Method for manufacturing metal additive product, metal additive manufacturing device, and metal additive product

Abstract

This method for manufacturing a metal additive product comprises a wire measurement step, a receiving step, a manufacturing condition generation step, and a first evaporation amount prediction step. In the wire measurement step, wire surface condition information related to the evaporation amount of a low-boiling-point metal in a raw material wire is measured. In the receiving step, manufacturing product information including the shape and material of the metal additive product is received. In the manufacturing condition generation step, manufacturing condition information for manufacturing the metal additive product is generated on the basis of the manufacturing product information. In the first evaporation amount prediction step, the evaporation amount of the low-boiling-point metal when the metal additive product indicated in the manufacturing product information is manufactured according to the manufacturing condition information is predicted on the basis of the wire surface condition information and the manufacturing condition information.

Description

Method for manufacturing a metal laminated object, metal laminated manufacturing apparatus, and metal laminated object

[0001] The present disclosure relates to a method for manufacturing a metal laminated object, a metal laminated manufacturing apparatus, and a metal laminated object, in which a laser beam is irradiated onto a wire made of an alloy to manufacture the metal laminated object.

[0002] Regarding the metal lamination molding of Al (aluminum) alloys, AlSi10Mg, A5183, etc., which are less likely to cause welding cracks, are widely used. In recent years, due to the need for weight reduction, metal lamination molding using high-strength Al alloys, that is, Al alloys of the A7000 series, has attracted attention. The A7000 series of Al alloys achieve high strength by containing Zn (zinc), Mg (magnesium), and Cu (copper) as additives. However, in metal lamination molding, it is known that cracks occur due to the precipitation of these elements at grain boundaries. Patent Document 1 discloses that a trace amount of additive is effective for suppressing cracks in Al alloys with high crack sensitivity such as the A7000 series. In Patent Document 1, for example, it is disclosed that Zr (zirconium) and Sc (scandium) have an effect of preventing casting cracks as crystal grain size refinement agents.

[0003] In addition, it is known that Zn and Mg, which are low-boiling metals, evaporate when molding using the A7000 series. In particular, since Zn has a lower boiling point than Mg, Zn evaporates considerably during molding. Therefore, Patent Document 1 discloses that when performing metal lamination molding using a powder of an A7000 series Al alloy, a powder raw material in which the amounts of Zn and Mg in the powder are about 20% higher than the component target values of the molded object is used.

[0004] Thus, in Patent Document 1, metal lamination molding is performed using a powder raw material with a uniformly 20% increase in the addition of low-boiling metals. However, in metal lamination molding, the evaporation amount of the metal changes depending on the molding conditions. Furthermore, since the evaporation amount also changes during molding, even if the addition amounts of Zn and Mg are uniformly increased in the powder raw material, the composition of the molded object will not be uniform, and the evaporation amounts of Zn and Mg will differ between molded samples.

[0005] Japanese Patent No. 7388670

[0006] Metal additive manufacturing methods include those using arcs or lasers as heat sources, and those using powders or wires as raw materials. When using arcs as the heat source, excessive heat input occurs, and when using powders as the raw material, the large surface area leads to overheating. Therefore, to suppress the evaporation of metals with lower boiling points than Al in Al alloys such as the A7000 series, the most suitable manufacturing method is one that uses a laser as the heat source and wires as the raw material. However, in metal additive manufacturing methods using this combination, no technology has been proposed to suppress the evaporation of metals with lower boiling points than Al during manufacturing compared to conventional methods.

[0007] This disclosure has been made in view of the above, and aims to provide a method for manufacturing a metal additively manufactured product that can suppress the amount of evaporation of the metal with a lower boiling point than Al in the metal additively manufactured product compared to conventional methods, when using an Al alloy containing a metal with a lower boiling point than Al as the raw material wire and irradiating the raw material wire with laser light to manufacture the metal additively manufactured product.

[0008] To solve the above-mentioned problems and achieve the objective, the present disclosure provides a method for manufacturing a metal additively manufactured product, which involves sequentially stacking fabricated layers formed by irradiating a raw material wire made of an Al alloy containing a low-boiling-point metal, a metal with a lower boiling point than Al, with a laser beam, and includes a wire measurement step, a receiving step, a fabrication condition generation step, and a first evaporation amount prediction step. The wire measurement step measures wire surface state information, which is information related to the evaporation amount of the low-boiling-point metal in the raw material wire. The receiving step receives fabricated product information, including the shape and material of the metal additively manufactured product. The fabrication condition generation step generates fabrication condition information, which is the conditions for fabricating the metal additively manufactured product, based on the fabricated product information. The first evaporation amount prediction step predicts the evaporation amount of the low-boiling-point metal when the metal additively manufactured product shown in the fabricated product information is fabricated according to the fabrication condition information, based on the wire surface state information and the fabrication condition information.

[0009] According to this disclosure, when an Al alloy containing a metal with a lower boiling point than Al is used as the raw material wire, and a metal additive manufacturing product is created by irradiating the raw material wire with laser light, the evaporation amount of the metal with a lower boiling point than Al in the metal additive manufacturing product can be suppressed compared to conventional methods.

[0010] Figure showing an example of the relationship between the characteristics of the wire material and the evaporation rate of Zn, a low-boiling-point metal. Block diagram showing an example of the configuration of a metal additive manufacturing apparatus according to Embodiment 1. Figure schematicly showing an example of the configuration of the molding section according to Embodiment 1. Block diagram showing an example of the configuration of the evaporation rate prediction section. Figure schematicly showing an example of a neural network used by the model generation section. Flowchart showing an example of the procedure for generating a trained model. Flowchart showing an example of the procedure for manufacturing a metal additive manufactured object according to Embodiment 1. Flowchart showing an example of the procedure for manufacturing a metal additive manufactured object according to Embodiment 1. Flowchart showing an example of the procedure for predicting the evaporation rate of a low-boiling-point metal. Block diagram showing an example of the configuration of a metal additive manufacturing apparatus according to Embodiment 2. Block diagram showing an example of the configuration of the evaporation rate prediction section. Flowchart showing an example of the procedure for manufacturing a metal additive manufactured object according to Embodiment 2. Flowchart showing an example of the procedure for predicting the evaporation rate of a low-boiling-point metal. Block diagram showing an example of the configuration of a computer system that realizes the calculation processing section of the metal additive manufacturing apparatus according to Embodiments 1 and 2.

[0011] The method for manufacturing a metal additively manufactured product, a metal additive manufacturing apparatus, and a metal additively manufactured product according to embodiments of this disclosure will be described in detail below with reference to the drawings.

[0012] Before describing the embodiments, let me briefly explain the background leading to this disclosure. As mentioned above, the challenge in additive manufacturing using Al alloys containing metals with lower boiling points than Al, such as the A7000 series, as raw materials is the evaporation of the metals with lower boiling points than Al. Hereafter, metals with lower boiling points than Al will also be referred to as low-boiling-point metals. Examples of low-boiling-point metals are Zn and Mg. Furthermore, this disclosure focuses on metal additive manufacturing using a laser as the heat source and wire as the raw material.

[0013] The inventors focused on wire materials from the viewpoint of suppressing evaporation. Figure 1 shows an example of the relationship between the properties of the wire material and the evaporation rate of Zn, a low-boiling-point metal. In this figure, "additive-free material" in the wire material indicates an Al alloy wire without grain size refiners, and "additive material" indicates an Al alloy wire with grain size refiners added. Here, the case where the wire material is A7075 material, which is an example of the A7000 series, is shown.

[0014] High-strength wires like the A7000 series are prone to breakage during the wire drawing process. In other words, during the wire drawing process, a round bar is stretched to create a wire, but the round bar undergoes work hardening during stretching, so if stretching continues, it becomes harder and more brittle. As a result, the round bar breaks during stretching, making wire manufacturing difficult. In particular, grain size refiners such as Zr and Sc have high melting points, so they do not dissolve during wire production and tend to grow into coarse particles, further increasing the risk of breakage compared to additive-free materials.

[0015] To prevent this, annealing is performed to remove the work hardening of the wire. However, since annealing is done in the atmosphere, the amount of oxygen on the surface increases. In other words, wires that have been annealed many times will have a higher amount of oxygen on their surface. The number of annealing cycles required varies depending on the material; materials that are not easily stretched will require more annealing cycles.

[0016] When metal additive manufacturing was performed using wires manufactured in this manner, it was found that Al alloys have a relatively low absorption rate for YAG (Yttrium Aluminum Garnet) lasers. Furthermore, the inventors' studies revealed that changes in the laser absorption rate of the raw wire surface have a drastic effect on the manufacturing conditions. In particular, the surface of Al alloy wires is covered with a native oxide film, and the more of this oxide film there is, the higher the laser absorption rate. As a result, the surface of the Al alloy wire is locally overheated, generating sputter or fumes, which in turn promotes the evaporation of low-boiling-point metals. Figure 1 shows that, regardless of whether or not a grain size refiner was added to the A7075 material, the more annealing cycles a wire material undergoes, the greater the surface oxygen content and the higher the laser absorption rate, resulting in a greater amount of Zn evaporation at the start of manufacturing. Note that when the wire material at the bottom of the table in Figure 1 is A7075 (additive material), it is shown that because the number of annealing cycles is low, the wire breaks during the wire drawing process, making it impossible to manufacture the wire and perform metal additive manufacturing.

[0017] As described above, we have found that the surface oxygen content of the wire or the laser absorptivity plays a significant role in controlling the evaporation of low-boiling-point metals. Below, we will describe a method for manufacturing metal additive manufacturing products, a metal additive manufacturing apparatus, and embodiments of metal additive manufacturing products that take into account the surface oxygen content of the wire or the laser absorptivity to suppress the evaporation of low-boiling-point metals.

[0018] Embodiment 1. Figure 2 is a block diagram showing an example of the configuration of a metal additive manufacturing apparatus according to Embodiment 1. The metal additive manufacturing apparatus 10 comprises a raw material wire measuring unit 11, a calculation processing unit 100, and a molding unit 20. The calculation processing unit 100 comprises a molded object information input unit 12, a molding condition generation unit 13, a performance data storage unit 14, an evaporation amount prediction unit 15, and a determination processing unit 16.

[0019] The raw wire measurement unit 11 measures wire surface state information, which is information indicating the surface state of the raw wire related to the evaporation amount of low-boiling-point metal in the raw wire of the molding unit 20, and outputs the wire surface state information to the molding condition generation unit 13. At least one of the following can be cited as wire surface state information: surface oxygen content and laser absorptivity. In Embodiment 1, the raw wire measurement unit 11 measures the wire surface state information at the start of molding. When the wire surface state information is surface oxygen content, an example of the raw wire measurement unit 11 is an energy dispersive x-ray spectroscopy (EDS) system attached to a scanning electron microscope (SEM), or an X-ray fluorescence analyzer. Furthermore, a handheld simple elemental analysis instrument may be used as the raw wire measurement unit 11. When the wire surface state information is laser absorptivity, an example of the raw wire measurement unit 11 is a spectrophotometer. The spectrophotometer measures the laser absorptivity in a specific wavelength range. When the laser oscillator is a YAG laser, the laser absorptivity is measured for wavelengths near 1064 nm, which is the wavelength range of the YAG laser. For this reason, it is sufficient to measure only the wavelength range near 1064 nm, and the laser absorptivity can be measured by transmitting only specific wavelengths using a dichroic mirror. Furthermore, the wire surface condition information may include heat input information that combines the boiling point of the low-boiling-point metal contained in the raw wire, the composition information of the low-boiling-point metal in the raw wire, and the amount of heat input per unit volume to the raw wire. However, in this case, the amount of heat input per unit volume to the raw wire is information calculated using the laser output value, laser diameter, etc., of the molding conditions described later.

[0020] The object information input unit 12 receives input of object information, including the shape and material of the metal additive manufactured object to be manufactured by the manufacturing unit 20. The object information input unit 12 may also be an input device such as a keyboard or mouse. In this case, the object information is directly input by the user, or the keyboard, mouse, etc., receives the input object information. In another example, the object information input unit 12 may be a storage medium interface that reads object information stored in a portable storage medium. In yet another example, the object information input unit 12 may be a communication interface that reads object information from another information processing device via communication. The object information input unit 12 outputs the object information to the manufacturing condition generation unit 13.

[0021] The molding condition generation unit 13 generates molding condition information for a metal additive manufactured object based on the manufactured object information. A metal additive manufactured object is formed by stacking a bead layer, which is a bead layer created by irradiating a raw wire with a laser beam and melting and solidifying it. The generated molding condition information is for the time when molding starts, but in Embodiment 1, it is assumed that molding will be performed using the molding conditions at the time when molding starts until the end of molding. The molding condition generation unit 13 generates molding environment condition information, which is various conditions for molding a metal additive manufactured object of the material and shape shown in the manufactured object information. The molding environment condition information is the conditions that define the environment in which metal additive manufacturing is performed in the molding unit 20. An example of the molding environment condition information is the laser output value, axis speed, wire supply speed, laser diameter, shielding gas amount, and stage rotation speed. The laser output value is information indicating the output of the laser beam from the laser oscillator 23. The axis speed is information indicating the movement speed of the processing head 26. The wire supply rate is the amount of wire 21 supplied by the wire nozzle 22 per unit time. The laser diameter is information indicating the diameter of the laser beam. The shielding gas amount is the flow rate of the shielding gas. The stage rotation speed is information indicating the rotation speed of the stage.

