Method of producing a three-dimensional object, control system and additive manufacturing apparatus

Abstract

A method of producing a three-dimensional object (28), the method comprising: providing an object model (96) of the three-dimensional object; providing, based on the object model, one or more candidate support structure models (98) of support structures (30) for the three-dimensional object; selecting (100), based on the object model, a removal strategy (42) for removing the one or more support structures from the three-dimensional object from among a plurality of candidate removal strategies; modifying (106) the candidate support structure models based on the selected removal strategy to provide modified support structure models (108); forming (112), by means of additive manufacturing, the three-dimensional object based on the object model and the one or more support structures supporting the three-dimensional object based on the modified support structure models; and removing (144) the one or more support structures from the three-dimensional object based on the selected removal strategy.

Description

Technical Field

[0001] This disclosure generally relates to additive manufacturing. Specifically, it provides a method for producing three-dimensional objects using additive manufacturing, a control system for additive manufacturing equipment, and additive manufacturing equipment for producing three-dimensional objects. Background Technology

[0002] Additive manufacturing (AM) (also known as 3D printing) is a manufacturing method implemented across various industries. When producing three-dimensional objects using additive manufacturing, support structures are typically used to support the 3D object, maintain its integrity during the build-up phase, and facilitate heat transfer. The support structures are printed in the same manner as the 3D object itself.

[0003] While the printing process itself is highly automated, many manual and delicate processes still involve additive manufacturing from raw materials to finished products. For example, support structures can be designed to improve functionality and reduce material consumption during the printing stage, but existing technologies do not offer any optimization for support structure removal. Most existing solutions require complex post-processing steps after the production of the 3D object. These steps may include removing support structures and smoothing the surface of the 3D object.

[0004] The support structure is typically relatively rigid and is usually removed manually using mechanical tools such as knives, pliers, or handheld rotary tools. Subsequent smoothing may include, for example, grinding or polishing. These post-processing steps are time-consuming and must be performed carefully, given the risk of damaging expensive 3D objects. Such post-processing steps are also not scalable. In some implementations, the cost exportable to post-processing can reach approximately one-third of the production cost.

[0005] WO 2019125970 A1 discloses an additive manufacturing system that uses a trained artificial intelligence module as part of a closed-loop control structure to adjust an initial set of build parameters during the process to improve part quality. The closed-loop control structure includes a slow control loop that considers images of the build layers during the process, and may include a fast control loop that considers melt pool monitoring data. Summary of the Invention

[0006] One object of this disclosure is to provide an improved method for producing three-dimensional objects using additive manufacturing.

[0007] Another object of this disclosure is to provide a method for producing three-dimensional objects by means of additive manufacturing, which improves the removal of support structures from three-dimensional objects.

[0008] Another object of this disclosure is to provide a method for producing three-dimensional objects using additive manufacturing, which can improve the quality of the three-dimensional objects.

[0009] Another object of this disclosure is to provide a cost-effective method for producing three-dimensional objects using additive manufacturing.

[0010] Another object of this disclosure is to provide a method for producing three-dimensional objects by means of additive manufacturing, which addresses some or all of the aforementioned objects in combination.

[0011] Another object of this disclosure is to provide a control system for additive manufacturing equipment that achieves one, several, or all of the foregoing objects.

[0012] Another object of this disclosure is to provide an additive manufacturing apparatus including a control system that addresses one, several, or all of the foregoing objects.

[0013] According to a first aspect, a method for producing a three-dimensional object is provided, the method comprising: providing an object model of the three-dimensional object; providing candidate support structure models for one or more support structures for the three-dimensional object based on the object model; selecting a removal strategy from a plurality of candidate removal strategies for removing one or more support structures from the three-dimensional object based on the object model; modifying the candidate support structure models based on the selected removal strategy to provide a modified support structure model; forming the three-dimensional object based on the object model and forming one or more support structures supporting the three-dimensional object based on the modified support structure models by means of additive manufacturing; and removing one or more support structures from the three-dimensional object based on the selected removal strategy.

[0014] This method can improve the design of support structures based on selected removal strategies (such as chosen removal techniques). By modifying candidate support structure models based on selected removal strategies to provide modified support structure models, the removal of the support structure can be improved and / or the size of the support structure can be reduced. Examples of improved removal include faster removal times and / or improved surface quality.

[0015] The object model may contain data defining the shape, size, texture, and / or material of a 3D object. The candidate support structure model and the modified support structure model may contain data defining the shape, size, texture, and / or material of one or more support structures. Each of the object model, candidate support structure model, and modified support structure model may be a CAD (Computer-Aided Design) model. The candidate support structure model may be provided based on one or more parameters of the object model.

[0016] The selection of a removal strategy can be automatic or manual. In any case, the selection of a removal strategy can be based on the object model. The selection of a removal strategy from multiple candidate removal strategies for removing one or more support structures from a 3D object can be additionally based on the candidate support structure model. In cases where the removal strategy is selected purely based on the candidate support structure model, the removal strategy is indirectly selected based on the object model, since the candidate support structure model is provided based on the object model. Therefore, the method includes selecting a removal strategy directly or indirectly based on the object model.

[0017] Modifying candidate support structure models to provide a modified support structure model may include modifying the shape of one or more of at least one support structure. According to one example, the modification may include modifying one or more interconnections between the support structure and a 3D object.

[0018] The method may also include adjusting the selected removal strategy. Adjustments can be made online and / or offline.

[0019] Three-dimensional objects and one or more supporting structures can be formed using various additive manufacturing techniques. Examples include selective laser sintering (SLS), selective laser melting (SLS), fused deposition modeling (FDM), stereolithography (SLA), and other material jetting techniques. The formation can include repeatedly forming a cured layer by irradiating predetermined portions of a powder layer with a light beam, allowing the powder to sinter or melt in the predetermined portions and subsequently solidify; and forming another cured layer by forming a new powder layer on the resulting cured layer and then irradiating predetermined portions of a metal powder layer with a light beam. The powder can include metals, ceramics, and / or plastics.

