JP2023181933A - Additive-manufactured article quality determination device and additive-manufactured article quality determination method - Google Patents

Abstract

To non-destructively analyze an internal defect position of an additive-manufactured article.SOLUTION: An additive-manufactured article quality determination device 1 comprises: a first assessment unit 14 which, from results of monitoring respective layers of an additive-manufactured article obtained by a shaping device equipped with a monitoring instrument, predicts generation of a defect in the respective layers; a second assessment unit 15 which predicts generation and/or disappearance of a defect in the respective layers from results of monitoring performed by the first assessment unit 14 with respect to a prescribed number of layers situated above the respective layers in the additive-manufactured article, and which reflects the prediction in generation prediction result information regarding a defect in the respective layers; and an output unit 11 which, from the generation prediction result information regarding a defect in the layers obtained by the second assessment unit 15, outputs defect position information about the additive-manufactured article.SELECTED DRAWING: Figure 3

Description

The present invention relates to an additive-shaped product quality determination device and an additive-shaped product quality determination method.

Additive manufacturing is used in the production of aftermarket parts and rapid prototyping of molds and specialized mechanical parts. However, quality assurance has not been established for additively shaped products. This is a factor that prevents manufacturers from adopting metal additive manufacturing for high value-added applications that require high reliability.

Regarding methods of confirming the quality of additive modeling, methods are being considered to estimate the quality by acquiring data using monitoring equipment such as cameras during modeling. For example, Patent Document 1 discloses an imaging device that images an area including a forming surface of an additive product in the middle of manufacturing, and image information and quality information acquired based on images of a plurality of layers of the forming surface of the additive product. It is said that it is possible to provide a quality estimating device for additive products that estimates quality by machine learning using as a training data set.

JP 2021-9126 Publication

From the viewpoint of quality confirmation, it is important to confirm the location of defects in additively manufactured products. For example, in the case of an additively shaped product used as a cooling water pipe, the airtightness of this pipe is very important. Therefore, it is necessary to check for defects around the piping and to conduct individual airtightness evaluation tests.

On the other hand, in the case of an additively shaped product used as cooling water piping, defects in areas other than the piping portion may not affect the quality. For this reason, it is very important to confirm the presence or absence of defects in additively shaped products and the location of the defects in order to confirm quality. In order to guarantee the quality of the additively molded product, for example, defect information of the additively molded product may be included in an attached document as quality information of the additively molded product.

Quality confirmation includes destructive testing and non-destructive testing such as X-ray CT (Computed Tomography). When confirming the position of a defect through a destructive inspection, the additively shaped product is destroyed, so it is not possible to inspect the additively shaped product that is applied to actual equipment. Furthermore, X-ray CT used in non-destructive inspection can only be applied to inspecting thin parts through which X-rays can pass.

Further, Patent Document 1 describes that it is possible to provide a device that estimates defects in an additively shaped product that is in the process of being manufactured based on information obtained from an image taken by an imaging device. With this device, it is possible to estimate the density of an additionally shaped article and the position of a defect from the relationship between the brightness of the shaped surface and the density of the shaped object. However, with analysis based on data obtained from comparing brightness and density, it is difficult to estimate the depth of the defect from the modeling surface or the location of the defect.

Therefore, an object of the present invention is to non-destructively analyze the internal defect position of an additively shaped article.

In order to solve the above-mentioned problems, the additive-fabricated product quality determination device of the present invention includes a method for predicting the generation of defects in each layer of an additive-fabricated product based on the monitoring results of each layer of the additive-fabricated product by an additive-fabrication device equipped with a monitoring device. one evaluation unit and the first evaluation unit predict generation and/or disappearance of defects in each layer from the monitoring results of a predetermined number of layers above each layer of the additively formed article, and A second evaluation unit that reflects the defect generation prediction result information of the layer, and outputs defect position information of the additively formed product based on the defect generation prediction result information of each of the layers obtained by the second evaluation unit. The invention is characterized by comprising an output section that performs.

The additive-printed product quality determination method of the present invention includes a step of predicting the generation of defects in each layer from the monitoring results of each layer of the additive-printed product by an additive-printing device equipped with a monitoring device; A step of predicting the generation and/or disappearance of defects in each layer from the monitoring results of a predetermined number of layers above the layer, a step of reflecting the defect generation prediction result information in each layer, and a step of reflecting the defect generation prediction result information in each layer. The method is characterized in that the step of outputting defect position information of the additively formed article is performed based on the generated defect generation prediction result information.
Other means will be explained in the detailed description.

According to the present invention, it is possible to non-destructively analyze the internal defect position of an additively shaped article.