[0022] Furthermore, the molding condition generation unit 13 generates molding path information that indicates the path for molding a metal additive manufacturing object of the material and shape indicated in the molded object information. In one example, the molding path information is a path calculated by CAM (Computer Aided Manufacturing) and is a movement path that moves the processing point. In one example, the molding path information is generated for each molding layer that makes up the metal additive manufacturing object. The molding path information is set to avoid interference with the components of the molding unit 20. Furthermore, the shortest route among the paths that can mold the molding layers of the metal additive manufacturing object is generated as the molding path information. The molding condition generation unit 13 outputs the molding condition information, including the generated molding environment condition information and molding path information, to the evaporation amount prediction unit 15.

[0023] If the determination processing unit 16, described later, determines that the molding condition information is within an acceptable range for predicting the evaporation amount of low-boiling-point metal, the molding condition generation unit 13 generates a numerical control (NC) program necessary for processing in the molding unit 20 based on the molding condition information. The NC program is a processing program that controls the molding unit 20 to perform processing, and can also be said to be information that includes the molding condition information. The molding condition generation unit 13 outputs the generated NC program to the molding unit 20. In one example, the molding condition generation unit 13 generates an NC program for each molding layer and outputs it to the molding unit 20.

[0024] The performance data storage unit 14 stores performance data, which is various data related to additive manufacturing. The performance data is data obtained during additive manufacturing and includes wire surface condition information, manufacturing condition information, and actual values ​​of the evaporation amount of low-boiling-point metal at the start of manufacturing. In other words, the performance data is data that associates the manufacturing condition information from past manufacturing with wire surface condition information and actual values ​​of the evaporation amount of low-boiling-point metal at the start of manufacturing. The evaporation amount of low-boiling-point metal at the start of manufacturing is obtained by performing compositional analysis on the metal additive manufactured object after manufacturing using ICP (Inductively Coupled Plasma) emission analysis, an EDS system attached to the SEM, etc. The performance data stored in the performance data storage unit 14 is read out as appropriate by the evaporation amount prediction unit 15.

[0025] The evaporation rate prediction unit 15 uses the actual data stored in the actual data storage unit 14 and the molding condition information generated by the molding condition generation unit 13 to predict the amount of evaporation of low-boiling-point metal in the molding layer when molding a metal additive manufacturing object indicated by the molded object information.

[0026] In Embodiment 1, the evaporation rate prediction unit 15 uses a pre-trained model, learned through machine learning, that shows the relationship between wire surface state information and molding condition information and the actual evaporation rate of low-boiling-point metal at the start of molding. The unit predicts the evaporation rate of low-boiling-point metal at the start of molding based on the wire surface state information and molding condition information. The evaporation rate prediction unit 15 outputs the predicted evaporation rate of low-boiling-point metal at the start of molding to the determination processing unit 16. The detailed configuration of the evaporation rate prediction unit 15 will be described later.

[0027] The determination processing unit 16 determines whether or not fabrication is possible with the fabrication conditions information based on whether the predicted evaporation amount of the low-boiling-point metal at the start of fabrication, input from the evaporation amount prediction unit 15, falls within a set tolerance range. More specifically, the determination processing unit 16 determines, based on the received fabrication conditions information, whether the desired physical properties can be obtained when fabrication is performed with these fabrication conditions information. The determination processing unit 16 outputs the determination result to the evaporation amount prediction unit 15. Examples of desired physical properties include whether a metal additive fabricated object with a composition conforming to the standard can be obtained, or whether the composition has not decreased compared to the raw wire. The desired physical properties and the tolerance range in which the desired physical properties can be obtained can be arbitrarily set by the user. The tolerance range is the range of evaporation amounts of the low-boiling-point metal that can obtain these desired physical properties.

[0028] The evaporation rate prediction unit 15 outputs molding condition information to the molding condition generation unit 13 or adjusts the molding condition information based on the determination result from the determination processing unit 16. Specifically, if the evaporation rate prediction unit 15 determines that the predicted evaporation rate of the low-boiling-point metal at the start of molding does not fall within the acceptable range, it adjusts the molding condition information and uses the adjusted molding condition information to predict the evaporation rate of the low-boiling-point metal again. In one example, the evaporation rate prediction unit 15 displays a molding condition information adjustment screen on a display unit (not shown) and requests input from the user of the metal additive manufacturing apparatus 10. The evaporation rate prediction unit 15 sets the input value entered on the molding condition information adjustment screen as the adjusted molding condition information. The evaporation rate prediction unit 15 uses the wire surface condition information and the adjusted molding condition information as input to predict the evaporation rate of the low-boiling-point metal again. The evaporation rate prediction unit 15 repeats the above process until the predicted evaporation rate of the low-boiling-point metal falls within the acceptable range.

[0029] The molding condition information adjustment screen displays the values ​​of the molding condition information generated by the molding condition generation unit 13, along with the item names. The molding condition information adjustment screen may also display the predicted value and the allowable range of the evaporation amount of low-boiling-point metal. Users who view the molding condition information adjustment screen adjust the molding condition values ​​so that the evaporation amount of low-boiling-point metal falls within the allowable range, specifically so that the evaporation amount of low-boiling-point metal is reduced. In one example, the user adjusts the laser output. In another example, if the raw wire does not melt or melts poorly when the laser output is reduced, the user adjusts the axis speed without reducing the laser output. Such adjustments of the molding condition information values ​​are made based on the user's experience or through trial and error.

[0030] Furthermore, if the evaporation rate prediction unit 15 determines that the predicted evaporation rate of the low-boiling-point metal at the start of molding falls within an acceptable range, it considers the molding condition information to be appropriate and outputs the molding condition information used to predict the evaporation rate of the low-boiling-point metal at the start of molding to the molding condition generation unit 13. This molding condition information may be the molding condition information generated by the molding condition generation unit 13 as is, or it may be the molding condition information adjusted by the evaporation rate prediction unit 15.

[0031] The molding unit 20 fabricates a metal additive manufacturing object by repeatedly executing a process to form a molded layer according to an NC program that includes molding condition information generated by the molding condition generation unit 13 or molding condition information adjusted by the evaporation amount prediction unit 15.

[0032] Figure 3 is a schematic diagram showing an example of the configuration of the molding unit according to Embodiment 1. In Embodiment 1, the metal additive manufacturing apparatus 10 manufactures metal additive manufactured objects using a so-called Directed Energy Deposition (DED) method. The molding unit 20 supplies material to a commanded position and forms a bead with the material melted using a beam. The bead is a solidified product obtained when the molten material solidifies. The metal additive manufacturing apparatus 10 manufactures metal additive manufactured objects by sequentially stacking molding layers, which are layers of beads, by the molding unit 20.

[0033] The beam, which is the heat source for melting the material, is a laser beam L, an electron beam, or an arc. In Embodiment 1, the heat source is assumed to be a laser beam L. In Embodiment 1, the material is assumed to be an Al alloy wire, more specifically an Al alloy wire containing a metal with a lower boiling point than Al, such as Zn or Mg.

[0034] In Figure 3, the X, Y, and Z axes are three axes perpendicular to each other. The X and Y axes are two horizontal axes. The Z axis is a vertical axis. In each of the X, Y, and Z axis directions, the direction indicated by the arrow is considered positive, and the direction opposite to the arrow is considered negative. The positive Z direction is considered to be the vertically upward direction. The bead layers are stacked in the positive Z direction.

[0035] The molding unit 20 includes a wire nozzle 22 for supplying a wire 21 which is the raw material wire, a laser oscillator 23 for outputting a laser beam L, a beam nozzle 24 through which the laser beam L that is emitted toward the processing point passes, a fiber cable 25 which is an optical transmission path, a processing head 26, a gas nozzle 27 for injecting shielding gas toward the processing point, a gas supply device 28, and a head drive device 33. The beam nozzle 24 and the gas nozzle 27 are attached to the processing head 26.

[0036] Furthermore, the molding unit 20 includes a stage 30 on which the base material is placed, and a rotation mechanism 31 for rotating the stage 30. The molding unit 20 creates a metal additive manufacturing object by layering beads on the base material. In addition, the molding unit 20 has a control device 35 which is an NC device.

[0037] The laser beam L output from the laser oscillator 23 propagates through the fiber cable 25 and enters the processing head 26. Inside the processing head 26, an optical system such as a collimating optical system or a focusing optical system is arranged. The diagram of the optical system is omitted. The central axis of the beam nozzle 24 attached to the processing head 26 coincides with the optical axis of the optical system. The central axis of the beam nozzle 24 is parallel to the Z-axis. The center line of the laser beam L emitted from the beam nozzle 24 is parallel to the Z-axis. The laser beam L passes through the optical system inside the processing head 26, passes through the beam nozzle 24, and is emitted from the processing head 26. The processing point is the irradiation position of the laser beam L and the position where the molten material is added. The position of the processing point is the position where the heat source and material are supplied, and is on the central axis of the beam nozzle 24.

[0038] The gas supply device 28 supplies shielding gas from a gas supply source to the gas nozzle 27. An example of a gas supply source is a gas cylinder. The gas supply source is not shown in the diagram. The gas supply device 28 supplies shielding gas to the gas nozzle 27 through piping 29. The molding unit 20 reduces oxidation of the molded object and cools the object by injecting shielding gas toward the processing point. The shielding gas is, in one example, an inert gas such as argon gas.

[0039] The wire nozzle 22 supplies the wire 21, which is fed out by the material supply mechanism 34, to the processing point. Figure 3 shows an example of a side supply method in which the wire 21 is supplied from a wire nozzle 22 positioned diagonally above the processing point. The molding unit 20 may also employ a center supply method in which the wire 21 is supplied from a wire nozzle 22 positioned directly above the processing point, rather than the side supply method. The wire nozzle 22 and the material supply mechanism 34 correspond to the raw material wire supply unit.

[0040] The head drive unit 33 moves the processing head 26. In one example, the head drive unit 33 includes a servo motor that moves the processing head 26 in the X-axis direction, a servo motor that moves the processing head 26 in the Y-axis direction, and a servo motor that moves the processing head 26 in the Z-axis direction. The illustration of each servo motor is omitted. The positional relationship between the processing head 26 and the wire nozzle 22 is fixed. The molding unit 20 moves the processing head 26 and the wire nozzle 22 within the stroke range by driving the head drive unit 33. By moving the processing head 26 and the wire nozzle 22, the molding unit 20 can move the processing point to any position within the stroke range. The laser oscillator 23, fiber cable 25, processing head 26, beam nozzle 24, and head drive unit 33 correspond to the laser beam irradiation unit.

[0041] The rotation mechanism 31 is an operating mechanism that enables the stage 30 to rotate around a first axis and around a second axis perpendicular to the first axis. In the rotation mechanism 31 shown in Figure 3, the first axis is parallel to the X axis, and the second axis is parallel to the Z axis. The rotation mechanism 31 includes a servo motor that rotates the stage 30 around the first axis and a servo motor that rotates the stage 30 around the second axis. The rotation mechanism 31 causes the stage 30 to perform rotational motion around each of the two axes by driving each servo motor. The individual servo motors are not shown.

[0042] The molding unit 20 rotates the stage 30 using the rotation mechanism 31, thereby changing the orientation of the base material placed on the stage 30. The molding unit 20 can change the orientation of the base material to an orientation suitable for processing.

[0043] The molding unit 20 moves the machining point by moving the machining head 26 relative to the stage 30. The molding unit 20 may also move the machining point by moving the stage 30 relative to the machining head 26. In this case, the molding unit 20 can move the stage 30 relative to the machining head 26 by configuring it so that at least one of the three axes is moved to the stage 30.

[0044] In Embodiment 1, the shaping unit 20 is a five-axis machining center capable of translational motion in the axial directions of the X-axis, Y-axis, and Z-axis and rotational motion about two axes. Note that the shaping unit 20 may be a three-axis machining center capable of translational motion in the axial directions of the X-axis, Y-axis, and Z-axis, or a four-axis machining center capable of translational motion in the axial directions of the X-axis, Y-axis, and Z-axis and rotational motion about one axis.

[0045] The control device 35 controls the entire shaping unit 20 according to the NC program. The NC program includes shaping condition information. As described above, the shaping condition information includes information necessary for shaping the shaped object, such as the laser output, which is the output of the laser beam L by the laser oscillator 23, the axis speed, which is the moving speed of the machining head 26, the wire supply speed, which is the amount of wire 21 supplied by the wire nozzle 22, the shield gas flow rate, which is the flow rate of the shield gas, and the rotation speed of the stage 30.

[0046] The control device 35 controls the head drive device 33 by outputting a position command to the head drive device 33. The head drive device 33 moves the machining head 26 along the preset shaping path information according to the position command. The control device 35 controls the rotation mechanism 31 by outputting a rotation command to the rotation mechanism 31. The rotation mechanism 31 rotates the stage 30 according to the rotation command.

[0047] The control device 35 controls the laser oscillator 23 by outputting a laser output command to the laser oscillator 23. The laser oscillator 23 outputs the laser beam L according to the laser output command. The control device 35 controls the material supply mechanism 34 by outputting a material supply command to the material supply mechanism 34. The material supply mechanism 34 supplies the wire 21 at a wire supply speed according to the material supply command. The control device 35 controls the gas supply device 28 by outputting a gas supply command to the gas supply device 28. The gas supply device 28 supplies the shield gas with a gas flow rate according to the gas supply command.

[0048] The shape condition information generated by the shape condition generation unit 13 shown in FIG. 2 includes various conditions for controlling each part of the shaping unit 20 by the control device 35. The various conditions included in the shape condition information are laser output, irradiation time of the laser beam L, laser diameter, timings of on and off of the laser beam L, wire supply speed, shield gas flow rate, axis speed, rotation speed of the stage 30, standby time between each shaping layer, and the like.

