JP4876599B2 - Quality detection method and apparatus for butt welds - Google Patents

Description

  The present invention relates to a method and an apparatus for detecting the quality of a butt weld, and in particular, a method for inspecting the quality of a weld after removing a raised portion caused by welding in a butt welding process such as a flash butt method. And an apparatus.

  Conventionally, in the cold rolling process and pickling process, the steel strip is continuously supplied to the process. Therefore, equipment that connects the preceding coil and the succeeding coil by welding without stopping the line with a looper or the like, specifically, a flash bat In general, a welding machine such as seam welding and spot welding and ancillary devices such as a build-up cutting means are provided. This build-up is that the molten metal protrudes from the cross-section as it is pressed in the longitudinal direction of the plate in the state where the steel strip butt is melted during welding, such as a bite, grindstone, etc. according to the shape of the build immediately after welding. It is common to cut in a planar shape with a cutting tool. By the way, the quality determination of the welding between coils by these welding machines is important for preventing troubles such as the fracture of the welded portion of the coil in the subsequent process. This is because, in general, the connected steel strip also passes through the same lines, pinch rolls, rolling rolls, and the like as the base material, so that the welded portion needs to have the same shape as the base material as much as possible.

  For the quality judgment of welded parts, the steel strip is temporarily stopped after the build-up cutting, and a visual inspection that observes the front and back surfaces of the welded part, a hammer test that confirms the strength of the welded part using a hammer, and an automatic inspection As a method, a method of discriminating based on welding power and welding time as disclosed in Patent Document 1, an ultrasonic probe is embedded in a welding electrode as disclosed in Patent Document 2. Various methods such as a method for determining the quality of a welding state based on the transmission intensity of ultrasonic waves, a method for determining the quality of a welding state based on a surface temperature measured immediately after welding, as disclosed in Patent Document 3. A method has been proposed.

  In addition, as another conventional technique, a plurality of methods for measuring the shape of the welded portion after the swell removal after welding and determining the quality of the welded portion based on the measured shape are proposed. That is, Patent Document 4 proposes a method for determining pass / fail according to the magnitude of the fluctuation component of the differential waveform of the welded portion shape after the swell removal detected by the optical cutting method. In Patent Document 5, a butt weld is proposed. A method has been proposed in which the amount of misalignment of the butt and the remaining bead are calculated and used for pass / fail judgment based on the welded portion shape detected by the two-dimensional distance meter from the front and back surfaces and the thickness information of the preceding and following materials.

JP-A-50-83245 JP-A-52-150760 JP 56-99082 A JP-A-1-209307 JP-A-5-154510

  However, the above visual inspection and hammer test cannot be performed in a welding machine in which welding electrodes are complicated, and it is necessary to stop the steel strip for a while after moving the steel strip to a position where the weld is exposed downstream of the welding machine. As a result, production efficiency is lowered, and the quality inspection is dependent on the subjectivity of the worker, so that there are problems of lack of reproducibility and objectivity.

  Moreover, in the method described in Patent Document 1, it is necessary to grasp in advance the ideal welding energy efficiency for each steel type and thickness of the steel strip to be welded, and to switch the discriminating factor for each of these factors even in actual operation. However, there is a problem that a change in welding efficiency due to contamination and deterioration of a welded part or a welder affects the correlation between welding energy and welded part quality.

  In addition, the method described in Patent Document 2 is costly because an ultrasonic probe is embedded for each electrode, whether the determination is made based on the difference between the time when the transmitted wave is minimized and the time when the energization is completed, Since the influence factor is not only the size of the weld estimated from the ultrasonic attenuation, there is doubt about the reliability of discrimination.

  In addition, the method described in Patent Document 3 is evaluated based on the maximum temperature of the surface and its variation, so that it is not only possible to accurately detect welding defects due to misunderstandings, etc., but also the field of view of the radiation thermometer has deviated from the weld line. There is a problem that it is difficult to detect with high reliability because it is greatly affected by emissivity changes due to errors in the case and the state of scale adhesion on the surface.

  In addition, the method described in Patent Document 4 calculates the welded portion shape after bulge removal by the optical cutting method, so there is an advantage that a wider measurement range can be obtained compared to the temperature method or the like, but the steel detected by the optical cutting method Even if the shape waveform on the surface of the band is a base plate portion, it contains minute fluctuations (high-frequency components) due to the influence of scattering, etc., so the minute fluctuation components are also emphasized by the differential operation, and the S / N for detecting defective welds is reduced. There was a problem that.

