Additive manufacturing build simulation data structure, additive manufacturing build simulation method, computer program, computer readable storage medium and apparatus

Abstract

The invention provides an additive manufacturing build simulation data structure, an additive manufacturing build simulation method, a computer program, a computer readable storage medium and a device. The data structure is assembled into a finite element analysis of the mechanical properties of an object to be additively manufactured, wherein the data structure comprises a matrix model representing an elasticity tensor, wherein the matrix elements of the elasticity tensor comprise material properties consisting of Young's modulus, shear modulus and / or Poisson's ratio of the object and a model function independent of the material properties, wherein the function takes into account a set of build process parameters, machine settings, voxel volume fill and / or lattice type used for the build simulation. The additive manufacturing build simulation method is for simulating the manufacturing of an object, which comprises (i) accessing the data structure, and (ii) performing the build simulation by executing a finite element algorithm based on the entries of the data structure, wherein all material properties are defined as constants in the matrix and only the elements of the model function are calibrated for a given material.

Description

Technical Field

[0001] This invention relates to an additive manufacturing build simulation data structure and a method for accessing and / or utilizing this new data structure for additive manufacturing build simulation. Furthermore, computer programs, computer-readable storage media, and apparatuses including said computer-readable storage media and / or including means for executing computer programs also belong to this invention. Background Technology

[0002] Additive manufacturing (AM), particularly powder bed methods, has proven useful and advantageous in the fabrication of prototypes or complex components, such as those with filamentary structures or functional cooling systems. Furthermore, additive manufacturing stands out due to its short process chain, which in turn enables material savings and exceptionally low lead times.

[0003] Additive manufacturing or 3D printing technologies include, for example, powder bed fusion methods such as selective laser melting (SLM) or laser powder bed fusion (LPBF), selective laser sintering (SLS), and electron beam melting (EBM). Furthermore, additive manufacturing methods involve "directional energy deposition (DED)," such as laser cladding, electron beam welding or plasma welding, metal inkjet forming (MIM), sheet lamination methods, or even thermal spraying (VPS, LPPS) methods.

[0004] The associated machine hardware for LPBF typically includes a manufacturing or building platform on which components are built layer by layer after a base material layer is supplied. The base material layer can then be melted, for example, by an energy beam from a laser, and subsequently cured. Layer thickness is determined by a recoater that automatically moves across the powder bed and removes excess material from the manufacturing plane or space. Typical layer thicknesses are, for example, between 20 μm and 40 μm. During manufacturing, the energy beam scans the surface and melts powder in selected areas predetermined by a CAD file based on the geometry of the object to be manufactured. This scanning or irradiation is preferably performed in a computer-aided manner, such as computer-aided manufacturing (CAM) instructions that can exist in the form of a dataset. The dataset or CAM file can be a computer program or a computer program product, or refers to a computer program or computer program product.

[0005] The geometric complexity of objects manufactured in this way usually requires thorough construction simulation and preparation.

[0006] For example, it is evident that overhanging object features require both thermal and mechanical support from a support structure, typically constructed in the form of a lattice. This reduces manufacturing time and material usage for the supports while also facilitating subsequent removal of the object, such as by cutting the supports. Although the support structure is not usually part of the object itself, its design is crucial because it has a significant impact on the success of the construction process.

[0007] To save materials, time, and cost, low-mass support structures are targeted. These low-mass support structures, or lattice support structures, inherently involve complex designs that require simulation and validation using so-called additive manufacturing. Therefore, accurate prediction of all emerging mechanical properties is necessary. Precise parameters are obtained by calibrating the materials using a large number of input variables. Correct calibration and application are consistently time-consuming and costly.

[0008] For example, EP 4 287 060 A1 describes a method for additive manufacturing of components with topology-optimized support structures based on thermo-mechanical process simulation.

[0009] Therefore, challenging AM designs and high-precision object construction settings are executed with the help of construction process simulation. Typically, the construction process of parts and supports is first simulated through initial guesses or experiments on the entire setup. Based on the results obtained, the orientation of the parts, materials, machine settings used, process parameters, support topology, support lattice, and / or pre-deformation are updated in several iterations until a feasible state is achieved.

[0010] AM (Ambimetrics) simulation methods typically employ so-called finite element analysis and multiphysics modeling (MPM) calculations, where components or objects are discretized using voxel or tetrahedral meshes. At least the design of supports often utilizes voxel meshes of a certain granularity for modeling. This "voxel approach" considers the linear dependence of material parameters based on voxel dimensions, which directly corresponds to voxel volume fill, fraction, or fill ratio. Specifically, material properties are linearly modeled in volume fill based on a third-order monomial in the voxel dimensions, i.e., the edge length of a voxel (cube).

