement light ML (measurement laser light) emitted from the second laser light source 22 is input to the optical interferometer 30. In the optical interferometer 30, a portion of the measurement light ML is reflected by the beam splitter 31 to become reference light, and another portion passes through the beam splitter 31 and is input to the processing head 10. The measurement light ML input to the processing head 10 is converted into parallel light by the collimating lens 14, reflected by the second mirror 12, reflected again by the dichroic mirror 13, reflected by the first mirror 11, passes through the lens 15, and is irradiated onto the workpiece 2. In this case, the measurement light ML is focused by the lens 15 onto the processing point 3 on the processing surface 2a of the workpiece 2.

[0075] The measurement light ML irradiated onto the workpiece 2 is reflected by the workpiece 2 and travels back along its propagation path to reach the optical interferometer 30. Specifically, the measurement light ML irradiated onto the workpiece 2 is reflected from the bottom surface of the keyhole 2b formed in the workpiece 2, passes through the lens 15 in the machining head 10, is reflected by the first mirror 11, is reflected by the dichroic mirror 13, is reflected by the second mirror 12, passes through the collimating lens 14 and is input to the optical interferometer 30. In this way, the reflected light of the measurement light ML reflected by the workpiece 2 is input to the optical interferometer 30.

[0076] The reflected light of the measurement light ML input to the optical interferometer 30 is reflected by the beam splitter 31 within the optical interferometer 30 and travels toward the detector 40. As a result, an optical interference signal is generated based on the optical path difference between the reflected light of the measurement light ML reflected by the workpiece 2 and the reference light of the measurement light ML reflected by the reference mirror 32. This optical interference signal is detected by the detector 40.

[0077] The optical interference signal detected by the detector 40 is input to the processing unit 50. The measurement unit 51 of the processing unit 50 measures the depth of the keyhole 2b (i.e., the penetration depth of the machining point 3) based on the optical interference signal detected by the detector 40. The penetration depth is the distance between the highest point of the melted portion of the workpiece 2 and the machined surface 2a.

[0078] In this embodiment, the laser processing apparatus 1 is configured such that the discriminant mode of the laser processing is determined by the discriminant unit 53 of the processing unit 50. Specifically, the discriminant unit 53 determines the discriminant mode of the laser processing based on a three-dimensional information image generated from the reflected light of the measurement light ML reflected from the workpiece 2.

[0079] This makes it possible to identify the defect mode in laser processing. Specifically, the discrimination unit 53 determines, based on the three-dimensional information image generated by the image generation unit 52, whether a gap defect mode, a penetration defect mode, or a dissimilar material penetration defect mode is occurring during laser processing.

[0080] In this case, the discrimination unit 53 can determine the defect mode using, for example, a machine learning model. Specifically, as shown in Figure 13, the discrimination unit 53 has a machine learning model that has been pre-generated by machine learning, and when a three-dimensional information image generated from the reflected light of the measurement light ML is input to this machine learning model during laser processing, the defect mode of the laser processing can be determined.

[0081] The machine learning model in the discrimination unit 53 is pre-generated by machine learning based on training data obtained by adding annotations indicating a defective mode to each of a plurality of previously acquired three-dimensional information images. The machine learning model is stored, for example, in a storage medium such as a memory mounted on the laser processing apparatus 1.

[0082] Furthermore, as shown in Figure 13, the discrimination unit 53 may output the type of defect mode and calculate numerical information related to that defect mode. For example, if the identified defect mode is a gap defect mode, the discrimination unit 53 may output the gap between the two workpieces 2 as numerical information. Also, if the identified defect mode is a penetration defect mode, the discrimination unit 53 may calculate the degree of penetration, which indicates the degree to which the measuring light ML penetrated the workpiece 2, as numerical information. Also, if the identified defect mode is a dissimilar material penetration defect mode, the discrimination unit 53 may calculate the degree of dissimilar material penetration, which indicates the degree to which the measuring light ML penetrated two dissimilar materials, as numerical information.