FIG. 2 is a schematic diagram of an additive modeling device. FIG. 3 is a schematic cross-sectional view showing a molten pool spanning both the N+1 layer and the N layer. FIG. 2 is a logical block diagram of an additionally shaped product quality determination device. FIG. 2 is a physical block diagram of an additive-molded product quality determination device. It is a flowchart of quality judgment processing. It is a flow chart of quality judgment processing concerning a modification. It is a flowchart of the determination process of the number of superimposed layers of an additively shaped article. It is a figure which shows the X-ray CT image of the Nth layer. It is a figure which shows the mapping image of the optical tomography of the Nth layer. It is a figure which shows the mapping image of the optical tomography of the (N+1)th layer. It is a figure which shows the mapping image of the optical tomography of the N+2nd layer. It is a figure which shows the mapping image of the optical tomography of the N+3rd layer. It is a figure which shows the mapping image of optical tomography. It is a graph showing the brightness and defect rate of optical tomography superposition.

Hereinafter, embodiments for carrying out the present invention will be described in detail with reference to the respective figures. The present invention is not limited to the embodiments mentioned here, and can be appropriately combined with known techniques or improved based on known techniques without departing from the technical idea of the invention. .

FIG. 1 is a schematic diagram of the additive modeling device 4.
As shown in FIG. 1, the additive modeling device 4 is roughly divided into an additive modeling unit 41 that performs additive modeling processing, an inspection device, a control device 430 that controls these, and a communication unit 431. The inspection device evaluates the powder layer and solidified layer formed by the additional modeling section 41. The control device 430 centrally controls the additional modeling section 41 and the inspection device. The communication unit 431 transmits the inspection data by the inspection device to the additional molded product quality determination device 1 shown in FIG. 3 .

The additive manufacturing apparatus 4 in the present invention is a three-dimensional metal additive manufacturing apparatus using a powder bed fusion method. The additive manufacturing device 4 irradiates energy to a powder layer covered with metal powder (raw material powder) that is the raw material for an additively molded object to form a two-dimensional solidified layer, and repeatedly stacks the layers to create an additively molded product. It manufactures.

In FIG. 1, a laser beam irradiation device is provided as a heat source supply device for solidifying a powder layer composed of raw material powder 414, and includes a laser oscillator 42, a process fiber 43, a galvano head 44, and a laser coaxial illumination 45. The heat source supply device is not particularly limited as long as it can melt and solidify the powder, and may be an electron beam irradiation device as well as a laser beam irradiation device.

The processing chamber 411 in which the additional shaped object 417 is manufactured has a gas supply pipe 412a and a gas exhaust pipe 412b, and has a configuration in which the atmosphere of the processing chamber 411 can be controlled. Atmosphere control is, for example, an inert gas atmosphere or a vacuum atmosphere when a laser beam is used as a heat source, and a vacuum atmosphere when an electron beam is used as a heat source.

The inside of the processing chamber 411 is divided into a raw material powder storage area 420, an additional modeling area 421 where a powder layer is melted and solidified to form a solidified layer, and a raw material powder recovery area 422 where surplus raw material powder is collected. The raw material powder storage area 420 is an area for storing the raw material powder 414 of the additively shaped object. The additive modeling region 421 is a region in which a powder layer is formed by laminating the raw material powder 414, and a solidified layer is formed by melting and solidifying the powder layer using a heat source supply device. The raw material powder collection area 422 is an area where surplus raw material powder is collected when forming a powder layer in the additional modeling area 421.

The powder supply machine 413 moves in the direction of the white arrow in FIG. 1 and supplies powder from the raw material powder storage area 420 to the additional modeling area 421. As the powder feeder 413, for example, a recoater, a coater, a squeegee, and a blade can be used. In the raw material powder storage area 420 and the additional modeling area 421, the sample stands 415a and 415b on which powder is placed have a configuration that allows them to be moved up and down in the vertical direction in FIG. Although not shown, the sample stage 415b on which additive modeling is performed may include a heater capable of heating the powder layer or the solidified layer. The heater is preferably one that can heat up to about 25°C to 650°C. Heating the powder layer or the solidified layer has the effect of improving the modeling speed by removing water in the raw material powder or reducing the amount of heat input by the beam, and reducing distortion by making the temperature distribution uniform.

In this embodiment, the inspection device includes a visible light imager 46, an infrared imager 47a, a plasma emission imager 47b, and a molten pool observation device 48. The visible light image capturing device 46 and the plasma emission capturing device 47b observe images of the powder layer and the solidified layer in the visible light region. The infrared image camera 47a takes infrared radiation images of the powder layer and the solidified layer, and observes the obtained thermal images. As described below, excessive heat input or insufficient heat input in additive manufacturing contributes to internal defects. When defects occur inside the solidified layer, the thermal conductivity decreases and the thermal diffusivity also decreases. Therefore, by analyzing the causes of internal defects, it is possible to estimate internal defects caused by changes in heat input. The molten pool observation device 48 observes the state when the powder layer is irradiated with a heat source and melted.

