JP5578005B2 - Welding bead inspection device, inspection method, and inspection program - Google Patents

Description

  The present invention relates to a technique for inspecting a weld bead.

  Patent Document 1 discloses a method for inspecting a weld bead in which two base materials are welded. In this inspection method, the surface shape of a welding location (a region including a weld bead and a base material) is measured. Next, based on the curvature of the measured shape, a weld bead is specified from the measured shape. When the weld bead is specified, the quality of the weld bead is determined based on the surface shape of the weld bead.

JP 2008-215839 A

  In order to determine the quality of the weld bead from the surface shape of the weld location, it is necessary to accurately identify which range of the weld locations is the weld bead. However, the surface shape often changes smoothly at the boundary between the weld bead and the base material. In addition, when the surface of the base material is a curved surface, there is often no significant difference in curvature between the weld bead and the base material. For this reason, even if it is going to specify a weld bead based on the curvature of a measurement shape etc. like the art of patent documents 1, a weld bead may not be specified correctly.

  The inventor of the present application has already proposed a technique for solving the above problem. In this technique, reference shape data indicating surface shapes before welding of two plate materials to be welded (first plate material and second plate material) is prepared in advance. The surface shape indicated by the reference shape data is a design shape (hereinafter referred to as a reference shape). The fitting area is set in the area where the sheet material is always exposed after welding (the area where the welding bead is not formed even if the manufacturing error in the welding process is maximized), out of the two plate material reference shapes. Has been. When inspecting the weld bead, the surface shape of the welded part is measured. Next, the reference shape of the first plate member and the reference shape of the second plate member are overlapped with the measurement shape so that the reference shape and the measurement shape in the fitting region accurately overlap. In the measurement shape, an area where the displacement of the shape is large with respect to both the reference shape of the first plate material and the reference shape of the second plate material is specified as the weld bead. According to this technique, the weld bead can be accurately specified by using the surface shape of each plate material before welding. Further, the thickness of the weld bead can be obtained from the reference shape and the surface shape of the weld bead. Therefore, the weld bead can be inspected more suitably.

  With this technique, if the fitting area can be set sufficiently wide, the weld bead can be accurately inspected. However, depending on the product, the area where the base material is exposed after welding is small, and the fitting area may not be set wide. If the fitting region cannot be set wide, the reference shape cannot be accurately overlapped with the measurement shape, and the weld bead may not be accurately specified.

  Therefore, the present specification provides a technique capable of setting a wide fitting region even when the region where the base material is exposed after welding is small.

The weld bead inspection device disclosed in this specification inspects a weld bead of a structure in which a first plate member having a curved surface shape and a second plate member having a curved surface shape are welded. This inspection apparatus has input means, storage means, and area specifying means. The input means receives measurement shape data describing a measurement shape obtained by measuring the surface shape of the structure in a cross section intersecting the weld bead. A memory | storage means memorize | stores the 1st board | plate material reference | standard shape data which describes the reference | standard shape of the surface shape before welding of the 1st board | plate material in the said cross section. The area specifying means specifies an area where the first plate material is exposed from the measurement shape based on the inclination angle distribution of the measurement shape and the inclination angle distribution of the reference shape of the first plate material.

The “reference shape” is a design shape of the first plate member. Since there is an error in the actual shape of the first plate member used in the structure, the reference shape and the actual shape may be slightly different.
The “tilt angle” means the inclination of the measurement shape at an arbitrary point on the measurement shape. For example, when the measurement shape is composed of a large number of coordinate points represented by the x-coordinate and the z-coordinate, the inclination of the measurement shape at an arbitrary coordinate point is the change rate dz / dx of the z-coordinate at that coordinate point. It is. The “tilt angle distribution” means a distribution of tilt angles at each point on the measurement shape.
Moreover, although this weld bead inspection apparatus specifies the area | region where the 1st board | plate material is exposed from measurement shape, you may specify the area | region where the 2nd board | plate material is exposed further from measurement shape. . If these can be specified, the remaining area corresponds to the weld bead. That is, the weld bead can be specified from the measured shape. In this case, the region where the second plate member is exposed may be specified in the same manner as the region where the first plate member is exposed, or may be specified by another method.

  In this inspection apparatus, the region where the first plate material is exposed is specified based on the inclination angle distribution of the measurement shape and the inclination angle distribution of the reference shape of the first plate material. The characteristic of the surface shape appears more prominently in the inclination angle distribution. Therefore, it is possible to accurately specify a region in which the shape substantially coincides with the reference shape of the first plate member from the measured shape based on the inclination angle distribution. Thereby, the area | region where the 1st board | plate material is exposed can be specified from measurement shape. The area | region where the specified 1st board | plate material is exposed can be used as a fitting area | region for overlapping the reference | standard shape of a 1st board | plate material. According to this technique, since the fitting area can be set based on the actually measured shape, the fitting area can be set widely. In other words, when the fitting area is set in advance, it is necessary to set the fitting area in the area where the first plate material is necessarily exposed even if the manufacturing error is maximized. . According to the technique of the present specification, since the region where the first plate member is exposed can be specified from the actually measured shape, a wide fitting region is set in the region where the first plate member is exposed without considering manufacturing errors and the like. be able to. For example, the entire region where the specified first plate material is exposed can be set as the fitting region.

  In the inspection apparatus described above, when the region specifying unit divides the measurement shape for each region where the inclination angle is close, and the inclination angle of each region, and the reference shape of the first plate material is divided for each region where the inclination angle is close Based on the inclination angle of each region, it is preferable to specify the region where the first plate member is exposed from the measurement shape.

  In this way, the measurement shape and the reference shape of the first plate material are divided for each region where the inclination angle is close, and it is necessary for identification by specifying the region where the first plate material is exposed based on the divided region. The amount of computation can be reduced.

