Method and apparatus for controlling multi-mode shaped beam in powder bed additive manufacturing

Abstract

Methods and apparatus for modulating a multi-mode shaped beam in powder bed additive manufacturing are provided. The method comprises the following steps: analyzing a geometric model of a component to obtain a plurality of feature regions which comprise at least one part of the component and correspond to different structure types; according to the structure type and in combination with material attributes and / or process targets, distributing corresponding light beam modes for the plurality of feature regions, wherein the light beam modes are modulated to have different light intensity distributions and / or phases; the light beam mode is applied to a light beam, so that the light beam is switched to the light beam mode corresponding to the light beam mode when the light beam scans a plurality of feature areas on a powder bed, matched light spot regulation and control are carried out on different feature areas in the additive manufacturing process, and therefore different forming requirements of various structure types are met in the same component at the same time.

Description

Technical Field

[0001] The present application relates to the field of additive manufacturing technology, and more particularly to a method, an optical path system, an additive manufacturing apparatus, an electronic device, and a computer-readable storage medium for regulating a multi-mode shaping light beam in powder bed additive manufacturing. Background Art

[0002] Additive manufacturing (AM), also known as 3D printing, is a process of constructing three-dimensional objects by layer-by-layer material deposition. Known powder bed-based additive manufacturing technologies, for example, are particularly suitable for manufacturing metal components. For example, the laser powder bed fusion (LPBF) process, because of its ability to manufacture metal components with high precision and high complexity, has been widely used in the fields of aerospace, medical, and mold manufacturing. The LPBF process generally includes the following steps: uniformly laying metal powder on a substrate, using a laser beam to scan through a galvanometer to melt and solidify the powder according to the cross-sectional profile of the component, after completing one layer of scanning, the substrate descends by one layer thickness, and repeating this process until a complete component is manufactured.

[0003] During the LPBF processing, the spot shape, energy distribution, focusing position, and scanning strategy of the laser beam have a decisive influence on the melt pool size, temperature gradient, cooling rate, and microstructure evolution. Existing LPBF processes usually use Gaussian spots. Because of its high central energy density and fast edge attenuation, the convection inside the melt pool is strong, it is difficult to maintain a stable temperature field, and it is easy to cause large fluctuations in solidification conditions, thus resulting in non-uniform heat input in different geometric features. For example, in thin-wall or slender structure regions, the high energy concentration of the Gaussian spot can cause over-melting, warping, or increased surface roughness; while in large-volume or thick-wall regions, the insufficient energy at the spot edge easily leads to poor fusion or pore defects.

[0004] To overcome the above deficiencies, the industry has tried to use beam shaping technologies, such as using diffractive optical elements (DOEs) to generate shaped spots with non-Gaussian modes, which can adjust the melt pool width and depth distribution to a certain extent and optimize the local thermal field. Currently, effective control of the geometric features and solidification behavior of the LPBF melt pool under different light field modes has been achieved, thereby realizing the regulation of the material microstructure.

[0005] However, existing beam shaping technologies usually only match a single geometric feature and it is difficult to simultaneously meet the different forming requirements of multiple structural types in the same component. Therefore, there is still an urgent need for a comprehensive solution that can adaptively allocate multiple spot modes for multiple geometric feature regions in complex components. Summary of the Invention

[0006] Embodiments of the present application provide a method, an optical path system, an additive manufacturing device, an electronic device, and a computer-readable storage medium for regulating a multi-mode shaping beam in powder bed additive manufacturing. The method dynamically allocates and switches multiple beam modes by analyzing the geometric model of the component and combining the structure type, material properties, and / or process objectives, so as to achieve precise spot regulation for different feature regions in additive manufacturing, optimize the melt pool morphology, reduce residual stress, and improve the forming quality.

[0007] In a first aspect, embodiments of the present application provide a method for regulating a multi-mode shaping beam in powder bed additive manufacturing, wherein the additive manufacturing manufactures a component by scanning a powder bed stacked layer by layer with a beam propagating along an optical path. The method includes: analyzing the geometric model of the component to obtain multiple feature regions corresponding to different structure types of at least a part of the component; allocating corresponding beam modes to the multiple feature regions according to the structure type and combining material properties and / or process objectives, and the beam modes are modulated to have different light intensity distributions and / or phases; applying the beam mode to the beam so that the beam switches to the corresponding beam mode when scanning the multiple feature regions on the powder bed.

[0008] According to a preferred embodiment of the first aspect, the structure type includes at least two of overhang, curved surface, porous, lattice, thin wall, thick wall, and connection structure, wherein the thickness of the thin wall is less than a first threshold, and the thickness of the thick wall is greater than a second threshold; the beam modes include at least two of Gaussian, annular, flat top, Bessel, vector, and saddle.

[0009] According to a preferred embodiment of the first aspect, the beam mode further includes a transition mode between at least two of the beam modes and / or a composite mode formed by superimposing or segmenting at least two of the beam modes.

[0010] According to a preferred embodiment of the first aspect, the method further includes: identifying the feature region to divide it into at least two grid cells; allocating different beam modes to the at least two grid cells according to the structure type and combining material properties and / or process objectives.

[0011] According to a preferred embodiment of the first aspect, the method further includes: allocating corresponding process parameters to the multiple feature regions according to the structure type and combining material properties and / or process objectives, and the process parameters include at least one of beam power, scanning speed, spot size, and scanning trajectory; applying the process parameters to the beam so that the beam switches to the corresponding process parameters when scanning the multiple feature regions on the powder bed.

[0012] According to a preferred embodiment of the first aspect, the process objectives include at least one reduction in thermal stress, residual stress, surface roughness, structural warping, porosity, and / or at least one improvement in density, build efficiency, energy efficiency ratio, mechanical property isotropy, dimensional accuracy, and interlayer bonding strength.

[0013] According to a preferred embodiment of the first aspect, the method further includes: respectively performing test scans on the plurality of feature regions in different beam modes to obtain test data characterizing the process objectives corresponding to the plurality of feature regions; determining the beam mode and / or process parameters for enabling the plurality of feature regions to achieve the process objectives according to the test data, so as to determine them as those to be allocated to the corresponding plurality of feature regions.

[0014] According to a preferred embodiment of the first aspect, the method further includes: simulating the temperature field, stress field, and molten pool morphology of the plurality of feature regions in different beam modes by using finite element simulation to obtain simulation data for characterizing the process objectives; determining the beam mode and / or process parameters for enabling the plurality of feature regions to achieve the process objectives according to the simulation data, so as to determine them as those to be allocated to the corresponding plurality of feature regions.

[0015] According to a preferred embodiment of the first aspect, the method further includes: constructing a machine learning model with the structure type, material property, beam mode, and / or process parameters of the plurality of feature regions as inputs and the achievement degree of the process objectives as outputs; training the machine learning model by using the test data and / or simulation data; predicting the achievement degree of the process objectives of the plurality of feature regions in different beam modes and / or process parameters by using the trained machine learning model, and determining the beam mode and / or process parameters to be allocated to the plurality of feature regions accordingly.

[0016] According to a preferred embodiment of the first aspect, the method further includes: capturing real-time data of the temperature field, stress field, and molten pool morphology of the plurality of feature regions during the scanning process; inputting the real-time data into the trained machine learning model for prediction, and making real-time adjustments to the beam mode and / or process parameters corresponding to the plurality of feature regions according to the prediction results.

[0017] According to a preferred embodiment of the first aspect, a variable aperture assembly and a variable focal length lens assembly are respectively arranged in the optical path. When applying a non-Gaussian beam mode to the beam, the method further includes: driving the variable aperture assembly and the variable focal length lens assembly to adjust the spot diameter and focal length to the target values in the current beam mode, so that the beam always maintains the best focusing state at different layer heights.

[0018] According to a preferred embodiment of the first aspect, the method for obtaining the target value includes: during the scanning process, using a ranging sensor integrated in the optical path to obtain the real-time distance from the reference position to the top layer of the powder bed, and combining with a pre-calibrated optical model to calculate the target value required for the current layer; or, before scanning, according to the geometric model and layer thickness of the component, calculating and storing in advance the target value required for each layer, generating a target value-layer number mapping table, and directly reading and applying the corresponding target value according to the layer number during the scanning process.

[0019] In a second aspect, an embodiment of the present application provides an optical path system for an additive manufacturing device, including: a beam emitter for generating and emitting a beam; a beam deflector for deflecting the beam to a powder bed stacked layer by layer along a preset scanning path to manufacture a component; a shaping unit disposed in the optical path between the beam emitter and the beam deflector for modulating the beam into multiple beam modes with different light intensity distributions and / or phases; and a control system connected to and controlling the beam emitter, the beam deflector, and the shaping unit to execute the method according to any one of the first aspect.