Note that this embodiment relates to an analysis method using a thermal image during modeling, and is a method that utilizes data in the stacking direction (for example, the N+1st layer relative to the Nth layer). The analysis method will be explained in detail below.

The control device 430 is connected to the additive modeling device 4 and the inspection device by wire or wirelessly, and controls their operations. The drive of the powder feeder 413 and sample stands 415a and 415b, and the operation of the laser oscillator 42 and galvano head 44 are also controlled and monitored by the control device 430.

The control device 430 also determines whether the powder layer and the solidified layer are good or bad based on the evaluation results of the inspection device. The control device 430 includes a visible light image processing section that processes images obtained from the visible light image capturing device 46 and the plasma emission capturing device 47b and determines the presence or absence of defects, and an image obtained from the infrared image capturing device 47a. It has an infrared image processing unit that processes images and determines the presence or absence of defects.

《Heterogeneity related to the quality of additively shaped products》
In addition to the characteristics of the additively shaped article, the quality of the additively shaped article includes the presence or absence of defects that occur during modeling and the presence of inhomogeneities such as specific metal structures. Here, the heterogeneity is classified into pores, cracks, structural heterogeneity, and geometric abnormalities.

Pores result from gases left within the additively printed article during additive manufacturing. In the gas atomization method used to manufacture metal powder, gas is used to manufacture the metal powder, so it is assumed that gas components remaining in the metal powder become pores in the additively formed article. In addition, excessive heat input during additive manufacturing causes pores to be formed in connection with metal melting phenomena such as Marangoni convection caused by changes in the surface tension of the molten pool, evaporation of elements, and keyhole formation.

If the energy of the molten pool is insufficient and the powder particles cannot be melted, voids are created due to unfused areas (LOF: Lack of Fusion). The voids in the unfused portion often have an irregular shape and may contain unfused powder.

Cracks are also one of the defects that occur in additively shaped products, and the voids of the pores and unfused portions described above remain in the form of cracks. Cracks can also occur if there is a difference in the coefficient of linear expansion of the substrate and the additively formed part, or if there is a large thermal gradient in the molten pool during solidification.

Next, we will discuss structural heterogeneity (anisotropic inclusions). Generally, when the amount of heat input is changed during modeling, the metal structure of the additively shaped article changes. As the temperature gradient of the molten pool changes, the solidification rate changes, which affects the microstructure such as the anisotropy of the additively formed product. Additionally, the atmosphere during modeling and impurities in the metal powder, such as the formation of oxides due to oxygen, also affect the microstructure. Changes in the metallographic structure are directly linked to changes in the mechanical properties of the additively shaped article.

Finally, we will discuss geometric anomalies. Geometric anomalies are related to dimensional changes and surface roughness. This relates to the stability of the melt value, and changes in the size and shape of the weld pool have a large effect on the dimensions and surface roughness of additively shaped products. A stable weld pool size and shape is required to minimize these geometric anomalies.

The pores (including LOF), cracks, structural heterogeneity, and geometric abnormalities described above are all thought to be formed due to melting phenomena during additive manufacturing. Therefore, by monitoring information related to the melting phenomenon during additive manufacturing, it is possible to determine the quality of the additively manufactured product.

《Monitoring data》
There are various methods of additive manufacturing. The powder bed fusion method involves irradiating a flat sheet of metal powder with a laser beam (L-PBF: Laser Powder Bed Fusion) or an electron beam (EBM: Electron Beam Melting). This is a manufacturing method.
The directed energy deposition method is a method of performing additive manufacturing while discharging metal powder, and includes LMD (Laser Metal Deposition) and DMP (Direct Metal Printing). As a representative example, we will discuss monitoring of the molten pool in powder bed fusion bonding using laser beam irradiation.

Melting in additive modeling is caused by laser irradiation. The molten pool is affected by modeling parameters such as laser output and scanning speed, as well as material properties such as the enthalpy required to melt the metal powder and laser reflectance. One of the changes in the molten pool is a change in the temperature of the molten pool. By monitoring information related to temperature changes in the molten pool, the status of the molten pool can be estimated. In powder bed fusion bonding using laser beam irradiation, a plasma plume consisting of reflected and scattered laser, infrared and other electromagnetic waves depending on temperature, and ionized gas and metal vapor is emitted from the molten pool. The electromagnetic waves and plasma plume emitted from the molten pool change depending on the state and temperature of the molten pool, so they are important indicators for detecting the formation of various defects related to the melting phenomenon and determining quality.