[0011] In conventional build simulation methods, the state or level of the structure existing within a given voxel (volume filling) is evaluated for the support. However, this often leads to lengthy and complex build simulations where partially filled voxels are practically useless and unusable. This is because any material-related information about the underlying structure and its dimensions within a given voxel is lost or cannot be tracked, as material parameters are treated in the same way as any other aspect or process parameter during the build simulation.

[0012] In other words, once a voxel is neither a perfectly solid nor a completely void space, the properties of the actual material deviate significantly or exhibit poor reproducibility. This is especially true for favorable lattice structures, where these conditions cannot be met because voxel meshes often need to be so fine that it can take days or even weeks of computation to make the model accurate, resulting in unreasonable and uncompetitive results. This is because the material dataset depends not only on specific material parameters but also on the chosen process parameters, the type of AM machine, voxel size, lattice type, and lattice volume fraction. Due to these numerous input parameters, new material data must be constantly calibrated, incurring costs. Summary of the Invention

[0013] Therefore, one object of the present invention is to provide an apparatus that helps to overcome the problems outlined, particularly those concerning simulation operation and complexity.

[0014] The purpose mentioned herein is achieved through the subject matter of the main aspects of the invention. Advantageous embodiments are the subject matter of other aspects of the invention.

[0015] One aspect of the invention relates to an additive manufacturing construction simulation data structure or data structure product, which is compiled or customized for finite element analysis of the mechanical and / or thermal properties of a (predefined) object to be additively manufactured. The object is preferably manufactured by a powder bed fusion method.

[0016] The data structure includes a matrix model representing an elastic tensor or a stress / strain tensor, wherein the matrix elements of the (elastic) tensor include material properties consisting of the object’s Young’s modulus, shear modulus and / or Poisson’s ratio or material properties that depend on the object’s Young’s modulus, shear modulus and / or Poisson’s ratio (or are parameterized by material properties), which may be predefined.

[0017] The data structure also includes model functions or coefficients independent of material properties, whereby the function alternatively takes into account a set of build process parameters, machine or 3D printer settings, voxel or volume fill ratios and / or lattice types used to build the simulation.

[0018] The term "process parameters" can refer to any additional material-independent parameters relating to settings used during build preparation and / or build operations. Specifically, these parameters relate to laser power, scanning speed, beam offset, beam focusing, beam wavelength, scanning pattern, or other hardware-related settings such as gauge settings.

[0019] The specific data structure in tensor form advantageously allows for the provision of input data that can be easily and conveniently read, accessed, edited, and modified by the relevant simulation program or software module. This presented data structure is particularly useful in finite element calculation or simulation environments. Furthermore, tensor form allows for the provision of data independent of a specific coordinate system.

[0020] It will be clear to those skilled in the art that the data structure is preferably not a graphical structure in the form of a matrix, but rather an abstract matrix having directional elements that indicate elasticity, stress, or strain or similar entries in a multilinear graph.

[0021] One advantage of the provided data structure is that it facilitates the provision of a general-purpose materials database or dataset, which largely eliminates the need for complex calculations of material parameters because the material data is provided realistically and can be processed separately from other construction process parameters, lattice, and hardware settings. In other words, a general and sustainable data structure can be provided, which can be further implemented in a user-defined manner. The data structure of this invention particularly overcomes the shortcomings of known voxel-filling methods for lattice structures; for example, the calculation and modeling of the data structure of this invention do not distinguish between materials and process parameters.

[0022] By implementing this user-defined material set for the voxel mesh within the AM processing model, the creation and calibration of new artificial material data for changing build settings becomes particularly obsolete. This significantly reduces the amount of material data required, accelerates the design process, and simplifies related data management.

[0023] Since the physical properties of certain lattice structures are well-known in the art, the proposed structures generally offer improved transferability of results, which also applies to and enables the morphological modeling of machine and process parameters. Typically, improved transferability reduces the total number of design iterations required. Furthermore, the design of material-efficient and generally advantageous lattice structures is no longer associated with the high workload of new material calibration.

