Method for detecting leaks in a fluid guide of a heat exchanger
Patent Information
- Application Number
- JP2023571584
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2021-06-11
- Filing Date
- 2022-06-10
- Publication Date
- 2025-01-08
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Suction leak detection for heat exchange devices like refrigerators and air conditioning units is time-consuming and prone to human error due to manual operation, often missing critical test sections.
A 3D sensor generates a point cloud of the device's outer surface, identifying target leak points, and a virtual model determines an approach path for a suction probe to automatically and accurately reach these points, avoiding collisions.
Enables rapid and reliable leak detection in fluid guides of heat exchange devices, reducing human error and improving efficiency in quality control processes.
Smart Images

Figure 00000000_0000_ABST
Abstract
Description
[Technical field]
[0001] The present invention relates to a leak detector for detecting a leak in a test object. [Background technology]
[0002] A leak detector is described, for example, in DE 10 2005 022 157 A1 (Inficon). This leak detector comprises a probe, the tip of which is placed against a defined area of a test object. The test object is filled with a test gas, for example helium. The leaking test gas is sucked in from the probe tip into a body and is fed to a test gas detector, which can be designed, for example, as a mass spectrometer.
[0003] When using a leak detector, a test object, e.g. an air conditioning system or a cooling group of a refrigerator, is filled with a test gas and the suction probe can be used to detect whether the test gas is leaking out of the test object. In quality inspection of industrial products, the probe tip is applied to a specific test part of the test object that may be leaking. The probe is then moved manually to the test part. During this process, it is difficult to ensure whether the probe tip has been moved to all the target test parts of the test object. The operator may accidentally miss some test parts or skip other test parts that he or she considers to be unimportant according to his or her own subjective judgment.
[0004] US Pat. No. 4,945,305 (Ascension) describes a position determination system consisting of a transmitter generating a pulsed DC magnetic field and a receiver mounted on the object. This method is highly suitable for determining the instantaneous position, comparing it with a desired position, and feedback. Non-magnetic objects in the line of sight between the transmitter and receiver do not cause disturbances. As the structure under test is stationary, residual disturbances from large magnetic metals such as compressor blocks and refrigeration machines can be negligible after calibration. This method has been extremely useful in realizing the present disclosure.
[0005] WO 2009 / 016160 A1 (Patent Document 3) discloses a leak detector with a main body connected to a probe by a hose. The tip of the probe is placed against a test area or part of the test object. If test gas leaks from the test object, this is detected by a gas detector in the main body. A position determination system is provided, which consists of a transmitter, a receiver arranged inside the probe, and a supply evaluation part. This monitors and checks whether the tip of the probe is located at the respective test part.
[0006] In the field of automation, it is known to collect digital data of an object with a 3D sensor to obtain a point cloud in a three-dimensional virtual space, each point being a surface point on the outer surface of the object. Robots, such as those used in the automation industry for the automated manufacturing of products such as cars, are controlled by software algorithms using the collected digital data. This is typically used to detect the entire perimeter or outer surface of an object in order to position, grasp, move or reposition the object or to paint the surface of the object with a paint spraying robot arm, for example in the vehicle manufacturing industry. [Prior art documents] [Patent documents]
[0007] [Patent Document 1] DE 102005022157 [Patent Document 2] U.S. Pat. No. 4,945,305 [Patent Document 3] International Publication No. 2009 / 016160 Summary of the Invention [Problem to be solved by the invention]
[0008] Suction leak detection for heat exchange devices such as refrigerators, air conditioning units, and heat pumps is a part of the quality inspection of heat exchange devices. An operator must visually identify the target parts to be inspected, such as the liquid guide tubes of the heat exchanger, and manually move the probe to each target part in turn. This type of suction leak detection for heat exchange devices is not only time-consuming, but also prone to human error, such as overlooking the target part or not getting the suction probe close enough to the target part.
[0009] SUMMARY OF THE PRESENT DISCLOSURE One object of the present invention is to more reliably and quickly detect a leak in a fluid guide of a heat exchanger. [Means for solving the problem]
[0010] The method according to the invention is defined by the features of independent claim 1.