  Further, in the method described in Patent Document 5, since the bead residue is defined as the difference between the distance value of the base plate portion and the highest point in the measurement range, cutting that is the most harmful factor for welding failure is the cutting failure. The step of the edge cannot be detected directly, or a deformation such as a bending of the butt portion is erroneously detected as a step, and even if there is a mistake, plate breakage will not occur if welding and cutting are normal There was a case, and there was a problem that the wrong amount could not be the main factor for the quality of the weld. The two-dimensional rangefinder used in this method generally measures the reflectance change of the target surface within the scanning range of the laser light while correcting the auto gain, but the cutting part after bead cutting is generally very strong in the mirror surface state. Therefore, since the reflectance is significantly different from the surrounding base plate, there is a problem in that the shape of the welded portion including the cutting surface cannot be correctly measured with a normal two-dimensional distance meter.

  The present invention has been made to solve the above-described conventional problems, and an object of the present invention is to improve the reliability of the quality determination of the butt weld.

The present invention relates to a method for inspecting the quality of a welded portion after removing a raised portion after butt welding, and scanning images of a plurality of positions by scanning slit light in a direction substantially orthogonal to the weld line in the direction of the weld line. imaged, in each of the image the imaging, a predetermined strength with, calculates the intensity distribution of the reflected light of the respective image, in the intensity distribution of the reflected light to calculate the height of the weld Budan surface based on light-section method The following area is obtained, and the difference between the maximum value and the minimum value of the height of the cross section within a predetermined area set with reference to the end of the area is calculated as the change amount of the cross section , and the change amount Is performed for all the captured images, and from the number of images in which the amount of change is equal to or greater than a predetermined value, the quality of the butt weld is determined. That solves the problem .

The present invention is also a weld quality inspection device that inspects the quality of the weld after removing the bulge after butt welding, and a light source that irradiates slit light in a substantially orthogonal direction with respect to the weld line; imaging means for capturing an image at a plurality of positions by scanning the slit light in the weld line direction, in each of the image the imaging, and calculates the height of the weld Budan surface based on light-section method, wherein each image Image processing means for calculating the intensity distribution of the reflected light, and obtaining an area having a predetermined intensity or less in the intensity distribution of the reflected light, and increasing the height of the cross section within the predetermined area set with reference to the end of the area. calculating the difference between the maximum value and the minimum value of the as the amount of change the cross-sectional shape is performed for all the image variation amount is the imaging of the process of determining whether or not a predetermined value or more, the amount of change Whether the number of images exceeds the specified value Characterized determination means for determining quality of the weld butt, further comprising a, there is provided a quality inspection apparatus for butt welds.

  In the present invention, since the three-dimensional shape of the region including the weld line after removal of the weld swell and the bead removal portion thereof are detected at the same time, the welding point fracture is not an indirect indicator such as welding energy, temperature, or ultrasonic attenuation. This makes it possible to discriminate according to the shape factor such as the uncut portion of the peripheral portion of the bead removal portion that directly affects the bend, and the reliability of the quality determination of the butt welded portion can be improved. In addition, since these inspections are automatically performed by numerical indexes, it is clear that there is no dependency on the inspector, which has been a problem in conventional visual observation and hammering tests, and that the reproducibility and objectivity are excellent.

  The present invention detects a large number of reflected light intensity distributions focusing on the shape distribution in the transverse direction of the object and the reflectance information of the surface at the same time and along the weld line by utilizing the apparatus configuration of the light cutting method, and welding The quality of the welded part is determined based on the three-dimensional shape of the part and the planar image information in the vicinity of the bead.

  In order to realize this, slit light extending in a direction substantially orthogonal to the bead portion immediately after the welding bead cutting (including light that is apparently considered as slit light by high-speed scanning of the laser spot light). , The diffuse reflected light (diffuse reflected light) is imaged as an image, and the position and brightness of the reflected light are detected. In general, the position of the reflected light is detected using a thinning method using binarization, and the brightness of the reflected light is detected by obtaining the maximum brightness in the brightness distribution in a direction substantially orthogonal to the slit light. Is possible.