[0024] The data structure of this invention also enables tools or entirely new workflows with less calibration and fewer build design iterations. Using the proposed method, material calibration work is significantly reduced, and the design process is simplified. By eliminating the additional work required to change the lattice or its density, the time advantage is particularly beneficial for the design of material-efficient lattice structures. As mentioned above, the "additional work" primarily stems from the material parameter calibration associated with conventional modeling methods, which inherently require modeling in the case of any new lattice setting or density change.

[0025] Therefore, the design process is much faster, and materials data management becomes simpler, resulting in significant cost advantages. This, in turn, allows for accelerated innovation and technology in industrial additive manufacturing.

[0026] In addition to the tensor matrix elements mentioned above, the data structure can also include additional matrix models representing the thermal conductivity tensor, where the matrix elements of the thermal conductivity tensor include material properties consisting of the object's thermal conductivity and / or thermal expansion properties. The structure of thermal conductivity values, as an element of the data structure, also helps to realize the aforementioned advantages in the environment of (multi-physics) FEM-based simulation methods.

[0027] Another aspect of the present invention relates to an additive manufacturing construction simulation method for simulating the manufacture of an object, which includes accessing or utilizing a described data structure.

[0028] Furthermore, the method includes, for example, performing a construction simulation based on entries or elements of a data structure during the relevant design process by executing a finite element algorithm, wherein all (i.e., mechanical and thermal) material properties are defined as (invariant) constants in a matrix, and for a given material, only the elements of the model function are actually modeled or calibrated. Preferably, for a given material, the elements of the model function only need to be calibrated once.

[0029] The construction simulation, the subject of this invention, may include an iterative adaptation procedure, according to which calibration is presented without requiring any (further) modeling or calculation of material properties. Therefore, it is clear that the use of a superior data structure also endows the proposed construction simulation method with the advantages described above. The matrix elements of the data structure are accessed and edited individually, for example, while the basic material properties remain unchanged, thereby forming artificial or user-defined material datasets or related libraries. The advantage of doing so is that the obtained simulation results can be stored and preserved without the need for multiple simulations as is required in conventional construction simulation modeling.

[0030] This method can simultaneously simulate or calculate the physical stress on the object to be manufactured and the temperature distribution in the object, and / or the thermal expansion that may cause changes in stress and temperature distribution.

[0031] In this embodiment, the object includes a lattice support structure. The advantages outlined according to this embodiment are primarily due to the importance of providing intelligent support strategies for complex components, such as employing topology optimization and several mesh refinement iterations.

[0032] A particular advantage of this invention is that the geometry or density of the lattice can be changed during the construction simulation or adaptation process. This does not significantly increase the modeling workload compared to conventional construction simulation methods. This is because, based on the advantages of the presented data structure, material parameter evaluations are set and presented as modeled independently of other process parameters (see model function f). ij For example, adaptive voxels or mesh structures with variable mesh sizes can be provided throughout a given component, where the mesh can be defined based on predefined loads and hotspots.

[0033] As mentioned above, the presented data structure can also be archived in a phenological database to serve as a basis or template for specific material, machine, and / or process parameters in upcoming construction process simulations.

[0034] In optimized simulation methods, the calibration of model functions is enhanced or evaluated through sophisticated data processing devices and / or artificial intelligence.

[0035] Another aspect of the present invention relates to a computer program (product) comprising means for causing a computer or associated analog processor to perform the steps of the described method when the program is executed by a computer or data processing device.

[0036] Another aspect of the invention relates to a non-transitory or non-temporary computer-readable storage medium, for example, on which data structures and / or computer programs (products) are stored.

[0037] In other words, a computer-readable storage medium can be a punched card, a (floppy disk) storage medium, a hard disk, a CD, a DVD, a USB (Universal Serial Bus) storage device, RAM (Random Access Memory), ROM (Read-Only Memory), and / or EPROM (Erasable Programmable Read-Only Memory). Preferably, the computer-readable storage medium can be RAM or ROM, with flash memory being particularly useful. A computer-readable storage medium can also be a data communication network that allows program code to be downloaded, such as the Internet or other systems.

[0038] Another aspect of the invention relates to an apparatus, controller unit, computer, or processor for providing data structures, particularly including a computer-readable storage medium having a computer program stored thereon, and including means for performing the described methods.

[0039] The device may include a controller, such as a programmable logic controller (PLC). The controller may include at least one processor and the computer-readable storage medium, wherein the computer-readable storage medium includes the computer program, which, when executed by the at least one processor, causes the controller to perform the methods described above.