Common monitoring devices in additive manufacturing include visible light monitoring, electromagnetic wave monitoring, and acoustic monitoring. Here, electromagnetic wave monitoring will be explained in detail.

Electromagnetic wave monitoring mainly observes infrared rays (wavelength from 700 nm) whose intensity changes depending on temperature, and uses optical tomography (OT), photodiodes, two-color thermometers, thermography, spectrometers, and high-speed cameras. and so on. Photodiodes and two-color thermometers used for detailed observation of the molten pool (melt pool monitoring) do not have spatial resolution, but they are installed coaxially with the laser and measured, and compared with the set laser irradiation pattern. By doing so, the spatial relationship can be determined. In addition to thermal radiation from the molten pool, it is also necessary to consider the reflection and scattering of laser light, but generally only the desired wavelength is observed using a spectral filter.

Optical tomography and thermography using CCD (Charge Coupling Device) cameras and CMOS (Complementary MOS) cameras are characterized by their spatial resolution. These CCD cameras and CMOD cameras can be installed coaxially with the laser to monitor electromagnetic waves.
Furthermore, a CCD camera or a CMOD camera can monitor the entire printing area from the top of the printing chamber. These are non-coaxial electromagnetic wave monitoring means. Also, like the photodiodes mentioned above, CCD cameras and CMOD cameras use spectral filters to select the wavelength range to observe.

In this way, information related to melting can be observed by coaxial or non-coaxial electromagnetic wave monitoring during additive modeling. Changes in the molten pool (temperature, convection, size, fusion depth, etc.) are related to defect formation, so it is possible to infer the defect location by arranging the relationship between defects formed in additively manufactured products and electromagnetic wave monitoring. becomes.

《Defect inspection of additively shaped products》
As mentioned above, dissolution in additive manufacturing is influenced by the modeling parameters and the physical properties of the metal powder to be modeled. The modeling parameters and the metal powder to be modeled vary depending on the target additively modeled product. Therefore, in electromagnetic wave monitoring, for example, the monitoring means cannot estimate quality using a uniform method, such as acquiring brightness information, determining a brightness threshold, and estimating the presence or absence of a defect. Therefore, the quality determination method used in the present invention requires confirming the formation of defects in the target material in advance in order to know the correlation between quality items (heterogeneity, material properties, residual stress, etc.) and monitoring data. . Therefore, defect inspection methods (destructive and non-destructive) will be described here.

Destructive methods include those that require cutting or grinding of additively shaped articles, and also include those that require specimen processing to obtain material properties. The above-mentioned heterogeneous parts (pores, cracks, structural heterogeneity, and geometric abnormalities) can also be confirmed by observing the structure at each position of the additively shaped article. For example, by polishing the cut cross section at a desired position and observing it with an optical microscope, scanning electron microscope, EPMA (electron beam probe microanalyzer), EBSD (electron backscatter diffraction method), etc., pores, cracks, etc. ), structural inhomogeneities and geometric abnormalities (deformations) inside the additively manufactured article can be confirmed. Geometric abnormalities (roughness) can also be evaluated using a roughness meter, laser microscope, etc.

In addition, material properties include mechanical properties (tensile tests, fatigue tests, hardness tests, impact tests, etc.) and corrosion resistance evaluation tests (immersion tests in various corrosive liquids, electrochemical methods such as official potential measurements), etc. It will be done.
Next, examples of nondestructive testing include X-ray CT, surface roughness measurement, and density measurement.

《Correspondence between heterogeneous areas and monitoring data》
Powder bed fusion bonding using laser beam irradiation performs additive modeling by stacking modeling layers of a certain thickness. The thickness T of the modeling layer is adjusted depending on the material and modeling parameters. Here, consider the modeling of the Nth layer (N is a natural number) and the N+1th layer or higher, which is the upper layer of the Nth layer. First, the Nth layer defect is formed during melting in the Nth layer modeling. Therefore, it is necessary to predict the defect position of the Nth layer from the monitoring data of the Nth layer. Generally, laser melting is performed to a thickness T or more of the modeling layer.

FIG. 2 is a schematic diagram showing the molten pool 21 spanning both the N+1 layer 23 and the N layer 22.
In the modeling of the N+1 layer 23, by melting up to the N layer 22, the N layer 22 and the N+1 layer 23 are tightly bonded. Therefore, the quality of the Nth layer 22 is affected by melting in the upper N+1th layer 23 and above. In the Nth layer 22, the disappearance and formation of a heterogeneous portion due to the melting of the N+1st layer 23 occurs, so this needs to be determined from the monitoring data of the N+1th layer 23. Similarly, it is necessary to consider the influence on defect formation in the Nth layer 22 over a predetermined number of layers in the layers above this, the N+2th layer and above.

