Defect detection system and defect detection method

Abstract

Provided is a defect detection system (1) for detecting a defect part in an additively manufactured object. The present invention has a first monitoring unit (21) that monitors a feature amount of a protrusion, a second monitoring unit (22) that monitors a feature amount different from that monitored by the first monitoring unit (21), and a determination unit (3) that detects a defect part on the basis of data acquired by the first monitoring unit (21) and data acquired by the second monitoring unit (22). When the feature amount of the protrusion is outside a prescribed range, the determination unit (3) detects the protrusion as an indication part, and when an abnormal part is detected by the second monitoring unit (22) on or beyond an (n+1)th layer at a specific coordinate position at which the indication part was detected on the built-up surface of an nth layer, the determination unit (3) detects the abnormal part as a defect candidate part. When a defect candidate part is detected for a prescribed number of layers at the specific coordinate position, it is determined that a defect part is present at the specific coordinate position.

Description

Defect detection system and defect detection method CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application is based on Japanese Application No. 2024-124112, filed on July 31, 2024, the contents of which are incorporated herein by reference.

[0002] The present disclosure relates to a defect detection system and a defect detection method.

[0003] In order to guarantee the quality of additively-manufactured objects, various detection systems have been proposed that detect defects in additively-manufactured objects by inspection during additive manufacturing. For example, the inspection system disclosed in Patent Literature 1 detects defects based on images of the surfaces of each layer in the additively-manufactured object, and determines that an abnormality has occurred when multiple defects occur consecutively at the same position in multiple layers.

[0004] Japanese Patent Application Laid-Open No. 2022-179347

[0005] Conventional inspection systems have insufficient inspection accuracy, making it difficult to adequately guarantee the quality of additively shaped objects.

[0006] The present disclosure aims to provide a defect detection system and a defect detection method that can improve inspection accuracy.

[0007] One aspect of the present disclosure is a defect detection system for detecting defects in an additively formed object obtained by stacking a large number of layers of material through additive manufacturing, the defect detection system having: a first monitoring unit that monitors feature values ​​of convex portions that may appear on the forming surface of each layer during additive manufacturing; a second monitoring unit that monitors each layer during additive manufacturing using feature values ​​different from those of the first monitoring unit; and a determination unit that detects the defect portion based on data acquired by the first monitoring unit and data acquired by the second monitoring unit, wherein the determination unit is configured to: detect the convex portion as a predictive portion when the feature value of the convex portion detected by the first monitoring unit is outside a predetermined range; detect the abnormal portion as a defect candidate portion when the second monitoring unit detects an abnormal portion from the (n+1)th layer onwards at a specific coordinate position where the predictive portion was detected on the forming surface of the nth layer; and determine that a defect portion has occurred at the specific coordinate position when the defect candidate portion has been detected for a predetermined number of layers at the specific coordinate position.

[0008] Another aspect of the present disclosure is a defect detection method for detecting defects in an additively formed object obtained by stacking a large number of materials by additive manufacturing, the defect detection method comprising: a primary detection step of monitoring the forming surface of each layer during additive manufacturing, and detecting a convex portion with a feature value outside a predetermined range as a predictive portion when the convex portion is detected; a secondary detection step of monitoring each layer during additive manufacturing using a feature value different from the feature value detected by the primary detection step, and detecting an abnormal portion as a defect candidate portion when an abnormal portion is detected on the forming surface of the nth layer from the (n+1)th layer onwards at a specific coordinate position where the predictive portion was detected; and a determination step of determining that a defect portion has occurred at the specific coordinate position when the defect candidate portion has been detected for a predetermined number of layers at the specific coordinate position.

[0009] The defect detection system is configured to detect a defect by using a predictive portion and a defect candidate portion, and is therefore capable of detecting a defect in an additively shaped object with high accuracy.

[0010] As described above, according to the above aspects, it is possible to provide a defect detection system and a defect detection method that can improve inspection accuracy.

[0011] The above and other objects, features, and advantages of the present disclosure will become more apparent from the following detailed description taken in conjunction with the accompanying drawings, in which Fig. 1 is an explanatory diagram of an additive manufacturing apparatus including a defect detection system according to a first embodiment, Fig. 2 is a conceptual diagram of a defect detection method according to the first embodiment, Fig. 3 is a cross-sectional explanatory diagram illustrating the height of a convex portion (a predictive portion) according to the first embodiment, Fig. 4 is a conceptual diagram of three-dimensional model data and two-dimensional slice data of a target object according to the first embodiment, Fig. 5 is a diagram illustrating the layout of a plurality of additively manufactured objects on a base plate as viewed from the stacking direction according to the first embodiment, Fig. 6 is a flow diagram of an additive manufacturing method including the defect detection method according to the first embodiment, and Fig. 7 is a perspective explanatory diagram illustrating a mechanism by which defects occur. FIG. 8 is a cross-sectional explanatory diagram explaining the mechanism by which defects occur, FIG. 9 is a cross-sectional photograph of an additively molded object obtained by a reproduction experiment, FIG. 10 is an explanatory diagram with auxiliary lines added to FIG. 9, FIG. 11 is a cross-sectional photograph of an additively molded object having a defect, FIG. 12 is an explanatory diagram of an additively molded apparatus including a defect detection system in embodiment 2, FIG. 13 is a flow diagram of an additive manufacturing method including a defect detection method in embodiment 3, FIG. 14 is a schematic diagram of a plurality of additively molded objects including an additively molded object having a defect in embodiment 3, viewed from the stacking direction, and FIG. 15 is a flow diagram of an additive manufacturing method including a defect detection method in embodiment 4.