[0040] In other words, at least one processor can be a microprocessor and / or a microcontroller and / or an FPGA (Field Programmable Gate Array) and / or a DSP (Digital Signal Processor).

[0041] The device may include means for providing data structures and performing the method.

[0042] The present invention also relates to implementation schemes that include combinations of features of several embodiments described in the embodiments.

[0043] The advantages and implementation methods of the described data structures and / or methods are effective, or equally applicable to the described computer programs and / or devices. Attached Figure Description

[0044] Furthermore, features and advantageous embodiments will become apparent from the following description of exemplary embodiments related to the accompanying drawings.

[0045] Figure 1 A schematic diagram of an additive manufacturing apparatus is shown, depicting an additive manufacturing part and its support structure.

[0046] Figure 2 For example, using Figure 1 A schematic diagram of the optimized support structure for the equipment manufacturing.

[0047] Figure 3 A schematic flowchart is shown to indicate the iterative nature of the additive manufacturing build simulation process in a simplified manner.

[0048] Figure 4 A simplified matrix indicating a flexible tensor as part of a data structure is shown according to an embodiment of the invention.

[0049] Figure 5 A simplified matrix indicating the thermal conductivity tensor of the data structure of the present invention is shown in another embodiment.

[0050] Figure 6 The qualitative process of scaling material properties as a function of voxel filling parameters is shown.

[0051] Figure 7 The illustration shows the modeling performance at the edges of the object to be manufactured, where sub-voxel elements of a given supervoxel are filled according to the path of the edge.

[0052] Figure 8 The diagram illustrates a scenario where the solution of the present invention is applied. Figure 3 A similar flowchart. Detailed Implementation

[0053] Similar elements, elements of the same kind, and elements with the same function may be given the same reference numerals in the accompanying drawings. The drawings are not necessarily drawn to scale and may be enlarged or reduced to allow for a better understanding of the illustrated principles. Rather, the drawings described should be understood in a broad sense and serve as a qualitative basis for allowing those skilled in the art to apply the presented teachings in a general manner.

[0054] As used herein, the term “and / or” should mean that each of the listed elements can be used alone or in combination with two or more of the additionally listed elements.

[0055] Figure 1 A schematic diagram of an additive manufacturing apparatus 10 for an additive manufacturing (AM) part or object 1 and its support structure 2 according to an embodiment of the present invention is shown.

[0056] 3D printing methods can be used to additively produce articles or objects 1 and supporting structures 2 of articles or objects 1 from modeling material 15. For this purpose, the method may include additively manufacturing part 1 by depositing and curing modeling material 15 in material layers. This may particularly include liquefied or molten deposited modeling material 15.

[0057] For example, the basic AM process can be a selective laser melting process (also known as laser powder bed fusion), in which the deposited modeling material 15 is melted by a continuous or pulsed laser beam 5a focused on a local melting point on the uppermost layer of the modeling material 15. This results in a molten pool or bath at the irradiation point, which can be several material layers deep, and the molten pool or bath can also melt some of the surrounding and already solidified modeling material 15. Next, the laser beam 5a moves further and the melt solidifies by cooling, thereby fusing the final material layer with the material layer below.

[0058] Modeling material 15 may be selected, for example, from the group consisting of metallic materials, combinations of metallic materials, and combinations of metallic alloys. In a particular example, modeling material 15 may be nickel, cobalt, iron, aluminum, titanium, and / or other metals and / or alloys or combinations of nickel, cobalt, iron, aluminum, titanium, and / or other metals.

[0059] In principle, the manufacturing equipment 10 can be configured for multi-component printing, enabling the simultaneous printing of combinations of different materials. Those skilled in the art will readily recognize that the term "material" 15 encompasses such combinations of materials. The modeling material 2 can be specifically provided in powder form.

[0060] Overall, the present invention provides many possibilities for liquefying the modeling material 15 by locally introducing heat into the deposited modeling material 15. Figure 1A specific implementation of the manufacturing apparatus 10 and the associated 3D printing method is based on the use of a laser beam 5a within the LPBF process. This laser beam 5a can be controlled very precisely and heat or energy can be locally focused onto a small dot. The laser beam 5a is controlled by a control unit 11 of the additive manufacturing apparatus 10, which may include a microprocessor or the like and may be integrated into and / or communicatively connected to a computing system.

[0061] The control unit 11 may be another device 3 or may be linked to another device 3, such as another computer, controller, or processor, and is preferably configured to perform a build simulation process subordinate to the present invention. As will be further described with reference to the following figures, device 3 is adapted to access the data structure of the present invention, which significantly improves the build simulation excellence for the manufacture of object 1.