[0020] In a third aspect, an embodiment of the present application provides an optical path system for an additive manufacturing device, including: a beam emitter for generating and emitting a beam; a beam deflector for deflecting the beam to a powder bed stacked layer by layer along a preset scanning path to manufacture a component; a shaping unit disposed in the optical path between the beam emitter and the beam deflector for modulating the beam into multiple beam modes with different light intensity distributions and / or phases; a variable aperture assembly disposed in the optical path between the shaping unit and the beam deflector for adjusting the spot diameter of the beam; a variable focal length lens assembly disposed in the optical path between the variable aperture assembly and the beam deflector for adjusting the focal length of the beam; a ranging sensor disposed in the optical path for measuring in real time the distance from the reference position to the top layer of the powder bed; and a control system connected to and controlling the beam emitter, the shaping unit, the variable aperture assembly, the variable focal length lens assembly, the beam deflector, and the ranging sensor to execute the method according to any one of the first aspect.

[0021] In a fourth aspect, an embodiment of the present application provides an additive manufacturing device including the optical path system according to the second aspect or the third aspect.

[0022] In a fifth aspect, an embodiment of the present application provides an electronic device, including: at least one processor; at least one memory; the at least one memory is coupled to the at least one processor and is used for storing instructions executed by the at least one processor, and when the instructions are executed by the at least one processor, the electronic device executes the method according to any one of the first aspect.

[0023] In a sixth aspect, an embodiment of the present application provides a computer-readable storage medium storing a computer program therein, and when the computer program is executed by a processor, the method according to any one of the first aspect is implemented.

[0024] It should be understood that the above general description and the following detailed description are only exemplary and do not limit the present application. Description of the Drawings

[0025] The drawings incorporated herein and forming a part of the specification illustrate one or more embodiments of the present application and, together with the description, are used to explain the principles of the present application and to enable those of ordinary skill in the relevant art to make and use the present application.

[0026] Figure 1 is a schematic diagram of the scanning of the optical path system in the powder bed according to an exemplary embodiment of the present application;

[0027] Figure 2 is a schematic flowchart of a method for regulating a multi-mode shaped beam according to an exemplary embodiment of the present application;

[0028] Figure 3A is a spot model diagram of different beam modes according to an exemplary embodiment of the present application, where (a)-(d) are flat-top, Gaussian, annular, and Bessel spots respectively;

[0029] Figure 3B is an organizational structure diagram of an In738 alloy specimen under different beam modes, where (a)-(d) are crystal orientation characterizations of the In738 alloy specimen under flat-top, Gaussian, annular, and Bessel modes respectively, and (e)-(h) are local average misorientation (KAM) characterizations of the In738 alloy specimen under flat-top, Gaussian, annular, and Bessel modes respectively;

[0030] Figure 4A is a schematic diagram of the first allocation scheme of the beam mode according to an exemplary embodiment of the present application;

[0031] Figure 4B is a schematic diagram of the second allocation scheme of the beam mode according to an exemplary embodiment of the present application;

[0032] Figure 5 is a schematic flowchart of a method for regulating a multi-mode shaped beam by using grid cell division according to an exemplary embodiment of the present application;

[0033] Figure 6 is a schematic flowchart of a method for regulating process parameters according to an exemplary embodiment of the present application;

[0034] Figure 7Schematic flow chart of a method for regulating a multi-mode shaped beam using a test scan according to an exemplary embodiment of the present application;

[0035] Figure 8 Schematic flow chart of a method for regulating a multi-mode shaped beam using finite element simulation according to an exemplary embodiment of the present application;

[0036] Figure 9 Schematic flow chart of a method for regulating a multi-mode shaped beam using machine learning according to an exemplary embodiment of the present application;

[0037] Figure 10 Schematic diagram of the scanning of another optical path system on a powder bed according to an exemplary embodiment of the present application;

[0038] Figure 11 Schematic diagram of the structure of an electronic device according to an exemplary embodiment of the present application;

[0039] Figure 12 Schematic diagram of the structure of a computer-readable storage medium according to an exemplary embodiment of the present application. Detailed implementation manners

[0040] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, the exemplary embodiments can be implemented in various forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are described so as to make the present application more thorough and complete, and to fully convey the concept of the exemplary embodiments to those skilled in the art. The features, structures, or characteristics described can be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to provide a deeper understanding of the embodiments of the present application.

[0041] The embodiments of the present application are not limited to specific additive manufacturing types and can be applied to various powder bed-based additive manufacturing processes, including Laser Powder Bed Fusion (LPBF), Electron Beam Melting (EBM), Selective Laser Melting (SLM), Selective Laser Sintering (SLS), and other additive manufacturing processes that use powder as the forming material and a light beam as the energy source. The "powder" described herein generally refers to the base material for forming the component, which is in the form of particles and allows for differences in morphology, particle size, and size distribution. In the preferred embodiment, a metal powder system is used, covering typical engineering materials such as stainless steel, aluminum alloy, titanium alloy, nickel-based alloy, cobalt-based alloy, copper, and copper alloy; in other alternative embodiments, it can also be extended to ceramic-based, polymer-based, or composite powder systems. In the following description, the "light beam" can refer to a laser beam or an electron beam; unless otherwise specified, the following embodiments will default to the "light beam" being exemplified by a laser beam.

[0042] Figure 1 An example of the optical path system 100 is shown. The optical path system 100 is applied in an additive manufacturing device and refers to the optical components arranged in the optical path along the propagation direction of the light beam, which can guide the light beam to the powder bed 110 stacked layer by layer for irradiation scanning, thereby manufacturing the component 111 layer by layer. The optical path system 100 has a light beam emitter 101 (such as a laser) and a light beam deflector 102 (such as a galvanometer). The light beam emitter 101 is used to generate and emit the light beam, and the light beam deflector 102 is used to deflect the light beam to the top layer of the powder bed 110 according to a preset scanning path to irradiate the powder, melt the powder, and form a solid structure after solidification. In some embodiments, the optical path system 100 may also have elements for further processing the light beam, such as optical elements like collimating mirrors, beam expanders, and focusing mirrors. Since these elements are known to exist in the optical path system 100, their structures and principles will not be elaborated further.

[0043] The optical path system 100 has a shaping unit 104. The shaping unit 104 is arranged in the optical path between the beam emitter 101 and the beam deflector 102, and is used to modulate the beam into multiple beam modes with different light intensity distributions and / or phases. The shaping unit 104 can be a switchable diffractive optical element (DOE) assembly. For example, by installing multiple prefabricated DOEs in the optical path, each DOE is translated or rotated laterally along the beam by a micro motor or a piezoelectric driver to switch to the center position of the optical path, so as to realize the rapid switching of multiple beam modes. The surfaces of each DOE can be etched through micro-nano processing to form a predetermined phase distribution pattern for modulating the light intensity distribution and / or phase characteristics of the incident beam. In an alternative embodiment, the shaping unit 104 can also be a programmable spatial light modulator assembly. For example, by integrating a spatial light modulator with a liquid crystal or MEMS micromirror array in the optical path, the spatial light modulator can dynamically generate any custom beam mode within a single device by sending a digital phase mask through a control chip. The spatial light modulator also supports superimposing or interpolating the phase distributions of two or more modes to achieve the transition between different beam modes or the output of a composite spot.

[0044] The optical path system 100 has a control system 103. The control system 103 is respectively connected to the beam emitter 101, the beam deflector 102 and the shaping unit 104, and is used to control each optical element according to the manufacturing task of the component 111. In addition, as the control center of the additive manufacturing equipment, the control system 103 can also be connected to other components in the equipment, such as Figure 1 the powder dispenser 112 shown for applying powder to the powder bed 110 in the form of layers and other components not shown, so as to achieve the overall control of the entire manufacturing process.

[0045] In some embodiments, the shaping unit 104 can also have a calibration module. The calibration module can perform on-line detection of the output spot (such as using a monitoring camera or a spot sensor), and continuously fine-tune the phase pattern or the element position through closed-loop feedback to correct the spot deviation caused by temperature drift, mechanical error or incident angle change, so that the spot shape is consistent with the target requirements.

[0046] Figure 2 An example of a method 200 for regulating a multi-mode shaped beam is shown, including steps 201-203 that make up the method 200. It should be understood that the steps of the method 200 can be configured to be executed by Figure 1 the control system 103, and the implementation of these steps will be described below with reference to Figure 1 and Figure 2 elaborate.