[0011] In other words, a) using a 3D sensor, a cloud of points in a three-dimensional virtual space is acquired from the heat exchange device, each point forming a surface point of an outer surface of the heat exchange device; b) after the point cloud is obtained, a structure corresponding to a tube of a predetermined outer diameter is searched for in the point cloud (the tube may correspond to a fluid guide of the heat exchange device in which a leak needs to be found); c) at least one target portion where the diameter or direction of the tube changes or where the tube ends is detected from the structure obtained in b), for example by an increase in diameter, which corresponds to a welding or soldering point or a welding or soldering point, which corresponds to a joint or junction with a further tube, which corresponds to a junction where the tube is joined to another tube by welding, soldering, etc., or which corresponds to an end of the tube, etc. (note that the change in direction of the tube can be a bending of the tube, whereby the direction of the main longitudinal axis (central axis) of the tube changes, and also the change in direction of the tube can be a bending of the outer wall of the tube, which target portion constitutes a weak point in the fluid guide to be detected, and therefore is of great importance for the detection of suction leaks during quality inspection in the automated manufacturing process of heat exchange devices), d) once the target portion is found, the method of the present invention determines an approach path for bringing a suction probe of a gas leak detector of a gas leak detection system into proximity with the target portion within the point cloud (which may typically be performed by calculating multiple approach paths within the point cloud in the virtual space prior to actually bringing an actual suction probe into proximity with the heat exchanger); e) Once an approach path has been determined, the suction probe is manually and physically brought close to the test portion of the heat exchange device along the approach path (wherein the test portion on the actual heat exchange device corresponds to the target portion obtained within the point cloud).
[0012] In other words, the present invention uses a 3D sensor to obtain a virtual model or simulation of at least a part of a heat exchange device, and locates leak candidate areas (tested parts on the actual heat exchange device) corresponding to each target part of the virtual model from the virtual model or simulation. In the present invention, before the suction probe actually physically approaches such test parts, a candidate approach path is determined using the virtual model instead of the actual heat exchange device. In this way, the suction probe can be physically approached to the test parts only after the approach path to each target part of the virtual model is determined.
[0013] As a result, the present invention can automatically detect the suction leak of the fluid guide of the heat exchange device more reliably and quickly. If the fluid guide is small, it may be hidden inside or at the back of the outer periphery of the heat exchange device. Specifically, in general, the location of the fluid guide structure to be inspected changes depending on the type of heat exchange device. Therefore, the present invention can automatically perform a reliable and rapid search even if the type of heat exchange device changes.
[0014] Suitably, the method is performed automatically by a leak detection system comprising said gas leak detector, preferably a leak detection system comprising said gas leak detector and a robot.
[0015] Advantageously, the point cloud in the three-dimensional virtual space is generated from image data obtained from an optical camera, preferably from at least two optical cameras of an imaging system. The 3D sensor may be an imaging system comprising at least one optical camera and at least one lighting device (e.g. LED, laser source, etc.).
[0016] After step a), the point cloud can be compared with digital reference data of the structure of the heat exchanger device that is to be located in the point cloud. This can be used to identify and / or locate the structure, e.g. a component, e.g. a frame, of the heat exchanger device, e.g. a compressor. Preferably, the digital reference data is pre-obtained CAD data of the heat exchanger device. The digital reference data can be stored in a memory of the inventive leak detection system.
[0017] In one embodiment, at least one search portion in the point cloud may be selected or obtained based on a comparison of the point cloud with the digital reference data. The search portion may be obtained using position data of the position of the heat exchange device in the point cloud, which may be obtained in step b). For example, once the structure corresponding to a tube of a given outer diameter is obtained in step b), the position of the structure may be obtained and stored. The target portion for which leak detection will be performed may be obtained by selecting the search portion using this position data and searching for the presence or absence of a diameter change for the obtained structure or tube.
[0018] The reference data may be generated by the 3D sensor of the leak detection system from the same heat exchange device or from a corresponding device of the same type.
[0019] Step c) may be performed by tracing and analysing the points found in step b) that correspond to said tubes in said point cloud.
[0020] In step d) the ideal position of the suction probe for gas leak detection in the tube or tested part can be calculated.
[0021] Step d) may include the sub-step of determining whether the determined approach path will lead to a collision between the suction probe and a structure of the heat exchanger device, and the sub-step of determining a further approach path if a collision is expected on the determined approach path.