[0049] Next, the details of the configuration of the evaporation amount prediction unit 15 when predicting the evaporation amount of the low-boiling point metal by machine learning will be described. FIG. 4 is a block diagram showing an example of the configuration of the evaporation amount prediction unit. The evaporation amount prediction unit 15 includes a learning unit 151, a learned model storage unit 152, and an inference unit 153.

[0050] The learning unit 151 uses the performance data stored in the performance data storage unit 14 to learn a learned model with the wire surface state information and the shape condition information as explanatory variables and the actual value of the evaporation amount of the low-boiling point metal at the start of shaping as the objective variable. The learned model storage unit 152 stores the learned model learned by the learning unit 151. The inference unit 153 estimates the predicted value of the evaporation amount of the low-boiling point metal at the start of shaping by inputting the wire surface state information from the raw material wire measurement unit 11 and the shape condition information from the shape condition generation unit 13 into the learned model.

[0051] The learning unit 151 includes a data acquisition unit 1511 and a model generation unit 1512.

[0052] The data acquisition unit 1511 acquires wire surface condition information, molding condition information, and actual values ​​of the evaporation amount of low-boiling-point metal at the start of molding as training data. Here, the wire surface condition information includes at least one of the following: the surface oxygen content of the wire 21, the laser absorptivity, and heat input information which is a combination of the boiling point of the low-boiling-point metal and the amount of heat input per unit volume to the wire 21. The heat input information is information which is a combination of the boiling point obtained from the type of low-boiling-point metal contained in the composition of the wire 21 when a metal additive manufacturing object was manufactured in the past, and the amount of heat input calculated from the molding conditions. The molding condition information is the value of the conditions set in the metal additive manufacturing apparatus 10 when the manufacturing layer was manufactured. The molding condition information includes at least one of the laser output value, axis speed, and wire supply speed. The actual value of the evaporation amount of low-boiling-point metal at the start of the molding process is calculated from the amount of low-boiling-point metal obtained by compositional analysis of the molded layer formed at the start of the molding process using a combination of wire surface condition information and molding condition information, and the amount of low-boiling-point metal in the wire 21 of the molding section 20.

[0053] The model generation unit 1512 learns a predicted value for the evaporation amount of low-boiling-point metal at the start of fabrication based on training data created based on a combination of wire surface state information and fabrication condition information output from the data acquisition unit 1511 and the actual value of the evaporation amount of low-boiling-point metal at the start of fabrication. That is, it generates a trained model that infers the optimal predicted value for the evaporation amount of low-boiling-point metal at the start of fabrication from the wire surface state information and fabrication condition information at the start of fabrication of the metal additive fabricated object and the actual value of the evaporation amount of low-boiling-point metal at the start of fabrication. Here, the training data is data that correlates the wire surface state information and fabrication condition information and the actual value of the evaporation amount of low-boiling-point metal at the start of fabrication. Embodiment 1 is characterized by including at least one of the wire surface state information that greatly affects the evaporation of low-boiling-point metal, namely the surface oxygen content of the wire 21 and the laser absorption rate, in the training data.

[0054] In this example, the learning unit 151, the trained model storage unit 152, and the inference unit 153 are built into the metal additive manufacturing apparatus 10 and used to learn a predicted value for the evaporation amount of low-boiling-point metal at the start of manufacturing in the metal additive manufacturing apparatus 10. However, Embodiment 1 is not limited to this example. In one example, the learning unit 151, the trained model storage unit 152, and the inference unit 153 may be connected to the metal additive manufacturing apparatus 10 via a network and may be separate devices from the metal additive manufacturing apparatus 10. Also, some parts of the learning unit 151, the trained model storage unit 152, and the inference unit 153 may be built into the metal additive manufacturing apparatus 10, while the remaining parts are connected to the metal additive manufacturing apparatus 10 via a network and may be separate devices from the metal additive manufacturing apparatus 10. Furthermore, the learning unit 151, the trained model storage unit 152, and the inference unit 153 may reside on a cloud server.

[0055] The model generation unit 1512 can use any known learning algorithm, such as supervised learning, unsupervised learning, or reinforcement learning. As an example, the case where a neural network is applied will be described.

[0056] In one example, the model generation unit 1512 learns a predicted value for the evaporation rate of the low-boiling-point metal at the start of the molding process by supervised learning, following a neural network model. Here, supervised learning is a method in which a pair of data consisting of inputs and labeled results is provided to the learning unit 151, which learns features in this learning data and infers the results from the inputs.

[0057] A neural network consists of an input layer made up of multiple neurons, a hidden layer (intermediate layer) also made up of multiple neurons, and an output layer also made up of multiple neurons. The intermediate layer can be one layer or two or more layers.

[0058] Figure 5 schematically shows an example of a neural network used by the model generation unit. In a three-layer neural network like the one exemplified in Figure 5, when multiple inputs are input from input layer X1 to input layer X3, these values ​​are multiplied by weights w11 to w16 and input to hidden layer Y1 to hidden layer Y2. When weights w11 to w16 are not individually distinguished, they are referred to as weight w1. Furthermore, the results from hidden layer Y1 to hidden layer Y2 are multiplied by weights w21 to w26 and output from output layer Z1 to output layer Z3. When weights w21 to w26 are not individually distinguished, they are referred to as weight w2. The output results from output layer Z1 to output layer Z3 vary depending on the values ​​of weights w1 and w2.

[0059] In Embodiment 1, the neural network learns a predicted value of the amount of evaporation of the low-boiling-point metal at the start of fabrication by so-called supervised learning, based on training data created from a combination of wire surface state information and fabrication condition information acquired by the data acquisition unit 1511 and the actual value of the amount of evaporation of the low-boiling-point metal at the start of fabrication.

[0060] In other words, the neural network learns by inputting wire surface state information and fabrication condition information into the input layer and adjusting weights w1 and w2 so that the result output from the output layer approaches the actual value of the evaporation amount of low-boiling-point metal at the start of fabrication.

[0061] The model generation unit 1512 generates and outputs a trained model by performing the learning described above.

[0062] Returning to Figure 4, the trained model storage unit 152 stores the trained model output from the model generation unit 1512. The trained model is a model in which wire surface state information and molding condition information are used as explanatory variables, and the actual value of the evaporation amount of low-boiling point metal at the start of molding is used as the objective variable.

[0063] The inference unit 153 comprises a data acquisition unit 1531 and a prediction unit 1532.

[0064] The data acquisition unit 1531 acquires wire surface condition information and molding condition information. The wire surface condition information is acquired from the raw wire measurement unit 11. The molding condition information is either the molding condition information input from the molding condition generation unit 13, or the molding condition information generated by the molding condition generation unit 13 that has been adjusted by the user.

[0065] The prediction unit 1532 infers a predicted value for the evaporation amount of low-boiling-point metal at the start of fabrication, obtained using a trained model. That is, by inputting wire surface condition information and fabrication condition information acquired by the data acquisition unit 1531 into this trained model, it is possible to output a predicted value for the evaporation amount of low-boiling-point metal at the start of fabrication, inferred from the wire surface condition information and fabrication condition information.

[0066] In Embodiment 1, it was explained that a pre-trained model learned by the model generation unit 1512 of the metal additive manufacturing apparatus 10 is used to output a predicted value for the evaporation amount of the low-boiling-point metal at the start of manufacturing. However, a pre-trained model may be obtained from an external source such as another metal additive manufacturing apparatus 10, and a predicted value for the evaporation amount of the low-boiling-point metal at the start of manufacturing may be output based on this pre-trained model.

[0067] In Embodiment 1, the case in which supervised learning is applied to the learning algorithm used by the model generation unit 1512 was described, but this is not the only case, and any applicable learning algorithm can be used.

[0068] Furthermore, the model generation unit 1512 may learn a predicted value for the evaporation amount of low-boiling-point metal at the start of manufacturing, based on the training data created for multiple metal additive manufacturing devices 10. In this case, the model generation unit 1512 may acquire training data from multiple metal additive manufacturing devices 10 used in the same area, or it may learn a predicted value for the evaporation amount of low-boiling-point metal at the start of manufacturing by utilizing training data collected from multiple metal additive manufacturing devices 10 operating independently in different areas. It is also possible to add or remove metal additive manufacturing devices 10 that collect training data from the target during the process. Moreover, the learning unit 1512 that has learned a predicted value for the evaporation amount of low-boiling-point metal at the start of manufacturing for one metal additive manufacturing device 10 may be applied to another metal additive manufacturing device 10, and the predicted value for the evaporation amount of low-boiling-point metal at the start of manufacturing for that other metal additive manufacturing device 10 may be relearned and updated.

[0069] Furthermore, the learning algorithm used in the model generation unit 1512 can be deep learning, which learns to extract the features themselves, or machine learning can be performed according to other known methods, such as genetic programming, functional logic programming, or support vector machines.

[0070] As described above, the molding condition information generated by the molding condition generation unit 13, or the adjusted molding condition information, is for the time when molding starts in the molding unit 20, that is, when the first molding layer is formed. In Embodiment 1, it is assumed that the evaporation rate of the low-boiling-point metal is the same from the start of molding to the end of molding. Therefore, in Embodiment 1, the molding conditions generated at the start of molding are used for molding the layers from the first layer to the final layer.

[0071] Next, the method for manufacturing a metal additive manufactured object using the metal additive manufacturing apparatus 10 will be described. The method for manufacturing a metal additive manufactured object according to Embodiment 1 is difficult to execute without a trained model. Therefore, until a certain amount of actual data is accumulated, the molding conditions are set according to experience or through trial and error, and the method for manufacturing a metal additive manufactured object is executed according to these molding conditions. This is the same as in the conventional method, so the explanation will be omitted. However, in this case, each time a metal additive manufactured object is produced, the combination of wire surface state information, molding condition information, and the actual value of the evaporation amount of low-boiling point metal at the start of molding is accumulated as actual data in the actual data storage unit 14. Then, a trained model is trained using the accumulated actual data, and the method for manufacturing a metal additive manufactured object according to Embodiment 1 is performed using this trained model. Here, the method for generating the trained model will be explained first, followed by the method for manufacturing a metal additive manufactured object.

[0072] Figure 6 is a flowchart showing an example of the procedure for generating a trained model. This process is performed by the learning unit 151 of the evaporation rate prediction unit 15. The data acquisition unit 1511 acquires wire surface condition information, molding condition information, and the actual value of the evaporation rate of the low-boiling-point metal at the start of molding (step S11). Although the wire surface condition information, molding condition information, and the actual value of the evaporation rate of the low-boiling-point metal at the start of molding are acquired simultaneously, it is sufficient if the wire surface condition information, molding condition information, and the actual value of the evaporation rate of the low-boiling-point metal at the start of molding are input in association. For this reason, the data for the wire surface condition information and molding condition information and the data for the actual value of the evaporation rate of the low-boiling-point metal at the start of molding may be acquired at different times. The data acquisition unit 1511 acquires the wire surface condition information, molding condition information, and the actual value of the evaporation rate of the low-boiling-point metal at the start of molding from the actual data storage unit 14.

[0073] Next, the model generation unit 1512 learns a predicted value for the evaporation amount of the low-boiling-point metal at the start of fabrication by so-called supervised learning, based on training data created from a combination of wire surface state information and fabrication condition information acquired by the data acquisition unit 1511 and the actual value of the evaporation amount of the low-boiling-point metal at the start of fabrication, and generates a trained model (step S12).

[0074] Subsequently, the model generation unit 1512 saves the generated trained model to the trained model storage unit 152 (step 13). This completes the method for generating a trained model.

[0075] Figures 7 and 8 are flowcharts illustrating an example of the procedure for manufacturing a metal additively fabricated object according to Embodiment 1. First, the raw wire measurement unit 11 measures wire surface condition information indicating the surface condition of the wire 21 supplied from the wire nozzle 22 (step S31). In this example, at least one of the surface oxygen content and laser absorptivity of the wire 21 is measured as wire surface condition information. When measuring the surface oxygen content of the wire 21, an example of the raw wire measurement unit 11 is an EDS system attached to a SEM, an X-ray fluorescence analyzer, or a handheld simple elemental analyzer. When measuring the laser absorptivity of the wire 21, an example of the raw wire measurement unit 11 is a spectrophotometer. Step S31 corresponds to the wire measurement process.

[0076] Next, the object information input unit 12 receives object information (step S32). The object information includes information indicating the material of the object and information indicating the shape of the object. This sets the material of the metal additive manufacturing object and the shape of the metal additive manufacturing object in the metal additive manufacturing apparatus 10. Step S32 corresponds to the receiving process.

[0077] Subsequently, the molding condition generation unit 13 generates molding condition information for molding a metal additive manufactured object based on the received molded object information (step S33). The molding condition generation unit 13 generates molding environment condition information, which is various conditions for molding a metal additive manufactured object of the material and shape indicated in the molded object information. The molding condition generation unit 13 also generates molding path information for the path to mold the metal additive manufactured object of the material and shape indicated in the molded object information. The molding path information is a path calculated by CAM. The molding condition generation unit 13 outputs the molding conditions, including the generated molding environment condition information and molding path information, to the evaporation amount prediction unit 15. Step S33 corresponds to the molding condition generation process.

[0078] The evaporation rate prediction unit 15 performs a low-boiling-point metal evaporation rate prediction process (step 34) to predict the amount of evaporation of the low-boiling-point metal when molding is started under the specified molding conditions, based on the wire surface condition information of the raw wire measured in step S31 and the molding condition information generated in step S33. In performing the prediction, the evaporation rate prediction unit 15 refers to past performance data stored in the performance data storage unit 14. That is, it predicts the amount of evaporation of the low-boiling-point metal at the start of molding based on the wire surface condition information, molding condition information, and the corresponding low-boiling-point metal evaporation values ​​from past molding operations. If heat input information is used as the wire surface condition information, the evaporation rate prediction unit 15 may obtain the boiling point of the low-boiling-point metal from the composition of the wire 21 used based on the molding conditions, and also calculate the amount of heat input per unit volume to the wire 21 from the molding condition information to generate heat input information. Here, we provide an example of using machine learning to predict the evaporation rate of a low-boiling-point metal at the start of fabrication, and therefore, we will further explain the details of the process in step S34. Step S34 corresponds to the first evaporation rate prediction process.

