A method for enhancing metallurgical bonding between layers of 3D printed aerospace engines

Abstract

The present invention discloses a method for enhancing metallurgical bonding between layers of 3D printed aerospace engines, comprising the following steps: S1, slicing a 3D model of an aerospace engine to construct a printing path graph structure; S2, constructing a lattice consistency map of the current printed layer; S3, aligning the lattice consistency maps of the current printed layer with those of the previous printed layer to generate a cross-layer lattice map sequence; S4, inputting the cross-layer lattice map sequence into a graph neural control body model, obtaining the disturbance sensitivity vector and disturbance score value of each node, and outputting a corrected control strategy vector; S5, performing a cross-layer disturbance projection operation based on the corrected control strategy vector to generate a control intervention instruction; S6, repeating steps S2 to S5 to form a complete lattice map evolution record sequence. The present invention utilizes lattice map modeling and a graph neural control body model to achieve metallurgical enhancement between printed aerospace engine layers, with advantages such as closed-loop control and structural continuity.

Description

Technical Field

[0001] The present invention relates to the field of metallurgical bonding control technology, and in particular to a method for enhancing metallurgical bonding between layers of 3D printed aerospace engines. Background Art

[0002] In the aerospace sector, the increasing complexity of engine structures and the extreme development of service conditions are placing higher demands on the manufacturing precision, microstructural properties, and overall service life of metal components. Metal 3D printing technology, particularly additive manufacturing processes such as selective laser melting (SLM) and electron beam melting (EBM), has become an important means of manufacturing key aerospace engine components due to its ability to achieve integrated forming of complex structures, high material utilization, and a wide range of adjustable parameters. However, during the layer-by-layer printing process, the quality of the interlayer metallurgical bond remains a major bottleneck restricting its reliability. Due to the uneven heat input conditions between different layers, the variability in cooling rates, and the incoordination of grain growth directions, metallurgical defects such as microcracks, pores, and grain boundary mismatches are easily formed in the interlayer region, thereby affecting the material's structural continuity and consistency of mechanical properties.

[0003] In existing technologies, researchers often use static parameter optimization or image monitoring of the printing process to improve interlayer bonding quality. Static optimization methods typically use experimental design to adjust parameters such as printing power, scanning speed, and scanning strategy in order to achieve a high-density macro-forming result. Image monitoring methods, on the other hand, rely on infrared thermal imagers, optical cameras, or laser reflective imaging devices to monitor the printing process online to identify possible melt pool instability, spatter anomalies, or surface texture defects. While these methods have improved printing stability to some extent, they generally suffer from limitations such as response lag, uncontrollable microstructure, and a lack of cross-layer structural understanding and intelligent feedback mechanisms.

[0004] In terms of microstructure modeling, some studies have attempted to incorporate machine learning models for pattern recognition in image data. However, most studies focus on static analysis of a single printed layer and lack systematic modeling of grain evolution as it progresses through the printed layer. In particular, existing methods for lattice-level structural control often fail to jointly model cross-layer lattice orientation consistency, thermal field conduction continuity, and stress evolution trends. This makes it difficult for control methods to provide actionable predictions and interventions based on microscopic metallurgical behavior.

[0005] Furthermore, traditional control methods rely primarily on feedforward control logic based on preset printing paths and parameters. These methods lack closed-loop adjustment capabilities based on the actual forming state, making it impossible to dynamically identify and locally adjust key control nodes within the path graph structure. Even some advanced systems that incorporate path replanning mechanisms often rely on compensating for macroscopic printing errors, failing to accurately map to lattice-level structural offsets, thermal-stress coupling disturbances, or microscopic defect predictions. Consequently, it is difficult to fundamentally enhance metallurgical continuity.

[0006] Therefore, how to provide a method for enhancing metallurgical bonding between layers of 3D printed aerospace engines is an urgent problem that needs to be solved by those skilled in the art. Summary of the Invention

[0007] One purpose of the present invention is to propose a method for enhancing metallurgical bonding between layers of 3D printing of aerospace engines. The present invention utilizes lattice consistency map construction and graph neural control body modeling methods to accurately perceive and dynamically model the evolution state of grains in each layer during the 3D printing process of aerospace engines. Combined with disturbance scoring and cross-layer projection mechanisms, the present invention outputs path-level control strategies and implements real-time closed-loop intervention, effectively improving the quality of metallurgical bonding between layers. The method has the advantages of high control accuracy, strong feedforward adjustment, good structural continuity, and adaptability to complex morphology paths.

