Method for performing quality control of component of wind turbine to be manufactured

Abstract

Method for performing quality control of a component of a wind turbine to be manufactured, in which geometric data are collected using a measuring device as a non-contact 3D scanner, which geometric data provide a 3D model of the actual shape of at least a portion of the component, a deviation between an actual shape and an ideal shape of the at least one portion of the component is detected using the geometric data.

Description

Technical Field

[0001] This invention relates to a method for performing quality control on components of a wind turbine to be manufactured. Background Technology

[0002] For components of wind turbines, particularly bearings or bearing assemblies, the tolerances of their shapes are crucial. This will be illustrated by the example of a ring in a contact ball bearing or contact roller bearing. Such a ring typically includes raceways in which the balls or rollers are guided. Therefore, the correct shape of the raceway profile ensures that excessive contact pressure is avoided. If the corresponding shape deviates too much from the ideal shape, bearing blockage or at least excessive wear will occur, thus shortening the bearing's life. Therefore, it is extremely important to perform quality control on the geometry or shape tolerances of the components. Currently, several methods for performing such quality control are known.

[0003] One approach involves using a coordinate measuring machine (CMM) to acquire the three-dimensional shape or geometry of a part, which requires direct or physical contact between the part and the CMM's sensor elements. During measurement, the part and sensor elements move relative to each other, with the displacement of the sensor elements used to collect corresponding data. However, this method is cumbersome, time-consuming, and costly.

[0004] Another method for testing bearing rings with raceways is the so-called bald slip test. Here, a line is drawn across the raceway profile using a black marker. Next, the ball is slid along the line by hand. Where the ball contacts the raceway, the line is worn away. The worn-away portion of the line is then measured using a protractor. However, this method is relatively inaccurate, typically by about + / - 2 degrees.

[0005] Another method is to measure the frictional torque on the component or within the raceway, where, in order to pass appropriate quality control, the torque must not exceed a given limit. However, this test is performed without an applied load and therefore is not under realistic conditions.

[0006] In the past, despite the performance of the described tests and compliance with the corresponding quality controls, failures, such as bearing blockages, have still been observed. Therefore, the object of this invention is to provide an enhanced concept for the corresponding quality control, particularly to ensure sufficient reliability to prevent failures. Summary of the Invention

[0007] According to the invention, this objective is achieved by a method as initially described, wherein geometric data is collected using a measuring device as a non-contact 3D scanner, which provides a 3D model of the actual shape of at least a portion of the component, wherein the geometric data is used to detect the deviation between the actual shape and the ideal shape of said at least a portion of the component.

[0008] Geometric data describes the geometry, or in other words, the extent or dimensions of a part, or more precisely, the extent or dimensions of the outer surface of the part, to provide a 3D model. Therefore, the 3D model describes the actual shape (i.e., the true shape) of the part. The acquired 3D model includes high data quality and allows for rapid processing. Thus, the present invention realizes a complex method for detecting unwanted deviations between the actual shape and the ideal shape, and therefore detecting possible default values ​​with respect to a given tolerance. In particular, there is the possibility of automating data processing to simplify quality control. This simplification is further achieved due to the fact that data collection by a 3D scanner requires very little effort, especially because the use of a non-contact 3D scanner eliminates the need for direct or physical contact between the part and the corresponding scanner's sensor elements.

[0009] Raw data obtained from the measuring device can be transferred to a computer for further processing, analysis, and evaluation. Software implemented on the computer can be used in this process. Geometric data can be raw data or acquired through processing or preparation of raw data.

[0010] The ideal shape specifies the reference shape of the part. The actual shape is equal to the ideal shape if deviations will not exist or if it will perfectly meet the tolerances. The ideal shape can be provided or defined by the shape of an ideal geometry, such as a cuboid, sphere, cylinder, cone, or a combination of these shapes.

[0011] The component can be a bearing or a component of a bearing in a wind turbine. As already described, and particularly for these components, it is crucial that the actual shape is within a given tolerance range. For example, the bearing can be a blade pitch bearing, a main bearing, or a yaw bearing. The blade pitch bearing acts as the mechanical interface between the hub and the blades of the wind turbine. The inner or outer ring of the bearing can be attached to the hub, and the blade can be attached to a corresponding other ring. The blade pitch bearing allows the corresponding blade to rotate about its longitudinal axis.