  The inspection apparatus described above arranges the reference shape of the first plate material in accordance with the region where the first plate material in the measurement shape is exposed, and the region where the weld bead is formed in the measurement shape and the first plate material It is preferable to further include a crank angle calculating means for calculating an angle between the shape of the region where the weld bead is formed and the reference shape of the first plate member at the boundary with the region where is exposed. In addition, since the area | region where the 1st board | plate material is exposed and the area | region where the weld bead is formed are adjacent, if the area | region where the 1st board | plate material is exposed is specified, the weld bead in a measurement shape will form. This means that the boundary between the region where the first plate material is exposed and the region where the first plate material is exposed is specified.

  Crank angle is closely related to overlap. Overlap is a problem in which the weld bead and the first plate material are not sufficiently melted and sufficient strength cannot be obtained at the weld location. If overlap occurs, the crank angle increases. According to this inspection device, it is possible to inspect the overlap based on the crank angle.

  The inspection apparatus described above includes welding bead specifying means for specifying a region where a weld bead is formed, a boundary point on the first plate member side of a region where a weld bead is formed, and a weld bead. Distance for calculating the distance between the straight line connecting the boundary point on the second plate member side of the formed area and the coordinate point on the measurement shape in the area where the weld bead farthest from the straight line is formed It is preferable to further include a calculation unit.

  Hereinafter, the distance is referred to as extra height. Extra height is closely related to overlap. If there is an overlap, the surplus height increases. According to this inspection apparatus, it is possible to inspect the overlap based on the surplus height.

The weld bead inspection apparatus of the present invention inspects a weld bead of a structure in which a first plate member and a second plate member are welded by overlapping a second plate member having a curved shape on a first plate member having a curved shape. The weld bead inspection device includes an input unit, a first storage unit, a first angle adjustment unit, and a region specifying unit. The input means receives measurement shape data describing a measurement shape obtained by measuring the surface shape of the structure in a cross section intersecting the weld bead. A 1st memory | storage means memorize | stores the 1st board | plate material reference | standard shape data which describes the 1st board | plate material reference | standard shape which is the reference | standard shape of the surface shape before welding of the 1st board | plate material in the said cross section. The first angle adjusting means adjusts the angle of the first plate material reference shape with respect to the measurement shape by rotating the first plate material reference shape. The area specifying means specifies an area where the first plate material is exposed from the measurement shape, based on the first plate material reference shape in which the measurement shape and the angle are adjusted.
The weld bead inspection device includes: a second storage unit that stores second plate material reference shape data describing a second plate material reference shape that is a reference shape of the surface shape of the second plate material before welding in the cross section; and a second plate material reference You may further have the 2nd angle adjustment means which adjusts the angle with respect to the measurement shape of a 2nd board | plate material reference | standard shape by rotating a shape. In this case, it is preferable that the area specifying unit specifies the area where the second plate material is exposed from the measurement shape based on the second plate material reference shape whose angle is adjusted with the measurement shape.
Further, the first angle adjusting means adjusts the position of the first plate material reference shape with respect to the measurement shape based on the inclination angle distribution of the measurement shape and the inclination angle distribution of the first plate material reference shape, and then the first plate material reference shape. The angle with respect to the measured shape may be adjusted.
In the structure, an end surface of the second plate member may be disposed on the first plate member, and a region including the end surface of the second plate member may be welded. In this case, it is preferable that the 2nd board | plate material reference | standard shape contains the reference | standard shape of the end surface shape of a 2nd board | plate material. Further, the area specifying means calculates the comparison shape obtained by connecting the extension line of the first plate material reference shape and the extension line of the end face shape of the second plate material reference shape, and on the measurement shape Difference value calculation means for calculating a difference value between coordinates between the coordinate point and the coordinate point on the comparative shape, and a region where the first plate member is exposed and a region where the second plate member is exposed based on the difference value It is preferable to have a means for specifying.
Also, the weld bead inspection method and the inspection program of the present invention can provide the same effects as those of the above-described inspection apparatus of the present invention .

1 is a block diagram of a weld bead inspection device 10. FIG. The perspective view of the welding component 30. FIG. The perspective view of the board | plate materials 32 and 34 before welding. The figure which shows the three-dimensional measurement shape 60 and the three-dimensional reference | standard shape 62. FIG. The flowchart which shows a weld bead inspection process. The flowchart which shows a weld bead inspection process. The flowchart which shows the detail of step S12. The figure which shows the two-dimensional measurement shape 70 and the two-dimensional reference | standard shape 72 which were divided into the area | region. Explanatory drawing of the other process which divides the two-dimensional measurement shape 70 into an area | region. The figure which shows the two-dimensional measurement shape 70 and the two-dimensional reference | standard shape 72 after fitting. Explanatory drawing of the process which specifies welding bead area | region WB. Explanatory drawing of crank angle (theta) and extra height Hc. The perspective view of the welded part by which fillet welding was carried out.

  The weld bead inspection apparatus according to the embodiment will be described with reference to the drawings. As shown in FIG. 1, a weld bead inspection device 10 according to the embodiment is connected to a surface shape measurement device 20. The weld bead inspection device 10 is configured by a general computer and includes an arithmetic device 12 and a storage device 14. The storage device 14 stores a welding bead inspection program, data indicating a reference shape of a plate material used for welding, and the like.