[0062] Once the build simulation has been executed (see below) Figure 3 and Figure 8 This requires specifying additional preparation steps, such as defining process parameters like layer thickness and irradiation parameters and vectors. When build preparation is complete, modeling material 15 is applied in powder form to a manufacturing table, plate, or platform 8 within the manufacturing chamber 12. Modeling material 15 is supplied from a supply chamber 9 adjacent to the manufacturing chamber 12. For this purpose, a powder delivery system including a material feeder 7, rollers, and / or recoater blades is used to deliver modeling material 15 in a stepwise manner to deliver each deposited layer of modeling material 15 into the manufacturing chamber 12. Modeling material 15 forms a so-called powder bed on the manufacturing plate 8 within the manufacturing chamber 12. The modeling material 15 placed on top of the powder bed is liquefied by localized laser irradiation of laser beam 5a, thereby producing a solid, continuous metal part 1 upon cooling. However, it should be understood that alternatively, or in addition to laser beam 5a, 3D printing method M may also use particle beams, such as electron beams, or other energy beams for liquefying modeling material 15.

[0063] Figure 1 The manufacturing apparatus 10 provides an energy source in the form of a laser 5, such as an Nd:YAG laser, a CO2 laser, etc. Figure 1 In this process, the laser 5 selectively (i.e., locally) transmits the laser beam 5a to certain portions of the powder surface of the powdered modeling material 15. For this purpose, an optical deflection device or scanner module, such as a movable or tiltable mirror 6, can be provided to deflect the laser beam 5a to certain portions of the powder surface of the modeling material 15 according to the tilt position of the laser beam 5a.

[0064] The manufacturing apparatus 10 may also include common components for adjusting such laser beam 5a, such as focusing optics or the like. Figure 1 (Not explicitly shown). At the point of impact of the laser beam 5a, i.e., the molten pool or melting point, the modeling material 15 is heated to the point that the powder particles locally melt and form agglomerates during cooling. Depending on the digital production model and build job preparation provided by the CAD system (“Computer-Aided Design”), the laser beam 5a scans the powder surface (see...). Figure 1 (Horizontal arrow in the image). After selective melting and local “fusion” of powder particles in the surface material layer of modeling material 15, excess and non-agglomerated metal modeling material 15 can be removed by the feeder or recoater 7.

[0065] Then, by means of manufacturing piston 13 (see...) Figure 1 (Middle arrow pointing downwards) The manufacturing platform 8 is lowered, and new modeling material 15 is transferred from the supply chamber 9 to the manufacturing chamber 12 by means of the material feeder 7 or another suitable device. As the manufacturing platform 8 is gradually lowered via the manufacturing piston 13, the modeling material 15 moves upward within the supply chamber 9 via the supply piston 14, allowing the material feeder 7 to move new modeling material 15 into the manufacturing chamber 12. Within the manufacturing chamber 12, the modeling material 15 can be preheated to a working temperature just below the melting temperature of the modeling material 15 by means of infrared light or a similar device, thereby accelerating the melting process or controlling thermal shock.

[0066] Using LPBF and similar methods as a manufacturing technique typically requires providing a support structure 2 for the object 1 during actual manufacturing to hold and stabilize the part on top of the manufacturing plate 8. This support structure 2 usually comprises one or more individual support members 2, which are then removed during post-processing of the part 1 after the printing process has been completed.

[0067] Especially for components 1 with large volumes, numerous supports are required to prevent deformation or potential malfunctions due to peeling or delamination from the manufacturing plate 8. To maintain the required material and thus keep the manufacturing process as low as possible, this support structure needs to be as light and thin as possible. The support structure 20 also plays a crucial role in heat dissipation from the manufacturing chamber through the manufacturing plate 8. If this heat dissipation is insufficient when considering the quasi-insulating thermal properties of loose powder particles, excessive heat or hot spots may occur during construction, leading to insufficient structural mass of the associated object 1.

[0068] In additive manufacturing methods similar to those described herein, the 3D design of part 1 necessarily requires a corresponding digital 3D model of the support structure 2. Based on such a model, the 3D sintered or "printed" metal part 1, together with its support structure 2, is created by modeling material 15 during the iterative generation process, wherein the surrounding powdered modeling material 15 can also be used to temporarily support the portion of the metal part 1 that has been constructed up to that point. Due to the continuous downward movement of the manufacturing stage 8, the metal part 1 is manufactured in a layered manner.