[0022] Step d) may be repeated until an approach path is determined in which no collision between the suction probe and the structure of the heat exchange device is expected. Alternatively or additionally, step d) may use previously obtained and stored digital data, e.g. CAD data, of the suction probe and / or the robot performing step e). In particular, at least step e) may be performed by a robot, which may be part of the leak detection system.
[0023] At least one of steps a) to d) may be performed by a software algorithm of a leak detection system comprising a leak detector, preferably a leak detection system comprising a leak detector and a robot as described above.
[0024] Step d) may include a substep of generating, in or by the cloud of points in the virtual space, a trajectory to be followed by a robot holding the suction probe when performing an actual leak detection.
[0025] At least steps b) to d) may be repeated for a number of target regions, after which step e) may be performed each time to subsequently approach the suction probe to a respective test portion corresponding to the target portion.
[0026] The method of the present invention may further comprise, after step e), f) suctioning the test area with the suction probe to perform a leak detection within the test portion of the tube and identify possible leaks in the fluid conducting structure; can be executed.
[0027] The suction probe may have a generally U-shaped sniffer tip with two suction arms extending distally, at least one of the suction arms including an inlet opening, and the U-shaped sniffer tip is brought close to the test portion with the two sniffer tip arms positioned on either side of the test tube.
[0028] If in step d) it is determined that the approach path will lead to a collision between the suction probe and a structure of the heat exchange device, the suction probe or the suction tip of the suction probe may be automatically reconfigured or deformed, or automatically replaced with another suction probe or suction tip.
[0029] The method of the present invention can also deal with other types of heat exchange devices, provided that information about the position of at least one target part and / or about the point at which the diameter or direction of at least one (further) tube changes and / or about the direction in which the search is performed within the target part and / or about the predetermined outer diameter searched for in step b).
[0030] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. [Brief description of the drawings]
[0031] [Figure 1] 1 is a schematic diagram of a leak detection system and a heat exchange device. [Diagram 2] FIG. 2 is a schematic diagram of a point cloud in a virtual space obtained from the heat exchange device of FIG. 1. [Diagram 3] FIG. 4 is a schematic diagram showing details of another heat exchange device. [Figure 4] FIG. 4 is a schematic diagram of a point cloud in a virtual space obtained from the heat exchange device of FIG. 3. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0032] In FIG. 1, a heat exchanger 12 in the form of a refrigerator is shown. In the present disclosure, the term "heat exchanger" refers to a system or device consisting of a heat exchanger, such as a refrigerator, an air conditioning system, a heat pump, etc. The heat exchanger comprises, at its lower rear side 14, a number of fluid guides 16, 18 in the form of tubes that are welded to the rear side of the refrigerator during its manufacture. As part of a fully automated quality control, a specific test part 20 must be identified in order to carry out a leak detection at said test part 20. The test part 20 should be considered as a part of the heat exchanger where tubes are attached, welded or soldered to each other, where a tube is terminated, where a tube is joined to another tube, or where tubes meet or intersect with each other. In the embodiment shown in FIG. 1, a first fluid guide 16 is arranged vertically and a second fluid guide 18 is joined to the first fluid guide 16 at one end. The portion where tubes 16, 18 are joined is considered to be a test portion 20, and is shown in FIG. 1 by a dashed circle.
[0033] 3 and 4 show a more general example in which two tubes 16, 18 of different diameters are joined together along a common longitudinal axis 17.
[0034] The object of the present invention is to automatically move the suction probe 22 of the suction leak detector 24 to the test portion 20 so that the suction tip 25 of the suction probe 22 is close enough to the test portion 20 to suck in gas leaking from any leaks in any of the piping 16, 18 in the test portion 20. In typical fashion, the suction probe 22 is connected to the gas leak detector 24 by a connecting hose 26.
[0035] The suction probe 22 is attached to the tip 28 of a robotic arm 30 of a robot 32 .
[0036] A 3D sensor 34 in the form of an imaging system 36 comprising two optical cameras 38, 40 and an illumination device 42 in the form of an LED lamp collects digital image data from the lower rear side 14 of the heat exchange device 12. In a generally known manner, the illumination device 42 illuminates the heat exchange device 12, in particular the lower rear side 14 of the heat exchange device 12. The cameras 38, 40 image the reflected light and the imaging system 36 generates digital image data from which a group 44 of points 46 in a three-dimensional virtual space 48 is obtained.