[0012] Embodiment 1 An embodiment of a defect detection system and a defect detection method will be described with reference to Figures 1 to 11. A defect detection system 1 of this embodiment is a system that detects defects in an additively shaped object obtained by stacking a large number of materials through additive manufacturing.

[0013] 1 , the defect detection system 1 includes a first monitoring unit 21, a second monitoring unit 22, and a determination unit 3. The first monitoring unit 21 monitors the feature amounts of convex portions (see reference numeral 61 in FIG. 3 ) that may appear on the molding surface of each layer during additive manufacturing. The second monitoring unit 22 monitors each layer during additive manufacturing using feature amounts different from those used by the first monitoring unit 21. The determination unit 3 detects defects based on data acquired by the first monitoring unit 21 and data acquired by the second monitoring unit 22.

[0014] When the feature amount of a convex portion detected by the first monitoring unit 21 is outside a predetermined range, the determination unit 3 detects the convex portion as a predictive portion 51. Furthermore, as shown in FIG. 2 , when an abnormal portion is detected by the second monitoring unit 22 on or after the (n+1)th layer at a specific coordinate position where the predictive portion 51 was detected on the building surface of the nth layer, the determination unit 3 detects the abnormal portion as a defect candidate portion 52. When defect candidate portions 52 are detected at a specific coordinate position for one layer including the (n+1)th layer or a predetermined number of consecutive layers, the determination unit 3 determines that a defect portion has occurred at the specific coordinate position. Here, n is a natural number.

[0015] Possible feature quantities of the convex portions include, for example, the height of the convex portions in the stacking direction, or the diameter of the convex portions as viewed from the stacking direction. In this embodiment, the feature quantity of the convex portions is the height of the convex portions. Note that the height of the convex portions is a local height, and is, for example, as shown in FIG. 3 , the amount of protrusion h of the convex portions 61 from the surrounding molding surface 62. In addition, in this specification, the coordinate position means a two-dimensional coordinate position as viewed from the stacking direction of the additively shaped object. In other words, it represents a position in the XY coordinate system (see FIG. 5 ), which will be described later. Furthermore, the specific coordinate position can be recognized as a position having a certain width (area).

[0016] The determination unit 3 detects an abnormality when the feature acquired by the second monitoring unit 22 is outside a predetermined range. The feature acquired by the second monitoring unit 22 is a physical quantity different from the feature of the convex portion acquired by the first monitoring unit 21. The feature acquired by the second monitoring unit 22 may be, for example, the emission intensity of the infrared spectrum generated from the printing surface. That is, for example, if there is a defect below the printing surface, the emission intensity of the infrared spectrum may be higher than that of a location without the defect. Furthermore, the emission intensity of this infrared spectrum may vary depending on the size of the defect, etc. Therefore, the second monitoring unit 22 acquires the emission intensity of the infrared spectrum generated from the printing surface, and when the value is outside a predetermined range (for example, when a value higher than a predetermined threshold is detected), the corresponding coordinate position is detected as an abnormality.

[0017] It should be noted that various other features than those described above can be considered as features obtained by the second monitoring unit 22, which will be described later.

[0018] As shown in FIG. 1 , the defect detection system 1 of this embodiment is provided in an additive modeling apparatus 10. In this embodiment, the additive modeling apparatus 10 is a powder bed type additive modeling apparatus. The additive modeling apparatus 10 has a base plate 11, a powder feeder 12, a recoater 13, a heat source supply device 14, and a control unit 15. The recoater 13 supplies powder supplied from the powder feeder 12 onto the base plate 11 to form a powder bed 101. The base plate 11 and the powder feeder 12 are each configured to be able to move up and down in the vertical direction. The raw material powder may be, for example, metal powder, resin powder, or ceramic powder.