[0079] Figure 9 is a flowchart showing an example of the procedure for predicting the evaporation rate of a low-boiling-point metal. This process is performed in the inference unit 153 of the evaporation rate prediction unit 15. First, the data acquisition unit 1531 acquires wire surface condition information and molding condition information (step S71). The wire surface condition information is the information measured by the raw material wire measurement unit 11 in step S31. The molding condition information is the information generated by the molding condition generation unit 13 in step S33.

[0080] Next, the prediction unit 1532 inputs wire surface state information and molding condition information into the learned model stored in the learned model storage unit 152 to obtain a predicted value for the evaporation amount of the low-boiling point metal at the start of molding (step S72). The learned model corresponds to the learned model for the start of molding.

[0081] Subsequently, the prediction unit 1532 outputs to the determination processing unit 16 a predicted value of the evaporation amount of the low-boiling-point metal at the start of molding, obtained from the trained model (step S73). With this, the low-boiling-point metal evaporation amount prediction process is completed, and the process returns to Figure 7.

[0082] Subsequently, the determination processing unit 16 determines whether the predicted evaporation amount of the low-boiling-point metal at the start of molding, obtained from the evaporation amount prediction unit 15, is within the acceptable range (step S35). Based on the received molding condition information, the determination processing unit 16 determines whether the desired physical properties can be obtained when molding is performed using the said molding condition information. An example of an indicator of desired physical properties is whether a metal additive manufactured product with a composition conforming to the standard can be obtained, or whether the composition of the low-boiling-point metal in the metal additive manufactured product has decreased by more than a predetermined value compared to the raw wire. The acceptable range for determining whether the desired physical properties are present can be arbitrarily set by the user. Step S35 corresponds to the determination process.

[0083] If the predicted evaporation amount of the low-boiling-point metal at the start of molding is not within the acceptable range (No in step S35), the determination processing unit 16 outputs to the evaporation amount prediction unit 15 that the predicted evaporation amount of the low-boiling-point metal at the start of molding is not within the acceptable range. In this case, the evaporation amount prediction unit 15 adjusts the molding condition information (step S36). In one example, the evaporation amount prediction unit 15 generates an adjustment screen on which the molding condition information can be changed, displays it on a display unit (not shown), and adjusts the molding condition information according to the changes in the molding condition information entered by the user on the adjustment screen. Step S36 corresponds to the first molding condition adjustment process. After that, the process returns to step S34. In step S34, the predicted evaporation amount of the low-boiling-point metal is predicted with the adjusted molding conditions. The process from step S34 to step S36 is repeatedly executed until the predicted evaporation amount of the low-boiling-point metal falls within a predetermined acceptable range and it is determined that molding is possible based on the molding condition information.

[0084] If the predicted value of the evaporation amount of the low-boiling-point metal at the start of molding is within the acceptable range (Yes in step S35), the determination processing unit 16 outputs to the evaporation amount prediction unit 15 that the predicted value of the evaporation amount of the low-boiling-point metal at the start of molding is within the acceptable range. In this case, the evaporation amount prediction unit 15 outputs molding condition information, which indicates that the predicted value of the evaporation amount of the low-boiling-point metal at the start of molding is within the acceptable range, to the molding condition generation unit 13 (step S37). If the molding condition information generated by the molding condition generation unit 13 is determined to be within the acceptable range, the evaporation amount prediction unit 15 outputs the molding condition information to the molding condition generation unit 13. If the molding condition information adjusted by the evaporation amount prediction unit 15 is determined to be within the acceptable range, the evaporation amount prediction unit 15 outputs the adjusted molding condition information to the molding condition generation unit 13.

[0085] The molding condition generation unit 13 uses the acquired molding condition information to generate an NC program that controls each part of the molding unit 20 for molding a metal additive manufactured object along the molding path indicated by the molding path information, and outputs the NC program to the molding unit 20 (step S38). The molding unit 20 molds the metal additive manufactured object according to the acquired NC program (step S39). Step S39 corresponds to the molding process. In this embodiment 1, the molded layers from the start of molding to the end of molding are molded according to the molding condition information generated in step S33 or the molding condition information adjusted in step S36. This completes the method for manufacturing a metal additive manufactured object.

[0086] Here, as an example, we will describe the manufacturing of a metal additively manufactured object using wire 21 made of A7075, an aluminum alloy that contains a large amount of Zn as a low-boiling-point metal. The composition of A7075 shall conform to the Japanese Industrial Standards (JIS). We will use wire 21 made of A7075 (unadditiveed material) conforming to the JIS standard, and wire 21 made of A7075 (additiveed material) to which Sc and Zr have been added in a range of 0.5% or less to reduce susceptibility to welding cracking. We will prepare wire 21 with different drawing conditions and manufacture a metal additively manufactured object. The wire 21 used here is shown in Figure 1. The drawing conditions correspond to the number of annealing cycles.

[0087] First, as explained in step S31, the raw wire measurement unit 11 measures the surface oxygen content and laser absorptivity, which are surface condition information of the raw wire. As shown in Figure 1, as the number of annealing cycles increases, the surface oxygen content tends to increase and the laser absorptivity of the wire 21 surface tends to increase. This trend is the same for both A7075 (additive-free material) and A7075 (additive-containing material), that is, regardless of the presence or absence of additives. However, A7075 (additive-containing material) breaks under conditions of a small number of annealing cycles, making it impossible to produce the desired wire 21. For this reason, the manufacture of metal additively fabricated objects using A7075 (additive-containing material) wire 21 with a small number of annealing cycles is not performed.

[0088] Next, as explained in step S32, the fabricated object information input unit 12 inputs information regarding the shape of the metal additive fabricated object. Also, as explained in step S33, the fabrication condition generation unit 13 generates fabrication condition information. Here, we assume that one layer is fabricated under constant conditions using all types of wire 21. The specific fabrication conditions are a laser output of 2000W, an axis speed of 900mm / min, and a wire supply speed of 50cc / h.

[0089] In the subsequent step S34, the evaporation rate prediction unit 15 predicts the amount of Zn evaporation at the start of molding based on the molding condition information generated in step S33. To make the prediction, (1) wire 21 information, (2) molding condition information, and (3) past molding results, which include Zn evaporation results linked to the information in (1) and (2), are used. The wire 21 information in (1) here further includes at least one of the surface oxygen content and laser absorptivity measured by the raw material wire measurement unit 11, as well as the composition of the wire 21 and the boiling point of the low-boiling-point metal. The molding condition information in (2) may include laser output, axis speed, and wire supply speed, but here it further includes laser diameter, shielding gas amount, molding path information, etc. The information in (1) and (2) is not limited to these. The prediction accuracy can be improved by including information such as the amount of heat input per unit volume to the wire 21 in (1) and information such as the laser diameter, shielding gas amount, and build path information in (2). Furthermore, prediction is not possible when there is no build history, and it is necessary to accumulate data from past builds, but the prediction accuracy can be improved by accumulating the build history data in the build history data storage unit 14 each time a build is performed.

[0090] The amount of Zn evaporation focused on here is calculated from the difference between the amount of Zn in the wire 21 and the amount of Zn in the metal additive manufacturing product. In calculating the actual evaporation amount stored in the performance data storage unit 14, compositional analysis is performed on the wire 21 and the metal additive manufacturing product. For compositional analysis, methods such as ICP emission spectrometry and SEM-EDS can be used.

[0091] As predicted in step S34, as shown in Figure 1, when fabrication is performed using wire 21 with a high number of annealing cycles, a high surface oxygen content, and a high laser absorption rate under the above fabrication conditions, the amount of Zn evaporation at the start of fabrication tends to increase. The Zn concentration of wire 21 for A7075 is specified to be between 5.1 wt% and 6.1 wt%, and here we assume the Zn concentration is 6 wt% based on the compositional analysis results of wire 21. Therefore, when fabrication is performed using wire 21 with a high or medium number of annealing cycles, the amount of evaporation exceeds 10% of the Zn concentration of wire 21. In other words, when fabrication is performed using wire 21 with a high or medium number of annealing cycles, the amount of Zn evaporation at the start of fabrication exceeds 0.6 wt%.

[0092] Subsequently, as explained in step S35, the determination processing unit 16 determines whether the predicted amount of Zn evaporation at the start of molding relative to the Zn concentration of the wire 21 is within the acceptable range. The acceptable range at this time is set to 10%. In the case of Figure 1, the determination processing unit 16 determines that the molding conditions need to be reviewed for those with "many" and "medium" annealing cycles. In this example, the acceptable range, which is the standard for the amount of Zn evaporation relative to the Zn concentration of the wire 21, is set to 10%, but the user can decide this arbitrarily.

[0093] In step S36, the evaporation rate prediction unit 15 adjusts the printing condition information for wires where the amount of Zn evaporation at the start of printing exceeds the allowable range. The adjustment of the printing condition information is based on the user's experience. Here, for wires 21 where the allowable range of Zn evaporation relative to the Zn concentration of the wire 21 exceeds 10%, the laser output is changed to 1600W.

[0094] For wire 21 where the amount of Zn evaporation at the start of printing exceeds the acceptable range, step S34 predicts the amount of Zn evaporation at the start of printing again when the laser output is reduced. In the subsequent step S35, the evaporation amount prediction unit 15 predicts that the predicted value of the amount of Zn evaporation at the start of printing is within the acceptable range, i.e., less than 10%. Here, we have given an example where the adjusted printing condition information is within the acceptable range, but if it exceeds the acceptable range, the process from step S34 to step S36 may be repeated many times until printing condition information is obtained in which the predicted value of the amount of Zn evaporation at the start of printing does not exceed the acceptable range.

[0095] Subsequently, as described in steps S37 to S38, an NC program is generated from the molding condition information where the predicted amount of Zn evaporation is within an acceptable range. Then, in step S39, molding is started according to the NC program, that is, according to the molding condition information where the amount of Zn evaporation is within an acceptable range. The obtained molding conditions are used not only for the first layer but also for the molding of the last layer. A metal additively manufactured object with the shape specified in the molded object information is then obtained.

[0096] By performing the above process, it is possible to obtain a metal additively manufactured product with less compositional unevenness of low-boiling-point metals in the wire 21 compared to conventional methods. In other words, the metal additively manufactured product obtained in this way has less evaporation of low-boiling-point metals from the wire 21. It also has the effect of enabling the fabrication of metal additively manufactured products using wire 21 to which Sc or Zr is added, which is difficult to balance with surface oxygen content and wire drawing processes.

[0097] Furthermore, in Embodiment 1, the amount of evaporation of low-boiling-point metal can be predicted according to the surface condition of the wire 21, and this can be fed back into the molding conditions.

[0098] Furthermore, it is desirable that the surface oxygen content of the wire 21 is 5 wt% or less, and that the difference in low-boiling-point metal composition between the metal additive manufactured product and the wire 21 in the above process is 10% or less. By setting the surface oxygen content of the wire 21 to 5 wt% or less, the amount of evaporation of low-boiling-point metal at the start of manufacturing falls within the acceptable range, and the metal additive manufactured product can be manufactured without revising the manufacturing conditions. However, even if the surface oxygen content of the wire 21 exceeds 5 wt%, the manufacturing conditions can be adjusted so that the amount of evaporation of low-boiling-point metal falls within the acceptable range. Here, the difference in low-boiling-point metal composition is expressed as a percentage of the difference between the concentration of low-boiling-point metal in the metal additive manufactured product and the concentration of low-boiling-point metal in the wire 21, relative to the concentration of low-boiling-point metal in the wire 21.

[0099] In the above description, the actual data in the actual data storage unit 14 is assumed to include wire surface condition information, molding condition information, and actual values ​​of the evaporation amount of low-boiling-point metal at the start of molding, but other information may also be included. For example, the actual data may include information on the molded object or the composition of the wire 21. Also, in the above description, an example was shown in which the molding environment condition information includes the laser output value, axis speed, and wire supply speed, but other information may also be included. For example, the molding environment condition information may further include at least one piece of information selected from the group of laser beam L irradiation time, laser diameter, on and off timings of the laser beam L, shielding gas flow rate, and stage 30 rotation speed.

[0100] The first embodiment describes a method for manufacturing a metal additively manufactured object, which involves sequentially stacking layers formed by irradiating a raw material wire made of an Al alloy containing a low-boiling-point metal, a metal with a lower boiling point than Al, with a laser beam L. The method includes a wire measurement step, a receiving step, a molding condition generation step, and a first evaporation amount prediction step. The wire measurement step measures wire surface state information, which is information related to the evaporation amount of the low-boiling-point metal in the raw material wire. The receiving step receives molded object information, including the shape and material of the metal additively manufactured object. The molding condition generation step generates molding condition information, which is the condition for manufacturing the metal additively manufactured object, based on the molded object information. The first evaporation amount prediction step predicts the evaporation amount of the low-boiling-point metal when the metal additively manufactured object shown in the molded object information is manufactured according to the molding condition information, based on the wire surface state information and the molding condition information. This has the effect of reducing the amount of evaporation of metals with lower boiling points than Al in the additively manufactured metal product when using an Al alloy containing metals with lower boiling points than Al as the raw material wire and irradiating the raw material wire with laser light.