[0069] For example, object 1 could be a turbine blade or turbine wheel made of metallic materials, such as in a heavy-duty industrial gas turbine engine used for power generation. The blade may specifically require three volumetric supports: one on each of the two distal wedges, and one below the central airfoil. As shown by the tree-like or lattice-like design of support 2 (the contact area indicated by number 4), the support often and preferably undergoes topology optimization to optimize material use, heat dissipation, and mechanical stability.

[0070] To meet these high requirements and quality demands, it is necessary to repeatedly process and access a vast library of materials, material properties, process parameters, and possibly other variables to provide excellent design and related construction simulations for the support components.

[0071] Because of these numerous input parameters, new material data is required periodically, and this data in turn needs to be calibrated and validated, resulting in significant cost and workload. Figure 3 The schematic and simplified flowcharts depicted should at least qualitatively indicate this situation.

[0072] The simulation methods discussed are typically set up for finite element analysis (FEM) of predefined geometries and / or materials to determine their mechanical and / or thermal properties. Finite element analysis (FEM) is a powerful approach for numerically solving (partial) differential equations in 2D or 3D space. In most cases, boundary value problems are modeled where the FEM subdivides or discretizes a given model system into smaller, simpler parts called finite elements. For discretization, voxel or tetrahedral meshes are typically implemented, providing a numerical domain with a finite number of points.

[0073] To meet quality standards, a large material library is required. Due to these numerous input parameters, new material datasets often need to be calibrated at considerable cost. Constructing the upper layers of the simulation BS (see...) Figure 3The top arrow in the diagram indicates the (time-wise) process from left to right. Within a regular build simulation, the Material Calibration (MC), as shown in the lower left layer, is required. This MC, in turn, requires the Calibration Print Test (CP). For calibration printing, material parameters, machine-related parameters, and process parameters (indicated by the PP) are typically considered in each iteration.

[0074] Only after this calibration print proves successful can the calibration simulation (CS) then consider and optimize the mesh size and lattice type (these parameters are not explicitly indicated but belong to step CS). Assuming the material calibration (MC) has proven successful, an actual build simulation is then performed, which requires generating an additive manufacturing build model (as indicated by the AM). The results are then evaluated and returned to, for example, the calibration print (CP), calibration simulation (CS), and additive manufacturing build model (AM). Clearly, this process is inherently complex, time-consuming, and cumbersome for AM designers.

[0075] Figure 4 A non-constraintive implementation of the data structure product DS of the present invention is illustrated by way of example. The data structure includes a matrix model representing the elastic tensor and / or stress-strain relationship (see σ = E : ε). The data structure of the present invention includes at least entries according to the matrix shown, namely material properties consisting of Young's modulus E, shear modulus G, and / or Poisson's ratio ν. Furthermore, the data structure DS includes a model function f. ij Or related function coefficients.

[0076] The elastic tensor can also be a yield tensor expressed as an inverse relationship between stress and strain, or a yield tensor expressed as an inverse relationship between stress and strain.

[0077] According to the present invention, the model function f ij Completely independent of material properties MP. Instead, the function takes into account a set of build process parameters PP used to build the simulation, machine or printer settings variables, voxel volume filling parameters Vf, and / or lattice type.

[0078] Based on such structural or functional data, the advantages described above in this invention can be achieved. Particularly advantageous is that the simulation construction (see below) allows access to the data structure and editing of individual (master) elements with unchanged fundamental material properties, thereby forming an artificial or user-defined dataset. The dataset of this invention has the potential to significantly simplify simulation construction because the material parameters do not need to be recalibrated during the process. Instead, a user-defined material dataset is provided, where only the model function f needs to be adjusted. ij The elements are modeled or calibrated, however, the model does not need to model the material properties.

[0079] The notation for tensors can also differ from that of tensors. Figure 4 The instructions in the text. In an exemplary embodiment of the invention, the material modeled with AM corresponds to the orthogonal anisotropic linear elastic material properties, where the material is essentially dependent on nine material parameters. For each dimension, independent Young's modulus Eij, shear modulus Gij, and Poisson's ratio ν are defined. ij The indicated material tensor (actually a fourth-order tensor) is further simplified and shown according to the so-called Kelvin or Voigt notation, where there are only three ν due to certain symmetry considerations. ij It is independent. Therefore, the dimension of a tensor can be reduced; however, reducing the dimension of a tensor is not a prerequisite for the presented data structure.