[0020] Support structures can be removed from a 3D object using various removal techniques, such as subtraction. Removing one or more support structures from a 3D object based on a selected removal strategy can include removing one or more support structures from the 3D object using either the selected removal strategy or a removal strategy adjusted based on the selected removal strategy.

[0021] The support structure is sacrificial; that is, it is not included in the target design of the 3D object.

[0022] The method may further include estimating one or more process values ​​associated with the selected removal strategy for a candidate support structure model; and modifying the candidate support structure model based on the one or more process values ​​to provide a modified support structure model. The process values ​​may be values ​​of various process parameters associated with the selected removal strategy.

[0023] The method may also include performing optimization of candidate support structure models to satisfy an objective function associated with one or more process values, thereby outputting an optimized support structure model; and using the optimized support structure model as a modified support structure model.

[0024] The selection of removal strategies and / or modification of candidate support structure models can be performed using machine learning (such as reinforcement learning). Machine learning can be performed using a machine learning system according to this disclosure. Furthermore, the selected removal strategy can be adjusted online and / or offline using machine learning.

[0025] The formation of a 3D object and the removal of one or more support structures can be performed using an additive manufacturing apparatus. In this case, the method may further include providing a digital twin of the additive manufacturing apparatus; and selecting a removal strategy and / or modifying candidate support structure models using the digital twin. The digital twin is a virtual model of the additive manufacturing apparatus. The digital twin may include one or more virtual components, each corresponding to a component of the additive manufacturing apparatus. The digital twin can be configured to render at least one virtual output parameter by simulating the selected removal strategy within the digital twin.

[0026] Candidate removal strategies can be associated with different removal trajectories. Therefore, two different candidate removal strategies can use the same removal technique but can be executed using different removal trajectories. As used herein, a removal trajectory includes a geometric path and optionally, velocity and acceleration profiles along that path. The removal trajectory is executed by, for example, an end effector of a robotic arm or a CNC (Computer Numerical Control) machine.

[0027] This method may, for example, include selecting a removal technique and removal trajectory from multiple candidate removal strategies with different removal trajectories. The method may also include adjusting the selected removal strategy by, for example, adjusting one or more process parameters based on a modified support structure model.

[0028] Candidate removal strategies can be associated with different tools. Therefore, two different candidate removal strategies can use the same removal trajectory but can be executed using different tools. Removal strategies executed using different tools are referred to herein as different removal techniques. The method may also include adjusting the selected removal strategy by modifying one or more removal trajectories and / or by adjusting one or more process parameters, for example, based on a modified support structure model.

[0029] Tools can include laser tools, jetting tools, machining tools, and / or vibratory tools. Jetting tools can be, for example, abrasive liquid jetting tools, such as abrasive water jetting tools or abrasive water suspension jets. Machining tools can be, for example, milling tools. Vibratory tools can be, for example, grippers including a vibration motor for vibrating the gripper. Three-dimensional objects can be held using grippers. Laser tools can be laser vaporization tools.

[0030] Interconnects can be provided between a 3D object and one or more supporting structures. Interconnects can have a thickness of less than 2 mm, such as less than 1 mm. In this way, the interconnects can be cut using laser vaporization cutting instead of laser melting cutting. Alternatively, or additionally, the interconnects can taper towards the 3D object.

[0031] Each support structure may include a body. The body may be wider than the associated interconnect.

[0032] Candidate removal strategies may include one or more vibration removal strategies, which include introducing vibration into one or more support structures to excite the eigenfrequency of the one or more support structures, thereby disconnecting one or more interconnections between the 3D object and each of the one or more support structures. The vibration may be induced in the 3D object and / or the substrate to be introduced into the one or more support structures.

[0033] According to a second aspect, a control system for an additive manufacturing apparatus is provided, the control system comprising at least one data processing device and at least one memory storing thereon at least one computer program, the at least one computer program including program code that, when executed by the at least one data processing device, causes the at least one data processing device to perform the following steps: providing an object model of a three-dimensional object; providing candidate support structure models for one or more support structures for the three-dimensional object based on the object model; providing a selection of a removal strategy for removing one or more support structures from the three-dimensional object from a plurality of candidate removal strategies based on the object model; modifying the candidate support structure models based on the selected removal strategy to provide a modified support structure model; instructing the formation of a three-dimensional object based on the object model by means of additive manufacturing and the formation of one or more support structures supporting the three-dimensional object based on the modified support structure models; and instructing the removal of one or more support structures from the three-dimensional object based on the selected removal strategy.

[0034] At least one computer program includes program code that, when executed by at least one data processing device, causes the at least one data processing device to perform or commands the various steps described herein. For example, at least one computer program may include program code that, when executed by at least one data processing device, causes the at least one data processing device to perform the step of selecting a removal strategy from a plurality of candidate removal strategies.

[0035] Commands to remove one or more support structures may include controlling a removal mechanism, such as a robotic arm or a CNC machine.

[0036] At least one computer program may include program code that, when executed by at least one data processing device, causes the at least one data processing device to perform the following steps: estimating one or more process values ​​associated with a selected removal strategy for a candidate support structure model; and modifying the candidate support structure model based on the one or more process values ​​to provide a modified support structure model.

[0037] At least one computer program may include program code that, when executed by at least one data processing device, causes the at least one data processing device to perform the following steps: optimizing a candidate support structure model to satisfy an objective function associated with one or more process values ​​to output an optimized support structure model; and using the optimized support structure model as a modified support structure model.