[0083] (modified version) Although the technology of this disclosure has been described above based on embodiments, this disclosure is not limited to the embodiments described above.

[0084] For example, in the above embodiment, a machine learning model was used to determine the defect mode using a three-dimensional information image, but this is not limited to this. Specifically, the discrimination unit 53 may determine the defect mode by extracting features from the three-dimensional information image. In this case, as shown in Figure 14, the discrimination unit 53 has a feature extraction unit that extracts features from the three-dimensional information image generated by the image generation unit 52. In this case as well, as described above, the discrimination unit 53 may output the type of defect mode and calculate numerical information related to that defect mode.

[0085] Furthermore, in the above embodiment, a two-dimensional information image shown in Figure 10 is first generated based on the optical interference signal obtained from the measurement light reflected from the workpiece 2, and the image generation unit 52 generates a three-dimensional information image from that two-dimensional information image. However, the system is not limited to this. For example, the image generation unit 52 may directly generate a three-dimensional information image from the optical interference signal obtained from the measurement light reflected from the workpiece 2. In other words, the three-dimensional information image only needs to be generated based on the reflected light of the measurement light reflected from the workpiece 2.

[0086] Furthermore, in the above embodiment, the defect mode was determined based on the three-dimensional information image generated by the image generation unit 52, but this is not limited to this. For example, the defect mode may be determined from the optical interference signal based on the peak frequency of occurrence of the keyhole depth 2b and / or the distribution information of that peak.

[0087] Furthermore, in the above embodiment, the processing described as the operation of the functional parts of the laser processing apparatus 1, such as the processing unit 50 and the control unit 60, can be executed by a computer. For example, a computer executes each of the above processes by running a program using hardware resources such as a processor (CPU), memory, and input / output circuits. Specifically, the processor executes each process by acquiring the data to be processed from memory or input / output circuits, performing calculations on the data, and outputting the calculation results to memory or input / output circuits. The processor may consist of a single semiconductor chip or may be physically composed of multiple semiconductor chips. When the processor consists of multiple semiconductor chips, each control of each embodiment may be realized by a different semiconductor chip. Also, the processing unit 50 and the control unit 60 may be composed of circuits. These circuits may constitute a single circuit as a whole, or they may be separate circuits. Furthermore, these circuits may each be a general-purpose circuit or a dedicated circuit.

[0088] Furthermore, the technology of this disclosure can be implemented as a laser processing method or a defect mode determination method. For example, the laser processing method according to this disclosure is a laser processing method that determines a defect mode of laser processing when laser processing a workpiece 2 by irradiating the workpiece 2 with a processing laser beam PL, and includes the step of determining the defect mode of laser processing based on a three-dimensional information image generated based on the reflected light of a measurement beam ML reflected from the workpiece 2. Furthermore, the defect mode determination method according to this disclosure is a defect mode determination method that determines a defect mode of laser processing when laser processing a workpiece 2 by irradiating the workpiece 2 with a processing laser beam PL, and includes the step of determining the defect mode of laser processing based on a three-dimensional information image generated based on the reflected light of a measurement beam ML reflected from the workpiece 2.

[0089] Furthermore, the laser processing method or defect mode determination method in the above embodiment may be implemented as a computer program implemented by a computer, or as a computer-readable recording medium storing the program. For example, the present disclosure may be a program that causes a computer to execute the laser processing method or defect mode determination method.

[0090] Furthermore, this disclosure also includes forms obtained by applying various modifications to the above embodiments that a person skilled in the art could conceive, and forms realized by arbitrarily combining the components and functions of the embodiments without departing from the spirit of this disclosure. In addition, this disclosure also includes any combination of two or more claims from the multiple claims described in the claims at the time of filing this application, provided that they are not technically contradictory. For example, if the cited claims described in the claims at the time of filing this application are made into a multi-claim or multi-multi-claim so as to refer to all of the higher-level claims without technically contradictory, then all combinations of claims included in that multi-claim or multi-multi-claim are also included in this disclosure. [Industrial applicability]

[0091] The technology disclosed herein is useful as a laser processing apparatus for performing laser processing such as laser welding. [Explanation of Symbols]