[0008] A method for enhancing metallurgical bonding between layers of 3D printed aerospace engines according to an embodiment of the present invention includes the following steps:

[0009] S1. Slice the 3D model of the aerospace engine component, generate the printing path data for each printing layer, and construct the printing path graph structure;

[0010] S2, collecting the infrared thermal image, optical texture map, laser reflection map and printing parameters of the current printing layer, and constructing the corresponding lattice consistency map based on the printing path map structure;

[0011] S3, aligning the lattice consistency maps of the current printing layer and the previous printing layer, establishing a cross-layer connection edge set, and generating a cross-layer lattice map sequence;

[0012] S4. Input the cross-layer lattice graph sequence into the graph neural control model, obtain the disturbance sensitivity vector and disturbance score value of each node, and output the correction control strategy vector of the current printing layer;

[0013] S5. Based on the corrected control strategy vector, perform a cross-layer disturbance projection operation to project the area to be corrected in the current printing layer into a cross-layer printing path graph structure through spatial mapping, and generate a control intervention instruction;

[0014] S6. Based on the control intervention instruction, repeatedly execute steps S2 to S5, iteratively complete the lattice consistency maps of all printed layers layer by layer, and finally form a complete lattice map evolution record sequence.

[0015] Optionally, the S1 specifically includes the spatial coordinate information, scanning order, scanning direction and scanning speed of each layer of printing path points, and constructs a printing path graph structure, in which the nodes represent path points, the edges represent the printing order relationship between adjacent path points, and the edge attributes include the scanning direction vector of the path segment, the geometric distance between the starting and ending points, and the scanning speed gradient.

[0016] Optionally, the S2 shown specifically includes:

[0017] S21, collecting the infrared thermal image corresponding to the current printing layer to obtain the two-dimensional image information of the thermal field distribution during the printing process;

[0018] S22, collecting the optical texture map and laser reflection map of the current printing layer to obtain the surface microstructure characteristics and energy coupling area distribution;

[0019] S23, extracting printing parameters of the current printing layer, wherein the printing parameters include printing power and cooling time interval;

[0020] S24, spatially registering the infrared thermal image, the optical texture image, the laser reflection image, and the printing parameters, performing pixel-level correction, and uniformly mapping them to the coordinate system of the printing path map structure;

[0021] S25. Construct a lattice consistency map based on the printing path map structure, wherein the nodes in the lattice consistency map are candidate lattice centers of the area corresponding to the path points, and the node attributes include crystal orientation estimation tensor, temperature gradient value and surface stress state vector.

[0022] Optionally, the S3 specifically includes:

[0023] S31, obtain the lattice consistency map of the current printing layer and the previous printing layer, perform preliminary registration on the lattice consistency map node set of the current printing layer and the previous printing layer, establish an initial matching node pair set based on the spatial coordinate position relationship, and preliminarily filter out nodes whose distance exceeds a preset threshold. Node pairs;

[0024] S32, calculating the grain direction difference, temperature thermal gradient direction difference and surface stress state similarity between the initial matching node pairs after preliminary screening, and performing weighted fusion to form a comprehensive matching score for evaluating the lattice consistency between cross-layer nodes;

[0025] S33, re-screening the comprehensive matching score to be higher than the set threshold Node pairs are used to construct a cross-layer connection edge set. The cross-layer connection edge attributes include crystal orientation difference, thermal continuity index and historical evolution mark.

[0026] S34, adding the cross-layer connection edge set to the lattice consistency map of the current printing layer to construct an extended lattice map containing the cross-layer connection structure;

[0027] S35, structurally integrating the extended lattice map with the lattice consistency map of the previous printed layer to form a cross-layer lattice map structure;

[0028] S36. Using the cross-layer lattice atlas structure as the lattice state of the current time step during the printing process to generate a cross-layer lattice atlas sequence.

[0029] Optionally, the S4 specifically includes:

[0030] S41, inputting the cross-layer lattice atlas sequence into a graph neural control volume model, wherein the graph neural control volume model includes a deformation fitting perception unit, a graph-graph coupling perturbation unit, and a residual control network;

[0031] S42, the deformation fitting perception unit extracts the lattice principal axis direction tensor of each node at different time steps and constructs a crystal direction evolution trajectory set ,in Indicates the Nodes at time step The lattice principal axis direction tensor of ;

[0032] S43, the deformation fitting perception unit uses Bezier surface fitting to solve and predict the crystal orientation tensor , and calculate the crystal orientation offset residual:

[0033] ;

[0034] in, Indicates the The crystal orientation offset residual of each node is represents the Frobenius norm;

[0035] S44, the graph-graph coupling perturbation unit embeds the node's crystal orientation offset residual, temperature gradient change, and stress tensor change into a perturbation sensitivity vector:

[0036] ;

[0037] in, Indicates the The disturbance sensitivity vector of each node, represents the activation function, 、 and represents the weight matrix, Indicates the offset, Indicates the The temperature gradient of each node changes, Indicates the Changes in stress tensor at each node;

[0038] S45, the residual control network is based on the disturbance feature vector , calculate the metallurgical disturbance score:

[0039] ;

[0040] in, Indicates the The metallurgical disturbance score of each node, 、 and represents the weight of the control factor, represents the Euclidean norm;

[0041] S46. Score all nodes To sort in descending order, select The metallurgical disturbance scoring nodes are used as the areas to be corrected, and the corresponding disturbance sensitivity vector is used to generate the local control strategy vector for each area to be corrected:

[0042] ;

[0043] in, Indicates the The local control strategy vector of the control target area, represents the control weight matrix, represents the control bias term;

[0044] The local control strategy vectors of all the areas to be corrected are aggregated and processed by boundary smoothing to form the correction control strategy vector of the current printing layer.