[0037] 2 and 4 are graphical representations of a simplified and reduced-level acquired point cloud 44. In these figures, the point cloud 44 is reduced to a two-dimensional matrix for illustrative purposes only. In reality, the point cloud 44 acquired by the 3D sensor 34 is three-dimensional, such as in the form of a three-dimensional grid or mesh. That is, each point 46 of the point cloud 44 includes three coordinates x, y, z and an amplitude value that represents, for example, the intensity of the reflected illumination captured by the cameras 38, 40.
[0038] In the present invention, the point cloud 44 is searched for structures corresponding to a tube of a predetermined outer diameter. For example, from the amplitude value of point 46, a value within a certain range and within a certain distance of a geometric structure of a tube of a predetermined outer diameter corresponding to the diameter of tubes 16, 18 can be searched for.
[0039] 2 and 4, the points 44 corresponding to the tubes 16, 18 are labeled 46a and are represented as filled circles, while the remaining points 46 are represented as unfilled circles. Lines 50 connect the points 46a to each other, thereby visualizing the structure or representation 15, 19 of the tubes 16, 18 of the heat exchanger 12 in a virtual space 48.
[0040] The term “virtual space” generally refers to the simulated environment of the heat exchange device 12 , the robot 32 , the robot arm 30 and the suction probe 22 , as well as the space between the probe 22 and the heat exchange device 12 .
[0041] In the method of the present invention, the next step is to search for changes in diameter from the obtained structures or points 46a. For example, by following the points 46a or lines 50 corresponding to the horizontal tube 18 on the left side from left to right in Fig. 2, the diameter of the structural representation 19 of the tube 18 changes as it merges and joins the structural representation 19 of the tube 18 in the virtual space 48 with the structural representation 15 of the vertical tube 16. For example, this can be found by analyzing the distance between pairs of points 46a and comparing it with each distance between further points 46a along the line 50. By analyzing from left to right, it appears as if the points 46a of the structural representation 15 of the vertical tube 16 increase the diameter of the structural representation 19 of the tube 18.
[0042] Similarly, when analyzing the structural representation 15 of the vertical tube 16 from bottom to top, the distance between the pairs of points 46a increases where the structural representation 15 of the vertical tube 16 merges with the structural representation 19 of the horizontal tube 18, appearing as if the diameter of the tube 16 is increasing.
[0043] 2 and 4, the portion where such an increase in diameter is found is marked with a dashed line as a portion of interest 52. This portion of interest 52 is stored in a memory of the leak detection system 56, which may be contained within the robot 32 and / or the 3D sensor 36, or within a separate computing device connected to the robot 32 and / or the 3D sensor 36. The leak detection system 56 includes the leak detector 24, the robot 32 and the 3D sensor 34.
[0044] The target portion 52 in the virtual space 48 corresponds to the test portion 20 of the heat exchange device 12 to which the suction probe 22 must be brought close to perform leak detection.
[0045] As a next step, an approach path 58 is determined for bringing the suction probe 22 closer to the test part 20 (for bringing the graphical representation of the suction probe 22 closer to the target part 52 in the point cloud 44). This is done by using CAD data of the suction probe 22 and its suction tip 25 and comparing the point cloud 44 with said CAD data. This allows a point 46b in the point cloud 44 to be identified which corresponds to the suction tip 25 or the suction probe 22. An approach path 58 is then calculated for bringing the suction tip 25 closer to the test part 20 (i.e. the target part 52 in the virtual space 48). This results in an approach path 58 that avoids collisions between the suction tip 25 and / or the suction probe 22 and / or the robot arm 30 or the connecting hose 26 and the heat exchanger 12 or any structure or part of the heat exchanger 12. This approach path 58 is indicated in FIG. 2 by four arrows and a point 46c.
[0046] Only after a collision-free approach path 58 has been determined in the virtual space 48 (i.e., the simulation environment) does the robot 32 or robot arm 30 actually approach the suction tip 25 to the actual test part 20 in the real world.
[0047] This allows leak detection in quality control and manufacturing industries to be performed completely automatically, much faster and much more reliably than conventional suction leak detection of heat exchange equipment performed by an operator.