[0101] Low-boiling-point metals such as Mg and Zn, which are contained in Al alloys, are added to increase the strength of the Al alloy. For example, in wrought alloys, Mg is contained in the A5000 and A6000 series, and Mg and Zn are contained in the A7000 series. However, the low-boiling-point metals that are the wire raw material evaporate during the molding process. The inventors have found that the amount of sputter or fume generated changes depending on the surface oxygen content of the wire 21. Based on this finding, by measuring the surface oxygen content of the wire 21 as wire surface condition information, it is possible to predict the amount of evaporation of low-boiling-point metals when molded under the specified molding conditions.

[0102] Furthermore, the inventors have also found that the laser absorptivity changes depending on the surface oxygen content of the wire 21. Based on this finding, the amount of evaporation of low-boiling-point metal when fabricated under the fabrication conditions can be predicted by measuring the laser absorptivity as wire surface state information. Note that the laser absorptivity and surface oxygen content of the wire 21 are correlated. In addition, information related to the laser absorptivity of the wire 21 includes heat input information, which is a combination of the boiling point of the low-boiling-point metal and the amount of heat input per unit volume to the wire 21. Therefore, the wire surface state information only needs to include at least one of the surface oxygen content of the wire 21, laser absorptivity, and heat input information. This makes it possible to suppress the amount of evaporation of low-boiling-point metal in the fabricated metal additive manufacturing product compared to conventional methods.

[0103] Furthermore, the manufacturing method for a metal additive manufactured product according to Embodiment 1 further includes a determination step and a manufacturing step. The determination step determines whether or not manufacturing is possible using the manufacturing condition information based on whether or not the predicted value of the evaporation amount of the low-boiling-point metal is within a set allowable range. The manufacturing step manufactures the product according to the manufacturing condition information if the predicted value of the evaporation amount of the low-boiling-point metal is within the allowable range and it is determined that manufacturing is possible using the manufacturing condition information. This makes it possible to manufacture the product under manufacturing conditions in which the predicted value of the evaporation amount of the low-boiling-point metal during manufacturing is within the allowable range.

[0104] Furthermore, the manufacturing method for a metal additive manufactured product according to Embodiment 1 further includes a first molding condition adjustment step in which the molding condition information is adjusted when the predicted value of the evaporation amount of low-boiling-point metal is not within the acceptable range and it is determined that molding is not possible based on the molding condition information. In this case, the first evaporation amount prediction step predicts the predicted value of the evaporation amount of low-boiling-point metal using the adjusted molding condition information. The first molding condition adjustment step, the first evaporation amount prediction step, and the determination step are repeatedly performed until the determination step determines that the predicted value of the evaporation amount of low-boiling-point metal is within the acceptable range and that molding is possible based on the molding condition information. As a result, molding is not performed under molding conditions where the predicted value of the evaporation amount of low-boiling-point metal during molding is outside the acceptable range, and the amount of evaporation of low-boiling-point metal in the resulting metal additive manufactured product can be suppressed compared to conventional methods.

[0105] Furthermore, in the manufacturing process of the metal additive manufacturing method according to Embodiment 1, the manufacturing layer from the start of manufacturing to the end of manufacturing is manufactured according to the manufacturing condition information generated in the manufacturing condition generation step or the manufacturing condition information adjusted in the first manufacturing condition adjustment step. This makes it possible to manufacture under manufacturing conditions in which the predicted value of the evaporation amount of low-boiling-point metal during manufacturing is within an acceptable range.

[0106] Furthermore, in the first evaporation rate prediction step in the metal additive manufacturing method according to Embodiment 1, a pre-trained model for the start of manufacturing, which infers a predicted value of the evaporation rate of low-boiling-point metal at the start of manufacturing from wire surface condition information and manufacturing condition information, is used to output a predicted value of the evaporation rate of low-boiling-point metal at the start of manufacturing from the wire surface condition information obtained in the wire measurement step and the manufacturing condition information generated in the manufacturing condition generation step. A pre-trained model for the start of manufacturing is generated using a set of wire surface condition information and manufacturing condition information related to the evaporation rate of low-boiling-point metal, and the actual value of the evaporation rate of low-boiling-point metal, as training data. By inputting the wire surface condition information and manufacturing condition information into this pre-trained model for the start of manufacturing to predict the evaporation rate of low-boiling-point metal, a more accurate prediction of the evaporation rate of low-boiling-point metal can be obtained compared to when prediction is made without using wire surface condition information.

[0107] The metal additive manufacturing apparatus 10 according to Embodiment 1 comprises a raw material wire supply unit, a laser beam irradiation unit, a wire measurement unit, a manufactured object information input unit 12, a manufacturing condition generation unit 13, an evaporation amount prediction unit 15, a determination processing unit 16, and a manufacturing unit 20. The raw material wire supply unit supplies a raw material wire composed of an Al alloy containing a low-boiling-point metal, which is a metal with a lower boiling point than Al. The laser beam irradiation unit irradiates the raw material wire with a laser beam L. The wire measurement unit measures wire surface state information, which is information related to the evaporation amount of the low-boiling-point metal in the raw material wire. The manufactured object information input unit 12 receives input of manufactured object information, including the shape and material of the metal additive manufacturing object to be manufactured. The manufacturing condition generation unit 13 generates manufacturing condition information, which is the condition for manufacturing the metal additive manufacturing object, based on the manufactured object information. The evaporation rate prediction unit 15 predicts the amount of evaporation of low-boiling-point metal when a metal additive manufactured object, as indicated in the manufactured object information, is manufactured according to the manufacturing condition information, based on the wire surface condition information and the manufacturing condition information. The determination processing unit 16 determines whether or not manufacturing is possible according to the manufacturing condition information based on whether or not the predicted amount of evaporation of low-boiling-point metal is within the set tolerance range. The manufacturing unit 20 irradiates the raw material wire supplied from the raw material wire supply unit with a laser beam L from the laser beam irradiation unit according to the manufacturing condition information to manufacture the manufactured layer. The manufacturing condition generation unit 13 outputs the manufacturing condition information that has been determined to be makerable by the determination processing unit 16 to the manufacturing unit 20. This has the effect of being able to manufacture a metal additive manufactured object with a reduced amount of evaporation of metal with a lower boiling point than Al compared to conventional methods, when using an Al alloy containing a metal with a lower boiling point than Al as the raw material wire and irradiating the raw material wire with laser light to manufacture a metal additive manufactured object.

[0108] Embodiment 2. In Embodiment 1, the amount of evaporation of low-boiling-point metal at the start of printing is predicted when printing is performed using the generated printing condition information. If this predicted value of the amount of evaporation of low-boiling-point metal falls within an acceptable range, printing is performed for all printing layers constituting the metal additive manufacturing product using the same printing conditions. As printing is repeated, due to heat accumulation during printing, etc., if N is a natural number of 2 or more, the amount of evaporation of low-boiling-point metal tends to increase as N increases. In other words, if the Nth layer is printed using the printing condition information at the start of printing, the amount of evaporation of low-boiling-point metal may exceed the acceptable range. For this reason, predicting only the amount of evaporation of low-boiling-point metal at the start of printing is insufficient to obtain a homogeneous metal additive manufacturing product. Therefore, Embodiment 2 describes a metal additive manufacturing apparatus 10 and a method for manufacturing a metal additive manufacturing product in which the predicted value of the amount of evaporation of low-boiling-point metal falls within an acceptable range throughout the entire printing process.

[0109] The amount of low-boiling-point metal evaporation during the manufacturing process is determined by the amount of low-boiling-point metal evaporation at the start of the manufacturing process, the effects of heat storage and heat dissipation, and the change in laser absorption rate due to oxidation of the underlying fabricated layer. In other words, during the manufacturing process, it is desirable to predict the amount of low-boiling-point metal evaporation by considering not only wire surface condition information and manufacturing condition information, but also the fabricated object condition information, which indicates the state of the underlying fabricated layer. In Embodiment 2, the underlying temperature information obtained by measuring the temperature of the underlying fabricated layer during the manufacturing process, and the fabricated object surface condition information, which is the surface condition information of the underlying fabricated layer, are measured.

[0110] The difference in heat storage and heat dissipation of the underlying build layer during the build process is considered one reason why the amount of low-boiling-point metal evaporation during the build process changes. Specifically, at the start of the build process, the object is built on a base material at room temperature, whereas during the build process, it is built on a build layer that constitutes a heated metal additive manufacturing object. Furthermore, it is thought that not only the temperature of the underlying build layer, but also the composition of the underlying build layer, i.e., the amount of low-boiling-point metal evaporation, has a significant impact on the newly formed build layer. Therefore, by considering not only the underlying temperature information, but also the amount of low-boiling-point metal evaporation in the underlying build layer, i.e., at the processing point when forming the new build layer, it is possible to improve the accuracy of predicting the amount of low-boiling-point metal evaporation in the new build layer. Accordingly, the build state information is defined as at least one of the following pieces of information: underlying temperature information, build surface state information, and the amount of low-boiling-point metal evaporation at the processing point. In Embodiment 2, the predicted value of the amount of low-boiling-point metal evaporation in the new build layer to be built is predicted from the wire surface state information, build condition information, and build state information. In Embodiment 2, "during printing" refers to the printing of the printing layers, excluding the printing of the first printing layer at the start of printing; in other words, it refers to the printing of the second and subsequent printing layers.

[0111] Figure 10 is a block diagram showing an example of the configuration of a metal additive manufacturing apparatus according to Embodiment 2. The metal additive manufacturing apparatus 10A according to Embodiment 2 includes a raw material wire measurement unit 11, a calculation processing unit 100, a molding unit 20, and a molded object state information measurement unit 50. The calculation processing unit 100 includes a molded object information input unit 12, a molding condition generation unit 13, a performance data storage unit 14A, an evaporation amount prediction unit 15A, and a determination processing unit 16A. Components identical to those in Embodiment 1 are denoted by the same reference numerals, and their descriptions are omitted.

[0112] The performance data storage unit 14A stores performance data, which is various data related to additive manufacturing. The performance data includes wire surface condition information, manufacturing condition information, and actual values ​​of the evaporation amount of low-boiling-point metal at the start of manufacturing, as well as the state of the manufactured object and actual values ​​of the evaporation amount of low-boiling-point metal during manufacturing. In other words, the performance data includes data from past manufacturing start times, associated with the manufacturing condition information, wire surface condition information, and actual values ​​of the evaporation amount of low-boiling-point metal at the start of manufacturing, and data from past manufacturing processes, associated with the manufacturing condition information, wire surface condition information, and actual values ​​of the evaporation amount of low-boiling-point metal during manufacturing. The state of the manufactured object is information that includes at least one of the following: substrate temperature information, surface condition information of the manufactured object, and the evaporation amount of low-boiling-point metal at the processing point when forming a new manufacturing layer. The data stored in the performance data storage unit 14A is read out as appropriate by the evaporation amount prediction unit 15A.

[0113] The evaporation rate prediction unit 15A not only predicts the amount of evaporation of low-boiling-point metal at the start of molding, but also predicts the amount of evaporation of low-boiling-point metal during molding based on molding condition information, wire surface condition information, and molded object condition information. The predicted amount of evaporation of low-boiling-point metal at the start of molding was explained in Embodiment 1, so that explanation will be omitted here. The evaporation rate prediction unit 15A predicts the amount of evaporation of low-boiling-point metal during molding using a trained model that shows the relationship between past molding condition information, wire surface condition information, and molded object condition information, learned by machine learning, and the actual amount of evaporation of low-boiling-point metal during molding. The molded object condition information includes the substrate temperature information and molded object surface condition information obtained by measurement by the molded object condition information measurement unit 50, and the amount of evaporation of low-boiling-point metal at the processing point when forming a new molded layer. The molded object surface condition information includes at least one of the surface oxygen content and laser absorption rate of the substrate molded layer. When forming a new build layer, the amount of evaporation of low-boiling-point metal at the processing point can be determined by using, in one example, the predicted value of the amount of evaporation of low-boiling-point metal predicted by the evaporation amount prediction unit 15A for the underlying build layer. The evaporation amount prediction unit 15A outputs the predicted value of the amount of evaporation of low-boiling-point metal during build to the determination processing unit 16A.

[0114] The determination processing unit 16A determines whether the predicted value of the evaporation amount of the low-boiling-point metal at the start of molding or during molding, which is input from the evaporation amount prediction unit 15A, falls within a set tolerance range. In one example, the evaporation amount of the low-boiling-point metal at the start of molding or during molding can be the amount relative to the Zn concentration of the wire 21. The determination method is the same as that described in Embodiment 1, so its explanation is omitted.

[0115] The molded object state information measurement unit 50 measures molded object state information that indicates the state of the molded layer that has already been molded and serves as the base for the next new molded layer to be formed. Here, the molded object state information measurement unit 50 includes a temperature measurement unit 51 and a molded object measurement unit 52.

[0116] The temperature measurement unit 51 measures the temperature of the metal additive manufacturing object being manufactured, that is, the temperature of the already formed manufacturing layer. The measured temperature becomes the substrate temperature information. An example of the temperature measurement unit 51 is a radiation thermometer. A radiation thermometer is a non-contact type thermometer that measures the temperature of the manufacturing layer. The location on the manufacturing layer where the temperature is measured can be selected as appropriate. Here, the temperature of the processing point is measured by the radiation thermometer. The radiation thermometer is mounted coaxially with the laser beam L emitted from the beam nozzle 24. Therefore, the radiation thermometer can measure the temperature of the processing point when the laser beam L is being emitted.