FIG. 2 shows a perspective view of the welded part 30 to be inspected. As shown in FIG. 3, the welded part 30 is a part obtained by welding a partially overlapped plate material 32 and a plate material 34 at the overlapped portion. That is, reference numeral 36 in FIG. 2 indicates a weld bead. The plate members 32 and 34 have curved surface shapes that are curved in a direction crossing the weld bead 36. The surface shape measuring device 20 measures the surface shape (two-dimensional shape in the xz plane of FIG. 2) of the welded part 30 in a cross section substantially orthogonal to the weld bead 36. This measurement is performed by a so-called light cutting method, irradiating linear light that crosses the weld bead 36 of the welded part 30, and an area 38 on which the linear light is projected is photographed. From the captured image, the cross-sectional shape of the region 38 is calculated. At this time, at least one of the light irradiation direction and the light photographing direction is set obliquely with respect to the surface of the welded part 30. Note that the measurement of the surface shape of the welded part 30 is not limited to the optical cutting method, and for example, a contact-type shape measuring instrument can also be used.
As shown in FIG. 2, the surface shape measuring apparatus 20 measures the surface shape in the xz plane at regular intervals along the direction (y direction) in which the weld bead 36 extends (a plurality of linear regions 38 in FIG. 2). Measure the surface shape with each of the). Hereinafter, the two-dimensional surface shape of the welded part 30 measured in each xz plane is referred to as a two-dimensional measurement shape. As illustrated in FIG. 4A, a three-dimensional surface shape 60 (hereinafter referred to as a three-dimensional measurement shape 60) of the welded part 30 is obtained by a large number of two-dimensional measurement shapes measured at regular intervals in the y direction. can get. In FIG. 4A, one of the many two-dimensional measurement shapes 70 is indicated by a dotted line. The three-dimensional measurement shape 60 is represented by a number of coordinate points defined by x, y, and z coordinates.
As is apparent from FIGS. 1 and 2, the second plate member 34 has a wide area exposed after welding, whereas the first plate member 32 has a small region exposed after welding.

  The arithmetic device 12 inspects the weld bead 36 by executing a weld bead inspection program stored in the storage device 14. When the weld bead inspection program is executed, the processing shown in the flowcharts of FIGS. 5 and 6 is performed by the arithmetic unit 12.

  In step S <b> 2, the arithmetic device 12 receives input of three-dimensional measurement shape data representing the three-dimensional measurement shape 60 of the welded part 30 from the surface shape measurement device 20. In step S <b> 4, the arithmetic device 12 reads the three-dimensional reference shape data from the storage device 14. The three-dimensional reference shape data is data representing a three-dimensional shape (hereinafter referred to as a three-dimensional reference shape) of the surfaces of the plate material 32 and the plate material 34 before welding. As illustrated in FIG. 4B, the three-dimensional reference shape 62 is represented by a large number of coordinate points defined by x, y, and z coordinates. The three-dimensional reference shape 62 is a design shape, and the actual shapes of the plate members 32 and 34 may differ from the three-dimensional reference shape 62 due to errors. In the following, in the three-dimensional reference shape 62, a portion representing the shape of the plate member 32 is referred to as a first plate member three-dimensional reference shape 64, and a portion representing the shape of the second plate member 34 is referred to as a second plate member three-dimensional reference shape 66. The first plate material three-dimensional reference shape 64 corresponds to the shape of the upper surface 32a of the first plate material 32 of FIG. The 2nd board | plate material three-dimensional reference | standard shape 66 is corresponded to the shape of the upper surface 34a and the end surface 34b of the 2nd board | plate material 34 of FIG. The first plate material three-dimensional reference shape 64 and the second plate material three-dimensional reference shape 66 are arranged in a positional relationship corresponding to the relative positions of the designed first plate material 32 and second plate material 34. The three-dimensional reference shape 62 is composed of a plurality of two-dimensional reference shapes 72 defined at regular intervals in the y direction. In FIG. 4B, one of many two-dimensional reference shapes 72 is indicated by a dotted line. The two-dimensional reference shape 72 is a two-dimensional shape in the xz plane. Hereinafter, a portion of the two-dimensional reference shape 72 corresponding to the shape of the first plate member 32 is referred to as a first plate member two-dimensional reference shape 74, and a portion corresponding to the shape of the second plate member 34 is referred to as the second plate member 2. This is referred to as a dimension reference shape 76.

  In step S <b> 6, the arithmetic device 12 adjusts the position of the second plate 3D reference shape 66 in the y direction with respect to the 3D measurement shape 60. Specifically, the following procedure is performed. As shown in FIG. 4B, a fitting region 68 is defined in the second plate material three-dimensional reference shape 66. The fitting area 68 is set to an area that is away from the welding location and is not deformed during welding. First, k coordinate points P1 to Pk on the second plate 3D reference shape 66 in the fitting region 68 are specified. Next, the coordinate points Q1 to Qk obtained by projecting the coordinate points P1 to Pk on the three-dimensional measurement shape 60 along the z axis are specified. Next, the difference e1 to ek in the z coordinate between the coordinate points Q1 to Qk corresponding to the coordinate points P1 to Pk is calculated. Next, the variance V of the differences e1 to ek is calculated. The variance V is expressed by the following mathematical formula.

  In the above (Equation 1), E (e) is an average value of the differences e1 to ek. It can be said that the smaller the variance V is, the more consistent the three-dimensional measurement shape 60 and the second plate material three-dimensional reference shape 66 are within the fitting region 68. In step S6, the arithmetic unit 12 specifies the position where the variance V is minimum while moving the second plate 3D reference shape 66 in the y direction, and moves the second plate 3D reference shape 66 to the specified position. Let

  In step S <b> 8, the arithmetic unit 12 adjusts the position of the first plate material 3D reference shape 64 in the y direction with respect to the 3D measurement shape 60. Here, the first plate 3D reference shape 64 is moved by the same movement amount as the movement amount of the second plate 3D reference shape 66 in step S6.

  In step S10, the arithmetic unit 12 adjusts the position in the x direction of the second plate material three-dimensional reference shape 66 with respect to the three-dimensional measurement shape 60. Here, as in step S6, the second plate 3D reference is set so that the variance of the difference in the z coordinate between the coordinate points in the fitting region 68 and the coordinate points projected onto the 3D measurement shape 60 is minimized. The shape 66 is moved.

  In step S <b> 12, the arithmetic unit 12 sets a fitting region 80 for aligning the first plate material 3D reference shape 64 and the 3D measurement shape 60. Step S12 is performed for each y coordinate. That is, the two-dimensional measurement shape 70 and the two-dimensional reference shape 72 in the xz cross section are extracted for each y coordinate, and step S12 is executed for each cross section. Hereinafter, the process of step S12 will be described using the two-dimensional shapes 70 and 72 illustrated in FIG. 4 as an example.