[0019] Examples of the heat source supply device 14 include a laser unit and an electron beam unit. A case where the heat source supply device 14 is a laser unit will be described below. The heat source supply device 14 irradiates a powder bed 101 on the base plate 11 with a laser beam L. The laser beam L emitted from the heat source supply device 14 is irradiated at a predetermined position on the powder bed 101 via a movable mirror 16. The control unit 15 controls the additive modeling device 10. That is, the control unit 15 controls the additive modeling device 10 to form an additive model in the powder bed 101 based on two-dimensional slice data 42 (see FIG. 4 ) converted from three-dimensional model data 41 of the desired model. That is, in each layer, the control unit 15 irradiates a predetermined position on the powder bed 101 with a laser beam L based on the two-dimensional slice data 42, thereby melting and solidifying the powder at the predetermined position. By repeating this process of melting and solidifying the powder by irradiating it with the laser beam in multiple layers, an additive model is obtained. The control unit 15 includes a processor, a memory for storing programs executed by the processor, and a communication unit for communicating with the outside.

[0020] FIG. 5 is a schematic diagram of an XY coordinate system viewed from the stacking direction in the case where multiple additively molded objects 6 are obtained by the additive molding apparatus 10. That is, multiple additively molded objects 6 are molded in parallel on the base plate 11 of one additive molding apparatus 10. Here, 20 additively molded objects 6 are molded, arranged five in the X direction and four rows in the Y direction. That is, FIG. 5 shows the solidified molded surfaces of 20 additively molded objects 6 during additive molding, viewed from the stacking direction. Note that in this specification, an unfinished additively molded object in the process of molding will also be referred to as the additively molded object 6, as appropriate. The molded surface refers to the surface formed by melting and then solidifying powder or the like when a heat source (e.g., laser light) is supplied to a powder bed or the like during additive manufacturing.

[0021] The defect detection system 1 of this embodiment is provided in an additive manufacturing apparatus 10, as shown in Fig. 1. For example, a first monitoring unit 21 and a second monitoring unit 22 are each arranged so as to be able to monitor a powder bed 101. The arrangement of the first monitoring unit 21 and the second monitoring unit 22 shown in Fig. 1 is an example, and the first monitoring unit 21 and the second monitoring unit 22 may be arranged in appropriate positions depending on the purpose. Information obtained by the first monitoring unit 21 and the second monitoring unit 22 is sent to the determination unit 3. The defect detection system 1 also has an output unit 17 that outputs the determination result in the determination unit 3. The output unit 17 may be a monitor that displays the determination result, a speaker that outputs the determination result as sound, or any other means.

[0022] The first monitoring unit 21 may be, for example, a line sensor, an image recognition device, or a three-dimensional shape measuring device. The second monitoring unit 22 may be, for example, a CCD camera, a CMOS camera, or a photodiode. Appropriate measuring means may be selected for the first monitoring unit 21 and the second monitoring unit 22 depending on the feature to be measured. The determination unit 3 includes a processor, a memory for storing programs executed by the processor, and a communication unit for communicating with the outside.

[0023] The defect detection method of this embodiment includes a primary detection step, a secondary detection step, and a determination step as follows.

[0024] In the primary detection step, the build surface of each layer is monitored during additive manufacturing. When a convex portion with a feature amount outside a predetermined range is detected, the convex portion is detected as a predictive portion 51 (see FIG. 2).

[0025] In the secondary detection step, each layer during additive manufacturing is monitored using a feature quantity different from the feature quantity detected in the primary detection step. Then, when an abnormality is detected in the (n+1)th layer or later at a specific coordinate position where a predictive portion 51 is detected on the building surface of the nth layer, the abnormality is detected as a defect candidate portion 52 (see FIG. 2).

[0026] As shown in Figure 2, the judgment step determines that a defect has occurred at a specific coordinate position when a defect candidate portion 52 is detected for one layer including the (n+1)th layer or for a predetermined number of consecutive layers at the specific coordinate position.

[0027] An example of an additive manufacturing flow incorporating the above-described defect detection method will be described with reference to the flow diagram shown in Figure 6. First, the control unit 15 receives 3D model data 41 (see Figure 4) of the object to be manufactured (step S1). Additive manufacturing is performed based on this manufacturing data. During this process, the first monitoring unit 21 acquires data on the manufacturing surface for each layer being manufactured (step S2). That is, data including height information of the manufacturing surface of each layer is acquired.

[0028] The determination unit 3 determines whether or not a convex portion that will become the predictor portion 51 (see FIG. 2 ) has been detected by the first monitoring unit 21 (step S3). If it is determined that the predictor portion 51 has been detected, the coordinate position of the predictor portion 51 is recorded as a specific coordinate position in the memory of the determination unit 3 (step S4). On the other hand, if it is determined that the predictor portion 51 has not been detected, the process proceeds to the next step S5 without recording the specific coordinate position.

[0029] Next, the powder for forming the next layer is recoated (step S5). Next, the determination unit 3 determines whether or not a specific coordinate position is recorded in memory (step S6). If a specific coordinate position is not recorded, the process proceeds to step S12, which will be described later. If a specific coordinate position is recorded, the data is acquired by the second monitoring unit 22 (step S7).