[0117] Furthermore, the radiation thermometer can also measure the temperature at the processing point, i.e., the position where the bead was formed, after the emission of the laser beam L has been stopped. Hereinafter, the position where the bead was formed will be referred to as the build point. The metal additive manufacturing apparatus 10A can observe the temperature history of the build point after the emission of the laser beam L has been stopped using the radiation thermometer. In addition, the radiation thermometer can also measure the interlayer temperature, i.e., the interpass temperature, by measuring the temperature of the substrate immediately before the emission of the laser beam L is started.

[0118] The temperature measurement unit 51 is not limited to a radiation thermometer, but may also be other thermometers such as a thermal viewer. Furthermore, the metal additive manufacturing apparatus 10A may be equipped with multiple temperature measurement units 51. In this case, it may be equipped with multiple temperature measurement units 51 of the same type, for example multiple radiation thermometers, or it may be equipped with multiple temperature measurement units 51 of different types, for example a radiation thermometer and a thermal viewer.

[0119] In Embodiment 2, the temperature history is the temperature value measured by the temperature measurement unit 51, in this example by a radiation thermometer, or a value calculated from this value. The maximum temperature is the highest temperature among the temperatures measured for each position of the fabricated layer. The interlayer temperature is the temperature of the substrate immediately before the laser beam L is emitted. The temperature measurement unit 51 outputs one of the following as substrate temperature information to the evaporation rate prediction unit 15A: temperature history, maximum temperature, interlayer temperature, etc. The user can appropriately set the information to be used as substrate temperature information in accordance with the input information used by the evaporation rate prediction unit 15A to predict the evaporation rate of low-boiling-point metals.

[0120] The fabricated object measurement unit 52 measures surface state information of the fabricated object, including at least one of the surface oxygen content and laser absorptivity of the fabricated layer that will serve as the base layer for forming the next new fabricated layer. When measuring the surface oxygen content of the fabricated layer, the fabricated object measurement unit 52 can be a small, portable elemental analysis instrument to measure the surface composition of the fabricated layer during fabrication. In one example, a commercially available handheld X-ray fluorescence device or the like is provided in the fabrication unit 20 as the fabricated object measurement unit 52. The fabricated object measurement unit 52 measures the surface oxygen content of the base fabricated layer after the fabrication of the fabricated layer is complete and before the next layer is additively fabricated.

[0121] When measuring the laser absorption rate of a fabricated layer, it is necessary to measure the surface of the fabricated layer during fabrication, so it is desirable that the fabricated object measurement unit 52 be small. In the example where the laser oscillator 23 is a YAG laser, a spectrophotometer is provided in the fabrication unit 20 as the fabricated object measurement unit 52 for the purpose of measuring only the laser absorption rate in the wavelength range of 1064 nm. In this case, the spectrophotometer measures the laser absorption rate by transmitting only a specific wavelength using a dichroic mirror.

[0122] Furthermore, the surface condition information of the fabricated object only needs to include at least one of the surface oxygen content of the fabricated layer and the laser absorptivity; therefore, the fabricated object measurement unit 52 only needs to be capable of measuring at least one of the surface oxygen content of the fabricated layer and the laser absorptivity.

[0123] Furthermore, if at least one of the substrate temperature information and the surface condition information of the molded object is used as the molded object state information, it is sufficient to have at least one of the temperature measurement unit 51 and the molded object measurement unit 52.

[0124] Next, the detailed configuration of the evaporation rate prediction unit 15A will be described. Figure 11 is a block diagram showing an example of the configuration of the evaporation rate prediction unit. Note that the same reference numerals are used for components identical to those in Figure 4 of Embodiment 1, and their descriptions are omitted. Figure 11 shows the case where the evaporation rate prediction unit 15A predicts the evaporation rate of a low-boiling-point metal when a metal additive manufacturing product consists of N layers, and N is 2 or more layers.

[0125] The evaporation rate prediction unit 15 in Figure 4 of Embodiment 1 predicts the amount of evaporation of the low-boiling-point metal at the start of the manufacturing process, while the evaporation rate prediction unit 15A in Figure 11 predicts the amount of evaporation of the low-boiling-point metal during the manufacturing process. Therefore, the information input to the evaporation rate prediction unit 15 in Figure 4 of Embodiment 1 is different from that input to the learning unit 15A. In other words, when predicting the amount of evaporation of the low-boiling-point metal during the manufacturing process, the learning unit 151 receives wire surface state information, manufacturing condition information, manufacturing state information during the manufacturing process, and the actual amount of evaporation of the low-boiling-point metal during the manufacturing process as learning data. Based on the learning data, a trained model is generated that infers the optimal predicted amount of evaporation of the low-boiling-point metal during the manufacturing process from the wire surface state information, manufacturing condition information, manufacturing state information during the manufacturing process, and the actual amount of evaporation of the low-boiling-point metal during the manufacturing process.

[0126] The trained model generated in Figure 4 is a model that infers the predicted amount of evaporation of low-boiling-point metal at the start of fabrication. On the other hand, the trained model generated in Figure 11 is a model that infers the predicted amount of evaporation of low-boiling-point metal during fabrication, excluding the start of fabrication. Therefore, in the following, the trained model generated in Figure 4 will be referred to as the trained model for the start of fabrication, and the trained model generated in Figure 11 will be referred to as the trained model for during fabrication.

[0127] The learned model storage unit 152 stores a learned model for use at the start of the printing process and a learned model for use during the printing process.

[0128] Furthermore, when predicting the evaporation rate of low-boiling-point metal during molding as shown in Figure 11, the inference unit 153 receives wire surface condition information, molding condition information, and molding condition information. The wire surface condition information is obtained from the raw wire measurement unit 11 and may be a value measured at the start of molding or a value measured in situ during molding. The molding condition information is generated by the molding condition generation unit 13 or adjusted by the evaporation rate prediction unit 15A. When at least one of the substrate temperature information and molding surface condition information is used as molding condition information, the molding condition information is a value measured in situ by the molding condition information measurement unit 50 during molding. Also, when using the evaporation rate of low-boiling-point metal at the processing point when forming a new molding layer as molding condition information, the molding condition information can be the predicted value of the evaporation rate of low-boiling-point metal predicted during the molding of the substrate molding layer.

[0129] The inference unit 153 then inputs wire surface state information, molding condition information, and molding state information into a pre-trained model for use during molding, and outputs a predicted value for the amount of evaporation of the low-boiling-point metal during molding, which is inferred from the wire surface state information, molding condition information, and molding state information.

[0130] With this configuration, the evaporation rate prediction unit 15A predicts the evaporation rate of low-boiling-point metals using a pre-trained model for the start of manufacturing when creating the first layer of a metal additive manufacturing structure. Furthermore, when creating the second and subsequent layers of a metal additive manufacturing structure, the evaporation rate prediction unit 15A predicts the evaporation rate of low-boiling-point metals using a pre-trained model for use during manufacturing.

[0131] Next, a method for manufacturing a metal additively manufactured object by generating manufacturing conditions in which the predicted evaporation amount of low-boiling-point metal for each layer to be fabricated falls within an acceptable range will be described. Figures 12 and 13 are flowcharts showing an example of the procedure for manufacturing a metal additively manufactured object according to Embodiment 2. Note that the process from step S31 to step S37 in Figure 7 of Embodiment 1 is the same in Embodiment 2, so its explanation will be omitted. However, the trained model used in Figure 7 is the trained model for the start of fabrication.

[0132] The molding condition generation unit 13 generates an NC program for forming one layer of molded material using the molding condition information acquired in step S37, and outputs the NC program to the molding unit 20 (step S91). The molding unit 20 forms the molded material according to the acquired NC program (step S92).

[0133] Subsequently, the process moves on to the formation of the second and subsequent layers. Here, we describe a generalized processing procedure for a metal additive manufacturing object consisting of N layers of built-in structures, where N is a natural number greater than or equal to 2, and i is a natural number between 2 and N, and the (i+1)th layer is being fabricated. Furthermore, here we give an example where substrate temperature information and surface condition information of the fabricated object are used as fabricated object state information. In other words, we give an example where the fabricated object state information measurement unit 50 has a temperature measurement unit 51 and a fabricated object measurement unit 52. The fabricated object surface condition information is at least one of the following: the surface oxygen content of the fabricated layer and the laser absorption rate.

[0134] The fabricated object state information measurement unit 50 measures the substrate temperature information and the fabricated object surface state information at the processing point of the (i+1)th layer after the fabrication of the i-th layer is completed (step S93). Specifically, the fabricated object measurement unit 52 moves to the fabrication start point of the (i+1)th layer and measures at least one of the surface oxygen content and laser absorptivity of the fabricated layer that serves as the substrate at that time. The temperature measurement unit 51 also measures the temperature of the fabricated layer that serves as the substrate at the fabrication start point. However, regarding temperature, if multiple temperature measurement units 51 are used, the measurement does not have to be limited to the fabrication start point, and a more accurate temperature history may be obtained based on the temperature measurement results at multiple points. In this case, the substrate temperature information becomes substrate temperature history information. The substrate of the (i+1)th layer is the uppermost layer that constitutes the metal additive fabricated object formed up to the point of the (i+1)th layer fabrication pass. When simply forming a build layer on top of a lower build layer, the base for the (i+1)th layer will be the ith layer. However, in the case of complex shapes, the base for the (i+1)th layer may not necessarily be the ith layer. Step S93 corresponds to the build object measurement process.

[0135] Next, the molding condition generation unit 13 outputs molding condition information to the evaporation amount prediction unit 15A (step S94). Based on the wire surface condition information of the wire 21 measured in step S31, the molded object condition information including the substrate temperature information and molded object surface condition information measured in step S91, and the molding condition information, the evaporation amount prediction unit 15A performs a low-boiling-point metal evaporation amount prediction process to predict the amount of evaporation of the low-boiling-point metal when molding the i+1th layer under the said molding conditions (step S95). Step S95 corresponds to the second evaporation amount prediction step.

[0136] Figure 14 is a flowchart showing an example of the procedure for predicting the evaporation rate of a low-boiling-point metal. This process is performed in the inference unit 153 of the evaporation rate prediction unit 15A. First, the data acquisition unit 1531 acquires wire surface condition information, molded object condition information, and molding condition information (step S111). The wire surface condition information is the information measured by the raw wire measurement unit 11 in step S31. The molded object condition information is the molded object surface condition information and substrate temperature information measured by the molded object condition information measurement unit 50 in step S93. The molding condition information is the information generated by the molding condition generation unit 13 in step S33. Here, the molding condition information includes the molding environment condition information and the i+1 layer molding path information generated by the molding condition generation unit 13.

[0137] Next, the prediction unit 1532 inputs wire surface state information, fabricated object state information, and fabrication condition information into the learned model for use during fabrication stored in the learned model storage unit 152, and obtains a predicted value for the evaporation amount of low-boiling point metal during fabrication (step S112).

[0138] Subsequently, the prediction unit 1532 outputs a predicted value for the evaporation amount of the low-boiling-point metal during molding, obtained from the pre-trained model for use during molding, to the determination processing unit 16A (step S113). With this, the low-boiling-point metal evaporation amount prediction process is completed, and the process returns to Figure 12.

[0139] Subsequently, the determination processing unit 16A determines whether the predicted value of the evaporation amount of the low-boiling-point metal during molding, obtained from the evaporation amount prediction unit 15A, is within the acceptable range (step S96). Step S96 corresponds to the second determination step. If the predicted value of the evaporation amount of the low-boiling-point metal during molding is not within the acceptable range (if the result is No in step S96), the determination processing unit 16A outputs to the evaporation amount prediction unit 15A that the predicted value of the evaporation amount of the low-boiling-point metal during molding is not within the acceptable range. In this case, the evaporation amount prediction unit 15A adjusts the molding condition information (step S97). This process is the same as step S36 in Figure 7. Step S97 also corresponds to the second molding condition adjustment step. After that, the process returns to step S95. In step S95, the predicted value of the evaporation amount of the low-boiling-point metal is predicted with the adjusted molding conditions. The processes from step S95 to step S97 are repeatedly executed until the predicted amount of evaporation of the low-boiling-point metal falls within a predetermined tolerance range and it is determined that the molding condition information allows for molding.

[0140] If the predicted value of the evaporation amount of the low-boiling-point metal during molding is within the acceptable range (Yes in step S96), the determination processing unit 16A outputs information to the evaporation amount prediction unit 15A indicating that the predicted value of the evaporation amount of the low-boiling-point metal during molding is within the acceptable range. The evaporation amount prediction unit 15A then outputs molding condition information, which indicates that the predicted value of the evaporation amount of the low-boiling-point metal during molding is within the acceptable range, to the molding condition generation unit 13 (step S98).

[0141] Next, the molding condition generation unit 13 uses the acquired molding condition information and the molding path information for the (i+1)th layer to generate an NC program for molding the (i+1)th layer, and outputs the NC program to the molding unit 20 (step S99). The molding unit 20 forms the (i+1)th layer according to the acquired NC program (step S100). Step S100 corresponds to the molding process.

[0142] Subsequently, the molding condition generation unit 13 determines whether the (i+1)th molding layer is the last molding layer (step S101). If it is not the last molding layer (if the result is No in step S101), the process returns to step S93. Then, the process from step S93 to step S101 is repeatedly executed until the Nth molding layer is formed. If the (i+1)th molding layer is the last molding layer in step S101 (if the result is Yes in step S101), the metal additive manufacturing process is completed.