[0080] The provided data structure is particularly excellent in its compatibility with the finite element algorithm presented in this paper.

[0081] Figure 4 Another example of a simplified (symmetric) thermal conductivity tensor is shown; the entire tensor or its elements can also be part of the data structure DS of this invention. In this case, the equations This paper provides an overview of the heat flux vector q (corresponding to stress in the mechanical equivalent), thermal conductivity κ (corresponding to mechanical yield or mechanical elasticity in the mechanical equivalent), and temperature gradient. The relationship between (strain or displacement corresponding to the structure in a mechanical equivalent) and κ, where κ is as follows: Figure 5 The matrix form shown is given.

[0082] For this purpose, the construction simulation method of the present invention can be a multiphysics construction simulation that considers both thermal and mechanical material properties (see below). Figure 6 ).

[0083] Figure 4 and Figure 5 The examples shown should simply illustrate that thermal and mechanical properties are typically aggregated or treated together in finite element analysis. For this purpose, the term "mechanical property" may be used synonymously with the thermomechanical properties and / or thermal and mechanical properties of a given material.

[0084] In other words, the (user-defined) material data structure of the present invention advantageously allows the specification of the properties of the principal elements for a selected set of input parameters. Each entry of the indicated tensor can, for example, be derived from an arbitrary model function f. ij Individual access. For each material (E) ij G ij ,ν ij All material properties MP are specific to each material and remain invariant to the mechanical yield tensor, while all other effects are determined by f. ijOverlay. This leads to a dependence on changes in the number of variables used to reduce material data. The material data then depends solely on mechanical (thermomechanical) properties. Grid type and density, as well as machine and grid size, all affect the function f. ij Or the function f ij Taken into account. However, the model function f ij The set only needs to be specified and calibrated once.

[0085] If inserted into a high-level voxel method, this results in the following contribution:

[0086] ,

[0087] E* depends on material data, voxel volume, lattice type, machine parameters, and material properties. On the other hand, Vf equals ( This is the ratio of the voxel volume (voxel size) to the maximum size of the corresponding voxel (voxel mesh size) (fill ratio).

[0088] Since the mechanical and thermal properties of lattice structures are generally well-known (see...), Figure 6 Therefore, this physical motivational knowledge can be applied. Consequently, lattice type, density, and voxel size, which are variables in material data, are eliminated.

[0089] in this case, Figure 6 The qualitative process of scaling material properties as a function of voxel volume fill Vf is illustrated. Referring to the figure, the numbers T and M represent thermal and mechanical properties, respectively; then, the straight lines represent approximate or scaled voxel properties.

[0090] As mentioned above, the outlined method can be used in the model function f ij The effects of build parameters and machine settings are also covered.

[0091] Figure 7 An example is shown of calculating or determining the voxel-filling density (see large cuboid) via daughter voxels (smaller cuboids). This example shows a voxel-filling density consisting of 5 daughter voxels (smaller cuboids). 3 = A large cuboid composed of 125 daughter elements.

[0092] Specifically, the path through the edge 17 of the object is shown to be contoured and filled if any part of the object passes through a given sub-voxel.

[0093] According to another example, in 70 sub-voxels such as Figure 7 Assuming the indicated fill condition, the remaining 55 sub-voxels may be left empty. This results in a voxel density of approximately 56%. The material properties in the solution for that given element are then reduced to only 56%.

[0094] and Figure 3 In comparison, the invention offers numerous advantages, such as Figure 8 The flowchart is shown schematically. Specifically, as shown... Figure 3 The charts indicated show that the actual evaluation of the results requires fewer input parameters, calculation steps, and parameter considerations.

[0095] Ideally, material properties should be considered at the end of the process, i.e., during the evaluation of results, rather than during the process as in conventional construction simulations. Figure 8 The process parameters PP in the lower left corner summarizes parameters such as support topology, grid size, grid type, machine settings, and other process parameters.

[0096] The presented data structure and the method for simulating additive manufacturing construction—accessing the data structure—are significantly improved through the inventive concept. Deploying the provided data structure and the model function f of this invention... ij A particular advantage of this method is that the lattice type and its density can be easily changed during construction preparation or construction simulation without requiring excessive modeling capabilities and computation time.

[0097] It is also conceivable that the provided data structure DS is archived in a phenological database to serve as the basis for material, machine, and / or process parameters in future construction process simulations.