[0030] Next, the determination unit 3 determines whether or not an abnormal portion (i.e., a defect candidate portion) has been detected at the specific coordinate position by the second monitoring unit 22 (step S8). If no abnormal portion has been detected at the specific coordinate position, the specific coordinate position is erased from memory (step S11), and the process proceeds to step S12, which will be described later. If it is determined in step S8 that an abnormal portion has been detected at the specific coordinate position, the determination unit 3 further determines whether or not the abnormal portion (i.e., a defect candidate portion) detected at the specific coordinate position has been detected for x consecutive layers (step S9), where x is a predetermined natural number.

[0031] If it is determined that the abnormal portion detected at the specific coordinate position is detected continuously for x layers, it is determined that a defect has occurred at the specific coordinate position (step S10). A message to that effect is output. This message can be output by various means, such as display on a monitor, audio output, or signal output to the control unit 15. If it is determined in step S9 that the abnormal portion detected at the specific coordinate position is not detected continuously for x layers, the process proceeds to step S12.

[0032] In this example flow, steps S2 to S4 correspond to the primary detection step, steps S6 to S8 correspond to the secondary detection step, and steps S9 and S10 correspond to the determination step.

[0033] Furthermore, in the method described in the above flow, the second monitoring unit 22 selectively monitors the coordinate positions where the predictor parts 51 are detected. That is, the second monitoring unit 22 does not monitor the presence or absence of abnormal parts throughout each layer, but selectively monitors the coordinate positions where the predictor parts 51 are detected (i.e., specific coordinate positions).

[0034] Various means can be considered as the means for detecting the convex portion by the first monitoring unit 21. The detection means can be broadly divided into non-contact and contact types. Examples of non-contact measurement means include optical methods such as triangulation, focusing, and time delay methods. These can generally be measured using a type of measuring device called a 3D sensor. Furthermore, examples of non-contact measurement methods that utilize magnetism include eddy current displacement sensors.

[0035] As a contact-type measuring means, for example, a differential transformer type contact displacement sensor may be used. Alternatively, the first monitoring unit 21 may be configured to detect convex portions by analyzing an image of the modeling surface taken by a camera using AI (i.e., artificial intelligence).

[0036] Next, various means are conceivable for detecting an abnormal portion by the second monitoring unit 22. The second monitoring unit 22 may acquire feature quantities by monitoring the surface of each layer. The second monitoring unit 22 may acquire feature quantities by monitoring the surface of each layer for at least a portion of the period from the start of formation of the layer to the completion of formation. Alternatively, the second monitoring unit 22 may acquire feature quantities by monitoring the heat source irradiation point of each layer. Here, the heat source irradiation point refers to a point where a physical change occurs due to irradiation with a heat source such as laser light. Examples of the heat source irradiation point include a point irradiated with laser light as a heat source and its vicinity. For example, a molten pool formed by laser light irradiation is a heat source irradiation point. In each case, the determination unit 3 may detect an abnormal portion when the acquired feature quantities fall outside a predetermined range.

[0037] Possible methods for monitoring the build surface of each layer include, for example, ultrasonic flaw detection, magnetic flaw detection, and active thermography. All of these methods obtain different measurement values ​​at positions where a specific internal defect exists and positions where it does not. By utilizing this, when a measurement value outside a specified range of measurement values ​​(feature values) is obtained at a specific coordinate position, the location can be detected as an abnormal part.

[0038] A possible method for monitoring the surface of each layer over at least a portion of the period from the start of formation to the completion of formation is, for example, a method using optical tomography. The at least a portion of the period from the start of formation to the completion of formation of each layer can be, for example, at least a portion of the period from the supply of powder (recoating) onto the base plate 11 through the irradiation of laser light to the melting and solidification of the powder. This entire period can be monitored, or a portion of this period can be monitored. Then, based on the temperature change profile measured by optical tomography during this period, it is possible to determine whether an abnormality exists at a specific coordinate position. In other words, even with this method, different measurement values ​​are obtained at positions where a specific internal defect exists and at positions where it does not.

[0039] Possible methods for monitoring the heat source irradiation points of each layer using the second monitoring unit 22 include optical, temperature, and acoustic measurement. Optical methods include, for example, using a photodiode to measure the blackbody radiation intensity of the heat source irradiation points, and using optical coherence tomography to measure the depth of keyholes formed in the molten pool. Temperature measurement methods include, for example, two-color temperature measurement using a photodiode. If a defect exists inside the weld, the temperature of the defect will be higher than the surrounding area due to reduced thermal conductivity. Based on this, abnormalities can be detected using the above methods. Abnormalities can also be detected by detecting abnormal sounds generated during melting. For example, abnormalities can be detected by detecting acoustic emission waves.

[0040] Conventional methods for detecting defects have had the problem of not necessarily being highly accurate. Therefore, the inventors of the present application have studied the causes of defects. As a result, they have found that spatter that may be generated when a laser beam is irradiated onto the powder bed 101 blocks the subsequent laser beam, which is thought to be a cause of defects.