Here, since the depth of the molten pool is finite, when considering the Nth layer defect, it is not necessary to consider the melting of all the upper layers. The depth of melting is affected by the temperature at the time of melting. In the monitoring data, deep melting has occurred in areas where high temperatures are expected, that is, where infrared rays are high in brightness. In areas where the temperature is expected to be low, that is, where the brightness of infrared rays is low, it is determined that the depth of melting is relatively shallow. In this way, by considering that the penetration depth varies depending on the brightness of infrared rays, it is possible to estimate the influence of the upper layer on the formation of defects in the Nth layer.

To predict defects, it is first necessary to perform additive manufacturing on a standard test piece under conditions similar to those used for actual additively manufactured products, and to collect monitoring data and quality measurement results. As mentioned above, additive modeling is performed with consideration to the melting depth. For example, consider a case where the position of a pore, which is an internal defect, is confirmed using X-ray CT and the relationship between the pore and monitoring data is sorted out. As mentioned above, the cause of pores is excessive heat input, that is, the molten pool becomes high temperature.

When the temperature of the molten pool becomes high, the brightness of infrared rays increases, so by examining the correspondence between the results of electromagnetic wave monitoring and the results obtained by X-ray CT, we can determine the presence or absence of pores or the probability of pore formation depending on the brightness. can be calculated. At this time, since the formation of pores in the Nth layer is affected by the formation of defects in the N+1st layer and above, an analysis is performed that takes into account the formation and disappearance of pores in the Nth layer from the monitoring data in the N+1th layer and above. By comparing and examining the pore formation position and monitoring data by X-ray CT, it is possible to obtain the relationship between the monitoring data of the Nth layer and its upper layer, which affects the pore formation in the Nth layer.

Although the relationship between the pore position and infrared monitoring data obtained by X-ray CT is illustrated here, the data relationship can be obtained in a similar manner even in the case of other heterogeneous parts and monitoring methods.

FIG. 3 is a block diagram showing the functional configuration of the additionally shaped product quality determination device 1. As shown in FIG.
The additive-molded product quality determination device 1 determines the quality of the additive-molded product from optical tomography data obtained during the modeling of the additive-molded product. The additive-shaped product quality determination device 1 includes, as functional configurations, an output section 11, a determination section 12, a correlation determination section 13, a first evaluation section 14, a second evaluation section 15, a data acquisition section 16, It has a defect position calculation section 17. The additive-fabricated product quality determination device 1 is a computer equipped with a processor to be described later, and by executing an additive-fabricated product quality determination program (not shown), embodies each functional unit and executes an additive-fabricated product quality determination method. do.

The first evaluation unit 14 predicts the generation of defects in each layer based on the monitoring results of each layer of the additively manufactured article by the additively molded device 4 equipped with a monitoring device. Here, the prediction of generation of defects in each layer is the first evaluation result. Monitoring results are provided from at least one of an optical tomography, a CCD camera, a CMOS camera, a two color thermometer, a thermography, a high speed camera, and a photodiode. The first evaluation unit 14 uses the brightness mapping image of the optical tomography of each layer of the additively shaped article as the monitoring result, but is not limited thereto, and may use a predetermined shape in the optical tomography as the monitoring result.

The second evaluation unit 15 predicts the generation and/or disappearance of defects in each layer from the monitoring results of a predetermined number of layers above each layer of the additively shaped article by the first evaluation unit 14, and Reflected in generation prediction result information. Here, the defect generation prediction result information for each layer is the second evaluation result.

The output unit 11 outputs defect position information of the additively shaped product based on the defect generation prediction result information of each layer obtained by the second evaluation unit 15. The determination unit 12 determines the quality of the additionally shaped article based on the defect information of the additionally shaped article outputted by the output unit 11. The quality data determined by the determining unit 12 includes at least one of internal defect position, internal defect size, density, defect rate, crack, structural heterogeneity, and geometric abnormality.

The correlation determination unit 13 determines the superposition of a predetermined number of layers that have a correlation with the defect of the Nth layer (N is a natural number) of this additionally shaped article that is higher than a threshold value.

The data acquisition unit 16 acquires monitoring data of all layers of additive modeling.
The defect position calculation unit 17 calculates defect position information of this additionally shaped article from the monitoring data.

FIG. 4 is a block diagram showing the hardware configuration of the additionally shaped product quality determination device 1. As shown in FIG.
The additional shaped product quality determination device 1 has a processor 101, a storage section 102, an operation section 103, a display section 104, and a communication section 105 as a hardware configuration. The processor 101, the storage section 102, the operation section 103, the display section 104, and the communication section 105 are connected by a bus 106.