  However, since the cutting band immediately after bead cutting is in a mirror surface state, the reflected light has a strong specular reflection component, most of which is in the regular reflection direction, the diffuse reflection light component is small, and the amount of light is small. On the other hand, since the base material part (material part) which is not a welded part is in a random surface state, the diffuse reflection component is strong and the amount of diffuse reflected light increases. Therefore, when the slit light is irradiated to the steel sheet in an oblique direction at a predetermined incident angle and observed (imaged) from the diffuse reflection direction with an imaging device such as a camera, the luminance distribution of the reflected image of the slit light is the base material portion ( The material portion is bright and the cutting portion is dark, and in some cases, the position of the reflected light from the bead cutting portion cannot be detected by a general binarization method.

  In such a case, it is preferable to use a thinning process based on a dynamic moment calculation that does not include a binarization process disclosed in Japanese Patent Application Laid-Open No. 2003-322513 filed by the present applicant. .

And, in the brightness data of the reflected light detected as described above, a region where the brightness of the reflected light of the slit light along the transverse direction (substantially orthogonal direction) of the object to be inspected (bead portion) is a bead cutting band, In the shape data obtained from the position data of the reflected light detected as described above, the maximum height of the cross section in the vicinity of the boundary between the cutting band and the base material (predetermined region set based on the calculated boundary position) And the difference between the minimum value and the minimum value is calculated as a change amount (or change rate) of the cross-sectional shape, and it is determined whether the change amount is equal to or greater than a threshold value. As a result, it is possible to accurately detect the presence or absence of a welded part shape defect without being affected by the bending of the butt base material.

  Imaging of the reflected light of the slit light is performed at a plurality of locations along the bead portion direction to obtain a plurality of image data. If these image data are subjected to the above-described arithmetic processing, the number of images that are equal to or greater than the threshold (corresponding to measurement locations) is counted, and if the value is a predetermined number, it is determined that the weld is abnormal. It is possible to determine the overall quality of the butt weld along the weld line. What is necessary is just to employ | adopt the determination method according to the point of interest in the conventional visual inspection, such as the presence or absence of the level | step difference generation | occurrence | production in a bead peripheral part, the ratio of the area | region where the level | step difference has generate | occur | produced, the change of cutting width, and the bending of cutting.

  Embodiments of the present invention will be described below with reference to the drawings.

  FIG. 1 is a schematic view showing an example of the configuration of a quality detection device for a butt weld to which the method according to the present invention is applied.

  In FIG. 1, 1 is a measuring head, 2 is an image processing device, 3 is a display device, and 4 is an object to be measured.

  The measuring head 1 collects a light-cut image of a welded portion. As an example of the configuration, the measuring head 1 includes a light source 10, a light source lens 11, a camera (imaging means) 12, a light receiving lens 13, a mirror 14, and the like. Appropriate measurement conditions such as the wavelength and the incident angle may be applied to those disclosed in Japanese Patent Application Laid-Open No. 2004-117053 filed by the present inventors. When the incident angle α of light from the light source and the angle β of the optical axis of the camera (imaging means), (α + β) is preferably approximately 90 °. Further, as the light source, scan light that scans spot light focused in a spot shape at high speed in a fan shape or in parallel by a reflection mirror or the like may be used. Preferably, however, the light emitted from the light source is linearly focused. The structure can be simplified by using a cylindrical light source and a slit light source in which these light source and lens are integrated. In addition, it is preferable that the short side width | variety of slit light is small enough compared with the level | step difference of welding. Here, when the method of scanning the point light source is adopted, the light source is turned on for at least a period of one scanning period or more, and the imaging cycle (exposure time) of the camera needs to be longer than that period.

  The positional relationship among the light source 10, the imaging means 12, and the welded portion 40 is as shown in FIG. 2, and the slit light 100 irradiated from the light source 10 is in the weld line direction of the welded portion 40 (for example, in the direction of conveying the steel plate in the continuous line). (Corresponding to the width direction, which is the orthogonal direction), the positional relationship is such that the light is irradiated in a substantially orthogonal direction.

  In consideration of the use environment of the apparatus, it is desirable that the measuring head 1 has a sealed structure except for the openings 16 and 17 that do not block the optical path, and protects the device from the heat of welding by cooling means such as air cooling (not shown). . Further, since the measuring head 1 collects light cutting images of the welded portion 40 at a plurality of locations along the weld line, the measuring head 1 is arranged in the width direction of the steel strip while facing the butt welded portion with a substantially constant gap by a moving mechanism (not shown). It is desirable to be able to move.