[0045] Optionally, the S5 specifically includes:

[0046] S51, extracting the spatial coordinates of all nodes in the area to be corrected based on the correction control strategy vector, and establishing a cross-layer disturbance projection relationship based on the spatial position correspondence between the printing path graph structure of the current printing layer and the printing path graph structure of the next printing layer;

[0047] S52, performing a cross-layer disturbance projection operation based on the cross-layer disturbance projection relationship, projecting the local control strategy vector corresponding to the to-be-corrected area of the current printing layer onto the to-be-corrected area corresponding to the next printing layer through a spatial mapping method, and calculating the projection weight based on the spatial distance between nodes, thereby achieving disturbance projection of the local control strategy vector on the to-be-corrected area corresponding to the next printing layer;

[0048] S53, performing boundary smoothing processing on the local control strategy vector after the disturbance projection in the next printing layer to eliminate the control discontinuity of the disturbance boundary, thereby forming a printing path graph structure of the next printing layer after disturbance optimization;

[0049] S54 , generating a control intervention instruction according to the next printing layer printing path diagram structure after disturbance optimization.

[0050] Optionally, the control intervention instruction includes the spatial position of the path point, the scanning direction, the laser power adjustment amount and the scanning timing adjustment information.

[0051] Optionally, the control rules of the control intervention instruction are as follows:

[0052] If the following conditions are met: Metallurgical Disturbance Score , the change in scanning path direction , laser power adjustment amount , scan timing variation , and the path points do not have abnormal cross-layer connectivity in the print path graph structure, then the standard print control instructions with the original parameters are output;

[0053] in Indicates the set metallurgical disturbance score threshold, Indicates the set scan path direction change threshold, Indicates the set laser power adjustment threshold. Indicates the set scan timing variation threshold;

[0054] If the metallurgical disturbance score is met , when it appears 、 、 If any of the above cases occurs and the path structure does not form a cross-layer continuous disturbance area, the corresponding single-parameter fine-tuning control intervention instruction is output, and only one of the path direction, laser power or scanning timing is adjusted;

[0055] If the metallurgical disturbance score is met , when it appears 、 、 If two or more of the above situations occur, and the path structure belongs to the node group of the disturbed connected area in the mapping relationship, a multi-parameter joint control intervention instruction is output to synchronously modify the path direction, laser power and scanning timing.

[0056] The beneficial effects of the present invention are:

[0057] First, unlike traditional approaches that rely on static parameter tuning and single-layer monitoring, this invention constructs a lattice consistency map for each printed layer, accurately capturing multi-physics parameters such as grain orientation estimation, temperature gradient distribution, and stress state within the print path region. This map is then mapped to a path graph structure, accurately expressing local metallurgical behavior in the form of a graph structure. This map is not only used for single-layer analysis, but also establishes structural alignment relationships across layers, generating a cross-layer lattice map sequence. This enables the modeling and storage of the continuous lattice evolution process for the first time, addressing the inability of existing technologies to express microscopic evolution trends.

[0058] Secondly, the graph neural control body model proposed in the present invention integrates the deformation fitting perception unit, the graph coupling perturbation mechanism and the residual scoring network, and can process the complex structural dependencies in the lattice map sequence in an end-to-end manner. By extracting the perturbation sensitivity vector and calculating the metallurgical perturbation score, the present invention can automatically identify the control key path units where crystal orientation mismatch or thermal-stress mutation may occur, and output targeted local control strategies, including multi-dimensional control means such as scanning direction adjustment, laser power compensation and path sequence optimization. This strategy generation method based on graph structure breaks through the limitations of traditional image classification or path heuristic correction, and truly realizes precise feedback and intervention for the tissue evolution process.

[0059] Furthermore, to ensure the control strategy's feedforward regulation capabilities, the present invention further designs a perturbation cross-layer projection mechanism. This spatially maps abnormal areas in the current printing layer onto the path structure of the next printing layer, pre-setting intervention operations before printing, thereby enhancing the foresight and stability of metallurgical continuity control. Through the fusion output of the control strategy vector, control intervention instructions can be dynamically generated, guiding the real-time update of the subsequent path map structure and printing parameters, thus establishing a data-driven, closed-loop control system for printing layers.