[0047] 201. Analyze the geometric model of the component 111 to obtain multiple feature regions corresponding to different structural types including at least a part of the component.

[0048] The control system 103 can read the geometric model of the component 111 to be manufactured from the storage unit. This geometric model can be generated by common computer-aided design (CAD) software and stored in STL, STEP or other formats. After the reading is completed, the control system 103 preprocesses the geometric model, such as meshing and voxelization. The meshing process discretizes the surface of the geometric model into a large number of triangular patches, and extracts the vertex coordinates, normal vectors and adjacency relationships of each patch; the voxelization process divides the volume of the geometric model into regular cubic cells to facilitate subsequent wall thickness measurement and internal structure analysis.

[0049] On this basis, the control system 103 performs structural feature analysis on the meshing results and voxel data. For thin-wall and thick-wall structures, the control system 103 samples multiple distance points inward and outward along the patch normal, and obtains the local wall thickness distribution through ray casting or distance field calculation. When the minimum wall thickness of a certain area is less than the preset first threshold, this area is marked as "thin wall"; when the maximum wall thickness of a certain area is greater than the preset second threshold, it is marked as "thick wall". It should be understood that the first threshold and the second threshold are used to distinguish between "thin wall" and "thick wall" structures, and their values should be determined in combination with the actual manufacturing size of the component 111 and the layer thickness of the used powder bed 110. Usually, the first threshold can be taken as 2 to 10 times the layer thickness (for example, when the layer thickness is 0.05 mm, the first threshold can be set to 0.1 mm to 0.5 mm) to ensure that the thin-wall area has sufficient energy concentration effect during processing and will not be overburned. The second threshold can be taken as 5% to 20% of the minimum cross-sectional size of the component (for example, for a component with a cross-sectional size of 10 mm, the second threshold can be set to 0.5 mm to 2 mm) to ensure that the thick-wall area can obtain sufficient energy input and avoid insufficient fusion. The above threshold ranges are only examples, and in actual applications, they can be optimized and adjusted according to the thermal physical properties of the material, the beam power, the scanning speed and the forming quality requirements.

[0050] For overhanging structures, the control system 103 calculates the angle between the patch normal and the building layer direction (Z-axis), and judges whether there is a sufficient support surface below the patch. When the inclination angle exceeds the preset value and there is no support below, this area is classified as "overhanging". The identification of porous and lattice structures can rely on the connectivity analysis of voxel data: the control system 103 detects internal closed or connected holes and classifies them as "porous"; for regularly arranged micro-channels or lattice-like units, they are marked as "lattice". The determination of curved surface structures is based on the patch curvature: when the average curvature radius of the patch group is less than the preset value, it is marked as "curved surface". The connection structure can include those areas that neither belong to any of the above categories nor are located at the transition parts of different geometric units, such as the transition bodies between beams and plates, columns and bases, such as ribs.

[0051] After the above structural type division, the control system 103 performs spatial clustering on the patches or voxel units of the same structural type to generate multiple non-overlapping feature regions. For parts with blurred boundaries or multi-type overlaps, they are refined according to the priority rules or morphological segmentation algorithms to ensure that each feature region has a unique type and clear boundaries. The control system 103 stores the identifiers, structural types, spatial ranges, and statistical features (such as average curvature, wall thickness distribution) of each feature region in the feature region database to provide input data for subsequent beam mode allocation and process parameter optimization.

[0052] 202. Allocate corresponding beam modes to multiple feature regions according to the structural type in combination with material properties and / or process objectives, where the beam mode is modulated to have different light intensity distributions and / or phases.

[0053] "Material properties" can refer to the powder materials used for the entire component 111, such as materials like stainless steel, aluminum alloy, titanium alloy, nickel-based alloy, cobalt-based alloy, copper and copper alloys, etc., or in the scenario of multi-material additive manufacturing, the different powder materials corresponding to each feature region. Further, "material properties" can be refined into thermophysical parameters related to the component 111 or each feature region, including thermal conductivity, laser absorption rate, specific heat capacity, melting temperature, and surface tension, etc. Thermophysical parameters will affect the transfer and distribution of laser energy in the powder and molten pool, and are important factors determining the molten pool morphology, temperature gradient, and solidification rate. Therefore, when allocating beam modes to each feature region, its material properties can be preferentially considered to ensure that the selected light intensity distribution and / or phase modulation scheme can match the thermal response characteristics of the material to achieve the best melting effect and microstructure control.

[0054] "Process objectives" refer to various performance or quality indicators desired during the additive manufacturing process, which can be selected or combined according to specific application requirements. For example, the process objectives can be the reduction of at least one of thermal stress, residual stress, surface roughness, structural warping, porosity and / or the improvement of at least one of density, build efficiency, energy efficiency ratio, mechanical property isotropy, dimensional accuracy, and interlayer bonding strength. Exemplarily, reducing local or overall thermal stress and residual stress to reduce the crack risk and deformation inside the component; reducing surface roughness and structural warping to improve the surface quality and geometric accuracy of the component; reducing porosity to improve the forming density and mechanical properties; improving build efficiency and energy efficiency ratio to reduce manufacturing time and energy consumption; improving the isotropy of mechanical properties to obtain a uniform mechanical response; improving dimensional accuracy and interlayer bonding strength to achieve the dimensional accuracy of the component and the firm bonding between layers. In practical applications, one or more of the above process objectives can be selected for optimization.

[0055] "Beam mode" refers to different irradiation patterns formed by modulating the light intensity distribution and / or phase of a light beam to control the morphology of the molten pool and the distribution of the thermal field. In some embodiments, the beam mode may include, but is not limited to, at least two types among Gaussian mode, annular mode, flat-top mode, Bessel mode, vector mode, and saddle mode. As shown in Figure 3A , (a)-(d) thereof respectively show the spot models of flat-top, Gaussian, annular, and Bessel modes. The Gaussian mode has the characteristics of the highest central energy density and fast edge attenuation, and is suitable for scenarios requiring deep penetration and rapid deposition; the light intensity of the annular mode is concentrated in the annular region, which is beneficial to reducing central overburn and improving surface flatness; the flat-top mode has a uniform light intensity distribution and is suitable for improving the forming density and surface quality; the Bessel mode has a long non-diffracting transmission distance and a large depth of focus, and is suitable for complex geometric structures sensitive to the depth of focus. In addition, the vector mode realizes precise control of the lateral diffusion of the molten pool through a directional light intensity distribution; the saddle mode forms high-energy regions at the center and symmetric sides to optimize the multi-directional thermal field distribution of complex structures.

[0056] Figure 3B shows the crystal orientation and local average orientation difference (KAM) characterization of In738 alloy specimens under different beam modes. As shown in Figure 3B , (a) and (e) thereof, under the action of the flat-top beam, the molten pool obtains a uniform energy distribution and maintains a stable heat flow field throughout the depth of the molten pool. The cooling rate is consistent throughout the molten pool area, and the columnar crystals have strong epitaxial growth characteristics, forming an orientation-consistent {001}<001> near single-crystal structure. It shows that the material has good high-temperature performance and creep performance; in the corresponding KAM characterization, the local KAM value only reaches below 3° in very few regions, and the average KAM is between 0-2°, indicating that both the residual stress and dislocation density are very low. As shown in Figure 3B , (b) and (f) thereof, due to the high central energy density and fast edge attenuation of the Gaussian beam, fine equiaxed crystals are formed at the top of the molten pool, while columnar crystals grow along the thermal gradient direction in the center of the molten pool, and the texture is significantly weakened. The number of fine equiaxed crystals is significantly increased compared with the flat-top light specimen. The KAM value reaches 4-5° in the central region and gradually decreases to ≤1° outward, indicating the concentration of residual stress under the high central thermal gradient. As shown in Figure 3B , (c) and (g) thereof, the annular beam provides the highest energy in the annular region, enhancing the stability of the molten pool. However, the grains in the center and the outer ring are coarser due to lower energy or cumulative overheating, forming a ring-like tissue characteristic. The KAM value in the annular region is about 3-4°, and the center and the outer ring are below 2°, indicating that the shear stress or thermal gradient is more significant at the annulus. As shown in Figure 3BIn (d) and (h), due to the non-diffracting property and long focal depth advantage of the Bessel beam, near equiaxed crystals that extend uniformly along the optical axis direction are generated, and the KAM value is overall high, reaching 4 - 5°, indicating large lattice distortion and residual stress.