  The image processing device 2 controls the operation of the measuring head 1 and generates a three-dimensional shape and luminance distribution image of the welded portion from a group of light cut images taken by the measuring head 1 at each position of the welded portion 40. Further, it is discriminated, and the configuration may be constituted by a light source power source 20, a camera power source 21, an image data conversion circuit 22, an image processing circuit 23, a discrimination circuit 24, and the like.

  Among these, the light source power source 20 and the camera power source 21 supply predetermined driving power to the slit light source 10 and the camera 12, respectively, and a known power source circuit having a voltage and capacity corresponding to the device may be used. .

  The image data conversion circuit 22 is an image input unit that converts the reflection intensity received by each pixel of the camera 12 into luminance information (voltage) into a numerical string corresponding to each two-dimensional pixel position.

  The image processing circuit 23 includes a cross-sectional shape calculation unit 230, a luminance distribution calculation unit 231, a shape data memory 232, and a luminance data memory 233. The cross-sectional shape calculation unit 230 receives the image data captured from the image data conversion circuit 22, calculates cross-sectional shape coordinates from the image data by a light cutting method, and stores the data in the shape data memory 232. The method for calculating the cross-sectional shape in the cross-sectional shape calculation unit is that the bead-cut shape of the light cut image is not affected by the difference in the brightness level between the cut portion and the non-cut portion of the light cut image. The thinning process based on the dynamic moment calculation disclosed in the above is performed to extract the light section line. As long as the difference in luminance level is not a problem, a thinning process using a general binarization process may be used. Then, a cross-sectional shape is calculated based on the extracted cutting line. Similarly, the luminance distribution calculation unit 231 inputs the image data captured from the image data conversion circuit 22, calculates the reflected light intensity (luminance) distribution data in which the slit light is reflected on the surface of the object to be inspected, Stored in the luminance data memory 233.

  The discrimination circuit 24 discriminates whether or not the welding point is good based on the cross-sectional shape data of the shape data memory 232 and the luminance distribution data of the luminance data memory 233. As an example of the configuration, as shown in FIG. It comprises a search circuit 240, a cutting part step detection circuit 241, a step generation determination circuit 242, a step generation region count counter 243, and a cutting defect determination circuit 244. Any of these may be configured by a logical operation element such as a comparison circuit, or may be configured by a computer having a program operation function corresponding thereto.

  The luminance change search circuit 240 calculates the area range of the bead peripheral edge (periphery of the bead cutting area edge) based on the luminance change of the luminance distribution data. In the luminance distribution data at each measurement position measured by scanning in the welding line direction, the boundary part of the area where the luminance is reduced near the center is calculated as the bead cutting part. The calculation is performed using a search method in the horizontal direction.

  Further, the processing of the cutting part step detection circuit 241, the step generation determination circuit 242, the step generation region count counter 243, and the cutting defect determination circuit 244 is performed on both ends of the bead cutting region calculated by the luminance change search circuit 240, respectively. Since the start point and end point of the cutting area are the same process, the start point will be described.

Specifically, the cutting step difference detection circuit 241 has a cutting shape (high height) in a coordinate region having a predetermined width in the direction orthogonal to the weld line with the coordinates of the bead peripheral edge starting point output from the luminance change search circuit 240 as the center. A) The maximum change amount of data (difference between the maximum value and the minimum value), that is, the maximum change amount of the unevenness is calculated.

  Then, the step generation determination circuit 242 determines whether the maximum change amount of the cutting shape data exceeds the threshold value T1.

  In the step difference region count counter 243, when the threshold value is exceeded, the count value is incremented, and the number of measurement points exceeding the threshold value is counted (counted). This process is performed for the start point and the end point of the cutting part.

  The cutting defect determination circuit 244 reads the value of the step generation area count counter 243, and the number of measurement points determined by the step generation determination circuit 242 to exceed the threshold T1 for determining whether the maximum change amount is good or not is The proportion of the total number of measurement points is calculated, and if this exceeds the second threshold value T2, a failure is output. It should be noted that if the total number of measurement points does not vary from measurement object to measurement object, the evaluation may be made simply with a count value exceeding the threshold T1, not the ratio to the total number.