[0038] The selection of removal strategies and / or modification of candidate support structure models can be achieved using machine learning (such as reinforcement learning). For this purpose, a machine learning system can be implemented within the control system. The machine learning system may include a machine learning agent. The machine learning agent can be configured to autonomously select removal techniques, generate removal trajectories, and generate process values ​​for the selected removal strategy. The machine learning agent can be trained offline before the removal process and / or online during the removal process. In each case, training can be performed using a digital twin. The machine learning system may also include a database storing historical process values ​​from previous printing and / or removal processes.

[0039] The formation of a three-dimensional object and the removal of one or more support structures can be performed using additive manufacturing equipment. In this case, at least one computer program may include program code that, when executed by at least one data processing device, causes the at least one data processing device to perform the following steps: providing a digital twin of the additive manufacturing equipment; and selecting a removal strategy and / or modifying candidate support structure models based on the digital twin.

[0040] Candidate removal strategies can be associated with different removal trajectories. Alternatively, or additionally, candidate removal strategies can be associated with different tools.

[0041] Tools may include laser tools, jetting tools, machining tools, and / or vibratory tools. Laser tools may be laser vaporization tools.

[0042] Candidate removal strategies may include vibration removal strategies, which involve introducing vibrations into one or more support structures to excite the eigenfrequency of one or more support structures, thereby disconnecting the 3D object from one or more interconnections between each of the one or more support structures.

[0043] According to a third aspect, an additive manufacturing apparatus for producing three-dimensional objects is provided. The additive manufacturing apparatus includes a control system according to this disclosure. The additive manufacturing apparatus of the third aspect can be of any type as described in the first aspect, and vice versa.

[0044] Additive manufacturing equipment may include a printer and a removal machine. Each of the printer and removal machine may include, for example, a robotic arm or a CNC machine. Alternatively, each of the printer and removal machine may be carried by a conventional robotic arm or a conventional CNC machine. Additive manufacturing equipment may also include fixtures for holding the three-dimensional object during the removal of support structures.

[0045] According to a fourth aspect, a method is provided for removing one or more support structures from a three-dimensional object produced by additive manufacturing. The method includes forming the three-dimensional object and one or more support structures by means of additive manufacturing; subjecting the three-dimensional object to vibrations at intrinsic frequencies of each support structure to cause the support structures to disconnect from the three-dimensional object. The method may further include forming counterweights on the support structures by means of additive manufacturing. In the case of forming multiple support structures, at least one support structure may have a first intrinsic frequency, and at least one of the support structures may have a second intrinsic frequency different from the first intrinsic frequency. Attached Figure Description

[0046] Other details, advantages, and aspects of this disclosure will become apparent from the following description taken in conjunction with the accompanying drawings, in which:

[0047] Figure 1 : A schematic side view of an additive manufacturing equipment;

[0048] Figure 2a : A side view schematically illustrating an example of a support structure removal strategy using laser tools;

[0049] Figure 2b : A side view schematically illustrating yet another example of a support structure removal strategy using laser tools;

[0050] Figure 2c : A perspective view schematically representing yet another example of a supporting structure;

[0051] Figure 2d : A perspective view schematically representing yet another example of a supporting structure;

[0052] Figure 2e : A perspective view schematically representing yet another example of a supporting structure;

[0053] Figure 2f : A side view schematically illustrating yet another example of a supporting structure;

[0054] Figure 3a : A side view schematically illustrating yet another example of a support structure removal strategy using a jetting tool;

[0055] Figure 3b : A side view schematically illustrating yet another example of a support structure removal strategy using a jetting tool;

[0056] Figure 4 : A side view schematically illustrating yet another example of a support structure removal strategy using machining tools;

[0057] Figure 5 : A side view schematically illustrating yet another example of a support structure removal strategy using vibration tools;

[0058] Figure 6a : A bottom view schematically illustrating yet another example of a support structure removal strategy that uses a movement trajectory;

[0059] Figure 6b : A bottom view schematically illustrating yet another example of a support structure removal strategy using a movement trajectory; and

[0060] Figure 7 : A schematic diagram representing a machine learning system. Detailed Implementation

[0061] The following describes a method for producing three-dimensional objects using additive manufacturing, a control system for additive manufacturing equipment, and additive manufacturing equipment for producing three-dimensional objects. The same or similar reference numerals will be used to denote the same or similar structural features.

[0062] Figure 1A schematic side view of additive manufacturing equipment 10 is shown. Additive manufacturing equipment 10 includes a printer 12. The printer 12 in this specific example includes a printing robot 14 having a printhead 16 (such as a laser source), a material reservoir 18, a delivery piston 20 in the material reservoir 18, a production chamber 22, a substrate 24 in the production chamber 22, and a leveling mechanism 26. New material (illustrated here as metal powder) can be introduced into the production chamber 22 by moving the delivery piston 20 upwards and moving the leveling mechanism 26.

[0063] exist Figure 1 In the process, a 3D object 28 and a support structure 30 for the 3D object 28 are being printed in a production chamber 22 using additive manufacturing. The support structure 30 supports the 3D object 28 during the printing process and conducts heat away from the 3D object 28. In this example, the support structure 30 is formed as an elongated pillar vertically below the 3D object 28.

[0064] To print the 3D object 28 and the support structure 30, the object model of the 3D object 28 and the support structure model of the support structure 30 are segmented into two-dimensional layers and then transformed into a set of instructions for execution by the print head 16. The support structure 30 and the 3D object 28 are then formed by adding material one layer at a time. The support structure 30 and the 3D object 28 can be produced, for example, by means of selective laser sintering (SLS), selective laser melting (SLS), fused deposition modeling (FDM), stereolithography (SLA), and other material jetting techniques.

[0065] The additive manufacturing apparatus 10 in this example also includes a removal robot 32. Figure 1 In this process, robot 32 carries a laser tool 34a. With the aid of the laser tool 34a, the support structure 30 can be removed from the three-dimensional object 28, as described below. Figure 1 As illustrated, the additive manufacturing equipment 10 also includes a jetting tool 34b, a machining tool 34c, and a vibrating tool 34d. A removal robot 32 is configured to automatically replace the laser tool 34a with any one of the jetting tool 34b, machining tool 34c, and vibrating tool 34d.