[0057] Figure 4A The first beam mode allocation scheme of the beam mode is shown. In the first beam mode allocation scheme, the control system 103 can allocate the beam mode according to the structural type and material property of each characteristic region.

[0058] Exemplarily, for the overhanging structure area, due to the lack of support below this area during the printing process, suspended molten pools and thermal overhanging effects are likely to occur, resulting in the drooping of the molten pool and unstable forming. Considering the absorption rates and thermal conductivities of typical metal powders such as stainless steel and titanium alloys, the control system 103 preferentially allocates a saddle-shaped beam or a vector beam to it. The saddle-shaped beam has high energy bands both in the center and on both sides, and can form three controlled melting points at the overhanging part, which can not only strengthen the heating of the central area, but also provide thermal "support" on both sides to reduce the drooping of the molten pool; the vector beam can enhance the energy in one-dimensional or two-dimensional directions and can directionally compensate for the heat loss in the overhanging area. For the curved surface structure area, the surface has a continuous curvature change, and local focal depth offset is likely to cause overburning or lack of fusion. Considering the high reflectivity of the metal powder and the change of the incident angle on the curved surface, the control system 103 selects a Bessel beam. The Bessel beam has a long focal depth and non-diffraction characteristics, and can keep the spot size and shape stable within a certain range, adapting to the focal point displacement caused by the undulation of the curved surface, so as to ensure uniform melting and consistent forming of the curved surface area. For the porous structure area, the internal pores in this area are connected and the channels are small, requiring rapid filling and reducing pore residues. Considering the thermal diffusion characteristics of high thermal conductivity materials such as copper and copper alloys near the pores, the control system 103 allocates a Gaussian beam to it. The Gaussian mode has concentrated central energy and a high peak, which can quickly generate a sufficient molten pool depth in the porous area, fill the fine channels and promote pore closure. At the same time, the rapid attenuation of the energy at its edge is beneficial to restricting the diffusion range of the molten pool and reducing the overheating of the surrounding microstructures. For the lattice structure area, its regular lattice-like microstructure has requirements for both local heating and supporting mechanical properties. The control system 103 allocates a flat-top beam to it. The flat-top mode has a uniform light intensity distribution, and can achieve stable melting in each lattice unit, reducing the appearance of hot spots and cold spots, so as to ensure the uniform density and mechanical consistency of the lattice structure. For the thin-walled structure area, due to the small wall thickness, it is easy to overburn or warp due to over-concentrated energy, and the thermal diffusion of the material is limited. The control system 103 preferably allocates a flat-top beam. The flat-top mode has a uniform light intensity distribution, and the energy is not concentrated in the center, reducing the risk of local overheating; at the same time, its wider spot diameter is beneficial to reducing the thermal gradient, thereby improving the surface roughness and dimensional accuracy of the thin-walled area. For the thick-walled structure area, a deeper molten pool and higher building efficiency are required to reduce the lack of fusion defects between layers. Considering the high heat capacity of the material in the thick-walled area, the control system 103 can also allocate a Bessel beam. Its long focal depth characteristic ensures that the focal point is not easily offset during deep melting, and the non-diffraction characteristic is beneficial to maintaining the molten pool depth, improving the fusion quality and building speed. For the connecting structure area, this area usually serves as a transition part between different geometric units, and it is necessary to control the heat input simultaneously to prevent stress concentration and cracking. The control system 103 can allocate an annular beam or a vector beam. The annular mode reduces the direct energy by leaving a blank in the center, which can reduce the thermal stress at the connection; the vector mode enhances the energy in a directional manner and directionally strengthens the fusion strength of the connection part, so that the transition area is firmly combined.

[0059] Figure 4B The second beam mode allocation scheme is shown. In the second allocation scheme, the control system 103 can allocate beam modes according to the structural types and process objectives of the respective characteristic regions.

[0060] Exemplarily, for process objectives mainly focused on reducing thermal stress and residual stress, in the overhang and connection structure regions where local temperature gradients and stress concentrations are likely to occur, the control system 103 preferentially allocates beam modes with annular or vector energy distributions. The annular beam has a low-energy band in the center, which can reduce the temperature peak in the central region and inhibit the generation of thermal stress; the vector beam adjusts the molten pool shape and cooling rate through directional energy enhancement, thereby reducing regional stress concentration and improving the residual stress distribution. In the porous and lattice structure regions, to reduce pore residue and ensure microstructural density, the control system 103 can take reducing porosity as the process objective and allocate Gaussian or flat-top beams. The high-energy characteristic at the center of the Gaussian beam is suitable for quickly filling pores, and the uniform energy distribution of the flat-top beam is beneficial to maintaining the uniform forming of the micro-grid structure. For thin-wall and curved surface structure regions, the main process objectives are usually to improve surface quality, dimensional accuracy, and suppress warping. The control system 103 allocates flat-top or Bessel beams for this purpose: the flat-top beam reduces the temperature gradient through uniform heating and improves the surface roughness; due to its long depth of focus characteristic, the Bessel beam can keep the focus stable and reduce local overburning or unfused defects caused by focal deviation in the curved surface and thin-wall regions. In the thick-wall regions, for example, when the process objective is to achieve construction efficiency and mechanical property consistency, the control system 103 therefore allocates Bessel beams or Gaussian beams with strong deep penetration ability to accelerate the formation of the molten pool and achieve interlayer fusion strength. At the same time, an appropriate energy distribution can reduce interlayer residual stress and improve the isotropy of the overall mechanical properties. In addition, for complex regions that require simultaneous optimization of multiple process objectives, such as functional gradient or local strengthening parts, the control system 103 can select beam modes that can provide compromise performance in multiple dimensions for the characteristic regions according to the target priority, such as alternately using flat-top beams and annular beams to balance surface quality and thermal stress control.

[0061] In the third allocation scheme, the control system 103 can allocate beam modes by means of multi-dimensional comprehensive evaluation according to the structural types, material properties, and process objectives of the respective characteristic regions.

[0062] Exemplarily, the control system 103 first establishes a vector for each feature region, which includes structure category identifiers (such as overhang, thin wall, thick wall, etc.), material thermal property parameters (such as thermal conductivity, absorptivity, specific heat capacity, melting temperature, etc.), and process target weights (such as reducing thermal stress, reducing residual stress, improving density, improving surface quality, etc.). Subsequently, based on the pre-calibrated optical-pool response model, the control system 103 calculates the influence of different beam patterns on the pool geometry, temperature field distribution, and cooling rate under this vector, and uses a multi-objective optimization algorithm (such as genetic algorithm, particle swarm optimization, or weighted summation method) to rank the comprehensive fitness of each pattern. During the optimization process, the control system 103 can dynamically adjust the weights of each process target to reflect the priority requirements in the manufacturing stage or the real-time feedback of the user. Finally, the control system 103 automatically selects the beam pattern with the highest score under the current conditions for allocation.

[0063] 203. Apply the beam pattern to the beam so that the beam switches to its corresponding beam pattern when scanning multiple feature regions on the powder bed.

[0064] The control system 103 can generate a layer-by-layer scanning plan according to the mapping relationship between the feature regions and the beam patterns obtained in step 202, where each scanning path segment corresponds to a feature region and its assigned beam pattern. During the actual scanning of the powder bed 110, the control system 103 drives the beam to move along a predetermined path through the beam deflector 102. When the beam is about to enter a new feature region, the control system 103 sends a mode switching instruction to the shaping unit 104 and triggers the switch using the position feedback or encoder signal of the beam deflector 102. For the shaping unit 104 using a DOE array, the micro motor or piezoelectric actuator translates or rotates the corresponding DOE element to the optical path center after receiving the switching instruction; for the shaping unit 104 using a spatial light modulator assembly, the control system 103 loads a new digital phase mask to the control chip of the spatial light modulator, so that the phase pattern is updated in the next refresh cycle. Through the above continuous switching process, the beam can always maintain the corresponding optimal pattern when scanning each feature region, realizing the targeted energy control of different structural regions.

[0065] In some embodiments, the beam pattern further includes a transition pattern between at least two beam patterns and / or a composite pattern formed by superimposing or segmenting at least two beam patterns.

[0066] The transition mode forms a continuously varying light intensity and phase gradient in the spatial or temporal dimension by interpolating or weighted superposition of the phase distributions or amplitude distributions of two or more adjacent beam modes, so that when the beam switches from one mode to another, the heat input to the molten pool does not change suddenly, avoiding stress concentration or geometric mismatch. Specifically, the control system 103 specifies a transition duration or a transition interval length before and after the mode switch, and sends multiple frames of phase masks or a multi-level DOE position sequence to the spatial light modulator or the DOE array, successively changing the phase pattern or the physical element position to achieve a predetermined transition curve.