  Then, the display device 3 alerts the worker with a screen or sound according to the output of the cutting failure determination circuit 244, or sends failure information to an external operation management device (business computer or the like) via a signal line (not shown). , And can be constituted by a known alarm device, CRT device, communication device, or the like.

  Hereinafter, the operation of this embodiment will be described using data of measurement examples.

  FIG. 4 shows the result of processing the light cut image in which the measurement head 1 irradiates slit light in the direction orthogonal to the weld line at an arbitrary position of the weld 40 and detects the reflected light by the image processing circuit 23. It is. 4, (a) is the cross-sectional shape data calculated by the cross-sectional shape calculating unit 230, and (b) is the luminance distribution data calculated by the luminance distribution calculating unit 231. The horizontal axis of FIG. 4A is a position coordinate orthogonal to the welding line, and the vertical axis is the height coordinate of the surface to be measured showing the cross-sectional shape. The cross-sectional shape of FIG. 4A may be calculated by a thinning process using a generally known binarization process, but here is disclosed in Japanese Patent Laid-Open No. 2003-322513, which has high detection accuracy. The method will be described below. The reflected light of the slit light is picked up by a two-dimensional image pickup means such as a CCD camera to obtain a two-dimensional light cut image shown in FIG. The pixel at the lower left corner of the two-dimensional mesh in FIG. 5 is set to X = X1, Y = Y1, the X axis in the width direction and the Y axis in the tube axis direction, and the coordinates of each pixel are X = Xi (i = 1, 2). ,..., M), Y = Yj (i = 1, 2,..., N), the pixel brightness is I (Xi, Yj), and the weighted average S (Xi) defined by the following equation is , X is determined as the Y coordinate of the light section line at Xi.

  S (Xi) = ΣYj I (Xi, Xj) / ΣI (Xi, Yj) (1)

  In order to eliminate error factors due to disturbance components such as indirect reflected light and back light generated in the formation part outside the slit light irradiation area, the following (1) to (3) are used for each X coordinate. The Y coordinate of the light section line is obtained.

  (1) The Y coordinate Y0 that is the maximum luminance in the Y-axis direction and the luminance I0 at that point are obtained.

  (2) Using a predetermined number of pixels Nw, 0 ≦ Y ≦ Y0−NwΔY, Y0 + NwΔY ≦ Y ≦ Yn (Yn is the Y coordinate of the representative point of the end pixel in the Y direction of the light section image. ΔY is the Y direction of one pixel) The maximum luminance I1 in the range of (length) is obtained.

  (3) An appropriate value between I0 and I1 (for example, average value (I0 + I1) / 2) is set as a threshold value J1, and in the range of Y in which the pixel luminance in the Y direction is larger than J1, the above equation (1) shows that S (Xi) is calculated.

  (4) With the relative movement of the slit light irradiation area in the tube axis direction, the above procedures (1) to (3) are repeated.

  By the above calculation, the position of the light cutting image of the bead cutting shape of the welded portion can be accurately detected without being affected by the difference in luminance level between the cutting portion and the non-cutting portion of the light cutting image.

  In addition, when the reflected light intensity of the non-cutting part is sufficiently high to cause a range over or when the noise from the formation part is so small that it can be ignored, the calculation of the cross-sectional line of the region where the reflected light intensity is high is performed. The threshold value for this is replaced with a predetermined fixed threshold value J2 that is set to be slightly smaller than the maximum value of the luminance range within a range that does not become less than the maximum luminance value of the formation portion and empirically, The above procedure may be as follows.

  (1) The Y coordinate (the average value when there are a plurality of values) Y0 and the luminance I0 at that point are obtained as the maximum luminance in the Y-axis direction.

  (2) Using a predetermined number of pixels Nw, 0 ≦ Y ≦ Y0−NwΔY, Y0 + NwΔY ≦ Y ≦ Yn (Yn is the Y coordinate of the representative point of the end pixel in the Y direction of the light section image. ΔY is the Y direction of one pixel) The maximum luminance I1 in the range of (length) is obtained.

  (3) When I0 is equal to or greater than a predetermined fixed threshold J2, S (Xi) is calculated in a range of Y in which the pixel luminance is equal to or greater than J2 in the Y-axis direction when X = Xi.

  (4) When I0 is lower than J2, an appropriate value between I0 and I1 (for example, average value (I0 + I1) / 2) is set to a threshold value J3 (corresponding to the above-mentioned J1), and the luminance in the Y direction is larger than J3. S (Xi) is calculated in the range of Y.