[0060] Finally, combined with the layer-by-layer lattice map reconstruction process, the present invention established a complete lattice map evolution record sequence, providing a long-term and effective data basis for printing quality traceability, model retraining and process adaptation, and significantly improving the micro-metallurgical bonding strength and overall structural reliability during the 3D printing process of key components of aerospace engines. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0062] Figure 1 This is an overall flow chart of the metallurgical bonding enhancement method for 3D printing interlayers of aerospace engines proposed by the present invention;

[0063] Figure 2 This is a flowchart for graph structure alignment and cross-layer connection construction of a method for enhancing interlayer metallurgical bonding in 3D printing of aerospace engines proposed in this invention;

[0064] Figure 3 This is a structural schematic diagram of the neural control body model of the interlayer metallurgical bonding enhancement method for 3D printing of aerospace engines proposed by the present invention. DETAILED DESCRIPTION

[0065] The present invention will now be described in further detail with reference to the accompanying drawings, which are simplified schematic diagrams that illustrate the basic structure of the present invention in a schematic manner.

[0066] refer to Figure 1-Figure 3 A method for enhancing metallurgical bonding between layers of 3D printed aerospace engines, comprising the following steps:

[0067] S1. Slice the 3D model of the aerospace engine component, generate the printing path data for each printing layer, and construct the printing path graph structure;

[0068] S2, collecting the infrared thermal image, optical texture map, laser reflection map and printing parameters of the current printing layer, and constructing the corresponding lattice consistency map based on the printing path map structure;

[0069] S3, aligning the lattice consistency maps of the current printing layer and the previous printing layer, establishing a cross-layer connection edge set, and generating a cross-layer lattice map sequence;

[0070] S4. Input the cross-layer lattice graph sequence into the graph neural control model, obtain the disturbance sensitivity vector and disturbance score value of each node, and output the correction control strategy vector of the current printing layer;

[0071] S5. Based on the corrected control strategy vector, perform a cross-layer disturbance projection operation to project the area to be corrected in the current printing layer into a cross-layer printing path graph structure through spatial mapping, and generate a control intervention instruction;

[0072] S6. Based on the control intervention instruction, repeatedly execute steps S2 to S5, iteratively complete the lattice consistency maps of all printed layers layer by layer, and finally form a complete lattice map evolution record sequence.

[0073] This method proposes an additive manufacturing process control mechanism that combines graph structure modeling and disturbance perception control. The printing path graph structure is generated by slicing layer by layer, and after each layer is completed, the lattice consistency map is constructed by combining multimodal printing data. Subsequently, cross-layer connections are established between the maps to generate a complete lattice evolution sequence. By modeling the evolutionary state of the graph nodes and edges through the graph neural control body model, the identification of key disturbance areas and the output of control strategies are realized, and disturbance projection and parameter updates are performed in subsequent printing layers, ultimately achieving closed-loop control of the entire printing process. This solution not only improves the metallurgical continuity of aerospace engine components during the layer-by-layer construction process, but also significantly enhances the adaptability of the printing process to microstructural evolution. While ensuring metallurgical quality, it also takes into account the real-time adjustment capabilities of forming path optimization and control instructions.

[0074] In this embodiment, the S1 specifically includes the spatial coordinate information, scanning order, scanning direction and scanning speed of each layer of printing path points, and constructs a printing path graph structure, in which the nodes represent path points, the edges represent the printing order relationship between adjacent path points, and the edge attributes include the scanning direction vector of the path segment, the geometric distance between the starting and ending points, and the scanning speed gradient.

[0075] During the path data construction phase, the printing points of each layer are abstracted as graph nodes, and a topological edge structure is constructed through sequence and spatial relationships. This not only preserves the structural information of the printing trajectory but also provides a basis for subsequent spatial alignment with thermal field images and lattice estimation results. By introducing edge attributes such as the scanning direction vector, path segment distance, and velocity gradient, the path graph has the ability to express parameter sensitivity, making subsequent mapping and control more physically interpretable. This graph structure expression opens up the connection channel between spatial position and printing parameters, laying the foundation for the modeling of microscopic grain behavior and the targeted output of control strategies, and improving the engineering expressiveness and controllability of the graph data structure throughout the additive manufacturing process.

[0076] In this embodiment, S2 specifically includes:

[0077] S21, collecting the infrared thermal image corresponding to the current printing layer to obtain the two-dimensional image information of the thermal field distribution during the printing process;

[0078] S22, collecting the optical texture map and laser reflection map of the current printing layer to obtain the surface microstructure characteristics and energy coupling area distribution;

[0079] S23, extracting printing parameters of the current printing layer, wherein the printing parameters include printing power and cooling time interval;

[0080] S24, spatially registering the infrared thermal image, the optical texture image, the laser reflection image, and the printing parameters, performing pixel-level correction, and uniformly mapping them to the coordinate system of the printing path map structure;

[0081] S25. Construct a lattice consistency map based on the printing path map structure, wherein the nodes in the lattice consistency map are candidate lattice centers of the area corresponding to the path points, and the node attributes include crystal orientation estimation tensor, temperature gradient value and surface stress state vector.