[0067] The composite mode generates a composite light spot with multiple light intensity distribution characteristics by superimposing two or more beam modes at the same time or projecting them in segments in space. For example, the control system 103 can load two DOEs with different phase patterns in the optical path at the same time, and generate a light intensity distribution with a flat top in the center and a ring in the periphery through the superimposed composite phase mask; or partition the spatial light modulator, and load different phase masks in each area to form a partitioned Gaussian and Bessel hybrid characteristic in the same light spot. The composite mode can meet multiple process objectives such as deep penetration, densification, and surface quality in a single scan for the complex requirements of specific structural regions (such as functional gradients or local strengthening).

[0068] Exemplarily, for example, for a lattice structure region, the conventional Gaussian mode can quickly fill the micro-cell units, while the annular mode provides uniform heating in the periphery, which is beneficial to improving the lattice connection. At this time, the control system 103 can assign a "Gaussian-annular" composite mode to the lattice region, that is, in the same scanning process, the central Gaussian distribution and the peripheral annular distribution are superimposed through the shaping unit 104 to achieve the dual effects of densification inside the micro-cell and strengthening of the boundary. Another example is that at the boundary of the region where the thin-walled structure switches to the curved surface structure, if directly jumping from the flat-top mode to the Bessel mode, it is easy to cause a sudden change in the heat input to the molten pool. At this time, the control system 103 can generate a "flat-top-Bessel" transition mode, and make the light intensity distribution slowly transition from the flat top to the Bessel form by loading intermediate phase masks in stages, so that the shape and temperature gradient of the molten pool change smoothly. Another example is that at the junction of the overhanging structure and the connecting structure, the control system 103 can adopt a "vector-annular" composite mode, while enhancing in the vector direction, using the annular peripheral energy band to reduce the central thermal stress.

[0069] Figure 5 Another example of the method 200 for regulating the multi-mode shaped beam is shown, including further steps 211 and 212 that make up the method 200.

[0070] 211. Identify the characteristic region and divide it into at least two grid units.

[0071] The control system 103 can divide each feature region into at least two functional grid units according to the geometric boundary of the feature region and the scanning layer thickness. For example, contour grid units and filling grid units. The contour grid units correspond to the boundary lines or near-boundary regions of the feature region, requiring high-precision melting and dimensional control; the filling grid units correspond to the larger filling portions inside the region, and relatively more attention is paid to the forming efficiency and density.

[0072] 212. Assign different beam patterns to the at least two grid units according to the structural type and in combination with the material properties and / or process objectives.

[0073] After the grid division is completed, the control system 103 assigns different beam patterns to the contour grid units and the filling grid units respectively in step 212 in combination with the structural type, material properties and / or process objectives of each grid unit. Specifically, the contour grid units are preferentially assigned a flat-top or vector pattern with a smaller spot diameter and a uniform light intensity distribution to achieve the control of boundary dimension accuracy and surface quality; the filling grid units can be assigned a Gaussian or Bessel pattern with a larger spot diameter and strong deep melting ability to improve the scanning speed and filling density. In addition, for grid units with special structures such as porous or lattice structures, the control system 103 can further subdivide between filling and contour and adopt an annular pattern or a composite pattern to optimize pore closure and connection strength. By implementing gridification and differential beam pattern assignment inside the feature region, the control system 103 can achieve dual optimization of boundary accuracy and internal performance in the same region.

[0074] Figure 6 Another example of the method 200 for regulating a multi-mode shaped beam is shown, including steps 221 and 222 that further constitute the method 200.

[0075] 221. Assign corresponding process parameters to a plurality of feature regions according to the structural type and in combination with the material properties and / or process objectives. Wherein the process parameters include at least one of beam power, scanning speed, spot size and scanning trajectory.

[0076] In addition to the beam pattern, the control system 103 can also assign and apply corresponding process parameters to each feature region. For example, the control system 103 can determine the optimal combination of process parameters such as beam power, scanning speed, spot size and scanning trajectory according to the structural type of each feature region and its material properties and / or process objectives. For example, for a thin-wall structure region, a lower beam power and a higher scanning speed can be assigned to reduce local heat input and prevent overburning; for a thick-wall structure region, a higher beam power and a lower scanning speed can be assigned to achieve sufficient melt depth and interlayer fusion strength; for a porous or lattice structure region, the control system 103 can preferably select a smaller spot size and an overlapping scanning trajectory to achieve pore closure and uniform density.

[0077] 222. Apply the process parameters to the light beam so that the light beam switches to the corresponding process parameters when scanning multiple feature regions on the powder bed.

[0078] After completing the process parameter allocation, the control system 103 issues the determined process parameters to the light beam emitter 101 and the light beam deflector 102, and combines with the mode switching instruction of the shaping unit 104 to achieve dynamic switching of multiple parameters. During the powder bed scanning process, the control system 103 can trigger the process parameter switching at the start of each feature region by real-time monitoring the scanning position and layer number information, so that the light beam power, scanning speed, spot size and trajectory mode are synchronously adjusted to the predetermined values of this region. In this way, the control system 103 can apply different energy inputs and scanning behaviors to different structural regions during the manufacturing process of the same component to achieve a balance between overall efficiency and local quality.

[0079] The control system 103 can compensate for the differences in molten pool behavior under different beam modes by optimizing process parameters, and reduce the non-uniform shrinkage stress during the cooling process by dynamically adjusting the spot shape and scanning strategy, etc. Exemplarily, when the control system 103 assigns a specific beam mode to a certain feature region, it can automatically optimize the following process parameters based on the light intensity distribution and phase characteristics of the mode: finely adjust the shape of the spot within the same region. For example, when using the Gaussian mode, the spot diameter can be slightly expanded to smooth the edge energy decay; in the Bessel mode, the intensity ratio of the zero-order and high-order Bessel components can be adjusted to optimize the ratio of the molten pool depth to width. For example, rotational scanning is adopted within the feature region, that is, within each layer or each grid cell, the scanning direction is changed according to a predetermined angle (such as 90°, 120° or randomly), so that the heat input direction continuously changes between adjacent scanning trajectories, thereby making the heat gradient more uniform and the residual stress distribution more balanced. In complex geometry or stress-sensitive regions, the control system 103 can further divide the feature region into multiple sub-regions and set different scanning starting points and scanning directions for each sub-region. For example, at the junction of thin walls and thick walls, the thin-wall sub-region can adopt linear scanning along the long axis of the thin wall, while the thick-wall sub-region scans along its short axis or diagonal direction to make the heat accumulation and cooling path more uniform and reduce the interface stress concentration. For thick-wall regions with a large heat capacity, the control system 103 can use a "preheating" strategy with a high power and a low speed at the initial stage of scanning, and then switch to a "filling" strategy with a low power and a high speed to control the temperature gradient and improve the efficiency. On the contrary, for thin-wall or overhanging regions, a stable molten pool can be formed first by using low power and slow scanning, and then the speed can be gradually increased to reduce the thermal shock. The control system 103 can freely switch the scanning trajectory between the contour and filling of the feature region. For example, after completing the boundary contour scanning, immediately perform internal filling along the opposite direction or a spiral trajectory to disperse the heat concentration and balance the stress. For porous or lattice structures, cross-filling or honeycomb scanning trajectories can also be adopted to further uniform the thermal field.

[0080] Figure 7 Another example of the method 200 for regulating a multi-mode shaped beam is shown, including further steps 231 and 232 that make up the method 200.

[0081] 231. Perform test scans on multiple feature regions respectively in different beam modes to obtain test data characterizing the corresponding process objectives of the multiple feature regions.

[0082] The control system 103 can perform test scans of different beam patterns on each feature region separately to obtain test data characterizing the achievement degree of the process objectives for each region. Specifically, the control system 103 can sequentially load different patterns for each feature region according to a predefined list of beam patterns, and perform short-range test scans under the same scan path and layer thickness conditions. During the test process, key process index data is synchronously collected, including the molten pool temperature field curve, molten pool geometric dimensions, cooling rate, surface roughness of the formed layer, microstructure orientation distribution, and local average misorientation (KAM), etc. All test data is recorded in the database in a one-to-one correspondence with the region and the pattern.

[0083] 232. Determine the beam pattern and / or process parameters that enable multiple feature regions to achieve the process objective based on the test data, and determine them as those to be allocated to the corresponding multiple feature regions.