[0041] That is, as described above, in additive manufacturing, a laser beam L is irradiated onto the powder bed 101 to melt and solidify powder at a predetermined coordinate position. This laser beam L melts multiple layers of powder. Here, as shown in Figures 7 and 8, spatter 71 (i.e., particles of material scattered from the molten pool 63) may be generated when the laser beam L is irradiated onto the powder bed 101. This spatter 71 may land on the surface of the powder bed 101 that is to be irradiated with the laser beam L thereafter.

[0042] In this case, when the laser beam L is irradiated onto the coordinate position, the spatter 71 blocks the laser beam L, and the powder 102 underneath is not sufficiently irradiated with the laser beam L. In this case, the powder 102 that is not fully melted remains underneath the spatter 71. The portion where the insufficiently melted powder 102 remains is considered to be a defective portion.

[0043] To confirm this mechanism, the following reproducible experiment was conducted. First, additive manufacturing was performed partway through, and then a powder bed was created. A small amount of granular foreign matter, which resembled spatter, was scattered on top of the powder bed. The material used for additive manufacturing was SKD alloy tool steel, which does not contain Ni. On the other hand, the scattered foreign matter was made of maraging steel, which contained Ni. Furthermore, SKD and maraging steel have similar melting points.

[0044] The powder bed containing the foreign matter was irradiated with laser light to melt and solidify the powder. Subsequently, recoating and laser light irradiation were repeated using a conventional method to perform additive manufacturing on multiple layers. The cross section of the resulting additively shaped object along the layering direction was then observed. The image is shown in Figure 9.

[0045] As can be seen from the figure, in the cross section where the defect appears, a foreign object with a different brightness compared to the surrounding area can be seen above the defect. When this area was analyzed using EDS (i.e., energy dispersive X-ray spectroscopy), Ni was detected, confirming that it was a foreign object made of the maraging steel described above. In this way, it was confirmed that foreign objects likened to spatters are the main cause of the defect.

[0046] Furthermore, when observing the grain boundaries around the defects, many downward-convex curves can be seen. Figure 10 shows auxiliary lines along these curves, indicated by dashed lines. These curves are traces of the molten pool formed by the laser beam irradiation. Closer examination of these traces reveals that they are regularly arranged in areas without foreign objects or defects, but the traces of the molten pool around the foreign objects and defects are irregular. In other words, no traces of the molten pool can be seen around the foreign objects and defects, especially directly below them. Therefore, it is believed that the laser beam was blocked by the foreign objects in these areas.

[0047] Similar patterns of molten pool traces were observed around defects in actual additively molded objects, as shown in Figure 11. These results support the mechanism of defect generation described above, and confirm that spatter is the main cause of defect generation.

[0048] Therefore, the inventors of the present application discovered that it is effective to first monitor the modeling surface of each layer during additive manufacturing using the first monitoring unit 21 and detect convex portions with a feature amount (height in this embodiment) that exceeds a predetermined range as predictive portions. Then, in this embodiment, the coordinate position of the predictive portion is recorded as the XY coordinate position (i.e., specific coordinate position) where the predictive portion exists.

[0049] It is conceivable that the height of the convex portion at this specific coordinate position will decrease from the n+1th layer onward due to new recoated powder (see FIG. 8 ). Therefore, from the n+1th layer onward at the specific coordinate position, the presence or absence of an abnormality is inspected by observation using other feature quantities (i.e., observation by the second monitoring unit 22). If a specific defect actually exists, an abnormal value is detected at the specific coordinate position. For example, it is conceivable that the emission intensity of the infrared spectrum described above will decrease. In other words, if the emission intensity of this infrared spectrum falls below a predetermined threshold, it is considered highly likely that a defect exists at the specific coordinate position. In this case, the specific coordinate position is detected as a defect candidate portion 52. Furthermore, if defect candidate portions 52 are detected consecutively across multiple layers, the probability that a defect exists becomes even higher. Therefore, when defect candidate portions 52 are detected consecutively for a predetermined number of layers (i.e., x number of layers), it is determined that a problematic defect exists. This determination criterion can be set appropriately depending on the guaranteed quality of the additively molded object, etc. For example, x may be set to 1, or x may be set to 2 or greater. From the viewpoint of suppressing the influence of noise on the data obtained by the second monitoring unit 22, it is preferable that x be set to 2 or greater.

[0050] Next, the effects of this embodiment will be described. The defect detection system 1 is configured to detect a defect portion by utilizing the predictor portion 51 and the defect candidate portion 52. Therefore, it is possible to detect a defect portion in the additively shaped object 6 with high accuracy.