Claims
1. A method for detecting a leak in a fluid guide (16, 18) of a heat exchange device (12), comprising the steps of: a) acquiring, using a 3D sensor (34), a point cloud (44) of the heat exchange device (12) in a three-dimensional virtual space (48), each point forming a surface point (46a) of an outer surface of the heat exchange device (12); b) searching within said point cloud (44) for a structure (50) corresponding to a tube of a given outer diameter; c) searching for target portions (52) in the obtained structure (50) where the diameter or direction of the tube changes; d) once the target portion (52) is found, determining an approach path (58) for bringing the suction probe (22) of the gas leak detector (24) into proximity with the target portion (52) within the point cloud (44); e) automatically physically approaching said suction probe (22) along said approach path (58) to a test portion (20) of said heat exchange device (12) that corresponds to said target portion (52) in said point cloud (44); A method comprising:
2. 2. The method of claim 1, wherein the method is performed automatically by a leak detection system (56) comprising the gas leak detector (24).
3. 3. The method of claim 1 or 2, wherein the point cloud (44) in the three-dimensional virtual space (48) is generated from image data obtained from at least two optical cameras (38, 40) of an imaging system (36).
4. The method according to any one of claims 1 to 3, wherein the 3D sensor (34) is an imaging system (36) comprising at least one optical camera and at least one lighting device (42).
5. 5. The method according to claim 1, wherein after step a), the point cloud (44) is compared with digital reference data of the structure of the heat exchange device (12) that is to be located in the point cloud (44).
6. The method of claim 5, wherein the digital reference data is pre-acquired CAD data for the heat exchange device (12).
7. The method of claim 5 or 6, wherein at least one search portion within the point cloud (44) is selected based on a comparison with the reference data.
8. 8. The method of claim 7, wherein the search portion is selected using position data of the location of the heat exchange device (12) within the point cloud (44) obtained in step b).
9. The method according to any one of claims 5 to 8, wherein the reference data is generated by the 3D sensor (34) from the same heat exchange device (12) or from a corresponding heat exchange device (12) of the same type.
10. 10. The method according to claim 1, wherein step c) is performed by tracing and analyzing the points (46a) in the point cloud (44) found in step b) that correspond to the tube.
11. The method according to any one of claims 1 to 10, wherein in step d) an ideal position of the suction probe (22) for gas leak detection in the pipe or test part (20) is calculated.
12. The method according to any one of claims 1 to 11, wherein step d) includes the sub-step of determining whether the determined approach path (58) will lead to a collision between the suction probe (22) and a structure of the heat exchange device (12), and the sub-step of determining a further approach path (58) if a collision is expected with the determined approach path (58).
13. The method according to any one of claims 1 to 12, wherein step d) is repeated until an approach path (58) is determined in which no collision between the suction probe (22) and the structure of the heat exchange device (12) is expected.
14. The method according to any one of claims 1 to 13, wherein step d) uses previously obtained and stored digital data, such as CAD data, of the suction probe (22) and / or the robot (32) performing step e).
15. The method according to any one of claims 1 to 14, wherein at least step e) is performed by a robot.
16. The method according to any one of claims 1 to 15, wherein at least one of steps a) to d) is performed by a software algorithm of a leak detection system (56) comprising a leak detector and a robot.
17. The method according to any one of claims 1 to 16, wherein step d) comprises the substep of generating a trajectory (46c) to be followed by a robot (32) holding the suction probe (22).
18. The method according to any one of claims 1 to 17, wherein at least steps b) to d) are repeated for a number of target regions, and step e) is then performed each time to subsequently bring the suction probe (22) close to a respective test portion corresponding to the target portion.
19. 19. The method according to any one of claims 1 to 18, further comprising, after step e), f) performing a leak detection in the test portion (20) of the tube by suctioning the test area with the suction probe (22) to identify possible leaks in the fluid conducting structures (16, 18); A method for performing.
20. 20. The method of any one of claims 1 to 19, wherein the suction probe (22) has a generally U-shaped sniffer tip with two distally extending suction arms.
21. 21. The method according to any one of claims 1 to 20, wherein if in step d) it is determined that the approach path (58) will lead to a collision between the suction probe (22) and a structure of the heat exchange device (12), the suction probe (22) or the suction tip (25) of the suction probe (22) is automatically reconfigured or deformed, or automatically replaced with another suction probe (22) or tip.