[0012] The components of a bearing as a part can be the rings of a contact ball bearing or a contact roller bearing. These bearings typically consist of one or more inner rings and one or more outer rings arranged concentrically with each other.

[0013] A ring may include at least one raceway extending along the circumference of the ring, wherein geometric data provides a 3D model of the ring's shape for at least the section of the ring in which the at least one raceway is arranged. The raceway may also be referred to as a guide groove and typically includes a longitudinal direction extending along the circumferential direction of the ring. At least one raceway is typically disposed on the outer surface of the inner ring, and at least one raceway is typically disposed on the inner surface of the outer ring. Balls or rollers are arranged between the raceways opposite each other to allow the ring to rotate relative to each other about its central axis. In particular, each of one or more rings includes two raceways extending parallel to each other.

[0014] Because the balls or rollers of a bearing are guided within the raceways, the corresponding sections of the ring, including one or more raceways, are prone to wear if the geometry does not conform to given tolerances. For contact ball bearings, the raceways typically have a circular shape. For contact roller bearings, the raceways typically have a rectangular shape.

[0015] The 3D scanner used can be an optical scanner or a laser scanner. Therefore, the 3D scanner detects radiation, particularly electromagnetic radiation, reflected or emitted by the component, and the corresponding data is used to acquire geometric data.

[0016] A 3D scanner can be an active scanner. In other words, a 3D scanner emits radiation that is reflected by a part, and the 3D scanner detects the reflected radiation to obtain geometric data. If the 3D scanner is a laser scanner, geometric data can be obtained using the corresponding pulse transmission time and / or phase difference.

[0017] A 3D scanner can be a passive scanner. In other words, a 3D scanner can detect radiation that is reflected or emitted by a part and is not originally generated or emitted by the 3D scanner. Radiation reflected by the part and not emitted by the 3D scanner can be sunlight. A 3D scanner that is an optical scanner can include several optical cameras to obtain geometric data of the part as a three-dimensional image.

[0018] During the collection of geometric data, the 3D scanner can be handheld or mounted on a stand. Preferably, a portable 3D scanner is used, eliminating the need to transport the part to the corresponding scanning location. During the scanning process, the handheld 3D scanner can be guided along the part by the scanner operator. Alternatively, the 3D scanner can be attached to a stand, particularly a tripod or other suitable holding device.

[0019] Preferably, the geometric data consists of a point cloud and / or a polygonal mesh. The points in the point cloud or the nodes in the polygonal mesh may include three coordinates corresponding to three spatial directions. The corresponding coordinates correspond to a specific reference coordinate system, where the origin of the coordinates may be the location of the 3D scanner or another suitable point in space, such as the center point of the part. For simplification, coordinate transformations may be performed, particularly to minimize the numbers describing the coordinates.

[0020] Preferably, the point cloud consists of data points arranged on a grid with a specified, particularly uniformly distributed, resolution. The data points can be provided by coordinates in a Cartesian coordinate system. The grid can be defined as values ​​uniformly or equally distributed about two coordinates (specifically the x and y coordinates), where a third coordinate (specifically the z coordinate) provides the position of the surface of the component. In other words, the data points can be uniformly distributed about the xy plane.

[0021] The geometric data of a 3D model providing the actual shape of at least a portion of a ring-shaped component can be transformed into adjusted geometric data, wherein the adjusted geometric data provides an adjusted 3D model of the actual shape of at least a portion of the component. To transform the geometric data into adjusted geometric data, the circumferential circles of the component can be converted into straight lines. In this embodiment, the component is particularly the ring of a bearing described above. The geometric data is transformed such that the circular shape of the component, particularly the ring, and therefore the circular shape of the raceway, is transformed into a straight line shape. The relative positions between data points (particularly points in the point cloud) and points on the circumferential lines can be maintained or preserved. The adjusted geometric data is suitable for simplified further processing or corresponding analysis.

[0022] At least one subset of data is extracted from geometric data or data obtained from geometric data (especially from adjusted geometric data), said at least one subset of data providing a 2D model of the shape of at least a portion of the cross-section of the component, wherein said at least one subset of data is used to detect deviations. The at least one subset of data provides slices or planes of the component, suitable for simplified further processing or corresponding analysis. Preferably, several subsets of data are extracted, wherein these subsets provide a 2D model of the shape of at least a portion of the cross-section of the component along a certain direction of the component, particularly a longitudinal range. If the component is annular, several subsets of data can be extracted from the adjusted geometric data such that these subsets provide a 2D model of the shape of at least a portion of the cross-section of the component along a circumferential range of the component.