[0051] As described above, the defect detection system 1 of this embodiment detects a predetermined convex portion as a predictive portion 51 based on the fact that spatters 71 are a major cause of defect occurrence. As a result, it is considered that there is a relatively high possibility that a defect will be formed at the specific coordinate position where the predictive portion 51 is detected. In other words, the predictive portion 51 can be considered to be a predictive portion of defect occurrence.

[0052] Therefore, by determining whether or not a defect exists based on whether or not a defect candidate portion 52 detected by other features is detected for a predetermined number of layers at the specific coordinate position where the predictive portion 51 is detected, it is possible to determine with a high probability whether or not a defect exists.

[0053] If an attempt is made to determine the presence or absence of defects without detecting the predictive portion 51, it is likely that, due to the influence of noise and the like, the presence of defects will be determined in locations where there are no defects, or that defects will not be determined in locations where there are defects. This makes it difficult to detect defects with high accuracy. Furthermore, even if an attempt is made to detect defects using only the features of the convex portions, highly accurate detection is unlikely because convex portions do not necessarily occur only at locations where defects exist. Furthermore, as described above, in layers after a convex portion has been detected, the height of the convex portion may be lower, making it difficult to detect (see FIG. 8 ). In such cases, it is also difficult to detect defects.

[0054] In contrast to this, if a predetermined convex portion is first detected as a predictor portion 51, and then an abnormal portion is detected at that specific coordinate position as a defect candidate portion 52, it is considered that there is a high possibility that a defect portion exists at the specific coordinate position. Therefore, the defect detection system 1 of this embodiment can detect defects with high accuracy.

[0055] In particular, when defect candidate portions 52 are detected at a specific coordinate position for one layer including the (n+1)th layer or for a predetermined number of consecutive layers, the determination unit 3 determines that a defect portion has occurred at the specific coordinate position. This allows for more accurate detection of defects.

[0056] Furthermore, the second monitoring unit 22 selectively monitors coordinate positions where the predictive portion 51 is detected. This can improve inspection efficiency. In particular, considering the mechanism of defect generation described above, it is considered that there is a low possibility that a defect of a predetermined size exists at a coordinate position where the predictive portion 51 is not detected. Therefore, by limiting monitoring by the second monitoring unit 22 to the specific coordinate position where the predictive portion 51 is detected, efficient inspection is possible.

[0057] As described above, according to the present embodiment, it is possible to provide a defect detection system and a defect detection method that can improve inspection accuracy.

[0058] 12 , this embodiment is a defect detection system 1 in which the determination unit 3 has a defect size estimation unit 31. The defect size estimation unit 31 estimates the size of the defect based on feature amount data of the convex portion detected by the first monitoring unit 21.

[0059] As described above, the main cause of the occurrence of defects is thought to be the adhesion of spatter 71. The size of the defects is thought to depend on the size of the spatter 71. Furthermore, the convex portions detected by the first monitoring unit 21 are thought to be spatter marks. Therefore, the size of the defects (e.g., volume, etc.) can be estimated based on the feature quantities of the convex portions, such as height and diameter.

[0060] Then, the defect size (size of the defective portion) obtained by the defect size estimation unit 31 is output by the output unit 17. In addition, it is also possible to evaluate the quality level of the manufactured additive-modeled object according to the estimated defect size, for example.

[0061] Note that, among the symbols used in the second and subsequent embodiments, the same symbols as those used in the previous embodiments represent the same components, etc. as those in the previous embodiments, unless otherwise specified.

[0062] In this embodiment, it is possible to increase the variety of subsequent measures based on the estimated defect size. In addition, the present embodiment has the same effects as the first embodiment.

[0063] (Embodiment 3) In this embodiment, as shown in Fig. 13 , when a defect is detected, the modeling of some additively molded objects 6 is stopped. As in the additive modeling method shown in Embodiment 1, when a plurality of additively molded objects 6 are manufactured in parallel, defect detection is performed in the manufacturing process of these plurality of additively molded objects 6. Then, as shown in Fig. 14 , when a defect is detected, the modeling of the additively molded object 6 including the defect is stopped at that stage (see step S13 in Fig. 13 ). On the other hand, the manufacturing of additively molded objects 6 in which no defect is detected continues.

[0064] In addition, when one additively shaped object 6 is manufactured, the additively shaped object 6 may be configured to be stopped at the stage when a defective portion is detected. The rest is the same as in the first embodiment.

[0065] In this embodiment, it is possible to improve material yield and productivity, and other effects similar to those of the first embodiment can be obtained.

[0066] (Embodiment 4) As shown in FIG. 15, this embodiment is an embodiment in which, when a defect is detected, the defect is repaired. That is, a repair process of step S14 in the same figure is provided. For example, when a defect is detected during additive manufacturing, it is conceivable to repair the defect by additionally irradiating a high-intensity laser beam at the specific coordinate position where the defect was detected. In other words, since it is believed that a sputtering mark exists at the specific coordinate position, a high-energy laser beam is irradiated thereto to re-melt it. This re-melts and re-solidifies the area around the defect, making it possible to eliminate the defect. The rest is the same as in embodiment 1.