[0084] For example, the control system 103 can use only the beam pattern as the optimization variable and compare the test results of each pattern in the same region. The control system 103 sorts the test data according to the process objectives (such as minimum thermal stress, highest density, or lowest surface roughness), and selects the beam pattern with the highest achievement degree in each feature region as the final allocation pattern. For example, the control system 103 can also use the beam pattern and process parameters (including beam power, scan speed, spot size, and scan trajectory) as optimization variables simultaneously. For each beam pattern, the control system 103 fine-tunes the process parameter combination in adjacent test scans, collects test data in more dimensions, and comprehensively evaluates the achievement degree of the pattern-parameter combination through a multi-dimensional scoring function. Finally, the control system 103 can select the beam pattern and process parameter combination that can maximize the objective function for each feature region as the actual scan configuration. In addition, the control system 103 can also independently test only the process parameters when needed to further fine-tune the processing conditions of each region.

[0085] Figure 8 Another example of the method 200 for regulating a multi-mode shaped beam is shown, including steps 241 and 242 that further constitute the method 200.

[0086] 241. Use finite element simulation to simulate the temperature field, stress field, and molten pool morphology of multiple feature regions under different beam patterns to obtain simulation data for characterizing the process objective.

[0087] The control system 103 can numerically simulate the thermo-mechanical behavior of each characteristic region under different beam modes based on the geometric models and material properties (such as thermal physical parameters) of multiple characteristic regions using finite element simulation. Specifically, the control system 103 can establish a heat conduction model that includes the characteristics of the beam mode, which converts the light intensity distribution and phase information of the beam mode into surface heat flux boundary conditions, and combines parameters such as the thermal conductivity, specific heat capacity, density, and latent heat of fusion of the material to solve the two-dimensional or three-dimensional temperature field evolution. Subsequently, based on the obtained temperature gradient field, the control system 103 calculates the stress field distribution of each characteristic region through thermo-stress coupling analysis, focusing on the peak thermal stress and the uniformity of the residual stress distribution; at the same time, it can also use the molten pool dynamics model or the phase field method to simulate the geometric shape change of the molten pool and obtain indicators such as the depth, width, and solidification rate of the molten pool. The above simulation process can be executed in parallel for beam modes such as Gaussian, annular, flat-top, Bessel, vector, and saddle-shaped, and the temperature field, stress field, and molten pool morphology data under each beam mode are stored in the simulation database.

[0088] 242. Determine the beam mode and / or process parameters that enable multiple characteristic regions to achieve the process target based on the simulation data, and determine them as those to be allocated to the corresponding multiple characteristic regions.

[0089] The control system 103 can compare the peak thermal stress, residual stress, and molten pool geometric uniformity of each beam mode in the same characteristic region according to the indicators in the simulation database, and select the beam mode that optimizes the process target (such as the minimum thermal stress, stable molten pool, or best surface flatness) as the allocation mode for this region. For scenarios that require optimizing multiple targets simultaneously, the control system 103 can use multi-objective sorting or weighted evaluation methods to score the comprehensive performance of each beam mode and select the mode with the highest score. In addition, the control system 103 can also incorporate process parameters (including beam power, scanning speed, spot size, and scanning trajectory) into the simulation optimization together with the beam mode. By parametrically scanning the main process parameter combinations under each beam mode, using finite element simulation to obtain temperature-stress-molten pool response data, and performing global optimization according to multi-dimensional objective functions (such as the weighted sum of thermal stress and residual stress, molten pool geometric consistency index, solidification rate difference, etc.), the optimal mode-parameter combination is determined.

[0090] Exemplarily, in finite element simulation, a typical heat conduction model can be described by the following partial differential equation:

[0091]

[0092] Among them, x, y, z represent the coordinate components in the three-dimensional space coordinate system, which are used to define the spatial position of the powder bed or the component; t represents the time variable, which is used to describe the evolution of temperature over time; ρ represents the material density; c p represents the specific heat capacity; k represents the thermal conductivity, represents the spatial gradient operator; represents at the spatial position and the absolute temperature at time t; is the gradient in space, representing the direction and magnitude of the change in temperature with spatial position; represents the laser heat source term, which is used to describe the heat power absorbed per unit volume at the spatial position (x, y, z) and time t.

[0093] Laser heat source term can be described as:

[0094]

[0095] Among them, η represents the absorption efficiency of the material to the laser; I0 represents the peak intensity of the incident laser; represents the normalized distribution function (such as Gaussian or annular) of the light spot on the x-y cross-section; represents the exponential decay function of the laser energy along the depth direction z, which is used to describe how the energy rapidly weakens with depth when the laser penetrates the material; represents the time window function, which is used to describe when the heat source reaches the coordinates (x, y), where is related to the scanning speed and the scanning trajectory.

[0096] By solving the above partial differential equation, the three-dimensional temperature field T(x, y, z, t) of the entire powder bed during the scanning process can be obtained, and then the temperature gradient , the molten pool size, and the subsequent thermal stress field distribution can be calculated.

[0097] Figure 9 Shows another example of the method 200 for regulating the multi-mode shaping beam, including further constituting steps 251-253 of the method 200.

[0098] 251. Using the structural type, material properties, beam mode, and / or process parameters of multiple characteristic regions as inputs and the degree of achievement of the process target as the output, construct a machine learning model.

[0099] The control system 103 can use the structural type identification of each feature region (such as overhang, thin wall, etc.), material properties (such as thermal conductivity, absorptivity, etc.), beam mode number, and / or process parameters (such as beam power, scanning speed, spot size, scanning trajectory) as model inputs, and the process target achievement degree (such as quantitative indicators such as residual stress, density, surface roughness, etc.) as outputs to construct a trainable machine learning model. Optional machine learning models include but are not limited to support vector machine (SVM), random forest, gradient boosting decision tree (GBDT), multi-layer perceptron neural network (MLP), convolutional neural network (CNN), etc.

[0100] 252. Train the machine learning model using test data and / or simulation data.

[0101] The control system 103 can use the test data obtained in step 231 and / or the finite element simulation data obtained in step 241 to train and validate the selected machine learning model. Taking the multi-layer perceptron neural network as an example, the control system 103 preprocesses the input features by normalization, encodes the structural type, material properties, mode number, and parameter values into numerical vectors; then designs a feedforward network with several hidden layers, and uses mean square error or cross-entropy as the loss function, and iteratively updates the network weights through backpropagation and gradient descent algorithms; after training, the control system 103 evaluates the prediction accuracy of the model through cross-validation and test sets, and adjusts the network structure or hyperparameters according to the evaluation results to obtain the optimal generalization ability.

[0102] 253. Use the trained machine learning model to predict the process target achievement degree of multiple feature regions under different beam modes and / or process parameters, and accordingly determine the beam mode and / or process parameters assigned to multiple feature regions.

[0103] Exemplarily, taking the beam mode as the optimization target, the control system 103 calls the trained model to input the structural type and material properties of each feature region, predicts the process target achievement degree under each beam mode, and automatically selects the mode number with the highest predicted value as the beam mode finally assigned to each feature region. In addition, the control system 103 can also take the combination of beam mode and process parameters as the joint optimization target, use the machine learning model to input the structural type, material properties, and multiple mode-parameter combinations, predict the corresponding process target achievement degree, and screen out the optimal mode-parameter combination from them as the processing configuration assigned to each feature region. Through the above machine learning-driven allocation method, the control system 103 can achieve the precise matching of beam mode and process parameters under complex and variable additive manufacturing conditions.

[0104] In some embodiments, the control system 103 may also capture real-time data of the temperature field, stress field, and molten pool morphology of multiple feature regions during the scanning process, input the real-time data into the trained machine learning model for prediction, and make real-time adjustments to the beam pattern and / or process parameters corresponding to the multiple feature regions according to the prediction results.

[0105] Exemplarily, during the actual scanning process, the control system 103 can use a variety of in-line sensors such as high-speed cameras, infrared thermal imagers, optical coherence tomography (OCT), or strain gauges arranged in the additive manufacturing equipment to capture the temperature field, stress field, and molten pool morphology data of each feature region in real time. The temperature field data can be obtained by the infrared thermal imager to obtain the surface temperature distribution, or the molten pool depth can be indirectly deduced by combining OCT with optical ranging; the stress field data can be measured by digital image correlation (DIC) technology or a strain gauge array; the molten pool morphology can be used by the high-speed camera to cooperate with the image processing algorithm to extract the molten pool width, depth, and edge curve features in real time. The captured real-time data can be preprocessed (denoising, calibration, and feature extraction) first, and then input into the trained machine learning model in the previous step according to the input format during training. This model quickly predicts the current process target achievement degree of each feature region based on the real-time data, such as the predicted residual stress level, porosity, or surface roughness, etc. Subsequently, the control system 103 dynamically adjusts the mode switching instruction of the shaping unit 104 and / or adjusts process parameters such as beam power, scanning speed, spot size, and scanning trajectory according to the deviation between the model output and the preset threshold or target value, so that the molten pool behavior is corrected in time and maintained within the optimal processing window. By combining closed-loop real-time feedback with machine learning prediction, the control system 103 can quickly respond to material and environmental disturbances and achieve dynamic control of each feature region.