The processor 101 is, for example, a CPU (Central Processing Unit), and centrally controls the additional molded product quality determination device 1 . The storage unit 102 becomes a work area for the processor 101. Further, the storage unit 102 is a non-temporary or temporary recording medium that stores various programs and data, and stores an additional molded product quality determination program (not shown). Examples of the storage unit 102 include ROM (Read Only Memory), RAM (Random Access Memory), HDD (Hard Disk Drive), and flash memory.

The operation unit 103 inputs operation data. The operation unit 103 includes, for example, a keyboard, a mouse, a touch panel, and a numeric keypad. The display unit 104 is, for example, a display or a printer, and displays data. The communication unit 105 connects to a network and transmits and receives data.

《Quality judgment of additively molded products in actual machine molding》
When creating additively shaped products, quality is determined by predicting non-uniform areas based on the relationship between monitoring data obtained using standard test pieces and the non-uniform areas.

FIG. 5 is a flowchart of the quality determination process.
First, the data acquisition unit 16 acquires monitoring data of all layers of additive modeling (step S10). This monitoring data is optical tomography data.
Next, the processor 101 repeats the process for each layer of the monitoring data of the additionally shaped article from steps S11 to S15.

Through this repeated process, the first evaluation unit 14 predicts the generation of the Nth layer heterogeneous portion from the Nth layer monitoring results (step S12). The information predicted here is also called the first evaluation result.

Then, the second evaluation unit 15 predicts the generation and/or disappearance of the Nth layer heterogeneous portion based on the monitoring results of the N+1th layer (step S13), and reflects this in the Nth layer defect generation prediction result information. (Step S14). The information reflected here is also called the second evaluation result.

The processor 101 determines whether the processing has been repeated for all layers of the monitoring data of this additionally shaped article (step S15). If the processor 101 has not repeated the process for all layers of the monitoring data of this additionally shaped article, it returns to the process of step S11.

The output unit 11 outputs heterogeneity information of the additionally shaped article based on the second evaluation results obtained for each layer (step S16). When the determination unit 12 determines the quality of the additionally shaped article based on this output result (step S17), the process of FIG. 5 is ended. Quality can be determined arbitrarily with respect to defect rate, defect size, density, defect position, number density of inclusions, material properties, etc.
The quality determination described here may be performed not only after modeling, but also at the same time as modeling.

Although the monitoring data from the N+1 layer and above is not described here, it is possible to use the monitoring data from the N+1 layer and above to predict the heterogeneous portion of the N layer, taking into account the melting depth at the time of melting. good.

FIG. 6 is a flowchart of quality determination processing according to a modification.
First, the data acquisition unit 16 acquires monitoring data of all layers of additive modeling (step S20).
Next, the processor 101 repeats the process for each layer of the monitoring data of the additionally shaped article from steps S21 to S25.

Through this repeated process, the first evaluation unit 14 predicts the generation of the Nth layer heterogeneous portion from the Nth layer monitoring results (Step S22). The information predicted here is also called the first evaluation result.

Then, the second evaluation unit 15 predicts the generation and/or disappearance of the heterogeneous portion of the Nth layer based on the monitoring results of the N+1st layer to the N+Mth layer (M is a natural number) (step S23), and It is reflected in the defect generation prediction result information (step S24). The information reflected here is also called the second evaluation result.

The processor 101 determines whether the processing has been repeated for all layers of the monitoring data of this additionally shaped article (step S25). If the processor 101 has not repeated the process for all layers of the monitoring data of this additionally shaped article, the process returns to step S21. After the processor 101 repeats the process for all layers of the monitoring data of the additionally shaped article, the process proceeds to step S26.

In step S26, the output unit 11 outputs heterogeneity information of the additively shaped article based on the second evaluation results obtained for each layer. When the determination unit 12 determines the quality of the additionally shaped article based on this output result (step S27), the process of FIG. 6 ends.

Further, it is preferable to determine in advance the number of layers above the Nth layer that will affect the Nth layer. As a result, the additive-printed product quality determination device 1 can calculate the suitable number of overlapping layers of the additive-printed product, output highly accurate heterogeneity information, and judge the quality of the additive-printed product with high accuracy. Become. Furthermore, in order to guarantee the quality of the additively shaped article, it is conceivable to use, for example, a superimposed image of multiple layers of the additively shaped article as an attached document to indicate the quality of the additively shaped article.

FIG. 7 is a flowchart of a process for determining the number of overlapping layers of an additively shaped article.
First, the data acquisition unit 16 acquires monitoring data of all layers of additive modeling (step S31). The defect position calculation unit 17 calculates defect position information of the additive-molded product from the monitoring data of all layers of the additive-molded product (step S32).

Next, the processor 101 repeats the variable P from 1 to a predetermined number from steps S33 to S36. Note that the variable P is a natural number.