[0067] In this embodiment, it is possible to improve the product yield, and other effects similar to those of the first embodiment can be obtained.

[0068] In the above embodiment, the case where monitoring by the second monitoring unit 22 is selectively performed for a specific coordinate position has been described, but it is also possible to perform monitoring by the second monitoring unit 22 over the entirety of each layer or over all of the coordinate positions where the next layer is to be formed. Furthermore, monitoring by the first monitoring unit 21 can also be performed over the entirety of the printing surface of each layer, or it is also possible to perform monitoring selectively over the coordinate positions where the next layer is to be printed.

[0069] The present disclosure is not limited to the above-described embodiments, and can be applied to various embodiments without departing from the spirit of the present disclosure.

[0070] Although the present disclosure has been described with reference to the embodiments, it is understood that the present disclosure is not limited to the embodiments or structures. The present disclosure also encompasses various modifications and modifications within the scope of equivalents. In addition, various combinations and forms, as well as other combinations and forms including only one element, more than one element, or less than one element, are also within the scope and spirit of the present disclosure.

[0071] The features of the present disclosure are as follows: [1] A defect detection system (1) for detecting defects in an additively formed object (6) obtained by stacking a large number of materials by additive manufacturing, comprising: a first monitoring unit (21) that monitors a feature amount of a convex portion that may appear on the forming surface of each layer during additive manufacturing; a second monitoring unit (22) that monitors each layer during additive manufacturing using a feature amount different from that of the first monitoring unit; and a determination unit (3) that detects the defect portion based on data acquired by the first monitoring unit and data acquired by the second monitoring unit, wherein the determination unit is configured to: detect the convex portion as a predictive portion (51) when the feature amount of the convex portion detected by the first monitoring unit is outside a predetermined range; detect the abnormal portion as a defect candidate portion (52) when the second monitoring unit detects an abnormal portion from the (n+1)th layer onwards at a specific coordinate position where the predictive portion was detected on the forming surface of the nth layer; and determine that a defect portion has occurred at the specific coordinate position when the defect candidate portion has been detected for a predetermined number of layers at the specific coordinate position. [2] The defect detection system according to [1], wherein the determination unit is configured to determine that a defect has occurred at the specific coordinate position when the defect candidate portion is detected at the specific coordinate position for one layer including the (n+1)th layer or for a predetermined number of consecutive layers. [3] The defect detection system according to [1] or [2], wherein the second monitoring unit acquires a feature amount obtained by monitoring the build surface of each layer, and the determination unit detects the abnormal portion when the feature amount is outside a predetermined range. [4] The defect detection system according to [1] or [2], wherein the second monitoring unit acquires a feature amount obtained by monitoring a heat source irradiation position on each layer, and the determination unit detects the abnormal portion when the feature amount is outside a predetermined range. [5] The defect detection system according to [1] or [2], wherein the second monitoring unit acquires a feature amount obtained by monitoring the surface of each layer for at least a portion of a period from the start of formation of the layer to the completion of formation, and the determination unit detects the abnormal portion when the feature amount is outside the predetermined range.[6] The defect detection system according to any one of [1] to [5], wherein the second monitoring unit selectively monitors coordinate positions where the defect precursor portion is detected. [7] The defect detection system according to any one of [1] to [6], wherein the determination unit has a defect size estimation unit (31) that estimates the size of the defect portion based on feature amount data of the convex portion detected by the first monitoring unit. [8] A defect detection method for detecting defects in an additively manufactured object obtained by stacking multiple layers of material through additive manufacturing, comprising: a primary detection step of monitoring the manufacturing surface of each layer during additive manufacturing, and, when a convex portion having a feature amount outside a predetermined range is detected, detecting the convex portion as a predictive portion (51), a secondary detection step of monitoring each layer during additive manufacturing using a feature amount different from the feature amount detected in the primary detection step, and, when an abnormal portion is detected on the manufacturing surface of the nth layer from the (n+1)th layer onwards at a specific coordinate position where the predictive portion was detected, detecting the abnormal portion as a defect candidate portion (52), and a determination step of determining, when the defect candidate portion is detected at the specific coordinate position for a predetermined number of layers, that a defect has occurred at the specific coordinate position. [9] The defect detection method according to [8], in the determination step, when the defect candidate portion is detected at the specific coordinate position for one layer including the (n+1)th layer or a predetermined number of consecutive layers, that a defect has occurred at the specific coordinate position.

[10] The defect detection method according to [8] or [9], wherein the secondary detection step detects the abnormal portion when a feature amount obtained by monitoring the build surface of each layer falls outside a predetermined range.

[11] The defect detection method according to [8] or [9], wherein the secondary detection step detects the abnormal portion when a feature amount obtained by monitoring a heat source irradiation point on each layer falls outside a predetermined range.