[0066] Manual post-processing to remove the 3D object 28 from the support structure 30 is labor-intensive and not scalable. However, the most suitable automatic removal strategy depends largely on the details of the 3D object 28 and the support structure 30.

[0067] In this example, each of the printing robot 14 and the removal robot 32 includes a robotic arm programmable on three or more axes, such as six or seven axes. One or both of the printing robot 14 and the removal robot 32 can be replaced, for example, by a CNC machine.

[0068] The additive manufacturing equipment 10 also includes a control system 36. The control system 36 includes a data processing device 38 and a memory 40. The memory 40 stores thereon a computer program that, when executed by the data processing device 38, causes the data processing device 38 to perform and / or commands the various steps described herein. Figure 1 As shown, the control system 36 of this example communicates with the printing robot 14, the removal robot 32, and the leveling mechanism 26.

[0069] Figure 2a A side view schematically illustrating an example of a support structure removal strategy 42a using laser tool 34a. Figure 2a An example of a three-dimensional object 28 and a support structure 30 is further shown. Although only one support structure 30 is shown, multiple such support structures 30 can be provided for the three-dimensional object 28.

[0070] The support structure 30 in this example includes a body 44 and an interconnect 46 between the body 44 and the three-dimensional object 28. The body 44 is wider than the associated interconnect 46. The interconnect 46 has a thickness 48. The body 44 is illustrated herein as a cylinder. The interconnect 46 is illustrated herein as a cylinder concentric with the body 44 and with a smaller diameter than the body 44. Therefore, the interconnect 46 has a reduced thickness relative to the body 44. The interconnect 46 is formed to provide the intended break point between the support structure 30 and the three-dimensional object 28. The interconnect 46 may have a maximum width of 2 mm or less, such as 1 mm or less.

[0071] Removal strategy 42a involves cutting the interconnect 46 by means of a laser beam 50 irradiated by a lens 52 from a laser tool 34a. Thus, the interconnect 46 is removed by subtraction.

[0072] Figure 2a Further schematically illustrated are examples of multiple process values ​​54 of the process parameters associated with the laser tool 34a when performing removal strategy 42a. Figure 2a In the process parameters, the laser tool 34a's execution speed 56, the distance between the laser tool 34a and the support structure 30 58, the focal length 60, and the intensity of the laser tool 34a 62 are included.

[0073] In removal strategy 42a, interconnect 46 is cut by laser vaporization. Laser tool 34a provides a high-energy-density laser beam 50 to heat interconnect 46. The temperature in interconnect 46 rises rapidly and reaches the boiling point of the metal within a very short time. The metal then begins to vaporize and form vapor. This vapor is ejected at very high speeds, forming cuts in the metal as it is ejected. The heat of vaporization of metal is typically large, therefore laser vaporization cutting requires high power and power density. To do this economically, the amount of metal to be removed must be as small as possible, and sufficient free (air, gas) volume must be available around the cutting point.

[0074] In removal strategy 42a, the process parameters of the laser tool 34a are adjusted so that metal evaporation occurs only within the Rayleigh length from the waist of the laser beam 50. Removing the support structure 30 using the laser tool 34a constitutes an example of the removal technique according to this disclosure. After removing the support structure 30 using the laser tool 34a, the laser tool 34a can be used for surface treatment of the three-dimensional object 28, such as polishing.

[0075] Figure 2b A side view schematically illustrating another example of support structure removal strategy 42b using laser tool 34a. The description will be relative to... Figure 2a The main differences. Figure 2b In this example, the support structure 30 includes a body 44 and multiple interconnects 46 between the body 44 and the three-dimensional object 28. The body 44 in this example is a relatively large and flat cube. The interconnects 46 in this example are parallel cylinders. The interconnects 46 form a non-interlaced perforated pattern for the laser tool 34a.

[0076] Figure 2c A perspective view schematically illustrating another example of the support structure 30 is shown. This example support structure 30 includes an elongated cubic body 44 and an elongated tapered interconnect 46 on top of the body 44. The body 44 is horizontally oriented. Furthermore, this example support structure 30 is thinned to facilitate laser vaporization cutting.

[0077] Figure 2d A perspective view schematically illustrating another example of a support structure 30. This example support structure 30 includes a cylindrical body 44 and an interconnect 46 in the form of a truncated cone. The interconnect 46 includes a base having the same diameter as the body 44. A bridge to the top of the three-dimensional object 28 has a diameter much smaller than the base.

[0078] Figure 2e A perspective view schematically illustrating another example of a support structure 30. This example of a support structure 30 includes a vertically oriented elongated cubic body 44 and interconnections 46 in the form of a truncated pyramid located at the top of the body 44.

[0079] Figure 2f A side view schematically illustrating another example of the support structure 30. In this example, the support structure 30 is a grid. The upper portions of the grid lines form interconnections 46 here.

[0080] Figure 3a A side view schematically illustrating yet another example of support structure removal strategy 42c using spray tool 34b. The description will be relative to... Figure 2a The main difference is that the jetting tool 34b can be, for example, an abrasive water jetting tool. The removal strategy 42c involves propelling an abrasive liquid jet 64 containing abrasive particles from the nozzle 66 of the jetting tool 34b to the interconnect 46. The interconnect 46 is thus cut by the liquid jet 64 to separate the support structure 30 from the three-dimensional object 28.

[0081] Figure 3a Examples of multiple process values ​​54 of the process parameters associated with the jetting tool 34b are further illustrated schematically when the removal strategy 42c is executed. In addition to the execution speed 56 of the jetting tool 34b and the distance 58 between the jetting tool 34b and the support structure 30, the process parameters of the removal strategy 42c also include the pressure 68 of the liquid jet 64 and the jet diameter 70.