[0023] Analysis of geometric data can involve fitting a function describing an ideal shape to the geometric data or data obtained from the geometric data, particularly to one or more subsets of data, where the best fit is used to obtain the deviation. Fitting can be performed by applying a method that minimizes the chi-square value. For the case of a contact ball bearing, the function can describe a circle in the region of the at least one raceway. For the case of a contact roller bearing, the function can describe a rectangle in the region of the at least one raceway. These geometries are the ideal shapes of the raceways. Therefore, preferably, a function describing the desired cross-section of the at least one raceway is fitted to the portion of the data subset describing the at least one raceway. In particular, if several parallel raceways are provided, the function can be defined segment by segment.

[0024] In a specific embodiment of the method according to the invention, the deviation can be obtained by the deviation between geometric data or data obtained from geometric data and the best-fit function. The deviation between the geometric data or data obtained from geometric data and the best-fit function can be provided by the value of the chi-square, particularly the value of the reduced chi-square. The larger the corresponding chi-square value, the greater the deviation at the corresponding location. In particular, if several fits are performed for different longitudinal positions of the component, a fit with a relatively large deviation between the corresponding data and the function indicates a large deviation between the actual shape and the ideal shape at the corresponding longitudinal position. Therefore, the potentially problematic locations or sections of the component can be identified using the deviation between the geometric data or data obtained from geometric data and the best-fit function.

[0025] Additionally or alternatively, the deviation can be obtained from at least one best-fit result of the best-fit function. In this embodiment, the result of at least one fitting parameter of the best-fit function, in particular the deviation between that result and the expected value of the corresponding fitting parameter corresponding to the ideal shape, serves as an indicator of the current deviation between the actual shape and the ideal shape. Attached Figure Description

[0026] Other objects and features of the invention will become apparent from the following detailed description taken in conjunction with the accompanying drawings. However, the drawings are merely schematic diagrams designed for illustrative purposes only and do not limit the invention. The drawings show: Figure 1 It is a wind turbine that includes a component as a blade pitch bearing, wherein a method according to an embodiment of the invention is performed in the context of quality control of this component. Figure 2 yes Figure 1 Detailed view of the blade pitch bearing component of a wind turbine. Figure 3This is a schematic diagram of a method according to an embodiment of the present invention. Figure 4 It includes components and is used in Figure 3 The arrangement of 3D scanners to acquire geometric data within the context of the method. Figure 5 It is a schematic diagram with data points as geometric data. Figure 6 It is possessed as a result of Figure 5 A schematic diagram of the data points of the adjusted geometric data obtained from the geometric data. Figure 7 It has the characteristics of Figure 6 A schematic diagram of the data subset obtained from the adjusted geometric data, along with the points of the best-fit function, and Figure 8 This is a schematic diagram of the best-fit results with several subsets of data. Detailed Implementation

[0027] Figure 1 A main sketch of a wind turbine 1, including a tower 2, is shown, with a nacelle 3 arranged on top of the tower. Furthermore, the wind turbine 1 includes a hub 4 on which several (i.e., three) blades 5 are arranged. To collect wind energy, the blades 5 interact with the wind, which in turn causes the hub 4 to rotate, and this rotation is transferred to a generator located within the nacelle 3.

[0028] exist Figure 1 In the diagram, one of the blades 5 is skipped, making the bearing 6 of the wind turbine 1, which is a contact ball bearing, visible. Alternatively, the bearing 6 could be a contact roller bearing. Specifically, the bearing 6 is a blade pitch bearing, which acts as a mechanical interface between the hub 4 and the corresponding blade 5. The bearing 6 allows the corresponding blade 5 to rotate or tilt about its longitudinal axis 7.

[0029] Figure 2 An enlarged three-dimensional cut view of bearing 6 is shown. Bearing 6 comprises two rings 8, namely an inner ring 8 and an outer ring 8. The rings 8 are arranged concentrically with each other, wherein these rings face each other on the outer surface of the inner ring 8 and the inner surface of the outer ring 8. On each of these surfaces, two adjacent raceways 9 are provided, which may be referred to as guide grooves. Balls, not shown in the figure, are arranged between the two opposing raceways 9. Each of the raceways 9 extends along the circumference of the corresponding ring 8. The two raceways 9 of each of the rings 8 extend parallel to each other. Thus, the rings 8 can rotate about each other, wherein one of the rings 8 is connected to the hub 4 and the other ring 8 is connected to the blade 5.