  FIG. 7 shows details of the processing performed in step S12. In step S60 of FIG. 7, the arithmetic unit 12 divides the two-dimensional measurement shape 70 into regions corresponding to the inclination angles. Specifically, first, a normal vector at each coordinate point of the two-dimensional measurement shape 70 is calculated. The normal vector is calculated by approximating the coordinate point group on the two-dimensional measurement shape 70 with a curve. Next, the starting coordinate point Rm is specified, and whether or not the angle between the normal vector of the coordinate point Rm and the normal vector of the adjacent coordinate point Rm + 1 is equal to or smaller than a predetermined reference angle. Determine whether. This determination can be made by calculating the inner product of the normal vectors. When the angle between the normal vectors is equal to or smaller than the reference angle, the coordinate point Rm and the coordinate point Rm + 1 are grouped. In this case, the same processing is then performed on the normal vector of the coordinate point Rm and the coordinate point (coordinate point Rm−1 or Rm + 2) adjacent to the grouped coordinate points Rm and Rm + 1. This process is repeated until there are no more coordinate points that can be grouped with the coordinate point Rm. Thereby, the coordinate point group grouped with the coordinate point Rm is divided into one area. When one region is divided, the above-described processing is performed on the coordinate points that have not yet been divided, and all the coordinate points on the two-dimensional measurement shape 70 are divided into regions according to the inclination angles. Thus, for example, as illustrated in FIG. 8, the two-dimensional measurement shape 70 is divided into regions S1 to Sn (regions S1 to S8 in the example of FIG. 7). The regions S1 to Sn are a collection of coordinate points that are approximated by normal vectors (that is, coordinate points that are approximated by an inclination angle). Further, normal vectors T1 to Tn (normal vectors T1 to T8 in the example of FIG. 7) are set in each of the regions S1 to Sn. The normal vectors T1 to Tn set here are vectors representing the inclination angles of the regions S1 to Sn. The normal vectors T1 to Tn may be vectors obtained by averaging the normal vectors of the coordinate point groups in the regions S1 to Sn, or may be vectors calculated by other processes.

  Step S60 may be performed by the following method. In this method, first, as shown in FIG. 9, a line segment L1 connecting the coordinate points K1 and K2 at both ends of the two-dimensional measurement shape 70 is drawn. Next, the coordinate point K3 on the two-dimensional measurement shape 70 farthest from the line segment L1 is specified. If the distance between the line segment L1 and the coordinate point K3 is larger than a predetermined threshold value, the line segments L2 and L3 are drawn along the coordinate points K1, K3, and K2. Next, the coordinate point K4 on the two-dimensional measurement shape 70 in the section K1-K3 farthest from the line segment L2, and the coordinate point on the two-dimensional measurement shape 70 in the section K3-K2 farthest from the line segment L3. Specify K5. Next, it is determined whether or not the distance between the line segment L2 and the coordinate point K4 is larger than the threshold value. If this distance is larger than the threshold value, a line segment is drawn along the coordinate points K1, K4, and K3, and the same processing is repeated. Further, it is determined whether or not the distance between the line segment L3 and the coordinate point K5 is larger than the threshold value. If this distance is larger than the threshold value, a line segment is drawn along the coordinate points K3, K5, and K2, and the same processing is repeated. By repeating the same process until the distance between the newly drawn line segment and the coordinate point farthest from the line segment within the section corresponding to the line segment becomes smaller than the threshold value, a broken line along the two-dimensional measurement shape 70 is formed. can get. Each line segment of the broken line thus obtained represents the inclination of the two-dimensional measurement shape 70. That is, the coordinate point group that substantially overlaps the line segment of the broken line is a collection of coordinate points that approximate the normal vector. The two-dimensional measurement shape 70 is divided into regions S1 to Sn according to the line segment in the broken line.

  In step S62, the arithmetic unit 12 divides the first plate material two-dimensional reference shape 74 into regions corresponding to the inclination angles as shown in FIG. Here, similarly to step S60, the first plate member two-dimensional reference shape 74 is divided into regions U1 to Uo (regions U1 to U4 in the example of FIG. 8). Further, normal vectors V1 to Vo (V1 to V4 in the example of FIG. 8) are set in the respective regions U1 to Uo. Note that data obtained by dividing the first plate 2D reference shape 74 into regions may be stored in the storage device 14 in advance, and the data may be read in step S62.

  In step S <b> 64, the arithmetic unit 12 associates each region of the first plate material two-dimensional reference shape 74 with each region of the two-dimensional measurement shape 70. Here, regions where normal vectors are approximated are associated with each other. That is, when the angle formed by the normal vector T and the normal vector V is smaller than the reference angle, regions having the normal vector are associated with each other. However, even if the normal vectors are approximate to each other, if the x coordinates are extremely different areas, these areas are not associated with each other. The angle between normal vectors can be determined by the inner product of the normal vectors. In the example of FIG. 8, the region S1 is associated with the region U1, the region S2 is associated with the region U2, the region S3 is associated with the region U3, and the region S4 is associated with the region U4. Each region (regions S1 to S4 in the example of FIG. 8) of the two-dimensional measurement shape 70 associated in step S64 corresponds to a region where the plate material 32 is exposed even after welding. As shown in FIG. 8, the arithmetic unit 12 fits the first plate material two-dimensional reference shape 74 in a range that overlaps with a region associated with the two-dimensional measurement shape 70 (regions S <b> 1 to S <b> 4 in the example of FIG. 8). Region 80 is set.