[0082] One advantage associated with the jetting tool 34b is that the spacing 58 can be relatively large, such as tens of millimeters, and still achieve a considerable cutting depth. The shape of the support structure 30 and / or the spacing 58 can be modified so that the shape of the support structure 30 matches the cutting shape produced by the jetting tool 34b.

[0083] Removing the support structure 30 by cutting the interconnect 46 from the jetting tool 34b using a liquid jet 64 constitutes another example of the removal technique according to the present disclosure. After removing the support structure 30 by the jetting tool 34b, the jetting tool 34b can be used for surface treatment of the three-dimensional object 28 by means of the liquid jet 64, such as polishing.

[0084] Figure 3b A side view schematically illustrating yet another example of support structure removal strategy 42d using jetting tool 34b. The description will be relative to... Figure 3a The main difference is that the support structure 30 in this example includes interconnects 46 that taper towards the three-dimensional object 28.

[0085] Instead of cutting the interconnect 46, removal strategy 42d includes advancing a liquid jet 64 toward the body 44, thereby pushing the body 44 to disconnect the interconnect 46. Figure 3bAs shown, the body 44 is relatively wide to provide a larger surface area for impact by the liquid jet 64. This larger surface area of ​​the body 44 makes it easier for the liquid jet 64 to hit its target. In this way, the positioning accuracy of the liquid jet 64 on the body 44 can be relaxed compared to positioning the liquid jet 64 on a smaller interconnect 46. Furthermore, the pressure 68 of the liquid jet 64 can be reduced when pushing the support structure 30 away from the three-dimensional object 28 compared to cutting the interconnect 46. When the support structure 30 is pushed in this way, the liquid from the jetting tool 34b does not necessarily contain abrasive particles.

[0086] Removing the support structure 30 by means of a liquid jet 64 from the spraying tool 34b constitutes another example of the removal technique according to the present disclosure.

[0087] Figure 4 A side view schematically illustrating another example of a support structure removal strategy 42e using machining tool 34c. Machining tool 34c is illustrated here as a milling tool. Removal strategy 42e involves machining the interconnect 46 to separate the support structure 30 from the three-dimensional object 28.

[0088] Figure 4 Examples of multiple process values ​​54 of process parameters associated with the machining tool 34c during the execution of removal strategy 42e are further illustrated. In addition to the execution speed 56, the process parameters of removal strategy 42e also include the rotational speed 72 of the machining tool 34c. Removing the support structure 30 by means of the machining tool 34c constitutes yet another example of the removal technique according to this disclosure.

[0089] Figure 5 A side view schematically illustrating another example of a support structure removal strategy 42f using vibration tool 34d. Removal strategy 42f constitutes an example of a vibration removal strategy according to this disclosure.

[0090] In this example, the three-dimensional object 28 is provided with three support structures 30a, 30b, and 30c. Alternatively, one, several, or each of the support structures 30a to 30c may be referred to by the reference numeral "30".

[0091] like Figure 5As shown, the first support structure 30a includes a first body 44a and a first counterweight 74a, the second support structure 30b includes a second body 44b and a second counterweight 74b, and the third support structure 30c includes a third body 44c and a third counterweight 74c. The counterweights 74a to 74c are printed together with the associated bodies 44a to 44c of the support structures 30a to 30c. Alternatively, one, several, or each of the bodies 44a to 44c may be designated by the reference numeral "44". Alternatively, one, several, or each of the counterweights 74a to 74c may be designated by the reference numeral "74".

[0092] The first counterweight 74a is larger than the second counterweight 74b and the third counterweight 74c. The third counterweight 74c is positioned closer to the three-dimensional object 28 than the second counterweight 74b. For this reason, each support structure 30a to 30c is provided with a unique intrinsic frequency. By causing the three-dimensional object 28 to vibrate at a frequency and direction that triggers one of these intrinsic frequencies, the associated support structures 30a to 30c can be detached from the three-dimensional object 28.

[0093] The vibration tool 34d is illustrated herein as a clamp including a vibration motor 76. The vibration motor 76 in this example includes actuators (not shown) on three orthogonal axes to allow triggering of the intrinsic frequencies of the support structures 30a to 30c in all possible directions. The actuators may be piezoelectric actuators.

[0094] Furthermore, each support structure 30a to 30c is connected to the substrate 24 via a corresponding substrate interconnect 78, and is connected to the three-dimensional object 28 via a corresponding interconnect 46. For example... Figure 5 As shown, the thickness of interconnect 46 is greater than the thickness of substrate interconnect 78. The vibratory tool 34d can thus disconnect the corresponding substrate interconnect 78 before interconnect 46 is disconnected.

[0095] Figure 5 The illustration further illustrates examples of multiple process values ​​54 of the process parameters associated with the vibration tool 34d when performing removal strategy 42f. The process parameters of removal strategy 42f include vibration frequency 80, vibration direction 82, and vibration amplitude 84.

[0096] Removal strategy 42f involves clamping the three-dimensional object 28 and subjecting it to vibration at a frequency 80 having a corresponding intrinsic frequency that triggers the support structures 30a to 30c and a relatively low amplitude 84. In this way, the substrate interconnect 78 can be disconnected. A vibration tool 34d can then lift the three-dimensional object 28, with the support structures 30a to 30c attached thereto, away from the substrate 24. The vibration tool 34d can then subject the three-dimensional object 28 to vibration at a frequency 80 having a corresponding intrinsic frequency that triggers the support structures 30a to 30c and a relatively high amplitude 84 to disconnect the interconnect 46 from the three-dimensional object 28. One advantage of this is that when the support structures 30a to 30c are disconnected from the three-dimensional object 28 by vibration, the three-dimensional object 28 can be lifted out of the production chamber 22, for example, positioned above the processing bay.