[0143] The above explanation shows that the fabricated object state information includes the fabricated object surface state information and the substrate temperature information, but it is not limited to this. For example, the fabricated object state information for the second layer and beyond may include a predicted value for the amount of low-boiling-point metal evaporation at the processing point when a new fabricated layer is formed. Specifically, the predicted value for the amount of low-boiling-point metal evaporation in the i-th fabricated layer may be used to predict the amount of low-boiling-point metal evaporation in the i+1th fabricated layer. In other words, the prediction for the next layer may be made based on information about the previous layers. In this case, the fabricated object state information during fabrication in the learning data input to the learning unit 151 in Figure 11 will include a predicted value for the amount of low-boiling-point metal evaporation at the processing point when a new fabricated layer is formed. Then, based on the training data, a trained model for use during metal additive manufacturing is generated that infers the optimal predicted amount of low-boiling-point metal evaporation during manufacturing from the state information of the manufactured object during manufacturing, which includes wire surface state information, manufacturing condition information, and predicted values ​​of the amount of low-boiling-point metal evaporation at the processing point when forming a new layer, as well as the actual value of the amount of low-boiling-point metal evaporation during manufacturing.

[0144] Furthermore, the fabricated object state information input to the inference unit 153 in Figure 11 includes a predicted value for the amount of low-boiling-point metal evaporation at the processing point when forming a new fabricated layer. The inference unit 153 then inputs the wire surface state information, fabrication condition information, and the fabricated object state information during fabrication, which includes the predicted value for the amount of low-boiling-point metal evaporation at the processing point when forming a new fabricated layer, into a pre-trained model for use during fabrication. The inference unit 153 then outputs a predicted value for the amount of low-boiling-point metal evaporation during fabrication, inferred from the wire surface state information, fabrication condition information, and the fabricated object state information during fabrication. This further improves the accuracy of the predicted value for the amount of low-boiling-point metal evaporation.

[0145] Furthermore, in the above description, the surface condition information of the molded object and the substrate temperature information are measured as the state information of the molded object in step S93. However, at least one of the following may be used: the surface condition information of the molded object, the substrate temperature information, and the predicted amount of evaporation of the low-boiling-point metal at the processing point when a new molded layer is formed.

[0146] Here, we will explain a specific example. However, the basic specific example is the same as in Embodiment 1. In Embodiment 1, we predicted the amount of Zn evaporation in the first layer at the start of molding, but in Embodiment 2, we predict the amount of Zn evaporation in the second layer and beyond.

[0147] First, using Embodiment 1, the evaporation rate of low-boiling-point metal at the start of the first layer of fabrication is predicted, and the layer is fabricated using fabrication condition information that sets the predicted evaporation rate of low-boiling-point metal at the start of fabrication within an acceptable range. After the first layer of fabrication is completed, the second and subsequent layers are fabricated using Embodiment 2. In step S93, at least one of the surface oxygen content and laser absorptivity, which are surface state information of the fabricated object at the start of fabrication of the second layer, and the interlayer temperature, which is substrate temperature information, are measured. In step S94, fabrication condition information for forming the second layer of fabrication is generated based on the fabricated object information.

[0148] In step S95, the amount of low-boiling-point metal to evaporate when creating the second layer is predicted based on wire surface condition information, fabricated object surface condition information, substrate temperature information, fabricated object condition information including a predicted value for the amount of low-boiling-point metal to evaporate when creating the first layer, and fabrication condition information.

[0149] Subsequently, as shown in steps S96 to S97, the molding condition information is adjusted as necessary based on the predicted evaporation rate of the low-boiling-point metal. Then, as shown in steps S98 to S100, the molding of the second layer is started. After the second layer is molded, the third layer is molded. That is, after the i-th layer is molded, the (i+1)th layer is molded.

[0150] The A7075 (additive) used as wire 21 in Embodiment 1 will be described. The amount of Zn evaporation at the start of molding is predicted in the same manner as in Embodiment 1. At the start of molding, the predicted value of the amount of Zn evaporation was 10% or less compared to wire 21, so the molding process proceeds as is. Subsequently, when molding the second layer, the processing head 26 is moved to the starting point of the second layer to measure the temperature of the first layer, i.e., the interlayer temperature. The molded object measurement unit 52 is also moved to the starting point of the second layer to measure at least one of the surface oxygen content and laser absorptivity of the first layer. Based on the substrate temperature information, including interlayer temperature, surface oxygen content and laser absorptivity, the predicted value of the amount of evaporation of the low-boiling-point metal predicted during the molding of the first layer i, the molding condition information, and the actual value of the amount of evaporation of the low-boiling-point metal obtained during past moldings, the amount of Zn evaporation of the second layer, which is the low-boiling-point metal, is predicted. If the predicted amount of Zn evaporation is 10% or less compared to wire 21, the process proceeds directly to the fabrication stage. By repeating these processes, the amount of Zn evaporation for each layer can be predicted.

[0151] Furthermore, the metal additive manufacturing product produced by the above process is made up of layers formed by irradiating a laser beam L onto a wire 21 made of an Al alloy containing a low-boiling-point metal, which is a metal with a lower boiling point than Al. In this metal additive manufacturing product, the difference in the component concentration of the low-boiling-point metal between the first layer, which is the initial manufacturing position of the metal additive manufacturing product, and any other layer is 10% or less. This is achieved by setting the tolerance range to 10%. As a result, the variation in the overall composition of the metal additive manufacturing product can be suppressed compared to conventional methods. Here, the difference in component concentration of the low-boiling-point metal is expressed as a percentage of the difference between the concentration of the low-boiling-point metal in any other layer and the concentration of the low-boiling-point metal in the first layer, relative to the concentration of the low-boiling-point metal in the first layer.

[0152] The manufacturing method for a metal additive manufactured product according to Embodiment 2 further includes a manufactured product measurement step and a second evaporation amount prediction step in addition to the manufacturing method for a metal additive manufactured product according to Embodiment 1. The manufactured product measurement step measures manufactured product state information indicating the state of the metal additive manufactured product at the processing point of the new manufactured layer when a new manufactured layer is formed on a metal additive manufactured product consisting of a manufactured layer formed in the manufacturing step. The second evaporation amount prediction step predicts the amount of evaporation of low-boiling-point metal when a new manufactured layer is formed according to the manufactured condition information generated in the manufacturing condition generation step, using wire surface state information and manufactured product state information. This has the effect of suppressing the amount of evaporation of metals with lower boiling points than Al in the metal additive manufactured product compared to conventional methods, when an Al alloy containing metals with lower boiling points than Al is used as the raw material wire and a metal additive manufactured product is formed by irradiating the raw material wire with laser light. Furthermore, in Embodiment 2, the amount of evaporation of low-boiling-point metal in each layer is predicted, and the manufacturing conditions are regenerated, so a homogeneous metal additive manufactured object can be obtained, and the problem of the amount of evaporation of low-boiling-point metal differing in the early, middle, and late stages of manufacturing can be solved.

[0153] As the fabrication process progresses, heat accumulates in the fabricated layers. Even if the fabrication conditions are based on those at the start of the process, the evaporation rate of low-boiling-point metals may exceed the acceptable range. In other words, the evaporation rate of low-boiling-point metals changes during the fabrication process. Therefore, by predicting both the evaporation rate of low-boiling-point metals at the start of the process and the evaporation rate during the process, the overall composition of the fabricated metal additive manufacturing object, or the properties based on that composition, can be made more homogeneous than before. In short, it is possible to manufacture more homogeneous metal additive manufacturing objects compared to conventional methods.

[0154] The manufacturing method for a metal additive manufactured product according to Embodiment 2 further includes a second determination step in which, in the manufacturing method for a metal additive manufactured product according to Embodiment 1, the manufacturer determines whether or not manufacturing is possible based on the manufacturing condition information, depending on whether or not the predicted value of the evaporation amount of the low-boiling-point metal is within an acceptable range. In this case, in the manufacturing step, if the second determination step determines that the predicted value of the evaporation amount of the low-boiling-point metal is within an acceptable range and that manufacturing is possible based on the manufacturing condition information, a new manufacturing layer is manufactured according to the manufacturing condition information. This makes it possible to manufacture under manufacturing conditions where the predicted value of the evaporation amount of the low-boiling-point metal is within an acceptable range, not only at the start of manufacturing but also during manufacturing.

[0155] The manufacturing method for a metal additive manufactured product according to Embodiment 2 further includes a second molding condition adjustment step in which, in the manufacturing method for a metal additive manufactured product according to Embodiment 1, the predicted value of the evaporation amount of low-boiling-point metal is not within the acceptable range and it is determined that molding is not possible based on the molding condition information. In this case, the second evaporation amount prediction step predicts the predicted value of the evaporation amount of low-boiling-point metal using the molding condition information adjusted in the second molding condition adjustment step. The second molding condition adjustment step, the second evaporation amount prediction step, and the second determination step are repeatedly performed until the predicted value of the evaporation amount of low-boiling-point metal is within the acceptable range and it is determined that molding is possible based on the molding condition information. As a result, molding is not performed under molding conditions where the predicted value of the evaporation amount of low-boiling-point metal during molding is outside the acceptable range, so the amount of evaporation of low-boiling-point metal in the resulting metal additive manufactured product can be suppressed compared to conventional methods.

[0156] The manufacturing method for a metal additive manufactured product according to Embodiment 2 is the same as the manufacturing method for a metal additive manufactured product according to Embodiment 1. In the second evaporation rate prediction step, a pre-trained model for use during manufacturing is used to infer a predicted value of the evaporation rate of low-boiling-point metal during manufacturing from wire surface state information, manufactured product state information, and manufacturing condition information. The model outputs a predicted value of the evaporation rate of low-boiling-point metal during manufacturing from wire surface state information obtained in the wire measurement step, manufactured product state information obtained in the manufactured product measurement step, and manufacturing condition information generated in the manufacturing condition generation step. A pre-trained model for use during manufacturing is generated using a set of wire surface state information, manufacturing condition information, and manufactured product state information related to the evaporation rate of low-boiling-point metal, and actual values ​​of the evaporation rate of low-boiling-point metal, as training data. By inputting wire surface state information, manufacturing condition information, and manufactured product state information into this pre-trained model for use during manufacturing to predict the evaporation rate of low-boiling-point metal during manufacturing, a more accurate prediction of the evaporation rate of low-boiling-point metal can be obtained compared to when prediction is made without using wire surface state information.

[0157] The manufacturing method for a metal additive manufactured product according to Embodiment 2 is the same as the manufacturing method for a metal additive manufactured product according to Embodiment 1, wherein the manufactured product state information includes at least one of the following: substrate temperature information including the temperature at the processing point when forming a new manufacturing layer, the surface oxygen content of the metal additive manufactured product at the processing point, the laser absorption rate of the metal additive manufactured product at the processing point, and a predicted value of the evaporation amount of low-boiling-point metal at the processing point. This makes it possible to further improve the accuracy of predicting the evaporation amount of low-boiling-point metal, taking into account the influence of the manufacturing substrate.

[0158] The metal additive manufacturing apparatus 10A according to Embodiment 2 further includes a material state information measurement unit 50 that measures material state information indicating the state of the metal additive manufacturing object at the processing point of the new material layer when forming a new material layer on top of a metal additive manufacturing object consisting of a material layer formed in the material layer unit 20, in addition to the metal additive manufacturing apparatus 10 according to Embodiment 1. The evaporation amount prediction unit 15A predicts the amount of evaporation of low-boiling-point metal when a new material layer is formed according to the material condition information generated in the material condition generation unit 13, using wire surface state information and material state information. As a result, when an Al alloy containing a metal with a lower boiling point than Al is used as the raw material wire, and a metal additive manufacturing object is formed by irradiating the raw material wire with laser light, the amount of evaporation of the metal with a lower boiling point than Al is suppressed compared to conventional methods, not only at the start of manufacturing but also during manufacturing when heat accumulation is affected. As a result, it is possible to manufacture a metal additive manufacturing object with a desired composition or physical properties based on a desired composition.

[0159] In Embodiments 1 and 2, Zn, which has a low boiling point among metals and is commonly used, was described as a low-boiling-point metal. However, Embodiments 1 and 2 can be applied to any element used in Al alloys that has a lower boiling point than Al. Examples of such elements include Li (lithium) and Mg.

[0160] The arithmetic processing unit 100 in the metal additive manufacturing apparatus 10 and 10A according to Embodiments 1 and 2 can be implemented by an information processing device, for example. The hardware configuration when the arithmetic processing unit 100 is implemented by a computer system, which is an information processing device, will be described below. The computer system functions as the arithmetic processing unit 100 when a program, which is a computer program describing the processing in the arithmetic processing unit 100, is executed on the computer system. Figure 15 is a block diagram showing an example of the configuration of a computer system that implements the arithmetic processing unit of the metal additive manufacturing apparatus according to Embodiments 1 and 2. As shown in Figure 15, this computer system includes a control unit 901, an input unit 902, a storage unit 903, a display unit 904, a communication unit 905, and an output unit 906, which are connected via a system bus 907.

[0161] In Figure 15, the control unit 901 is, in one example, a processor such as a CPU (Central Processing Unit), and executes a program that describes the processing in the arithmetic processing unit 100 of Embodiments 1 to 4. The input unit 902 is, in one example, composed of a keyboard, mouse, etc., and is used by the user of the computer system to input various information. The storage unit 903 includes various types of memory such as RAM (Random Access Memory), ROM (Read Only Memory), and storage devices such as a hard disk, and stores the program that the control unit 901 should execute, necessary data obtained in the process of processing, etc. The storage unit 903 is also used as a temporary storage area for the program. The display unit 904 is composed of a display, liquid crystal display panel, etc., and displays various screens to the user of the computer system. In one example, the input unit 902 and the display unit 904 may be configured as a touch panel in which the input unit 902 and the display unit 904 are integrally formed. The communication unit 905 is a receiver and transmitter that perform communication processing. The output unit 906 is a printer, speaker, etc. Note that Figure 15 is just one example, and the configuration of the computer system is not limited to the example shown in Figure 15.