[0030] Next, embodiments of the method according to the present invention will be described. Figure 3A flowchart of the method is shown, wherein a specific embodiment of the method includes steps 10 to 16. The purpose of the method is to perform quality control on component 17 of the wind turbine 1 to be manufactured, wherein component 17 is a member or bearing 6, i.e., the inner ring 8. The corresponding method can also be applied to other components of the wind turbine 1, particularly to the outer ring 8.

[0031] Regarding the first step 10 of the method, an inner roller 8 or corresponding component 17 is provided, along with a non-contact 3D scanner 18. Exemplarily, the portable 3D scanner 18 is mounted on a stand 19, which serves as a tripod, with the 3D scanner 18 positioned next to component 17. Figure 4 This scenario is illustrated in the image, where only a portion of ring 8 is shown. Instead of being mounted on bracket 19, the 3D scanner 18 can be handheld by the operator during subsequent procedures.

[0032] According to this embodiment, the 3D scanner 18 is an active scanner, but it can also be a passive scanner. Basically, the 3D scanner 18 can be an optical scanner or a laser scanner. (See reference...) Figure 4 In the scenario shown, the 3D scanner 18 is an active laser scanner. The 3D scanner 18 includes a radiation source 20 as a laser source, wherein, during the next step 11 of the method, corresponding laser radiation 21 is emitted by the radiation source 20. The emitted radiation 21 is reflected by the component 17, wherein the reflected radiation 23 is detected by the radiation or laser detector 22 of the 3D scanner 18. By performing pulse transmission time and phase difference analysis, geometric data 24 providing a 3D model of the actual shape of the component 17 is obtained, and this geometric data is transferred to a computer for further processing.

[0033] Next, refer to Figure 5 This figure illustrates a schematic diagram or coordinate system with geometric data 24. The 3D model provided by geometric data 24 consists of a point cloud. Additionally or alternatively, geometric data 24 or the corresponding 3D model may consist of a polygonal mesh. (See figure from...) Figure 5 It becomes apparent that the shape of the point cloud corresponds to the actual or real shape of component 17, wherein, in particular, the raceway 9 becomes visible. Thus, the geometric data 24 provides a 3D model of the segment of ring 8 including the raceway 9, such that the geometric data 24 provides the geometry of the raceway 9.

[0034] In the next step 12, the geometric data 24 undergoes a coordinate transformation so that the origin of the coordinate system corresponds to the center point of ring 8. In other words, the axis of the coordinate system on which the geometric data 24 is based is offset.

[0035] Regarding the next step 13, the geometric data 24 is transformed into adjusted geometric data 25. The adjusted geometric data 25 provides an adjusted 3D model of component 17 or, correspondingly, raceway 9. To transform geometric data 24 into adjusted geometric data 25, the circumferential circles extending along the longitudinal direction of raceway 9 are removed, or more precisely, transformed into straight lines. In other words, geometric data 24 is adjusted so that the circumferential lines of ring 8 or, correspondingly, raceway 9 become straight lines, wherein the relative positions between points in the point cloud and points on the corresponding lines are preserved. Therefore, Figure 6 The result of this coordinate transformation (adjusted geometry 25) is shown in the figure, which illustrates a schematic diagram or coordinate system with the adjusted geometry 25. As will become apparent, the adjusted geometry 25 consists of data points arranged on a grid with a specified and equally or uniformly distributed resolution.

[0036] Next, during step 14, several data subsets 26 are extracted from the adjusted geometric data 25. Each of the data subsets 26 provides a 2D model of the cross-sectional shape of the ring 8 or corresponding raceway 9. Specifically, the data subsets 26 provide cross-sectional cuts through the ring 8 or correspondingly through the raceway 9, wherein these cuts are arranged to be equally spaced along the longitudinal extent of the raceway 9 and thus along the circumferential extent of the ring 8. Figure 7 A schematic diagram or coordinate system with one of the data subsets 26 is shown.