[0082] In the process of constructing the lattice consistency atlas, the hierarchical thermal field information is obtained through infrared thermal images, the microscopic surface morphology is obtained through optical texture images, and the energy coupling area is captured through laser reflection images. Then, the printing parameters are combined to act on the atlas node feature construction, thus realizing multi-source perception of the printing state. After completing the image registration in a unified coordinate system, the atlas node can accurately correspond to the center of the lattice candidate structure in the printing path point area, and give it properties such as crystal orientation estimation, temperature gradient and stress state. This construction method significantly improves the fit of the atlas to the actual state of the lattice, and provides a basic data structure for subsequent cross-layer structure alignment and perturbation analysis, with clear physical interpretation and executability.

[0083] In this embodiment, S3 specifically includes:

[0084] S31, obtain the lattice consistency map of the current printing layer and the previous printing layer, perform preliminary registration on the lattice consistency map node set of the current printing layer and the previous printing layer, establish an initial matching node pair set based on the spatial coordinate position relationship, and preliminarily filter out nodes whose distance exceeds a preset threshold. Node pairs;

[0085] S32, calculating the grain direction difference, temperature thermal gradient direction difference and surface stress state similarity between the initial matching node pairs after preliminary screening, and performing weighted fusion to form a comprehensive matching score for evaluating the lattice consistency between cross-layer nodes;

[0086] S33, re-screening the comprehensive matching score to be higher than the set threshold Node pairs are used to construct a cross-layer connection edge set. The cross-layer connection edge attributes include crystal orientation difference, thermal continuity index and historical evolution mark.

[0087] S34, adding the cross-layer connection edge set to the lattice consistency map of the current printing layer to construct an extended lattice map containing the cross-layer connection structure;

[0088] S35, structurally integrating the extended lattice map with the lattice consistency map of the previous printed layer to form a cross-layer lattice map structure;

[0089] S36. Using the cross-layer lattice atlas structure as the lattice state of the current time step during the printing process to generate a cross-layer lattice atlas sequence.

[0090] For the lattice consistency maps between different printed layers, potential cross-layer node connection relationships are screened through preliminary coordinate alignment and physical property similarity scores. Then, edge connections are constructed based on the matching scores and thermal-mechanical continuity indicators, thereby establishing a set of graph structure sequences with directional and evolutionary properties. While maintaining spatial topological continuity, this sequence truly reflects the transmission path of lattice direction offsets and thermal coupling disturbances between layers. By cross-layer expansion of each layer of the map and integrating it into a time series sequence, it can provide structured time series input for the subsequent graph neural control body model, effectively improving the ability to predict the development trend of metallurgical deviations, and has the characteristics of strong cross-layer structural continuity and high dynamic evolution expression ability.

[0091] In this embodiment, the S4 specifically includes:

[0092] S41, inputting the cross-layer lattice atlas sequence into a graph neural control volume model, wherein the graph neural control volume model includes a deformation fitting perception unit, a graph-graph coupling perturbation unit, and a residual control network;

[0093] S42, the deformation fitting perception unit extracts the lattice principal axis direction tensor of each node at different time steps and constructs a crystal direction evolution trajectory set ,in Indicates the Nodes at time step The lattice principal axis direction tensor of ;

[0094] S43, the deformation fitting perception unit uses Bezier surface fitting to solve and predict the crystal orientation tensor , and calculate the crystal orientation offset residual:

[0095] ;

[0096] in, Indicates the The crystal orientation offset residual of each node is represents the Frobenius norm;

[0097] S44, the graph-graph coupling perturbation unit embeds the node's crystal orientation offset residual, temperature gradient change, and stress tensor change into a perturbation sensitivity vector:

[0098] ;

[0099] in, Indicates the The disturbance sensitivity vector of each node, represents the activation function, 、 and represents the weight matrix, Indicates the offset, Indicates the The temperature gradient of each node changes, Indicates the Changes in stress tensor at each node;

[0100] S45, the residual control network is based on the disturbance feature vector , calculate the metallurgical disturbance score:

[0101] ;

[0102] in, Indicates the The metallurgical disturbance score of each node, 、 and represents the weight of the control factor, represents the Euclidean norm;

[0103] S46. Score all nodes To sort in descending order, select The metallurgical disturbance scoring nodes are used as the areas to be corrected, and the corresponding disturbance sensitivity vector is used to generate the local control strategy vector for each area to be corrected:

[0104] ;

[0105] in, Indicates the The local control strategy vector of the control target area, represents the control weight matrix, represents the control bias term;

[0106] The local control strategy vectors of all the areas to be corrected are aggregated and processed by boundary smoothing to form the correction control strategy vector of the current printing layer.