[0106] It should be understood that compared with the traditional Gaussian beam, the shaped beam has obvious advantages in controlling the molten pool size and temperature gradient. However, in actual operation, researchers found that when directly scanning with a Gaussian beam, as the forming height of the component increases, the surface quality basically remains stable without obvious defocusing or forming deterioration; while when adding a DOE in the optical path, the depth range of the shaped spot in the Z-axis direction, that is, the focal depth, is greatly compressed, and it only maintains the best shape and focusing state within a very narrow axial interval. As the layer height of the component continues to increase, the powder bed surface is prone to deviate from the focal plane of the optical system, and the shaped spot is defocused, resulting in sudden changes in the molten pool geometry and an increase in the surface roughness of the formed surface, affecting the forming consistency of subsequent layers.

[0107] To solve this technical problem, Figure 10Another example of the optical path system 100 is shown. The optical path system 100 also has a variable aperture assembly 105 and a variable focal length lens assembly 106 respectively connected to the control system 103 through a high-speed bus. The variable aperture assembly 105 is arranged in the optical path between the shaping unit 104 and the beam deflector 102 for adjusting the spot diameter of the beam; the variable focal length lens assembly 106 is arranged in the optical path between the variable aperture assembly 105 and the beam deflector 102 for adjusting the focal length of the beam.

[0108] In a specific implementation manner, the variable aperture assembly 105 can be composed of an adjustable aperture mechanism, an aperture position sensor, and a spot control unit. For example, the adjustable aperture mechanism is composed of a set of concentrically installed blades, and its blades are driven by a micro stepping motor or a piezoelectric actuator, and can change the aperture opening diameter under millisecond-level response, so as to control the beam diameter passing through the aperture. The aperture position sensor uses an optical encoder or an inductive displacement sensor to detect the blade opening in real time, and feeds back the current aperture diameter to the control system 103 to achieve closed-loop control. The spot control unit integrates a motor driver and an aperture control algorithm, generates a corresponding drive signal according to the target spot diameter command issued by the control system 103, and performs position closed-loop correction. The variable focal length lens assembly 106 can be composed of a variable focal length lens group, a focal length measurement device, and a lens control unit. For example, the variable focal length lens group is composed of two or more movable lens elements, and the lens group spacing is adjusted by a micro linear actuator (such as a piezoelectric push rod or a stepping motor) to achieve continuous adjustment of the focal length range. The lens material is preferably optical glass or a silicon-based microlens with low dispersion and high transmittance. The focal length measurement device integrates an optical encoder or a laser interferometer in the lens group for real-time measurement of the lens spacing or the focal point position, and feeds back the measurement result to the control system 103. The lens control unit includes a motion controller and a focal length calculation module, receives the target focal length command issued by the control system 103, inversely solves the required lens spacing in combination with the optical model, and drives the actuator to adjust, while performing closed-loop error correction.

[0109] It should be understood that in some embodiments of the present application, an additive manufacturing device is also provided. The device has an optical path system 100 with any of the described structures, such as Figure 1 or Figure 10 The shown optical path system 100 can be used as part of the additive manufacturing device of the present application.

[0110] In some embodiments, the optical path system 100 may further include a ranging sensor connected to the control system 103. The ranging sensor is disposed in the optical path and is configured to measure in real time the distance from the reference position to the top layer of the powder bed. During the scanning process, the control system 103 may utilize the ranging sensor integrated in the optical path to obtain the real-time distance from the reference position to the top layer of the powder bed, and in combination with a pre-calibrated optical model, calculate the target values required for the current layer, and then drive the variable aperture assembly 105 and the variable focal length lens assembly 106 to adjust the spot diameter and the focal length to the target values in the current beam mode, so that the beam always maintains the best focus state at different layer heights.

[0111] Exemplarily, the ranging sensor may be arranged along the center line of the optical path and is configured to measure in real time the distance from the optical path reference position (reference position) to the top layer of the powder bed, denoted as Z(t). Combining with the pre-calibrated optical model, the spot radius ω and the focus offset Δf required for the current layer can be calculated:

[0112]

[0113] wherein, represents the spot radius at the designed focal plane; represents the designed focal length; represents the Rayleigh length, represents the laser wavelength; represents the height of the top layer of the powder bed measured at time t.

[0114] The control system 103 may directly convert the calculated spot radius and focus offset into variable aperture opening diameter and variable focal length lens group spacing commands according to the above formulas, and respectively send them to the variable aperture assembly 105 and the variable focal length lens assembly 106 to achieve real-time adjustment of the spot diameter and the focal length.

[0115] In some embodiments, the control system 103 may also calculate and store in advance the target values required for each layer according to the geometric model of the component and the layer thickness before scanning, generate a target value - layer number mapping table, and directly read and apply the corresponding target values according to the layer number during the scanning process, and then drive the variable aperture assembly 105 and the variable focal length lens assembly 106 to adjust the spot diameter and the focal length to the target values in the current beam mode, so that the beam always maintains the best focus state at different layer heights.

[0116] Exemplarily, before scanning, the control system 103 calculates the absolute height of the top layer of the powder bed for each layer according to the geometric model of the component 111 and the predetermined layer thickness Δh:

[0117]

[0118] where denotes the initial height of the top surface of the powder bed on the first layer, and n is the current layer number. Subsequently, the control system 103 uses the same optical model to calculate offline the spot radius ω and focus offset Δf required for each layer:

[0119]

[0120] where 、 and have the same meanings as in the above model. Store all the of all layers in the lookup table indexed by the layer number. During the additive manufacturing process, the control system 103 reads the corresponding spot radius and focus offset from the lookup table according to the current layer number n, and converts these values into the aperture opening setting value of the variable aperture assembly 105 and the lens group spacing setting value of the variable focus lens assembly 106, and issues them to the two assemblies in real time. Through the pre-calculation of "layer number - focus parameters" and the direct lookup table method, it is possible to achieve fast and accurate focusing and spot adjustment for each layer without online ranging, and ensure that the shaped beam is in the best focused state during the scanning of each layer.

[0121] Figure 11 FIG. shows a structural example of the electronic device 300. The electronic device 300 includes a processor 301 (which can be one or more) and a memory 302 (which can be one or more); the memory 302 is coupled to the processor 301 and is used to store instructions executed by the processor 301. When the instructions are executed by the processor 301, the electronic device 300 is caused to execute the method of the embodiment of the present application. Among them, the memory 302 can be a separate device independent of the processor 301 or integrated in the processor 301.

[0122] The processor 301 can be a general-purpose processor (such as a microprocessor), a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, and can also complete the decoding and execution of instructions through the combination of hardware logic circuits and software modules. The memory 302 can be a volatile memory (such as SRAM, DRAM, SDRAM, DDR SDRAM, etc.) or a non-volatile memory (such as ROM, PROM, EPROM, EEPROM, flash memory, etc.), or any combination of the two, and is used to store the operating system, application programs, data, and the execution code of the above method.

[0123] In some embodiments, the electronic device 300 further includes an input interface 303 and an output interface 304, both of which are controlled by the processor 301 and communicate with other external devices or chips. The input interface 303 can be used to receive instructions and data from external sensors, ranging modules, or host computers; the output interface 304 can send control signals or feedback information to the optical path system 100 and other actuators. The processor 301, the memory 302, and each interface can be integrated on a single chip or distributed among multiple chips or modules, and the specific form can be selected according to the system integration requirements.

[0124] It should be understood that the above configuration of the electronic device 300 is for illustrative purposes only and should not be construed as a limitation on the embodiments of the present application. Any type of processor and storage medium capable of implementing the above-described processing, storage, and communication functions can be included within the scope of protection of the present application.

[0125] Figure 12 The structural example of a computer-readable storage medium 400 is shown. A computer program is stored in the computer-readable storage medium 400, and when the computer program is executed by a processor, it implements the method described in the embodiments of the present application.