  As described above, the fitting region 80 is set in the first two-dimensional measurement shape 74 by executing steps S60 to S64. In this welded part 30, since the area where the plate material 32 is exposed is small, a wide fitting area cannot be set in advance on the plate material 32 side like the fitting area 68 on the plate material 34 side. However, by comparing the two-dimensional measurement shape 70 and the first plate material two-dimensional reference shape 74 based on the inclination angle distribution in this way, the region where the plate material 32 is exposed can be accurately specified. By using the region thus identified as the fitting region 80, the fitting region 80 can be set most widely within the range where the plate 32 is exposed. Therefore, it is possible to accurately perform the position adjustment described later. By performing steps S60 to S64 for each cross section of each y-coordinate, a fitting region 80 is set in the three-dimensional measurement shape 60 as shown in FIG.

  If the fitting area | region 80 is set by step S12, the arithmetic unit 12 will adjust the position of the x direction with respect to the three-dimensional measurement shape 60 of the 1st board | plate material 3D reference | standard shape 64 in step S14. Here, similarly to step S6, the variance of the difference between the z coordinates of the coordinate points in the fitting region 80 and the coordinate points projected on the three-dimensional measurement shape 60 along the z direction is minimized. One plate 3D reference shape 64 is moved in the x direction.

  By the processing so far, the positions of the three-dimensional reference shape 62 in the x direction and the y direction are adjusted. In step S16, the arithmetic unit 12 determines whether or not the movement amount (position adjustment amount) of the three-dimensional reference shape 62 in steps S6, S8, S10, and S14 is zero. If the movement amount is not zero in any of steps S6, S8, S10, and S14, steps S6 to S14 are repeated again. That is, steps S6 to S14 are repeated until the movement amount becomes zero in all steps S6, S8, S10, and S14. Thereby, the positions of the three-dimensional reference shape 62 in the x direction and the y direction are optimized.

  In step S18, the arithmetic unit 12 adjusts the position in the z direction of the second plate material two-dimensional reference shape 76 with respect to the two-dimensional measurement shape 70 for each y coordinate. That is, the second plate 2D reference shape 76 and the 2D measurement shape 70 in the xz cross section are extracted for each y coordinate, and the position of the second plate 2D reference shape 76 in the z direction is adjusted in each xz cross section. Hereinafter, position adjustment in each xz cross section will be described. First, the coordinate point of the 2nd board | plate 2D reference | standard shape 76 in the fitting area | region 68 is specified, and the coordinate point which projected the specified coordinate point on the two-dimensional measurement shape 70 along az direction is specified. Then, the average value of the z coordinates of the coordinate points of the second plate material two-dimensional reference shape 76 in the fitting region 68 and the average value of the z coordinates of the coordinate points projected on the two-dimensional measurement shape 70 are calculated. And the 2nd board | plate material two-dimensional reference | standard shape 76 is moved to az direction so that both average values may correspond.

  In step S20, the arithmetic unit 12 adjusts the position in the z direction of the first plate material two-dimensional reference shape 74 with respect to the two-dimensional measurement shape 70 for each y coordinate. That is, the first plate member two-dimensional reference shape 74 and the two-dimensional measurement shape 70 in the xz section are extracted for each y coordinate, and the position of the first plate member two-dimensional reference shape 74 in the z direction is adjusted in each xz section. Hereinafter, position adjustment in each xz cross section will be described. First, the coordinate point of the first plate 2D reference shape 74 in the fitting region 80 is specified, and then the coordinate point obtained by projecting the specified coordinate point on the 2D measurement shape 70 along the z direction is specified. Then, the average value of the z coordinates of the coordinate points of the second plate material two-dimensional reference shape 76 in the fitting region 80 and the average value of the z coordinates of the coordinate points projected on the two-dimensional measurement shape 70 are calculated. And the 1st board | plate material two-dimensional reference | standard shape 74 is moved to az direction so that both average values may correspond.

  In step S <b> 22, the arithmetic unit 12 finely adjusts the position of the second plate material two-dimensional reference shape 76 with respect to the two-dimensional measurement shape 70 for each y coordinate. That is, the second plate material two-dimensional reference shape 76 and the two-dimensional measurement shape 70 in the xz section are extracted for each y coordinate, and the position of the second plate material two-dimensional reference shape 76 is finely adjusted in each xz section. Below, the fine adjustment of the position in each xz cross section is demonstrated. First, the arithmetic unit 12 specifies the coordinate point of the second plate 2D reference shape 76 in the fitting region 68, and for each specified coordinate point, the coordinate point on the 2D measurement shape 70 closest to the coordinate point (hereinafter referred to as the coordinate point). ). Then, an ICP (Iterative Closest Point) algorithm is executed based on the coordinate points of the second two-dimensional reference shape 76 in the fitting region 68 and their closest points. The translation and rotation of the second plate 2D reference shape 76 are repeatedly executed by the ICP algorithm. As a result, the position of the second plate 2D reference shape 76 is determined so that the coordinate points of the second plate 2D reference shape 76 in the fitting region 68 and these closest points are most consistent. When the second plate 2D reference shape 76 is rotated during the execution of the ICP algorithm, the average point of the coordinate points of the second plate 2D reference shape 76 in the fitting region 68 can be used as the rotation center.

  In step S24, the arithmetic unit 12 finely adjusts the position of the first plate material two-dimensional reference shape 74 with respect to the two-dimensional measurement shape 70 for each y coordinate. That is, the first plate member two-dimensional reference shape 74 and the two-dimensional measurement shape 70 in the xz section are extracted for each y coordinate, and the position of the first plate member two-dimensional reference shape 74 is finely adjusted in each xz section. Below, the fine adjustment of the position in each xz cross section is demonstrated. First, the arithmetic unit 12 specifies the coordinate point of the first plate 2D reference shape 74 in the fitting region 80, and for each specified coordinate point, the coordinate point on the 2D measurement shape 70 closest to the coordinate point (that is, the coordinate point) , Recent points). Then, the ICP algorithm is executed based on the coordinate points of the first two-dimensional reference shape 74 in the fitting region 80 and their closest points. The translation and rotation of the first two-dimensional reference shape 74 are repeatedly executed by the ICP algorithm. As a result, the position of the first plate 2D reference shape 74 is determined so that the coordinate points of the first plate 2D reference shape 74 in the fitting region 80 and these closest points are most closely matched. Note that when the first plate member two-dimensional reference shape 74 is rotated during the execution of the ICP algorithm, the average point of the coordinate points of the first plate member two-dimensional reference shape 74 in the fitting region 80 can be used as the rotation center.