[0107] With the help of a graph neural control body model, the input lattice map sequence is converted into perturbation scores and strategy generation results for each path control unit. The crystal orientation offset residual is calculated through a deformation fitting perception unit, and a multidimensional perturbation feature vector is constructed by combining thermal gradient and stress tensor changes. After screening with a perturbation scoring function, key control areas are identified. A control transformation function is used to output scanning angle perturbations, power adjustments, and sequence offset values, which are aggregated to form a control strategy vector set for the printing layer. This strategy has the advantages of strong physical feature perception, good spatial distribution adaptability, and fine control granularity. It realizes a closed-loop mapping from physical data drive to path parameter optimization, significantly enhancing the organizational control capabilities of the printing process.

[0108] In this embodiment, the S5 specifically includes:

[0109] S51, extracting the spatial coordinates of all nodes in the area to be corrected based on the correction control strategy vector, and establishing a cross-layer disturbance projection relationship based on the spatial position correspondence between the printing path graph structure of the current printing layer and the printing path graph structure of the next printing layer;

[0110] S52, performing a cross-layer disturbance projection operation based on the cross-layer disturbance projection relationship, projecting the local control strategy vector corresponding to the to-be-corrected area of the current printing layer onto the to-be-corrected area corresponding to the next printing layer through a spatial mapping method, and calculating the projection weight based on the spatial distance between nodes, thereby achieving disturbance projection of the local control strategy vector on the to-be-corrected area corresponding to the next printing layer;

[0111] S53, performing boundary smoothing processing on the local control strategy vector after the disturbance projection in the next printing layer to eliminate the control discontinuity of the disturbance boundary, thereby forming a printing path graph structure of the next printing layer after disturbance optimization;

[0112] S54 , generating a control intervention instruction according to the next printing layer printing path diagram structure after disturbance optimization.

[0113] For the generated corrected control strategy vector, cross-layer intervention control is transferred through disturbance region identification and path point spatial projection. This mechanism establishes a spatial mapping relationship between the spatial position and control vector of the key area of the current printing layer and the path structure of the next printing layer. The control instructions are then projected onto the target layer's corrected area, enabling feedforward transfer of the control strategy and continuous influence. Subsequent boundary smoothing and local control fusion mechanisms ensure the continuity and stability of control adjustments. Ultimately, these disturbance results are encapsulated as control intervention instructions, which can be directly parsed and executed by the printing control system, significantly improving the predictability and control accuracy of printing parameter adjustments.

[0114] In this embodiment, the control intervention instruction includes the spatial position of the path point, the scanning direction, the laser power adjustment amount and the scanning timing adjustment information.

[0115] In this embodiment, the control rules of the control intervention instruction are as follows:

[0116] If the following conditions are met: Metallurgical Disturbance Score , the change in scanning path direction , laser power adjustment amount , scan timing variation , and the path points do not have abnormal cross-layer connectivity in the print path graph structure, then the standard print control instructions with the original parameters are output;

[0117] in Indicates the set metallurgical disturbance score threshold, Indicates the set scan path direction change threshold, Indicates the set laser power adjustment threshold. Indicates the set scan timing variation threshold;

[0118] If the metallurgical disturbance score is met , when it appears 、 、 If any of the above cases occurs and the path structure does not form a cross-layer continuous disturbance area, the corresponding single-parameter fine-tuning control intervention instruction is output, and only one of the path direction, laser power or scanning timing is adjusted;

[0119] If the metallurgical disturbance score is met , when it appears 、 、 If two or more of the above situations occur, and the path structure belongs to the node group of the disturbed connected area in the mapping relationship, a multi-parameter joint control intervention instruction is output to synchronously modify the path direction, laser power and scanning timing.

[0120] Example 1

[0121] In order to verify the feasibility of the present invention in implementation, the present invention is applied to the 3D printing process of a certain type of high thrust-to-weight ratio aerospace engine nozzle assembly. The nozzle assembly is made of Inconel 718 high-temperature alloy material and is constructed layer by layer through a laser selective melting process. The overall structure has extremely high interlayer thermal stress concentration and micro-grain boundary mismatch risks. Under traditional processes, the interlayer metallurgical defect rate is always maintained between 2% and 3%, which seriously restricts the achievement of nozzle life indicators. This embodiment selects the middle contraction section of the nozzle structure with a large aspect ratio and drastic geometric mutation as the key area, deploys the method of the present invention in this area, and executes the entire process of lattice consistency map construction, cross-layer map structure alignment, disturbance perception modeling and printing path closed-loop adjustment.

[0122] During the printing process, the system first collects infrared thermal images, optical texture images, and laser reflection images after each layer is printed, and combines the position information, power, and speed parameters of each scanning path point to construct a lattice consistency map. Subsequently, the map is structurally aligned and cross-layered with the map of the previous printed layer to generate a map sequence and input it into the graph neural control body model for disturbance sensitivity assessment. After printing the first layer, the system identified abnormal areas with an average crystal orientation offset residual of 3.1 degrees and a temperature gradient mutation of 18.4 K / mm, with a disturbance score of up to 0.85. Based on this, the system adjusted the scanning angle, reduced the power, and locally delayed the scanning timing, and executed the printing after generating a control intervention instruction. In the subsequent iteration process from the second to the fifth layer, the system continuously constructed the lattice map and output the strategy. The response of the disturbance area gradually decreased, the control adjustment amplitude decreased, and the metallurgical state tended to be stable.