  (5) The above procedures (1) to (4) are repeated with relative movement of the slit light irradiation area in the tube axis direction.

  From the position data X and Y coordinates of the light cutting line of the slit light detected as described above, the cross section of the bead cutting shape of the welded portion based on the geometrical positional relationship of the pixel address, light source, imaging means, and inspection object Calculate the shape.

Specifically, based on the direction angle definition in which the direction perpendicular to both the weld line direction and the direction orthogonal to the weld line is 0 °, the incident angle of the light source is α, and the light receiving angle of the imaging means is β. In this case, the optical cutting line position (Xi, S (Xi)) on the image is expressed by the following equation: xi = Xi (2)
yj = ΔY × S (Xi) × cos α / sin (α + β) (3)
To calculate the coordinates (xi, yj) of the cross-sectional shape of the bead cutting. Note that ΔY is the actual length of one pixel in the y direction.

  The horizontal axis of FIG.4 (b) is a position coordinate orthogonal to a welding line, and a vertical axis | shaft is a value of reflected luminance. This reflection luminance is data obtained by calculating the maximum value (maximum luminance) of the luminance distribution in the y direction at each coordinate position in the image x direction in the light section image as shown in FIG. 5 obtained by imaging with the camera. It is.

  In the measurement in this example, welding was performed such that one of the steel strips to be abutted was intentionally wavy and a butt failure was partially generated, and the measurement was performed after removing the swell. As a result, since the base material is deformed (warped) from the outside of the welded portion to the vicinity of 20 mm in addition to defective cutting of the welded portion at the measurement location, if the maximum value search is performed only with the shape data, the baseboard warpage is performed. The peak position (A) and the recess (A) due to the tip shape of the cutting tool may be erroneously detected as the highest point and the lowest point, but the present invention shows that it can be accurately detected without erroneous detection. .

  The data as shown in FIG. 4 is output for each light cut image taken at each position where the measuring head 1 moves along the weld line, and the data is output to the determination circuit 24.

  Next, the operation of the discrimination circuit 24 will be described with reference to FIG.

  First, in step S1, an average value of the entire luminance of the luminance distribution data is calculated. In step S2, the barycentric position of the pixel range in which the luminance is lower than the average value obtained in step S1 is calculated, and the coordinate point is set to A as shown in FIG. This point A is a starting point for the luminance change search, and is a point for searching for the changing point by moving the coordinates from here to the left and right in the figure.

  In step S3, the minimum luminance is obtained while the luminance is lower than the average value obtained in step S1 of the luminance distribution, and the luminance value is set to V. On the other hand, with respect to the point A obtained in step S2, the maximum luminance in the left region is obtained in step S4, the luminance value is set to L, and the maximum luminance in the right region is obtained in step S5. Let the value be R.

  Based on these V, L, and R, a luminance threshold TL for calculating the left change point and a luminance threshold TR for calculating the right change point are obtained, and the change point is determined based on these values. In step S6, TL = (V + L) / 2 is set, and the coordinate point where the luminance value corresponding to each coordinate first exceeds TL is set as the luminance change point PL by moving from the point A to the left side. Similarly, in step S7, TR = (V + R) / 2 is set, and the coordinate point at which the luminance value corresponding to each coordinate first exceeds TR is set as the luminance change point PR by moving to the right from the point A.

Thus, when the example which the brightness | luminance change search circuit 240 computed regarding the measurement data of FIG. 4 is shown, the coordinate PL of the left edge of a cutting strip is the vertical line written in FIG. Then, the coordinates PR obtained here, that is, the shape of the cutting portion (height ) within the range of 1 mm width (2 mm width) in the width direction of the welded portion (in the direction orthogonal to the weld line) centering on the left end of the cutting range. ) (The difference between the maximum value and the minimum value) is calculated as 0.025 mm as shown in the figure. Similarly, the coordinate PR of the right edge of the cutting band is calculated as the position of the line in FIG. 4, and the step amount is calculated as 0.074 mm. Here, the threshold values should be determined in relation to the operation, but in the present embodiment, T1 = 0.06 mm and T2 = 5% are preferable values.