  With the above processing, the alignment of the three-dimensional reference shape 62 with the three-dimensional measurement shape 60 is completed. As a result, data obtained by superimposing the three-dimensional reference shape 62 and the three-dimensional measurement shape 60 is completed. In the cross section at each y-coordinate, as shown in FIG. 10, data obtained by superimposing a two-dimensional reference shape 72 on a two-dimensional measurement shape 70 (hereinafter referred to as comparative shape data) can be generated. In FIG. 10, the two-dimensional measurement shape 70 is indicated by a solid line, the two-dimensional reference shape 72 is indicated by a dotted line, and the overlapping portion is indicated only by a solid line. As shown in FIG. 10, the second plate 2D reference shape 76 is arranged so that the fitting region 68 substantially matches the two-dimensional measurement shape 70, and the fitting region 80 substantially matches the two-dimensional measurement shape 70. One plate material two-dimensional reference shape 74 is arranged. In steps S10, S14, S18, S20, S22, and S24 described above, the first plate member two-dimensional reference shape 74 and the second plate member two-dimensional reference shape 76 are separately adjusted in position, so the first plate member two-dimensional reference shape is adjusted. 74 and the second plate 2D reference shape 76 are discontinuous.

  When proceeding to step S26 of FIG. 6, the arithmetic unit 12 connects the first plate material two-dimensional reference shape 74 and the second plate material two-dimensional reference shape 76 as shown in FIG. Specifically, in FIG. 10, a straight line is drawn by extending the first plate material two-dimensional reference shape 74 to the second plate material two-dimensional reference shape 76 side. Further, a straight line is drawn by extending the line 42b of the second plate material two-dimensional reference shape 76 (a line representing the end surface 34b of the plate material 34 in FIG. 3) downward. Then, as shown in FIG. 11A, the first plate material two-dimensional reference shape 74 and the second plate material two-dimensional reference shape 76 are connected at the intersection Cu between these straight lines.

  In step S28, as shown in FIG. 11B, the arithmetic unit 12 subtracts the z coordinate of the two-dimensional reference shape 72 from the z coordinate of the two-dimensional measurement shape 70 in each x coordinate of the comparison shape data. Δz is calculated. The difference value Δz is calculated for the region between the fitting region 68 and the fitting region 80. The difference value Δz represents the amount of change in the position of the surface before and after welding (the amount of change in the z direction). A region where the difference value Δz is positive is a region raised by welding, and a region where the difference value Δz is negative is a region recessed by welding.

In step S30, based on the difference value Δz and the threshold value z _TH , the arithmetic unit 12 divides the region on the x coordinate into the upper deformation region WU, the lower deformation region WL, and the non-deformation region as shown in FIG. The deformation area WN is classified. The threshold value z _TH is a value determined in advance according to the assumed difference value Δz. A region where the difference value Δz is greater than the threshold value z _TH is identified as an upper deformation region WU. A region where the difference value Δz is smaller than the threshold value −z _TH is identified as the downward deformation region WL. A region where the absolute value of the difference value Δz is equal to or less than the threshold value z _TH is identified as a non-deformed region WN. The upper deformation area WU indicates an area raised by welding, the lower deformation area WL indicates an area recessed by welding, and the non-deformation area WN indicates an area where the change in surface shape is small before and after welding. Note that when there is no deformation region (WU or WL), it means that there is almost no deformation before and after welding (that is, no weld bead is formed), so an error is reported.

  In step S32, the arithmetic unit 12 integrates the absolute value of the difference value Δz for each region specified in step S30. And the area | region where the calculated integral value is the largest is specified as welding bead area | region WB. In the example of FIG. 11B, the upper deformation area WU4 is specified as the weld bead area WB1.

In step S34, the arithmetic unit 12 is sandwiched between the weld bead region WB and the deformation region (WU or WL), and is there a non-deformation region WN in which the width in the x direction is less than the reference distance? Determine whether or not. The reference distance is a predetermined value. If such a non-deformation region WN exists (YES in step S34), in step S36, the non-deformation region WN and a deformation region adjacent to the non-deformation region WN (deformation region that is not the weld bead region WB). ) Is specified as part of the weld bead area WB (added to the weld bead area WB). Steps S34 and S36 are repeated until NO is determined in step S32. As a result, the final weld bead region WB is specified.
For example, in the example of FIG. 11B, the upper deformation area WU4 is specified as the weld bead area WB1 in step S32. Since the width of the non-deformation region WN3 between the weld bead region WB1 and the lower deformation region WL2 is less than the reference distance, YES is determined in step S34, and the lower deformation region WL2 and the non-deformation region WN3 are welded in step S36. It is incorporated into the area WB. That is, weld bead region WB1 expands to weld bead region WB2. Next, step S34 is executed again. At this time, there is no deformation region outside the weld bead region WB2, and the condition of step S34 is not satisfied. Therefore, it is determined as NO in Step S34. For this reason, in the example of FIG. 11B, the region from the lower deformation region WL2 to the upper deformation region WU4 is specified as the weld bead region WB.