[0123] In this embodiment, with the dynamic execution of the control strategy, the crystal orientation residual gradually decreased from 3.1° to 2.1°, the temperature gradient disturbance decreased from 18.4 K / mm to 14.2 K / mm, and the stress tensor variation range was also smoothed, ultimately resulting in a significant drop in the metallurgical defect rate from the initial 2.5% to 0.9%. The average adjustment amount of the control instruction was reduced from the initial 1.00 to 0.68, indicating that the system can adaptively control complex lattice changes, avoiding unnecessary intervention while ensuring metallurgical consistency. After printing, X-ray tomography was used to verify the continuity of the grain boundaries. No obvious microcracks or interlayer pores were found inside the nozzle contraction section. The surface quality and grain structure showed good consistency, and the conditions for direct assembly application were met.

[0124] This example effectively demonstrates the adaptability and control accuracy of the method in printing highly complex aerospace engine parts. Its graph-driven intervention control mechanism, based on lattice-level perception, breaks through the bottleneck of traditional static parameter setting and image monitoring. This not only improves micrometallurgical quality but also brings new capabilities to aerospace-grade printing processes, including structural control, path adjustment, and evolution traceability, demonstrating its high engineering application and promotion value.

[0125] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A method for enhancing metallurgical bonding between layers of 3D printed aerospace engines, characterized in that: The steps include: S1. Slice the 3D model of the aerospace engine component, generate the printing path data for each printing layer, and construct the printing path graph structure; S2, collecting the infrared thermal image, optical texture map, laser reflection map and printing parameters of the current printing layer, and constructing the corresponding lattice consistency map based on the printing path map structure; S3, aligning the lattice consistency maps of the current printing layer and the previous printing layer, establishing a cross-layer connection edge set, and generating a cross-layer lattice map sequence; S4. Input the cross-layer lattice graph sequence into the graph neural control model, obtain the disturbance sensitivity vector and disturbance score value of each node, and output the correction control strategy vector of the current printing layer; S5. Based on the corrected control strategy vector, perform a cross-layer disturbance projection operation to project the area to be corrected in the current printing layer into a cross-layer printing path graph structure through spatial mapping, and generate a control intervention instruction; S6. Based on the control intervention instruction, repeatedly execute steps S2 to S5, iteratively complete the lattice consistency maps of all printed layers layer by layer, and finally form a complete lattice map evolution record sequence.

2. The method for enhancing metallurgical bonding between layers of 3D printed aerospace engines according to claim 1, characterized in that: The S1 specifically includes the spatial coordinate information, scanning order, scanning direction and scanning speed of each layer of printing path points, and constructs a printing path graph structure, where the nodes represent path points, the edges represent the printing order relationship between adjacent path points, and the edge attributes include the scanning direction vector of the path segment, the geometric distance between the starting and ending points, and the scanning speed gradient.

3. The method for enhancing metallurgical bonding between layers of 3D printed aerospace engines according to claim 1, characterized in that: The S2 shown specifically includes: S21, collecting the infrared thermal image corresponding to the current printing layer to obtain the two-dimensional image information of the thermal field distribution during the printing process; S22, collecting the optical texture map and laser reflection map of the current printing layer to obtain the surface microstructure characteristics and energy coupling area distribution; S23, extracting printing parameters of the current printing layer, wherein the printing parameters include printing power and cooling time interval; S24, spatially registering the infrared thermal image, the optical texture image, the laser reflection image, and the printing parameters, performing pixel-level correction, and uniformly mapping them to the coordinate system of the printing path map structure; S25. Construct a lattice consistency map based on the printing path map structure, wherein the nodes in the lattice consistency map are candidate lattice centers of the area corresponding to the path points, and the node attributes include crystal orientation estimation tensor, temperature gradient value and surface stress state vector.

4. The method for enhancing metallurgical bonding between layers of 3D printed aerospace engines according to claim 1, characterized in that: The S3 specifically includes: S31, obtain the lattice consistency map of the current printing layer and the previous printing layer, perform preliminary registration on the lattice consistency map node set of the current printing layer and the previous printing layer, establish an initial matching node pair set based on the spatial coordinate position relationship, and preliminarily filter out nodes whose distance exceeds a preset threshold. Node pairs; S32, calculating the grain direction difference, temperature thermal gradient direction difference and surface stress state similarity between the initial matching node pairs after preliminary screening, and performing weighted fusion to form a comprehensive matching score for evaluating the lattice consistency between cross-layer nodes; S33, re-screening the comprehensive matching score to be higher than the set threshold Node pairs are used to construct a cross-layer connection edge set. The cross-layer connection edge attributes include crystal orientation difference, thermal continuity index and historical evolution mark. S34, adding the cross-layer connection edge set to the lattice consistency map of the current printing layer to construct an extended lattice map containing the cross-layer connection structure; S35, structurally integrating the extended lattice map with the lattice consistency map of the previous printed layer to form a cross-layer lattice map structure; S36. Using the cross-layer lattice atlas structure as the lattice state of the current time step during the printing process to generate a cross-layer lattice atlas sequence.