  For example, when the level difference occurrence threshold T1 is 0.06 mm, it is determined that the left side is normal but the right side is defective. Then, the counter value of the step occurrence area counting counter 243 is incremented. Then, the step occurrence determination circuit 242 scans in the welding line direction and determines whether the step difference amount calculated on both edges of the cutting part calculated above exceeds a preset threshold for each of the measured image groups. This may be determined by associating a numerical value obtained in the present invention with a defect that occurs in operation or by past operation experience. For example, when the number of images exceeding the level difference occurrence threshold T1 is 5% or more of the whole, the butt weld may be determined to be defective.

  Next, the measurement using the apparatus of the present embodiment was performed on a butt weld after bead cutting, which was different from the sample of FIG. The cutting part shape and the brightness distribution of the slit light (graph corresponding to FIG. 4 above) within the range of 2 mm width in the direction perpendicular to the weld line from the cutting start point according to this measurement example are as shown in FIG. The difference in level was 0.030 mm at the left edge and 0.032 mm at the right edge. At the measurement position shown in FIG. 8, the level difference at both edges of the cutting part is equal to or less than T1 determined above (for example, T1 = 0.05 mm), and is determined to be harmless. For this sample, none of the measured image groups of the butt welds have a step amount calculated in the same manner for both edges of the cut part exceeding T1, and the butt weld is determined to be harmless. It was.

  In the above description, for simplification of explanation, the step shape on the start point side of the cutting area is discriminated, but the end point side of the cutting area is discriminated in exactly the same procedure. It is good also as a final result to go and the one where a defect determination score is large. Further, when both the start point and the end point exceed the threshold value, each 1 count may be combined to make a total of 2 counts, or may be 1 count according to the gist that the measurement location is abnormal. This setting may be made so that any one can be selected as appropriate depending on the measurement target, the operation conditions, and the operation method of the apparatus.

Sectional drawing including a partial block diagram which shows an example of a structure of embodiment of this invention The perspective view which similarly shows the principal part The figure which shows an example of the detailed structure of a discrimination circuit among Examples A graph explaining the operation of a measurement example using the present embodiment for a welded portion in which a cutting defect has occurred due to warpage of the butt portion The figure which shows the example of an optical cross-section image imaged with the camera The flowchart which shows the process sequence which searches the cutting area edge part of this invention Similarly, a graph showing an example of processing results A graph explaining a measurement example of the present embodiment for a normal welded portion

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Measuring head 2 ... Image processing apparatus 3 ... Display apparatus 4 ... Measured object 10 ... Laser light source 12 ... Camera 20 ... Light source power supply 21 ... Camera power supply 22 ... Image data conversion circuit 23 ... Image processing circuit 24 ... Discrimination circuit 40 ... Butt weld

Claims (2)

In the method of inspecting the quality of the welded part after removing the raised part after butt welding,
With respect to the welding line, the slit light in the substantially orthogonal direction is scanned in the welding line direction to capture images at a plurality of positions,
In each of the image the imaging, and calculates the height of the weld Budan surface based on light-section method, to calculate the intensity distribution of the reflected light of the respective image,
A region having a predetermined intensity or less in the intensity distribution of the reflected light is obtained, and the difference between the maximum value and the minimum value of the cross-section height within the predetermined region set with reference to the end of the region is determined as the cross-sectional shape. The amount of change is calculated and the process of determining whether the amount of change is equal to or greater than a predetermined value is performed for all the captured images. From the number of images where the amount of change is equal to or greater than the predetermined value, A quality inspection method for a butt weld, wherein the quality is determined. A welded part quality inspection device that inspects the quality of the welded part after removing the raised part after butt welding,
A light source that irradiates slit light in a substantially orthogonal direction with respect to the welding line;
Imaging means for scanning the slit light in the direction of the welding line and imaging images at a plurality of positions;
In each of the image the imaging, and calculates the height of the weld Budan surface based on light-section method, and image processing means for calculating the intensity distribution of the reflected light of the respective image,
A region having a predetermined intensity or less in the intensity distribution of the reflected light is obtained, and the difference between the maximum value and the minimum value of the cross-section height within the predetermined region set with reference to the end of the region is determined as the cross-sectional shape. The amount of change is calculated and the process of determining whether the amount of change is equal to or greater than a predetermined value is performed for all the captured images. From the number of images where the amount of change is equal to or greater than the predetermined value, A determination means for determining pass / fail;
An apparatus for inspecting the quality of a butt weld, characterized by comprising:

Images