[0097] Removing the support structure 30 by means of a vibratory tool 34d constitutes another example of the removal technique according to this disclosure. In an alternative example, a vibratory motor 76 is positioned in the substrate 24. Alternatively, one, several, or each of the tools 34a to 34d is designated by the reference numeral "34".

[0098] Using each of the removal strategies 42a to 42f, the support structure 30 is removed from the 3D object 28 with great accuracy, resulting in improved surface quality of the 3D object 28 at the location of the interconnect 46. The automated removal of the support structure 30 also significantly reduces the costs associated with additive manufacturing.

[0099] Figure 6a The diagram schematically illustrates a bottom view of yet another example of a support structure removal strategy 42g using removal trajectory 86a. Removal trajectory 86a can be used with each of the removal techniques in removal strategies 42a to 42f. Figure 6a In this process, the support structure 30 is positioned within a matrix. The removal trajectory 86a involves the sequential removal of the support structure 30, typically along a spiral path. When the removal strategy 42g is executed, the removal robot 32 follows the path of the removal trajectory 86a.

[0100] Figure 6b The bottom view schematically illustrates yet another example of support structure removal strategy 42h using removal trajectory 86b. Removal trajectory 86b can be used in each of the removal strategies 42a to 42f. Removal trajectory 86b comprises the sequential removal of support structure 30, typically along a zigzag path.

[0101] One or both of removal trajectories 86a and 86b may alternatively be designated by the reference numeral "86". By modifying removal trajectory 86, access to interconnect 46 can be improved. Alternatively, one, some, or all of removal strategies 42a to 42h may also be designated by the reference numeral "42".

[0102] As detailed above, various designs and removal strategies 42 for the support structure 30 are available. Therefore, manually finding the most suitable support structure 30 and the most suitable removal strategy 42 is extremely challenging. A manually customized removal strategy 42 may work well for one implementation but not for another. Due to the large number of different designs, manually customized removal strategies 42 are therefore inconvenient and expensive. Furthermore, implementing a custom removal strategy 42 generated by an expert does not guarantee the quality of the 3D object 28.

[0103] Figure 7 A schematic diagram illustrating an example of a machine learning system 88 and the various steps in a method for producing a 3D object 28 is provided. The machine learning system 88 can, for example, be implemented in a control system 36. This specific example of the machine learning system 88 includes a digital twin 90, a machine learning agent 92, and a database 94.

[0104] Digital twin 90 is a virtual model representing additive manufacturing equipment 10 or one or more sections thereof. Specifically, digital twin 90 can simulate the interaction between removal robot 32 and 3D object 28 and support structure 30. Digital twin 90 can also model tools 34, 3D object 28, vision system, pliers, fixtures, etc.

[0105] The digital twin 90 includes multiple virtual components corresponding to multiple mechanical and / or electrical components of the additive manufacturing equipment 10. The digital twin 90 takes into account the effects of the respective virtual components and their interactions with other virtual components. The digital twin 90 mimics the geometry and dynamic behavior of the environment encountered during the removal of the support structure 30. The digital twin 90 can perform various simulations, such as different removal strategies 42, and provide one or more virtual output parameters from such simulations. The virtual output parameters may, for example, correspond to process values ​​54. The virtual output parameters may correspond to process values ​​54 measured by sensors on the additive manufacturing equipment 10, and / or may correspond to process values ​​54 not measured by sensors on the additive manufacturing equipment 10.

[0106] Database 94 stores virtual output parameters from digital twin 90 and process values ​​54 from additive manufacturing equipment 10. In this way, experience regarding the behavior of additive manufacturing equipment 10 can be acquired. Database 94 may, for example, contain information about the optimal profile of the support structure 30 for different types of printing materials and / or the abrasive parameterization of the jetting tool 34b. Database 94 records past successful process values ​​54 so that they can be reused when similar needs arise and / or synthesized into new process values ​​54 that can be used to initialize machine learning agent 92.

[0107] Machine learning agent 92 employs algorithms to automatically build and improve mathematical models through experience and the use of sample data. Machine learning agent 92 is trained using sample data. Sample data may include, for example, virtual output parameters from digital twin 90 and / or information from database 94. Machine learning agent 92 is configured to learn and predict optimal removal techniques, optimal designs for support structure 30, optimal process values ​​54, and optimal removal trajectories 86 for different designs of 3D object 28 and associated support structure 30.

[0108] The machine learning agent 92 and the digital twin 90 operate in sync, refining the removal process until it meets some pre-specified performance criteria, such as accuracy, execution speed 56, computation time, or surface roughness. The machine learning agent 92 is configured to supply input data to the digital twin 90. Examples of input data include removal trajectories 86, process values ​​54, an object model of the 3D object 28, and candidate support structure models of the support structure 30. The digital twin 90 is configured to react dynamically to the input data from the machine learning agent 92.

[0109] Digital twin 90 can query with varying levels of accuracy. For example, digital twin 90 can simulate the interaction of the tip of machining tool 34c with higher accuracy than a vision system, which can simply rely on the geometry of 3D object 28 and support structure 30 regardless of the material. Digital twin 90 can provide information about the execution of a specific removal strategy 42 and its process value 54. Digital twin 90 can, for example, receive a binary query about whether machining tool 34c collides with 3D object 28 or support structure 30. The corresponding binary output parameters from digital twin 90 can be used by machine learning agent 92 as a reward (if machining tool 34c does not collide) or a penalty (if machining tool 34c collides). A further example of a query from machine learning agent 92 to digital twin 90 is the time taken to execute full removal strategy 42 and the accuracy of removing support structure 30. The corresponding response output from digital twin 90 can be used in the reward / penalty function of machine learning agent 92.

[0110] The method includes step 96 of providing an object model of the three-dimensional object 28. The object model may also be referred to by the reference numeral "96". The object model 96 may be a CAD model of the three-dimensional object 28 and may contain information about its materials.