[0092] 1. Laser processing device 2. Object to be processed 2a Machining surface 2b Keyhole 3 processing points 3a Melting pool 10 Machining heads 11 First Mirror 12 Second Mirror 13 Dichroic Mirror 14. Collimating lenses 15 lenses 16. First Inlet 17 Second Inlet 21. First laser light source 22. Second laser light source 30 Optical interferometer 31 Beam Splitter 32 Reference Mirror 40 detectors 50 Processing Unit 51 Measurement Unit 52 Image Generation Unit 53 Discrimination part 60 Control Unit 71 First drive unit 72 Second drive unit Laser light for PL processing ML measurement light

Claims

1. A laser processing apparatus that performs laser processing on a workpiece by irradiating the workpiece with processing laser light, A laser light source that emits the aforementioned processing laser light, The system includes a discrimination unit that determines the defect mode of the laser processing based on a three-dimensional information image generated from the reflected light of the measurement light reflected from the workpiece, The three-dimensional information image is an image showing the distribution of the depth of the keyhole formed in the workpiece when the processing laser light is irradiated, the intensity of the reflected light, and the frequency of occurrence of the keyhole depth at the intensity of the reflected light. Laser processing equipment.

2. The system includes an optical interferometer that generates an optical interference signal based on the optical path difference between the reflected light and the reference light, The three-dimensional information image is generated from the optical interference signal. The laser processing apparatus according to claim 1.

3. The optical interferometer generates the optical interference signal by wavelength-sweep type optical coherence tomography. The laser processing apparatus according to claim 2.

4. The discrimination unit determines the defect mode of the laser processing by inputting the three-dimensional information image generated from the reflected light during the laser processing into a machine learning model that was previously generated by machine learning based on training data obtained by adding annotations indicating the defect mode to each of the multiple three-dimensional information images acquired in advance. A laser processing apparatus according to any one of claims 1 to 3.

5. The discrimination unit identifies one failure mode from among the multiple failure modes, The multiple defect modes include a gap defect mode in which a gap occurs between two workpieces when two workpieces are stacked and laser processed, a penetration defect mode in which the measuring light penetrates the workpiece, and a dissimilar material penetration defect mode in which the measuring light penetrates from one of the two dissimilar materials to the other when the workpiece includes two different dissimilar materials. A laser processing apparatus according to any one of claims 1 to 3.

6. The aforementioned discrimination unit is If the identified defect mode is the gap defect mode, the gap between the two workpieces is output. If the identified defect mode is the penetration defect mode, the degree of penetration, which indicates the degree to which the measuring light penetrated the workpiece, is calculated. If the identified defect mode is the dissimilar material penetration defect mode, the degree of dissimilar material penetration, which indicates the degree to which the measurement light penetrated the two dissimilar materials, is calculated. The laser processing apparatus according to claim 5.

7. A defect mode determination system for determining a defect mode of laser processing when irradiating a workpiece with processing laser light and performing laser processing on the workpiece, The system includes a discrimination unit that determines the defect mode of the laser processing based on a three-dimensional information image generated from the reflected light of the measurement light reflected from the workpiece, The three-dimensional information image is an image showing the distribution of the depth of the keyhole formed in the workpiece when the processing laser light is irradiated, the intensity of the reflected light, and the frequency of occurrence of the keyhole depth at the intensity of the reflected light. Malfunction mode detection system.

8. A method for determining a defect mode in laser processing when irradiating a workpiece with processing laser light to perform laser processing on the workpiece, The step includes determining a defect mode of the laser processing based on a three-dimensional information image generated from the reflected light of the measurement light reflected from the workpiece, The three-dimensional information image is an image showing the distribution of the depth of the keyhole formed in the workpiece when the processing laser light is irradiated, the intensity of the reflected light, and the frequency of occurrence of the keyhole depth at the intensity of the reflected light. Method for identifying malfunction modes.

Citation Information

Patent Citations

  • Device for measuring the depth of the weld seam in real time

    JP2018501964A