[0126] A computer-readable storage medium includes a computer program for executing a computer process on a computing device. In some embodiments, the computer-readable storage medium is provided using a signal-bearing medium 401. The signal-bearing medium 401 may include one or more program instructions, which when run by one or more processors can provide the functions or partial functions described above for the method embodiments of the present application. Thus, for example, one or more features in the method can be borne by one or more instructions associated with the signal-bearing medium 401. In addition, Figure 12 the program instructions in also describe example instructions. In some examples, the signal-bearing medium 401 may include a computer-readable medium 402, including a non-volatile storage carrier. Specifically, a hard disk drive, a compact disc (CD), a digital video disc (DVD), a digital tape, a read-only memory (ROM), or a flash memory chip (such as NOR / NAND type memory) can be selected during implementation, and its data storage characteristics meet the requirements for long-term retention of computer programs.

[0127] In some embodiments, the signal-bearing medium 401 may include a computer-recordable medium 403, including a rewritable carrier. Specifically, during implementation, it covers random access memory (RAM), rewritable compact disc (CD-RW / DVD-RW), solid-state drive (SSD), and phase change memory (PCM), which support dynamic updating of program instructions through a read-write controller.

[0128] In some embodiments, the signal-bearing medium 401 may include a communication medium 404, such as, but not limited to, digital and / or analog communication media (e.g., fiber optic cables, waveguides, wired communication links, wireless communication links, etc.).

[0129] The signal-bearing medium 401 may be conveyed by a wireless form of the communication medium 404 (e.g., a wireless communication medium compliant with the IEEE 802.11 standard or other transmission protocols). One or more program instructions may be, for example, computer-executable instructions or logic implementation instructions.

[0130] It should be understood that the units and algorithm steps of the examples described in the embodiments disclosed herein can be implemented in electronic hardware, or in a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. A person skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0131] Those skilled in the art will readily conceive of other embodiments of this application after considering the specification and practicing the content disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application, which follow the general principles of this application and include well-known knowledge or conventional technical means in the technical field not disclosed in this application. The specification and embodiments are only regarded as exemplary, and this application is not limited to the precise structures described above and shown in the drawings, and various modifications and changes can be made without departing from its scope.

Claims

1. A method for regulating a multi-mode shaping beam in powder bed additive manufacturing, wherein the additive manufacturing manufactures a component by scanning a powder bed stacked layer by layer with a beam propagating along an optical path, characterized in that, The method includes: Analyzing the geometric model of the component to obtain a plurality of characteristic regions corresponding to different structural types, including at least a part of the component; Assigning corresponding beam modes to the plurality of characteristic regions according to the structural type, in combination with material properties and / or process objectives, where the beam modes are modulated to have different light intensity distributions and / or phases; Applying the beam mode to a beam so that the beam switches to its corresponding beam mode when scanning the plurality of characteristic regions on a powder bed.

2. The method according to claim 1, wherein The structural type includes at least two of overhang, curved surface, porous, lattice, thin wall, thick wall, and connection structure, where the thickness of the thin wall is less than a first threshold, and the thickness of the thick wall is greater than a second threshold; The beam modes include at least two of Gaussian, annular, flat top, Bessel, vector, and saddle shape.

3. The method according to claim 2, wherein The beam mode further includes a transition mode between at least two of the beam modes and / or a composite mode formed by superimposing or segmenting at least two of the beam modes.

4. The method according to claim 1 or 2 or 3, characterized in that The method further includes: Identifying the characteristic regions to divide them into at least two grid cells; Assigning different beam modes to the at least two grid cells according to the structural type, in combination with material properties and / or process objectives.

5. The method according to claim 1, characterized in that, The method further includes: Assigning corresponding process parameters to the plurality of characteristic regions according to the structural type, in combination with material properties and / or process objectives, where the process parameters include at least one of beam power, scanning speed, spot size, and scanning trajectory; Applying the process parameters to a beam so that the beam switches to its corresponding process parameters when scanning the plurality of characteristic regions on a powder bed.

6. The method according to claim 1 or 5, characterized in that, The process objectives include at least one of reducing thermal stress, residual stress, surface roughness, structural warping, porosity, and / or enhancing at least one of density, build efficiency, energy efficiency ratio, mechanical property isotropy, dimensional accuracy, and interlayer bonding strength.

7. The method according to claim 6, wherein The method further includes: Performing test scans on the plurality of characteristic regions with different beam modes respectively to obtain test data characterizing the corresponding process objectives of the plurality of characteristic regions; Determining the beam mode and / or process parameters that enable the plurality of characteristic regions to achieve the process objectives according to the test data, so as to determine them as those to be assigned to the corresponding plurality of characteristic regions.

8. The method according to claim 7, wherein The method further includes: Simulating the temperature field, stress field, and molten pool morphology of the plurality of characteristic regions under different beam modes using finite element simulation to obtain simulation data for characterizing the process objectives; Determining the beam mode and / or process parameters that enable the plurality of characteristic regions to achieve the process objectives according to the simulation data, so as to determine them as those to be assigned to the corresponding plurality of characteristic regions.

9. The method according to claim 8, characterized in that, The method further includes: Constructing a machine learning model with the structural type, material properties, beam mode, and / or process parameters of the plurality of characteristic regions as inputs and the degree of achievement of the process objectives as outputs; Training the machine learning model using the test data and / or simulation data; Predict the process target achievement degrees of the multiple feature regions under different beam patterns and / or process parameters by using the trained machine learning model, and accordingly determine the beam patterns and / or process parameters allocated to the multiple feature regions.

10. The method according to claim 9, wherein The method further includes: Capturing real-time data of the temperature field, stress field, and molten pool morphology of the multiple feature regions during the scanning process; Inputting the real-time data into the trained machine learning model for prediction, and making real-time adjustments to the beam patterns and / or process parameters corresponding to the multiple feature regions according to the prediction results.

11. The method according to claim 2, characterized in that A variable aperture assembly and a variable focal length lens assembly are respectively arranged in the optical path. When applying a non-Gaussian beam pattern to the beam, the method further includes: Driving the variable aperture assembly and the variable focal length lens assembly to adjust the spot diameter and focal length to the target values under the current beam pattern, so that the beam always maintains the best focusing state at different layer heights.

12. The method according to claim 11, wherein The obtaining method of the target values includes: During the scanning process, using a distance measuring sensor integrated in the optical path to obtain the real-time distance from the reference position to the top layer of the powder bed, and combining with a pre-calibrated optical model to calculate the target values required for the current layer; or before the scanning, calculating and storing the target values required for each layer in advance according to the geometric model and layer thickness of the component, generating a target value-layer number mapping table, and directly reading and applying the corresponding target values according to the layer number during the scanning process.

13. An optical path system for an additive manufacturing device, characterized in that, The optical path system includes: A beam emitter for generating and emitting a beam; A beam deflector for deflecting the beam to the powder bed stacked layer by layer along a preset scanning path to manufacture a component; A shaping unit arranged in the optical path between the beam emitter and the beam deflector for modulating the beam into multiple beam patterns with different light intensity distributions and / or phases; A control system connected to and controlling the beam emitter, the beam deflector, and the shaping unit to execute the method according to any one of claims 1 to 10.

14. An optical path system for an additive manufacturing device, characterized in that, The optical path system includes: A beam emitter for generating and emitting a beam; A beam deflector for deflecting the beam to the powder bed stacked layer by layer along a preset scanning path to manufacture a component; A shaping unit arranged in the optical path between the beam emitter and the beam deflector for modulating the beam into multiple beam patterns with different light intensity distributions and / or phases; A variable aperture assembly arranged in the optical path between the shaping unit and the beam deflector for adjusting the spot diameter of the beam; A variable focal length lens assembly arranged in the optical path between the variable aperture assembly and the beam deflector for adjusting the focal length of the beam; A distance measuring sensor arranged in the optical path for measuring the distance from the reference position to the top layer of the powder bed in real time; A control system connected to and controlling the beam emitter, the shaping unit, the variable aperture assembly, the variable focal length lens assembly, the beam deflector, and the distance measuring sensor to execute the method according to any one of claims 1 to 12.

15. An additive manufacturing device, characterized in that, Including the optical path system according to claim 13 or 14.

16. An electronic device, characterized in that, Including: At least one processor; At least one memory; The at least one memory is coupled to the at least one processor and is configured to store instructions executable by the at least one processor, which when executed by the at least one processor, cause the electronic device to perform the method according to any one of claims 1 to 12.

17. A computer-readable storage medium, characterized in that, A computer program is stored in the computer-readable storage medium, and when the computer program is executed by a processor, the method according to any one of claims 1 to 12 is implemented.

Images