  If it determines with NO by step S34, the arithmetic unit 12 will calculate the crank angle of a weld bead in step S38. As shown in FIG. 12, the crank angle θ is determined based on the surface shape of the weld bead region WB and the first plate material two-dimensional at the boundary between the weld bead region WB and the region W32 where the plate material 32 is exposed (hereinafter referred to as the first plate material region). It is an angle between the reference shape 74. In step S38, first, the arithmetic unit 12 extracts the coordinate point group of the two-dimensional measurement shape 70 in the weld bead region WB in the vicinity of the boundary between the weld bead region WB and the first plate material region W32. Next, a straight line 78 shown in FIG. 12 is calculated by linearly approximating the extracted coordinate point group by a robust estimation method such as LMedS (least square median estimation) or Hough transform. Then, the crank angle θ between the straight line 78 and the first plate material two-dimensional reference shape 74 is calculated. In addition, if the rise of the weld bead is too large, the coordinates of the two-dimensional measurement shape 70 in the weld bead region WB in the vicinity of the boundary portion may be abnormal values. However, if the robust estimation method described above is used, the crank angle θ can be accurately calculated while suppressing the influence of the abnormal value even when the abnormal value is included.

  In step S40, as shown in FIG. 12, the arithmetic unit 12 determines the boundary point Ca between the first plate material region W32 and the weld bead region WB, and the boundary between the region W34 where the plate material 34 is exposed and the weld bead region WB. A line segment CaCb connecting the point Cb is drawn. Next, the coordinate point Cc in the weld bead region WB farthest from the line segment CaCb is specified. Then, a distance Hc between the line segment CaCb and the coordinate point Cc is calculated. Hereinafter, the distance Hc is referred to as extra height Hc.

  The crank angle θ and the extra height Hc are closely related to a defect called overlap. The overlap is a problem in which the weld bead and the plate material 32 are not sufficiently melted and a sufficient strength cannot be obtained at the welded portion. When the overlap occurs, the crank angle θ increases and the surplus height Hc increases. According to this inspection apparatus, it is possible to inspect the overlap based on the crank angle θ and the surplus height Hc.

  In step S42, the arithmetic unit 12 measures other parameters of the weld bead. Thereby, the characteristics of the other weld beads of the overlap described above are evaluated. For example, in step S42, the dimensions (width, thickness, etc.) of the weld bead, the width of the gap between the plate member 32 and the plate member 34, the presence or absence of an undercut region (concave region), the depth of the undercut region, the pit ( The presence / absence of a minute dent formed on the surface of the weld bead is calculated.

  In step S44, the arithmetic unit 12 determines the quality of the weld bead based on the various parameters of the weld bead calculated up to step S42. If it is determined that the weld bead is defective, the administrator is notified. By the above processing, the arithmetic unit 12 ends the welding bead inspection processing.

  As described above, the weld bead inspection apparatus according to the embodiment is based on the distribution of the inclination angle (that is, the normal vector) of the measurement shape and the distribution of the inclination angle of the reference shape of the plate 32. The region where the plate member 32 is exposed is specified. Since the characteristic of the surface shape of the plate member 32 appears more significantly in the tilt angle, the region where the plate member 32 is exposed can be accurately specified from the measured shape based on the tilt angle. The weld bead inspection device performs alignment between the reference shape of the plate material 32 and the measurement shape using the entire region where the specified plate material 32 is exposed as a fitting region. Thus, by specifying the area where the plate member 32 is exposed based on the measurement result and setting the entire specified area as the fitting area, the fitting area can be secured most widely. For this reason, according to this weld bead inspection apparatus, even if the area | region where the board | plate material 32 is exposed is narrow, a fitting area | region can be ensured to the maximum within the narrow area | region. For this reason, this weld bead inspection device can accurately align the reference shape of the plate 32 and the measurement shape. As a result, the weld bead can be accurately inspected even if the region where the plate member 32 is exposed is narrow. Further, even when the region where the plate member 32 is exposed is wide, there is little change in the surface shape of the plate member 32, and it may be difficult to align the reference shape of the plate member 32 and the measurement shape. According to the weld bead inspection apparatus of the embodiment, since the fitting area can be set as wide as possible, even in such a case, the reference shape and the measurement shape of the plate member 32 are aligned using the wide fitting area. Can be done accurately.

  In the above-described embodiment, the case where the plate members 32 and 34 are welded with being overlapped in the thickness direction has been described. However, as shown in FIG. 13, the welded parts that are fillet welded (T-shaped joints) are inspected. This technique can also be used in some cases.

Specific examples of the present invention have been described in detail above, but these are merely examples and do not limit the scope of the claims. The technology described in the claims includes various modifications and changes of the specific examples illustrated above.
The technical elements described in this specification or the drawings exhibit technical usefulness alone or in various combinations, and are not limited to the combinations described in the claims at the time of filing. In addition, the technology illustrated in the present specification or the drawings achieves a plurality of objects at the same time, and has technical utility by achieving one of the objects.

10: Weld bead inspection device 12: Arithmetic device 14: Storage device 20: Surface shape measuring device 30: Welding part 32: Plate material 34: Plate material 34b: End surface 36: Weld bead 38: Linear region 42b: Line 60: Three-dimensional measurement Shape 62: Three-dimensional reference shape 64: First plate member three-dimensional reference shape 66: Second plate member three-dimensional reference shape 68: Fitting region 70: Two-dimensional measurement shape 72: Two-dimensional reference shape 74: First plate member two-dimensional reference shape 76: Second plate 2D reference shape 78: Straight line 80: Fitting region

Claims (12)