[0111] The method also includes a step 98 of providing a candidate support structure model. The candidate support structure model can also be referred to by the reference numeral "98". The candidate support structure model 98 is a model of a support structure 30 used to support the three-dimensional object 28. The candidate support structure model 98 is provided based on the object model 96. For example, for a larger three-dimensional object 28, more support structures 30 are typically required. The candidate support structure model 98 can be provided by a machine learning agent 92. Furthermore, the candidate support structure model 98 can be a CAD model of the support structure 30 and can include information about its materials. Step 98 may also include determining the optimal orientation of the three-dimensional object 28 during additive manufacturing.

[0112] The method also includes a step 100 of selecting a removal strategy 42 from a plurality of candidate removal strategies 42 (such as removal strategies 42a to 42h). The selection of removal strategy 42 is performed here based on object model 96 and candidate support structure model 98. In this example, the selection of removal strategy 42 is performed by machine learning agent 92. The selection of removal strategy 42 may include first selecting the type of removal technique and subsequently generating a removal trajectory 86 and / or process value 54 for the selected removal technique.

[0113] For a specific object model 96 of the 3D object 28, one removal technique may be more suitable than another. For example, this may be related to the accessibility due to factors such as the size of the tool 34, the material strength of the 3D object 28, the structural stiffness of the 3D object 28 (if the material can be wetted), etc. Furthermore, depending on the center of gravity of the 3D object 28 determined based on the object model 96, one removal trajectory 86 may be more suitable, for example, than another. As a further example, the required intensity 62 of the laser tool 34a, the required pressure 68 of the jet tool 34b, the required rotational speed 72 of the machining tool 34c, or the required frequency 80 of the vibrating tool 34d can be considered in light of the object model 96 of the 3D object 28.

[0114] The method also includes step 102, which estimates one or more process values ​​54 of process parameters associated with the selected removal strategy 42 of the candidate support structure model 98. If process values ​​54 have already been generated in step 100, step 102 may include estimating additional process values ​​54. In any case, the digital twin 90 may be used for such estimation. Estimation may include performing the selected removal strategy 42 in the digital twin 90 and collecting rewards / penalties from the digital twin 90.

[0115] The method in this example also includes step 104, which modifies the selected removal strategy 42 based on the estimate in step 102. The modification can be performed using a machine learning agent 92 or through an optimization algorithm.

[0116] The modified removal strategy 42 is then provided as offline feedback to step 100. Steps 100 and 102 can then be repeated using the modified removal strategy 42. Step 104 can also be repeated. Thus, based on the initially selected removal strategy 42, which includes a specific removal trajectory 86 and process value 54 (both of which can be conservatively selected, pre-programmed, or obtained through any other heuristic), the selected removal strategy 42 can be progressively improved.

[0117] The method also includes step 106 of modifying the candidate support structure model 98 based on the selected removal strategy 42 to provide a modified support structure model 108. This modification may be performed by a machine learning agent 92. Alternatively or additionally, the modification may include optimization of the candidate support structure model 98, for example, to satisfy an objective function associated with the process value 54, thereby providing an optimized support structure model that can then be used as the modified support structure model 108.

[0118] The modified support structure model 108 may differ from the candidate support structure model 98, for example, in the number of support structures 30, the positioning of the support structures 30 relative to the 3D object 28, and / or the design of the support structures 30 (such as interconnects 46). Modifying the candidate support structure model 98 to provide the modified support structure model 108, based on rewards / penalties collected from the digital twin 90, can be performed by the machine learning agent 92.

[0119] The method also includes a step 110 of adjusting the selected removal strategy 42 based on the modified support structure model 108. Moreover, this modification of the selected removal strategy 42 can be performed by a machine learning agent 92 based on rewards / penalties collected from the digital twin 90. This process is deployed to the additive manufacturing apparatus 10 when the machine learning agent 92 fully executes the modified support structure model 108 and the selected removal strategy 42 given one or more predetermined criteria (e.g., from the user).

[0120] The method also includes an additive manufacturing step 112. In step 112, the support structure 30 is formed based on a modified support structure model 108, and the three-dimensional object 28 is formed based on an object model 96.

[0121] The method also includes step 114 of removing the support structure 30 from the 3D object 28 based on a selected removal strategy 42, optionally as adjusted in step 110. The removal strategy 42 is thus deployed to a real-world environment, here in the case of the removal robot 32.

[0122] The method further includes step 116, which involves online inspection of the support structure 30 and / or the 3D object 28 during the removal of the support structure 30 in step 114. The online inspection in step 116 may include sensory feedback from the additive manufacturing equipment 10. This sensory feedback may be sent to a digital twin 90, as indicated by arrow 118. The digital twin 90 provides this feedback to a machine learning agent 92. The online feedback is then provided to the removal process in step 114, as indicated by arrow 120. Therefore, during the removal of the support structure 30, the machine learning agent 92 can passively learn from the actual performance of the removal robot 32 and can adjust the removal strategy 42 in real time.

[0123] The method also includes a final inspection step 122. The method also includes a step 124 of updating the database 94. For example, sensory feedback from the additive manufacturing equipment 10 during removal step 114 can be sent to and stored in the database 94.

[0124] Removal strategy 42 can be performed in batches of printed 3D objects 28. In the first batch, removal strategy 42 is executed, and sensory feedback is used to improve the digital twin 90. In this way, the discrepancy between the expected behavior of the removal robot 32 and the digital twin 90 is reduced by changing the model adopted by the digital twin 90.