Through this repeated process, the first evaluation unit 14 and the second evaluation unit 15 calculate superimposition information of the monitoring results from the Nth layer to the N+Pth layer (step S34). Then, the correlation determination unit 13 calculates the correlation between the defect occurrence area of the X-ray CT image and the superimposition information of the monitoring results.

In step S36, the processor 101 determines whether the variable P has been repeated from 1 to a predetermined number. If the processor 101 has not repeated the variable P from 1 to the predetermined number, the process returns to step S33. After the processor 101 repeats the variable P from 1 to a predetermined number, the processor 101 proceeds to step S37.

In step S37, when the correlation determination unit 13 determines the superposition of a predetermined number of layers whose correlation with the occurrence of defects is higher than the threshold value, the process of FIG. 7 ends.
Hereinafter, the present disclosure will be explained in more detail with reference to Examples. Note that the present disclosure is not limited to these examples.

《Modeling and monitoring equipment》
Molding was performed by powder bed fusion bonding using laser beam irradiation, and monitoring was performed using an optical tomography monitoring system. Inside the modeling area, a sCMOS (scientific CMOS) camera installed diagonally above records the brightness generated from the molten pool by laser irradiation from vertically above. The sCMOS camera is installed diagonally above the modeling surface, but the distance and angle are corrected in the analysis program and converted to look like an image viewed from directly above. In subsequent measurements, the luminance integrated value was used. The integrated brightness value is the integrated value of the brightness in one pixel of the image. Note that the unit of optical tomography data is expressed in Gv (Gray value).

《Building conditions》
In this build, only the in-skin condition was applied, in which the build area was filled with laser irradiation. The laser output and construction speed were set to conditions that resulted in a higher heat input than the appropriate range. Here, the appropriate range is experimentally defined as a condition under which the internal defect rate is less than a predetermined rate.

《X-ray CT》
In order to determine defects in additive-molded products from optical tomography data, it is necessary to obtain X-ray CT results that serve as a reference for specifying the defect size and position. The obtained CT data was reconstructed into three-dimensional data, and then analyzed using the defect/inclusion analysis module of the analysis/visualization software. The inventors then obtained the defect rate in addition to the defect location and defect distribution.

《Density measurement》
In order to confirm the accuracy of the defect rate obtained by X-ray CT, the inventors calculated the defect rate from relative density measurement using an immersion method. The immersion method used Archimedes' principle. The inventors obtained a defect rate of 2.5% from a comparison of the obtained density and true density. Since the defect rate was 2.8% in the X-ray CT analysis, it was determined that although an error of about 11% was observed, the defect information could be obtained generally accurately with the X-ray CT.

《Correlation between optical tomography and quality》
The X-ray CT was output as slice data with a thickness of a predetermined pitch, and eight slices were stacked so that the thickness was the same as that of one layer.

FIG. 8 is a diagram showing an X-ray CT image of the Nth layer.
Black portions in this X-ray CT image indicate defects.

9A to 9D are brightness mapping images in optical tomography of the Nth layer to the N+3rd layer.
Comparison of the high-brightness part of optical tomography and X-ray CT in each layer from the Nth layer to the N+3th layer shows that many defects are also present in the low-brightness part.

FIG. 10 is a diagram showing an image in which four layers of luminance mapping images obtained by optical tomography are stacked.
This image is an image in which four layers of luminance mapping images obtained by optical tomography are stacked. With this image, it is possible to show, by density, areas where heat has been continuously applied over multiple layers. This image is an image for quality evaluation of the modeled product, and is preferably included in the document attached to the modeled product as modeling quality information.

FIG. 11 is a graph showing the brightness and defect rate of optical tomography superimposition.
The vertical axis of the graph indicates the defect rate. The horizontal axis of the graph indicates the brightness of optical tomography. From this graph, it can be confirmed that the higher the brightness, the higher the probability that it is a defect.
In this way, by analyzing the monitoring data in consideration of the melting depth and then performing actual additive manufacturing, it is possible to predict the heterogeneous portion of the additively manufactured product.

(Modified example)
The present invention is not limited to the embodiments described above, and includes various modifications. For example, the embodiments described above are described in detail to explain the present invention in an easy-to-understand manner, and are not necessarily limited to those having all the configurations described. It is possible to replace a part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. Furthermore, it is also possible to add, delete, or replace some of the configurations of each embodiment with other configurations.

Part or all of the above configurations, functions, processing units, processing means, etc. may be realized by hardware such as an integrated circuit. Each of the above configurations, functions, etc. may be realized by software by a processor interpreting and executing a program for realizing each function. Information such as programs, tables, and files that implement each function can be stored in a storage device such as memory, hard disk, SSD (Solid State Drive), or recording medium such as a flash memory card or DVD (Digital Versatile Disk). can.