[0162] Here, we will describe an example of the operation of a computer system until a program becomes executable. In a computer system with the above configuration, for example, a program is installed in the storage unit 903 from a CD-ROM or DVD-ROM set in a CD (Compact Disc)-ROM drive or DVD (Digital Versatile Disc)-ROM drive (not shown). When the program is executed, the program read from the storage unit 903 is stored in the main memory area of ​​the storage unit 903. In this state, the control unit 901 performs processing as the arithmetic processing unit 100 of Embodiments 1 and 2 according to the program stored in the storage unit 903.

[0163] In the above description, a program describing the processing in the arithmetic processing unit 100 is provided using a CD-ROM or DVD-ROM as the recording medium. However, the system is not limited to this, and depending on the configuration of the computer system, the capacity of the program to be provided, a program provided via a transmission medium such as the Internet via the communication unit 905 may also be used.

[0164] The molded object information input unit 12, the molding condition generation unit 13, the evaporation amount prediction units 15, 15A, and the determination processing units 16, 16A shown in Figures 2 and 10 are realized by the execution of a program stored in the storage unit 903 shown in Figure 15 by the control unit 901 shown in Figure 15. The storage unit 903 shown in Figure 15 is also used to realize the molded object information input unit 12, the molding condition generation unit 13, the evaporation amount prediction units 15, 15A, and the determination processing units 16, 16A. In addition, the input unit 902 or the communication unit 905 shown in Figure 15 is also used to realize the molded object information input unit 12. Furthermore, the input unit 902 and the output unit 906 shown in Figure 15 are also used to realize the evaporation amount prediction units 15, 15A. The actual data storage units 14, 14A are realized by the storage unit 903 shown in Figure 15.

[0165] Next, we will describe comparative examples in which a metal additive manufacturing product is produced without using the metal additive manufacturing apparatus 10, 10A and the metal additive manufacturing product manufacturing method according to Embodiments 1 and 2, and comparative examples in which a metal additive manufacturing product is produced using the metal additive manufacturing apparatus 10A and the metal additive manufacturing product manufacturing method according to Embodiment 2.

[0166] (Comparative Example) In the comparative example, a wall-shaped metal additive manufacturing object with a height of 100 mm is manufactured using an Al alloy wire 21 containing Zn, without using the metal additive manufacturing apparatus 10, 10A and the method for manufacturing the metal additive manufactured object according to Embodiments 1 and 2. In this case, as the manufacturing progresses, the amount of Zn evaporation tends to increase due to the effect of heat accumulation. Elemental analysis of the position of the last manufactured layer showed that the amount of Zn evaporation had increased by 30% compared to the start of manufacturing. In other words, in the comparative example, there is unevenness in the Zn content within the metal additive manufactured object depending on the location. Therefore, there is a possibility that the desired physical properties cannot be obtained for the metal additive manufactured object as a whole.

[0167] (Example) In this example, a metal additive manufacturing apparatus 10A and a method for manufacturing a metal additive manufactured object according to Embodiment 2 are used to manufacture a wall-shaped metal additive manufactured object with a height of 100 mm using a wire 21 made of Al alloy containing Zn. In this case, a metal additive manufactured object with little variation in the internal composition is obtained. Therefore, the desired physical properties can be obtained for the entire metal additive manufactured object.

[0168] Furthermore, the threshold for the evaporation amount of the low-boiling-point metal, which defines the acceptable range, is set to 20% or less, and the metal additive manufactured object is produced according to the flow described above. In this case, if it is predicted that the threshold will be exceeded based on the predicted value, the manufacturing condition information is changed, such as by setting a pause time or lowering the laser output. This makes it possible to produce a metal additive manufactured object with less variation in the amount of Zn evaporation inside. In this case, elemental analysis of the position of the last manufactured layer showed that the amount of Zn evaporation had increased by 18% compared to the start of manufacturing. From this, it can be seen that by using the metal additive manufacturing apparatus 10A and the method for manufacturing a metal additive manufactured object according to Embodiment 2, the fluctuation in the concentration of the low-boiling-point metal is suppressed in the manufacturing of Al alloys containing low-boiling-point metals compared to the comparative example. As a result, a more homogeneous metal additive manufactured object can be produced compared to the comparative example.

[0169] The configurations shown in the above embodiments are merely examples, and it is possible to combine them with other known technologies, combine different embodiments, and omit or modify parts of the configuration without departing from the gist of the invention.

[0170] 10, 10A Metal additive manufacturing apparatus, 11 Raw material wire measurement unit, 12 Molded object information input unit, 13 Molding condition generation unit, 14, 14A Performance data storage unit, 15, 15A Evaporation amount prediction unit, 16, 16A Judgment processing unit, 20 Molding unit, 21 Wire, 22 Wire nozzle, 23 Laser oscillator, 24 Beam nozzle, 25 Fiber cable, 26 Processing head, 27 Gas nozzle, 28 Gas supply device, 29 Piping, 30 Stage, 31 Rotation mechanism, 33 Head drive device, 34 Material supply mechanism, 35 Control device, 50 Molded object state information measurement unit, 51 Temperature measurement unit, 52 Molded object measurement unit, 100 Calculation processing unit, 151 Learning unit, 152 Learned model storage unit, 153 Inference unit, 901 Control unit, 902 Input unit, 903 Storage unit, 904 Display unit, 905 Communication unit, 906 Output unit, 907 System bus, 1511, 1531 Data acquisition unit, 1512 Model generation unit, 1532 Prediction unit, L Laser beam.

Claims

1. A method for manufacturing a metal additive manufacturing product, comprising sequentially stacking layers formed by irradiating a raw material wire made of an Al alloy containing a low-boiling-point metal which is a metal with a lower boiling point than Al with a laser beam, the method comprising: a wire measurement step of measuring wire surface state information which is information related to the amount of evaporation of the low-boiling-point metal in the raw material wire; a receiving step of receiving manufactured product information including the shape and material of the metal additive manufacturing product; a manufacturing condition generation step of generating manufacturing condition information which is the conditions for manufacturing the metal additive manufacturing product based on the manufactured product information; and a first evaporation amount prediction step of predicting a predicted value of the amount of evaporation of the low-boiling-point metal when the metal additive manufacturing product shown in the manufactured product information is manufactured according to the manufacturing condition information, based on the wire surface state information and the manufacturing condition information.

2. A method for manufacturing a metal additively manufactured product according to claim 1, further comprising: a determination step of determining whether or not it is possible to manufacture the product according to the manufacturing condition information based on whether or not the predicted amount of evaporation of the low boiling point metal is within a set tolerance range; and a manufacturing step of manufacturing the product according to the manufacturing condition information when it is determined that the predicted amount of evaporation of the low boiling point metal is within the tolerance range and it is possible to manufacture the product according to the manufacturing condition information.

3. The method for manufacturing a metal additively manufactured product according to claim 2, further comprising a first molding condition adjustment step for adjusting the molding condition information if the predicted value of the evaporation amount of the low boiling point metal does not fall within the allowable range and it is determined that the molding condition information does not allow for manufacturing, wherein the first evaporation amount prediction step predicts the predicted value of the evaporation amount of the low boiling point metal using the adjusted molding condition information, and the first molding condition adjustment step, the first evaporation amount prediction step, and the determination step are repeatedly performed until the determination step determines that the predicted value of the evaporation amount of the low boiling point metal falls within the allowable range and that the molding condition information allows for manufacturing.

4. The method for manufacturing a metal additively manufactured product according to claim 3, characterized in that, in the manufacturing step, the manufactured layer from the start of manufacturing to the end of manufacturing is manufactured according to the manufacturing condition information generated in the manufacturing condition generation step or the manufacturing condition information adjusted in the first manufacturing condition adjustment step.

5. A method for manufacturing a metal additive manufacturing product according to any one of claims 2 to 4, further comprising: a manufacturing product measurement step for measuring manufacturing product state information indicating the state of the metal additive manufacturing product at the processing point of the new manufacturing layer when forming a new manufacturing layer on the metal additive manufacturing product consisting of a manufacturing layer formed in the manufacturing step; and a second evaporation amount prediction step for predicting the amount of evaporation of the low boiling point metal when the new manufacturing layer is formed according to the manufacturing condition information generated in the manufacturing condition generation step, using the wire surface state information and the manufacturing product state information.

6. The method for manufacturing a metal additive manufacturing product according to claim 5, further comprising a second determination step of determining whether or not it is possible to manufacture using the manufacturing condition information based on whether or not the predicted amount of evaporation of the low boiling point metal is within the allowable range, wherein in the manufacturing step, if the second determination step determines that the predicted amount of evaporation of the low boiling point metal is within the allowable range and that it is possible to manufacture using the manufacturing condition information, the new manufacturing layer is manufactured according to the manufacturing condition information.

7. The method for manufacturing a metal additively manufactured product according to claim 6, further comprising a second molding condition adjustment step for adjusting the molding condition information if the predicted value of the evaporation amount of the low boiling point metal does not fall within the allowable range and it is determined that the molding condition information does not allow for manufacturing, wherein the second evaporation amount prediction step predicts the predicted value of the evaporation amount of the low boiling point metal using the molding condition information adjusted in the second molding condition adjustment step, and the second molding condition adjustment step, the second evaporation amount prediction step, and the second determination step are repeatedly performed until the predicted value of the evaporation amount of the low boiling point metal falls within the allowable range and it is determined that the molding condition information allows for manufacturing.

8. The method for manufacturing a metal additive manufacturing product according to any one of claims 5 to 7, characterized in that, in the second evaporation amount prediction step, a trained model for use during manufacturing is used to infer a predicted value of the evaporation amount of the low boiling point metal during manufacturing from the wire surface state information obtained in the wire measurement step, the manufactured object state information obtained in the manufactured object measurement step, and the manufacturing condition information generated in the manufacturing condition generation step, thereby outputting a predicted value of the evaporation amount of the low boiling point metal during manufacturing.

9. The method for manufacturing a metal additive manufacturing product according to any one of 5 to 8, characterized in that the molded product state information includes at least one of the following: substrate temperature information including the temperature at the processing point when forming the new molded layer; surface oxygen content of the metal additive manufacturing product at the processing point; laser absorption rate of the metal additive manufacturing product at the processing point; and predicted value of the evaporation amount of the low boiling point metal at the processing point.

10. The method for manufacturing a metal additive manufacturing product according to any one of claims 1 to 9, characterized in that, in the first evaporation amount prediction step, a trained model for the start of manufacturing is used to infer a predicted value of the evaporation amount of the low boiling point metal at the start of manufacturing from the wire surface state information obtained in the wire measurement step and the manufacturing condition information generated in the manufacturing condition generation step, thereby outputting a predicted value of the evaporation amount of the low boiling point metal at the start of manufacturing.

11. The method for manufacturing a metal additive manufacturing product according to any one of claims 1 to 10, characterized in that the wire surface state information includes at least one of the following: the surface oxygen content of the raw material wire, the laser absorptivity, and the boiling point of the low-boiling-point metal, the composition information of the low-boiling-point metal in the raw material wire, and the heat input information which combines the amount of heat input per unit volume to the raw material wire.

12. A method for manufacturing a metal additive manufacturing product according to any one of claims 1 to 11, characterized in that the surface oxygen content of the raw material wire is 5 wt% or less, and the difference in the composition of the low-boiling-point metal between the metal additive manufacturing product to be formed and the raw material wire is 10% or less.

13. A raw material wire supply unit that supplies a raw material wire made of an Al alloy containing a low-boiling-point metal which is a metal with a lower boiling point than Al; a laser beam irradiation unit that irradiates the raw material wire with a laser beam; a wire measurement unit that measures wire surface state information which is information related to the evaporation amount of the low-boiling-point metal in the raw material wire; a molded object information input unit that receives input of molded object information including the shape and material of the metal additive manufactured object to be manufactured; a molded condition generation unit that generates molded condition information which is the condition for manufacturing the metal additive manufactured object based on the molded object information; an evaporation amount prediction unit that predicts the amount of evaporation of the low-boiling-point metal when the metal additive manufactured object shown in the molded object information is manufactured according to the molded condition information, based on the wire surface state information and the molded condition information; and a determination processing unit that determines whether or not it is possible to manufacture according to the molded condition information based on whether or not the predicted amount of evaporation of the low-boiling-point metal is within a set tolerance range. A metal additive manufacturing apparatus comprising: a molding unit that irradiates a raw material wire supplied from a raw material wire supply unit with a laser beam from a laser beam irradiation unit according to the molding condition information to form a molded layer, wherein the molding condition generation unit outputs the molding condition information that has been determined to be molten by the determination processing unit to the molding unit.

14. The metal additive manufacturing apparatus according to 13, further comprising a molded object state information measuring unit that measures molded object state information indicating the state of the metal additive manufacturing object at the processing point of the new molded layer when a new molded layer is formed on the metal additive manufacturing object consisting of a molded layer formed in the molding unit, wherein the evaporation amount prediction unit predicts a predicted value of the evaporation amount of the low boiling point metal when the new molded layer is formed according to the molding condition information generated in the molding condition generation unit, using the wire surface state information and the molded object state information.

15. A metal additive manufacturing product comprising layers of fabricated material formed by irradiating a raw material wire made of an Al alloy containing a low-boiling-point metal, which is a metal with a lower boiling point than Al, with a laser beam applied to the wire, characterized in that the difference in the component concentration of the low-boiling-point metal between the first fabricated layer, which is the initial fabrication position of the metal additive manufacturing product, and any other fabricated layer other than the first layer, is 10% or less.

Images