[0125] In the second batch of printed 3D objects 28, removal strategy 42 is executed, and the reward / penalty function of machine learning agent 92 is used to improve removal strategy 42. Machine learning agent 92 continues to improve digital twin 90 and removal strategy 42 until removal strategy 42 is satisfactory given some predefined criteria. Digital twin 90 is thus continuously or periodically updated and / or retuned to better match the behavior of removal robot 32. If the training processing steps do not damage 3D objects 28, a batch of printed 3D objects 28 used for training purposes can be reprocessed when removal robot 32 has reached a satisfactory performance level. With the help of machine learning system 88, the removal of support structure 30 can be automated and improved without requiring experts to tune the parameterization of additive manufacturing equipment 10.

[0126] Database 94 contains all historical information, such as process parameters for different design options. When the same or similar design options are present, the same process value 54 is used for the initialization of the machine learning agent 92. When different design options are present, database 94 infers the most likely process value 54 and uses this parameterization for the initialization of the machine learning agent 92. Database 94 can be updated offline using computing resources to make more efficient use of the control system 36.

[0127] While this disclosure has been described with reference to exemplary embodiments, it is to be understood that the invention is not limited to what has been described above. For example, it is to be understood that the dimensions of the parts can be varied as needed. Therefore, the invention is intended to be limited only by the scope of the appended claims.

Claims

1. A method for producing a three-dimensional object (28), the method comprising: - Provide the object model (96) of the three-dimensional object (28); - Based on the object model (96), provide one or more candidate support structure models (98) for the three-dimensional object (28) and support structures (30). - Based on the object model (96), select (100) a removal strategy (42) from a plurality of candidate removal strategies (42) for removing the one or more support structures (30) from the three-dimensional object (28). - For the candidate support structure model (98), estimate (102) one or more process values ​​(54) associated with the selected removal strategy (42), wherein the one or more process values ​​(54) include values ​​of various process parameters associated with the selected removal strategy (42); - Modify (106) the candidate support structure model (98) based on the one or more process values ​​(54) to provide a modified support structure model (108). - Optimize the candidate support structure model (98) to satisfy the objective function associated with the one or more process values ​​(54) to output an optimized support structure model; as well as - Use the optimized support structure model as the modified support structure model (108). - Using additive manufacturing, the three-dimensional object (28) is formed (112) based on the object model (96), and the one or more support structures (30) supporting the three-dimensional object (28) are formed based on the modified support structure model (108); and - Based on the selected removal strategy (42), remove (144) the one or more support structures (30) from the three-dimensional object (28).

2. The method according to any one of the preceding claims, wherein the selection (100) of the removal strategy (42) and / or the modification of the candidate support structure model (98) are performed by means of machine learning.

3. The method according to any one of the preceding claims, wherein the formation (112) of the three-dimensional object (28) and the removal (114) of the one or more support structures (30) are performed by means of additive manufacturing equipment (10), and wherein the method further comprises: - Provide a digital twin (90) of the additive manufacturing equipment (10); as well as - Select (100) the removal strategy (42) and / or modify (106) the candidate support structure model (98) by using the digital twin (90).

4. The method according to any one of the preceding claims, wherein the candidate removal strategy (42) is associated with different removal trajectories (86).

5. The method according to any one of the preceding claims, wherein the candidate removal strategy (42) is associated with a different tool (34).

6. The method according to claim 5, wherein the tool (34) comprises a laser tool (34a), a jetting tool (34b), a machining tool (34c), and / or a vibration tool (34d).

7. The method according to any one of the preceding claims, wherein the candidate removal strategy (42) comprises one or more vibration removal strategies (42f), the one or more vibration removal strategies (42f) comprising introducing vibrations into one or more support structures in the support structure (30) to excite the eigenfrequency of the one or more support structures (30) such that one or more interconnections (46) between the three-dimensional object (28) and each of the one or more support structures (30) are disconnected.

8. A control system (36) for an additive manufacturing apparatus (10), the control system (36) comprising at least one data processing device (38) and at least one memory (40), wherein at least one computer program is stored in the at least one memory (40), the at least one computer program comprising program code, the program code, when executed by the at least one data processing device (38), causing the at least one data processing device (38) to perform the following steps: - Provides an object model (96) for three-dimensional objects (28); - Based on the object model (96), provide one or more candidate support structure models (98) for the three-dimensional object (28) and support structures (30). - Based on the object model (96), a selection (100) of a plurality of candidate removal strategies (42) for removing the one or more support structures (30) from the three-dimensional object (28) is provided. - For the candidate support structure model (98), estimate (102) one or more process values ​​(54) associated with the selected removal strategy (42), wherein the one or more process values ​​(54) include values ​​of various process parameters associated with the selected removal strategy (42); - Modify (106) the candidate support structure model (98) based on the one or more process values ​​(54) to provide a modified support structure model (108). - Optimize the candidate support structure model (98) to satisfy the objective function associated with the one or more process values ​​(54) to output an optimized support structure model; as well as - Use the optimized support structure model as the modified support structure model (108). - The command uses additive manufacturing to form (112) the three-dimensional object (28) based on the object model (96), and to form one or more support structures (30) supporting the three-dimensional object (28) based on the modified support structure model (108); and - The command removes one or more support structures (30) from the three-dimensional object (28) based on the selected removal strategy (42).

9. The control system (36) according to claim 8, wherein the selection (100) of the removal strategy (42) and / or the modification of the candidate support structure model (98) are performed by means of machine learning.

10. The control system (36) according to claim 8 or 9, wherein the formation (112) of the three-dimensional object (28) and the removal of the one or more support structures (30) are performed by means of an additive manufacturing apparatus (10), and wherein the at least one computer program includes program code that, when executed by the at least one data processing device (38), causes the at least one data processing device (38) to perform the following steps: - Provide a digital twin (90) of the additive manufacturing equipment (10); and - Select (100) the removal strategy (42) and / or modify (106) the candidate support structure model (98) by using the digital twin (90).

11. An additive manufacturing apparatus (10) for producing three-dimensional objects (28), the additive manufacturing apparatus (10) comprising a control system (36) according to any one of claims 8 to 10.

Images