In each embodiment, the control lines and information lines are those considered necessary for explanation, and not all control lines and information lines are necessarily shown in the product. In reality, almost all components may be considered interconnected.

1 Additionally shaped product quality determination device 101 Processor 102 Storage unit 103 Operation unit 104 Display unit 105 Communication unit 106 Bus 11 Output unit 12 Judgment unit 13 Correlation determining unit 14 First evaluation unit 15 Second evaluation unit 16 Data acquisition unit 17 Defect position Calculation unit 21 Molten pool 22 Nth layer 23 N+1st layer 4 Additive modeling device 41 Additive modeling unit 411 Processing chamber 412a Gas supply pipe 412b Gas exhaust pipe 413 Powder feeder 414 Raw powder 415a Sample stage 415b Sample stage 417 Additive shaped object 42 Laser oscillator 420 Raw material powder storage area 421 Additional modeling area 422 Raw material powder collection area 430 Control device 43 Process fiber 44 Galvano head 45 Laser coaxial illumination 46 Visible light imager 47a Infrared imager 47b Plasma emission imager 48 Molten pool observation device

Claims (15)

a first evaluation unit that predicts the generation of defects in each layer based on the monitoring results of each layer of the additively manufactured product by the additively molded device equipped with a monitoring device;
The generation and/or disappearance of defects in each layer is predicted from the monitoring results of a predetermined number of layers above each layer of the additively shaped article by the first evaluation unit, and the defects in each layer are determined. a second evaluation unit that reflects the generated prediction result information;
an output unit that outputs defect position information of the additively shaped product based on defect generation prediction result information of each layer obtained by the second evaluation unit;
An additively formed product quality determination device comprising: a determination unit that determines the quality of the additively shaped product based on the defect position information of the additively shaped product outputted by the output unit;
The additionally shaped product quality determination device according to claim 1, further comprising: The quality data determined by the determination unit includes at least one of internal defect position, internal defect size, density, defect rate, crack, structural heterogeneity, and geometric abnormality.
3. The additionally shaped product quality determination device according to claim 2. The first evaluation unit uses a brightness mapping image of optical tomography of each layer of the additively shaped article as the monitoring result.
The additionally shaped product quality determination device according to claim 1, characterized in that: The first evaluation unit takes a predetermined shape in optical tomography of each layer of the additively shaped article as the monitoring result.
The additionally shaped product quality determination device according to claim 1, characterized in that: The second evaluation unit generates defects in each layer from the monitoring results from the first layer to the layer M layers (M is a natural number of 2 or more) above each layer of the additively formed product. or/and predicting the disappearance and reflecting it in the generation prediction result information of the defects in each layer;
The additionally shaped product quality determination device according to claim 1, characterized in that: The monitoring results are provided from at least one of an optical tomography, a CCD camera, a CMOS camera, a two-color thermometer, a thermography, a high-speed camera, and a photodiode.
The additionally shaped product quality determination device according to any one of claims 1 to 6. comprising a correlation determination unit that determines the superposition of a predetermined number of layers whose correlation with the defect of the Nth layer (N is a natural number) of the additively shaped product is higher than a threshold;
The additionally shaped product quality determination device according to claim 1, characterized in that: comprising a defect position calculation unit that calculates defect position information of the additionally shaped product from the monitoring results;
9. The additionally shaped product quality determination device according to claim 8. predicting the generation of defects in each layer based on the monitoring results of each layer of the additively manufactured article by an additively molded device equipped with a monitoring device;
predicting the generation and/or disappearance of defects in each layer based on the monitoring results of a predetermined number of layers above each layer of the additively formed article;
reflecting the defect generation prediction result information of each layer;
outputting defect position information of the additively shaped product based on defect generation prediction result information obtained in each layer;
A method for determining the quality of additively manufactured products, characterized by performing the following steps. a step of determining the quality of the additively shaped article based on the defect position information of the additively shaped article;
The method for determining quality of an additively shaped product according to claim 10, further comprising: The quality data determined for the additively shaped article includes at least one of internal defect location, internal defect size, density, defect rate, cracks, structural heterogeneity, and geometric abnormality.
The method for determining the quality of an additively shaped product according to claim 11. a step of using a brightness mapping image of optical tomography of each layer of the additively shaped article as the monitoring result;
11. The method for determining the quality of an additively shaped product according to claim 10. a step of determining a predetermined shape in optical tomography of each layer of the additively shaped article as the monitoring result;
The method for determining the quality of an additively shaped product according to claim 10. Predicting the generation and/or disappearance of defects in each layer from the monitoring results from the first layer to the layer M layers (M is a natural number of 2 or more) above each layer of the additively shaped product. a step of reflecting the defect generation prediction result information in each layer;
The method for determining the quality of an additively shaped product according to claim 10.

Images