[0098] As mentioned herein, object 1 can specifically refer to a component or article of a complex shape, such as a filamentary portion with a structure. Preferably, the component is made of a high-performance material, such as a material with high strength and / or high thermal resistance. In particular, the component can form part of a steam turbine component or a gas turbine component, such as blades, impellers, shrouds, protective covers such as heat shields, tips, sections, inserts, injectors, seals, transitions, burners, nozzles, filters, orifices, bushings, distributors, domes, superchargers, cones, spray guns, plates, resonators, pistons, or any corresponding retrofit kit. Alternatively, the component can refer to another component or a similar component.

[0099] Correspondingly, the number of process parameters describing or comprehensively characterizing layers used for structurally complex objects or components, such as combustor components of gas turbines, can exceed one hundred. The following quantities can also be understood as process parameters: layer thickness, molten pool geometry, thermal shock per unit volume or per unit area, laser wavelength, shadow distance (i.e., distance to adjacent scan lines), beam offset, beam velocity, molten pool size, beam point geometry, beam angle, type of purge gas, purge gas flow rate, possible exhaust gas flow rate, valve status, ambient pressure set before or during the build operation, substrate condition, etc.

[0100] It will be apparent to those skilled in the art that these embodiments and items merely depict examples of a variety of possibilities. Therefore, the embodiments shown herein should not be construed as limiting these features and configurations. Any possible combination and configuration of the described features can be selected according to the scope of the invention.

Claims

1. An additive manufacturing construction simulation data structure (DS), said additive manufacturing construction simulation data structure (DS) is compiled into a finite element analysis (FEM) for the mechanical properties (M, T) of an object (1) to be additively manufactured, wherein, The data structure (DS) includes a matrix model representing the elastic tensor (E), wherein the matrix elements of the elastic tensor (E) include: The material properties of the object (1), consisting of Young's modulus (E), shear modulus (G), and / or Poisson's ratio (ν), and Model function (f) independent of the material properties (MP) ij ), where the function (f ij The simulation takes into account a set of construction process parameters (PP), machine settings, voxel volume filling (Vf), and / or grid type.

2. The data structure (DS) according to claim 1 further includes another matrix model representing the thermal conductivity tensor (κ), wherein, The matrix elements of the thermal conductivity tensor (κ) include material properties (MP) consisting of the thermal conductivity (κ) of the object (1).

3. An additive manufacturing construction simulation method, the additive manufacturing construction simulation method being used to simulate the manufacturing of the object (1), the method comprising the following steps: (i) Accessing the data structure (DS) as described in claim 1 or 2, and (ii) The constructed simulation is achieved by performing a finite element method (FEM) based on entries of the data structure (DS), wherein all material properties (MP) are defined as constants in the matrix, and for a given material, only the model function (f) is calibrated. ij () elements.

4. The method according to claim 3, wherein, The simulation includes an iterative adaptation procedure, according to which any modeling that does not involve material properties (MS) is calibrated.

5. The method according to claim 3 or 4, wherein, The matrix elements of the data structure (DS) are accessed and edited individually while the material properties (MP) remain unchanged, thereby forming a user-defined material dataset.

6. The method according to any one of claims 3 to 5, wherein the method simultaneously simulates the physical stress on the object (10) to be manufactured and the temperature distribution (T) and / or thermal expansion in the object, a multi-physical process.

7. The method according to any one of claims 3 to 6, wherein, The object includes a lattice support structure (2).

8. The method according to claim 7, wherein, The geometry of the lattice (2) or the density of the lattice (2) may be changed during the construction simulation process.

9. The method according to any one of claims 3 to 8, wherein, The data structure (DS) is archived in a phenology database to serve as the basis for material, machine, and / or process parameters in future construction process simulations.

10. The method according to any one of claims 3 to 9, wherein, The model function (f) ij The calibration of ( ) is enhanced or evaluated through sophisticated data processing devices and / or artificial intelligence.

11. A computer program comprising a method for causing the computer to perform the steps of any one of the methods according to claims 3 to 10 when the program is executed by the computer.

12. A computer-readable storage medium having a data structure (DS) according to claim 1 and / or a computer program according to claim 11 stored on the computer-readable storage medium.

13. An apparatus (2), the apparatus (2) particularly comprising a computer-readable storage medium having a computer program according to claim 12 stored on the computer-readable storage medium, and comprising means for performing the method according to any one of claims 3 to 10.

Images