[0037] Next, in step 15, a function describing the ideal shape of component 17 or corresponding raceway 9 is fitted to each of the data subset 26, where the best fit result 27 is obtained. Figure 7 The best-fit function is indicated by a small circle. This function is defined segment by segment and describes the ideal cross-sectional shape of raceway 9, i.e., each raceway is circular. For the case where component 17 is a contact roller bearing, the function can describe a rectangle. Each of the best-fit results 27 includes best-fit parameters, such as the radius and center point of the fitted circle, and the corresponding chi-square value.

[0038] Next, in step 16, output Figure 8 The schematic diagram shown is used to evaluate the results of quality control. In this diagram, the best-fit result 27 is illustrated. Specifically, curves are plotted showing the relationship between the values ​​of the best-fit parameters and the chi-square value and the position of the corresponding data subset 26, with respect to the longitudinal range of raceway 9 and therefore the circumferential range of ring 8. Locations of large profile inhomogeneities in raceway 9 can be identified by these locations on the x-axis of the schematic diagram; that is, locations where there is a large deviation between the best-fit result 27 and the corresponding value expected for the ideal shape, and / or where the chi-square value is large or correspondingly exceeds a given threshold.

[0039] Although the present invention has been described in detail with reference to preferred embodiments, the invention is not limited to the disclosed examples, and those skilled in the art can derive other variations from these examples without departing from the scope of the invention.

[0040] Regardless of the use of grammatical terms, individuals who identify as masculine, feminine, or other genders are included in this term.

Claims

1. A method for performing quality control on components (17) of a wind turbine (1) to be manufactured, wherein, A measuring device (18) as a non-contact 3D scanner is used to collect geometric data (24), which provides a 3D model of the actual shape of at least a portion of the component (17), wherein the geometric data (24) is used to detect the deviation between the actual shape and the ideal shape of the at least one portion of the component (17).

2. The method according to claim 1, characterized in that, The component (17) is the bearing (6) of the wind turbine or a component of the bearing (6).

3. The method according to claim 2, characterized in that, The bearing (6) is a blade pitch bearing, a main bearing, or a yaw bearing.

4. The method according to claim 2 or 3, characterized in that, The component of the bearing (6) that serves as part (17) is a ring (8) of a contact ball bearing or a contact roller bearing.

5. The method according to claim 4, characterized in that, The ring includes at least one raceway (9) extending along the circumference of the ring (8), wherein the geometry (24) provides a 3D model of the shape of the ring (8) for at least the section of the ring (8) in which the at least one raceway (9) is arranged.

6. The method according to claim 5, characterized in that, The ring (8) includes two raceways (9) that extend parallel to each other.

7. The method according to any one of the preceding claims, characterized in that, The 3D scanner (18) used is an optical scanner or a laser scanner.

8. The method according to any one of the preceding claims, characterized in that, During the collection of the geometric data (24), the 3D scanner (18) is either handheld or mounted on a stand (19).

9. The method according to any one of the preceding claims, characterized in that, The geometric data (24) or the data obtained from the geometric data (24) consists of point clouds and / or polygon meshes.

10. The method according to claim 9, characterized in that, The point cloud consists of data points arranged on a grid with a specified, particularly uniformly distributed, resolution.

11. The method according to any one of the preceding claims, characterized in that, The geometric data (24) of the 3D model providing the actual shape of at least a portion of the annular component (17) is transformed into adjusted geometric data (25), wherein the adjusted geometric data (25) provides an adjusted 3D model of the actual shape of at least a portion of the component (17), wherein, in order to transform the geometric data (24) into the adjusted geometric data (25), the circumferential circle of the component (17) is transformed into a straight line range.

12. The method according to any one of the preceding claims, characterized in that, At least one data subset (26) is extracted from the geometric data (24) or from the data obtained from the geometric data (24), the at least one data subset providing a 2D model of the shape of at least a portion of the cross section of the component (17), wherein the at least one data subset (26) is used to detect the deviation.

13. The method according to claim 12, characterized in that, Extract several data subsets (26), wherein the data subsets (26) provide a 2D model of the shape of at least a portion of the cross section of the component (17) along a certain direction, particularly the longitudinal range.

14. The method according to any one of the preceding claims, characterized in that, The function describing the ideal shape is fitted to the geometric data (24) or the data obtained from the geometric data (24), wherein the deviation is obtained using the result of the best fit (27).

15. The method according to claim 14, characterized in that, The deviation is obtained by the deviation between the geometric data (24) or the data obtained from the geometric data (24) and the best-fit function and / or by at least one best-fit result (27) of the best-fit function.

Images