An inspection apparatus of weld bead welded structure first plate member and the overlapping of the second plate member and the second plate member having a curved surface shape on the first plate member on which has a curved shape,
An input means for receiving measurement shape data describing a measurement shape obtained by measuring the surface shape of the structure in a cross section intersecting the weld bead;
First storage means for storing first plate material reference shape data describing a first plate material reference shape which is a reference shape of a surface shape of the first plate material before welding in the cross section;
A first angle adjusting means for adjusting an angle of the first plate material reference shape with respect to the measurement shape by rotating the first plate material reference shape;
An area specifying means for specifying an area where the first plate material is exposed from the measurement shape based on the first plate material reference shape in which the measurement shape and the angle are adjusted ,
A weld bead inspection device characterized by comprising: Second storage means for storing second plate material reference shape data describing a second plate material reference shape which is a reference shape of the surface shape of the second plate material before welding in the cross section;
A second angle adjusting means for adjusting an angle of the second plate material reference shape with respect to the measurement shape by rotating the second plate material reference shape;
Further comprising
The area specifying means specifies the area where the second plate material is exposed from the measurement shape based on the second plate material reference shape adjusted in the measurement shape and angle.
The weld bead inspection device according to claim 1. The first angle adjusting means adjusts the position of the first plate material reference shape with respect to the measurement shape based on the inclination angle distribution of the measurement shape and the inclination angle distribution of the first plate material reference shape, and then measures the first plate material reference shape. The weld bead inspection device according to claim 1, wherein an angle with respect to the shape is adjusted. In the structure, the end surface of the second plate member is disposed on the first plate member, and the region including the end surface of the second plate member is welded,
The second plate material reference shape includes the reference shape of the end surface shape of the second plate material,
Area identification means
A comparison shape calculation means for calculating a comparison shape obtained by connecting an extension line of the first plate material reference shape and an extension line of the end surface shape of the second plate material reference shape;
A difference value calculating means for calculating a difference value of coordinates between a coordinate point on the measurement shape and a coordinate point on the comparison shape;
Means for identifying a region where the first plate member is exposed and a region where the second plate member is exposed based on the difference value;
The weld bead inspection device according to any one of claims 1 to 3, wherein: A first plate member and the inspection method of the weld bead of the structure obtained by welding the second plate overlapping the second plate member having a curved surface shape on the first plate member on which has a curved shape,
On the computer,
An input step for inputting measurement shape data describing a measurement shape obtained by measuring the surface shape of the structure in a cross section intersecting the weld bead;
A first storage step of storing first plate material reference shape data describing a first plate material reference shape which is a reference shape of a surface shape of the first plate material before welding in the cross section;
A first angle adjustment step of adjusting an angle of the first plate material reference shape with respect to the measurement shape by rotating the first plate material reference shape;
An area specifying step for specifying an area where the first plate material is exposed from the measurement shape based on the first plate material reference shape in which the measurement shape and the angle are adjusted ,
The inspection method characterized by making it perform. On the computer,
A second storage step of storing second plate material reference shape data describing a second plate material reference shape which is a reference shape of a surface shape of the second plate material before welding in the cross section;
A second angle adjusting step for adjusting an angle of the second plate material reference shape with respect to the measurement shape by rotating the second plate material reference shape;
Is executed further,
In the region specifying step, the computer is caused to specify a region where the second plate material is exposed from the measured shape based on the second plate material reference shape in which the measurement shape and the angle are adjusted.
The inspection method according to claim 5. In the first angle adjustment step, the computer is caused to adjust the position of the first plate reference shape relative to the measurement shape based on the inclination angle distribution of the measurement shape and the inclination angle distribution of the first plate reference shape, and then the first plate reference shape The inspection method according to claim 5, wherein an angle with respect to the measurement shape is adjusted. In the structure, the end surface of the second plate member is disposed on the first plate member, and the region including the end surface of the second plate member is welded,
The second plate material reference shape includes the reference shape of the end surface shape of the second plate material,
In the region identification step, the computer
A comparison shape calculation step for calculating a comparison shape obtained by connecting an extension line of the first plate material reference shape and an extension line of the end surface shape of the second plate material reference shape;
A difference value calculating step for calculating a difference value of coordinates between a coordinate point on the measurement shape and a coordinate point on the comparison shape;
Identifying a region where the first plate member is exposed and a region where the second plate member is exposed based on the difference value;
The inspection method according to claim 5, wherein the inspection method is executed. A program used for inspecting a weld bead of a structure in which a first plate member and a second plate member are welded by overlapping a second plate member having a curved shape on a first plate member having a curved shape,
On the computer,
An input step for inputting measurement shape data describing a measurement shape obtained by measuring the surface shape of the structure in a cross section intersecting the weld bead;
A first storage step of storing first plate material reference shape data describing a first plate material reference shape which is a reference shape of a surface shape of the first plate material before welding in the cross section;
A first angle adjustment step of adjusting an angle of the first plate material reference shape with respect to the measurement shape by rotating the first plate material reference shape;
An area specifying step for specifying an area where the first plate material is exposed from the measurement shape based on the first plate material reference shape in which the measurement shape and the angle are adjusted ,
A program characterized by having executed. On the computer,
A second storage step of storing second plate material reference shape data describing a second plate material reference shape which is a reference shape of a surface shape of the second plate material before welding in the cross section;
A second angle adjusting step for adjusting an angle of the second plate material reference shape with respect to the measurement shape by rotating the second plate material reference shape;
Is executed further,
In the region specifying step, the computer is caused to specify a region where the second plate material is exposed from the measured shape based on the second plate material reference shape in which the measurement shape and the angle are adjusted.
The program according to claim 9. In the first angle adjustment step, the computer is caused to adjust the position of the first plate reference shape relative to the measurement shape based on the inclination angle distribution of the measurement shape and the inclination angle distribution of the first plate reference shape, and then the first plate reference shape The program according to claim 9 or 10, wherein an angle with respect to the measured shape is adjusted. In the structure, the end surface of the second plate member is disposed on the first plate member, and the region including the end surface of the second plate member is welded,
The second plate material reference shape includes the reference shape of the end surface shape of the second plate material,
In the region identification step, the computer
A comparison shape calculation step for calculating a comparison shape obtained by connecting an extension line of the first plate material reference shape and an extension line of the end surface shape of the second plate material reference shape;
A difference value calculating step for calculating a difference value of coordinates between a coordinate point on the measurement shape and a coordinate point on the comparison shape;
Identifying a region where the first plate member is exposed and a region where the second plate member is exposed based on the difference value;
The program according to any one of claims 9 to 11, wherein the program is executed.

Images