[12] The defect detection method according to [8] or [9], wherein the secondary detection step detects the abnormal portion when a feature amount obtained by monitoring the surface of each layer over at least a portion of a period from the start of formation of the layer to the completion of formation falls outside a predetermined range.

[13] The defect detection method according to any of [8] to

[12] , wherein the secondary detection step selectively monitors coordinate positions where the predictive portion is detected.

[14] The defect detection method according to any one of [8] to

[13] , wherein, when the defect is detected, the size of the defect is estimated based on feature amount data of the convex portion detected in the primary detection step.

Claims

1. A defect detection system (1) for detecting defects in an additively formed object (6) obtained by layering a large number of materials by additive manufacturing, comprising: a first monitoring unit (21) that monitors the characteristic quantities of convex portions that may appear on the forming surface of each layer during additive manufacturing; a second monitoring unit (22) that monitors each layer during additive manufacturing using characteristic quantities different from those monitored by the first monitoring unit; and a determination unit (3) that detects the defect portion based on data acquired by the first monitoring unit and data acquired by the second monitoring unit, wherein the determination unit is configured to: detect the convex portion as a predictive portion (51) when the characteristic quantities of the convex portion detected by the first monitoring unit are outside a predetermined range; detect the abnormal portion as a defect candidate portion (52) when the second monitoring unit detects an abnormal portion from the (n+1)th layer onwards at a specific coordinate position where the predictive portion was detected on the forming surface of the nth layer; and determine that a defect has occurred at the specific coordinate position when the defect candidate portion has been detected for a predetermined number of layers at the specific coordinate position.

2. A defect detection system as described in claim 1, wherein the judgment unit is configured to judge that a defect has occurred at the specific coordinate position when the defect candidate portion is detected for one layer including the (n+1)th layer or for a predetermined number of consecutive layers at the specific coordinate position.

3. A defect detection system as described in claim 1 or 2, wherein the second monitoring unit acquires feature values ​​obtained by monitoring the build surface of each layer, and the judgment unit detects the abnormal part when the feature value falls outside a predetermined range.

4. A defect detection system as described in claim 1 or 2, wherein the second monitoring unit acquires characteristic quantities obtained by monitoring the heat source irradiation points on each layer, and the judgment unit detects the abnormal part when the characteristic quantities fall outside a predetermined range.

5. A defect detection system as described in claim 1 or 2, wherein the second monitoring unit acquires characteristic quantities obtained by monitoring the surface of each layer for at least a portion of the period from the start of formation of the layer to the completion of formation, and the judgment unit detects the abnormal part when the characteristic quantity falls outside a predetermined range.

6. A defect detection system according to claim 1 or 2, wherein the second monitoring unit selectively monitors the coordinate position where the precursor portion is detected.

7. A defect detection system as described in claim 1 or 2, wherein the judgment unit has a defect size estimation unit (31) that estimates the size of the defect based on feature data of the convex portion detected by the first monitoring unit.

8. A defect detection method for detecting defects in an additively formed object obtained by stacking a large number of materials through additive manufacturing, comprising: a primary detection step of monitoring the forming surface of each layer during additive manufacturing, and when a convex portion with a feature value outside a predetermined range is detected, detecting the convex portion as a predictive portion (51); a secondary detection step of monitoring each layer during additive manufacturing using a feature value different from the feature value detected in the primary detection step, and when an abnormal portion is detected on the forming surface of the nth layer from the (n+1)th layer onwards at a specific coordinate position where the predictive portion was detected, detecting the abnormal portion as a defect candidate portion (52); and a determination step of determining that a defect has occurred at the specific coordinate position when the defect candidate portion has been detected for a predetermined number of layers at the specific coordinate position.

9. A defect detection method as described in claim 8, wherein in the determination step, when the defect candidate portion is detected at the specific coordinate position for one layer including the (n+1)th layer or for a predetermined number of consecutive layers, it is determined that a defect portion has occurred at the specific coordinate position.

10. A defect detection method according to claim 8 or 9, wherein the secondary detection step detects an abnormal portion when a feature value obtained by monitoring the build surface of each layer falls outside a predetermined range.

11. A defect detection method according to claim 8 or 9, wherein the secondary detection step detects an abnormal portion when a feature value obtained by monitoring the heat source irradiation point on each layer falls outside a predetermined range.

12. A defect detection method as described in claim 8 or 9, wherein the secondary detection step detects an abnormal portion when a feature value obtained by monitoring the surface of each layer for at least a portion of the period from the start of formation of the layer to the completion of formation falls outside a predetermined range.

13. A defect detection method according to claim 8 or 9, wherein the secondary detection step selectively monitors coordinate positions where the precursor portion is detected.

14. A defect detection method according to claim 8 or 9, wherein when the defect is detected, the size of the defect is estimated based on feature data of the convex portion detected in the primary detection step.

Images