5. The method for enhancing metallurgical bonding between layers of 3D printed aerospace engines according to claim 1, characterized in that: The S4 specifically includes: S41, inputting the cross-layer lattice atlas sequence into a graph neural control volume model, wherein the graph neural control volume model includes a deformation fitting perception unit, a graph-graph coupling perturbation unit, and a residual control network; S42, the deformation fitting perception unit extracts the lattice principal axis direction tensor of each node at different time steps and constructs a crystal direction evolution trajectory set ,in Indicates the Nodes at time step The lattice principal axis direction tensor of ; S43, the deformation fitting perception unit uses Bezier surface fitting to solve and predict the crystal orientation tensor , and calculate the crystal orientation offset residual: ； in, Indicates the The crystal orientation offset residual of each node is represents the Frobenius norm; S44, the graph-graph coupling perturbation unit embeds the node's crystal orientation offset residual, temperature gradient change, and stress tensor change into the perturbation sensitivity vector: ； in, Indicates the The disturbance sensitivity vector of each node, represents the activation function, 、 and represents the weight matrix, Indicates the offset, Indicates the The temperature gradient of each node changes, Indicates the Changes in stress tensor at each node; S45, the residual control network is based on the disturbance feature vector , calculate the metallurgical disturbance score: ； in, Indicates the The metallurgical disturbance score of each node, 、 and represents the weight of the control factor, represents the Euclidean norm; S46. Score all nodes To sort in descending order, select The metallurgical disturbance scoring nodes are used as the areas to be corrected, and the corresponding disturbance sensitivity vector is used to generate the local control strategy vector for each area to be corrected: ； in, Indicates the The local control strategy vector of the control target area, represents the control weight matrix, represents the control bias term; The local control strategy vectors of all the areas to be corrected are aggregated and processed by boundary smoothing to form the correction control strategy vector of the current printing layer.

6. The method for enhancing metallurgical bonding between layers of 3D printed aerospace engines according to claim 1, characterized in that: The S5 specifically includes: S51, extracting the spatial coordinates of all nodes in the area to be corrected based on the correction control strategy vector, and establishing a cross-layer disturbance projection relationship based on the spatial position correspondence between the printing path graph structure of the current printing layer and the printing path graph structure of the next printing layer; S52, performing a cross-layer disturbance projection operation based on the cross-layer disturbance projection relationship, projecting the local control strategy vector corresponding to the to-be-corrected area of the current printing layer onto the to-be-corrected area corresponding to the next printing layer through a spatial mapping method, and calculating the projection weight based on the spatial distance between nodes, thereby achieving disturbance projection of the local control strategy vector on the to-be-corrected area corresponding to the next printing layer; S53, performing boundary smoothing processing on the local control strategy vector after the disturbance projection in the next printing layer to eliminate the control discontinuity of the disturbance boundary, thereby forming a printing path graph structure of the next printing layer after disturbance optimization; S54 , generating a control intervention instruction according to the next printing layer printing path diagram structure after disturbance optimization.

7. The method for enhancing metallurgical bonding between layers of 3D printed aerospace engines according to claim 1, characterized in that: The control intervention instruction includes the spatial position of the path point, the scanning direction, the laser power adjustment amount and the scanning timing adjustment information.

8. The method for enhancing metallurgical bonding between layers of 3D printed aerospace engines according to claim 1, characterized in that: The control rules of the control intervention instructions are as follows: If the following conditions are met: Metallurgical Disturbance Score , the change in scanning path direction , laser power adjustment amount , scan timing variation , and the path points do not have abnormal cross-layer connectivity in the print path graph structure, then the standard print control instructions with the original parameters are output; in Indicates the set metallurgical disturbance score threshold, Indicates the set scan path direction change threshold, Indicates the set laser power adjustment threshold. Indicates the set scan timing variation threshold; If the metallurgical disturbance score is met , when it appears 、 、 If any of the above cases occurs and the path structure does not form a cross-layer continuous disturbance area, the corresponding single-parameter fine-tuning control intervention instruction is output, and only one of the path direction, laser power or scanning timing is adjusted; If the metallurgical disturbance score is met , when it appears 、 、 If two or more of the above situations occur, and the path structure belongs to the node group of the disturbed connected area in the mapping relationship, a multi-parameter joint control intervention instruction is output to synchronously modify the path direction, laser power and scanning timing.

Images