Artificial intelligence-based ultra-micro three-dimensional printing repair method and system for metal parts

Abstract

The application provides a kind of metal component ultra-micro three-dimensional printing repair method and system based on artificial intelligence.The method comprises: generating component repair basic data, fusion representation based on component repair basic data, generating high temperature and high pressure service damage field;According to the high temperature and high pressure service damage field, the damage mechanism type, the repairable boundary and the bearing associated area of the target repair area are determined, and the partition repair strategy is generated;Based on ultra-micro deposition unit, repair process knowledge graph and artificial intelligence repair decision model, ultra-micro three-dimensional printing repair path and corresponding process parameters are generated;Collecting molten pool state data and matching with predicted forming state, the printing path and process parameters of subsequent ultra-micro deposition unit are updated according to the matching result;According to the detection data after repair, printing process record and re-service constraint, the intelligent ultra-micro three-dimensional printing repair processing result is generated.The application can improve the damage identification accuracy, improve the deposition control accuracy and enhance the re-service reliability.

Description

Technical Field

[0001] This application relates to the field of metal additive manufacturing repair technology, and in particular to an artificial intelligence-based method and system for repairing metal parts using ultra-micro 3D printing. Background Technology

[0002] High-temperature and high-pressure metal components are widely used in critical equipment such as aero-engines, energy equipment, petrochemical pipelines, and high-pressure valves. They endure long-term exposure to high temperatures, pressure cycles, corrosive media, and complex loads, making them prone to damage such as erosion wear, thermal fatigue cracks, creep cavitation, oxidation corrosion, localized thinning, and residual stress concentration. Due to the high material value, long manufacturing cycle, and stringent service safety requirements of these components, using 3D printing for precise repair of locally damaged areas has become an important technological direction for extending component lifespan and reducing replacement costs.

[0003] Existing additive manufacturing repair methods for metal parts typically determine the defect area based on geometric scanning results, generate a repair path according to preset printing parameters, and assess the repair quality by combining molten pool monitoring or post-repair inspection. Some solutions introduce artificial intelligence models to process defect identification, recommend process parameters, or predict forming quality, thereby improving the automation level of the repair process.

[0004] However, existing technologies are mostly based on geometric repair or single defect detection, which makes it difficult to uniformly characterize service conditions, material degradation, internal defects, residual stress, and load transfer status. This results in insufficient matching between the repair path and the actual damage mechanism of high-temperature and high-pressure components. At the same time, existing printing parameters are usually configured layer by layer or path as a whole, lacking a refined path generation and real-time correction mechanism for ultra-micro deposition units, making it difficult to take into account the requirements of interface bonding, thermal impact control, and load recovery. In addition, the post-repair quality evaluation is not sufficiently correlated with re-service constraints, making it difficult to reliably judge the service adaptability of the repaired area under high-temperature and high-pressure conditions. Summary of the Invention

[0005] In view of this, the present application provides an artificial intelligence-based method and system for repairing metal parts by ultra-micro 3D printing, in order to solve the problems of insufficient damage mechanism characterization, rough ultra-micro deposition correction, and disconnection between the existing technology and the re-service verification.

[0006] A first aspect of this application provides an artificial intelligence-based method for ultra-micro 3D printing repair of metal parts, comprising: acquiring service condition data, inspection data, material state data, and geometric morphology data of a target high-temperature and high-pressure metal part, and mapping them to the same part coordinate system to generate basic data for part repair; based on the basic data for part repair, performing fusion characterization of geometric defects, material degradation, internal defects, residual stress, and load transfer state to generate a high-temperature and high-pressure service damage field; determining the damage mechanism type, repairable boundary, and load-bearing associated region of the target repair area according to the high-temperature and high-pressure service damage field, and generating a partitioned repair strategy; dividing the target repair area into multiple ultra-micro deposition units according to the partitioned repair strategy, and configuring deposition sequence, material constraints, and thermal input constraints for each ultra-micro deposition unit; generating an ultra-micro 3D printing repair path and corresponding process parameters based on the ultra-micro deposition units, a repair process knowledge graph, and an artificial intelligence repair decision model; during the ultra-micro 3D printing repair process, collecting molten pool state data and performing deviation matching with the predicted forming state, and updating the printing path and process parameters of subsequent ultra-micro deposition units according to the matching results; and generating an intelligent ultra-micro 3D printing repair processing result based on post-repair inspection data, printing process records, and re-service constraints.

[0007] A second aspect of this application provides an artificial intelligence-based ultra-micro 3D printing repair system for metal parts, comprising: an acquisition module for acquiring service condition data, inspection data, material state data, and geometric morphology data of a target high-temperature and high-pressure metal part, and mapping them to the same part coordinate system to generate basic data for part repair; a fusion module for fusing and characterizing geometric defects, material degradation, internal defects, residual stress, and load transfer state based on the basic data for part repair to generate a high-temperature and high-pressure service damage field; and a determination module for determining the damage mechanism type, repairable boundary, and load-bearing associated region of the target repair area based on the high-temperature and high-pressure service damage field, and generating a partitioned repair strategy; and dividing the area into zones. The system comprises four modules: a module for dividing the target repair area into multiple ultra-micro deposition units according to a zoning repair strategy, and configuring deposition sequence, material constraints, and thermal input constraints for each ultra-micro deposition unit; a repair module for generating ultra-micro 3D printing repair paths and corresponding process parameters based on ultra-micro deposition units, a repair process knowledge graph, and an artificial intelligence repair decision model; a matching module for collecting molten pool state data and matching it with the predicted forming state during ultra-micro 3D printing repair, and updating the printing paths and process parameters of subsequent ultra-micro deposition units based on the matching results; and a generation module for generating intelligent ultra-micro 3D printing repair processing results based on post-repair inspection data, printing process records, and re-service constraints.

[0008] The above-described technical solutions adopted in the embodiments of this application can achieve the following beneficial effects: By acquiring service condition data, inspection data, material state data, and geometric morphology data of the target high-temperature and high-pressure metal component and mapping them to the same component coordinate system, basic data for component repair is generated. Based on the basic data for component repair, geometric defects, material degradation, internal defects, residual stress, and load transfer status are fused and characterized to generate a high-temperature and high-pressure service damage field. The damage mechanism type, repairable boundary, and load-bearing related area of ​​the target repair area are determined according to the high-temperature and high-pressure service damage field, generating a zonal repair strategy. The target repair area is divided into multiple ultra-micro deposition units according to the zonal repair strategy, and deposition sequence, material constraints, and thermal input constraints are configured for each ultra-micro deposition unit. Based on the ultra-micro deposition units, repair process knowledge graph, and artificial intelligence repair decision model, an ultra-micro 3D printing repair path and corresponding process parameters are generated. During the ultra-micro 3D printing repair process, molten pool state data is collected and matched with the predicted forming state to update the printing path and process parameters of subsequent ultra-micro deposition units based on the matching results. Based on post-repair inspection data, printing process records, and re-service constraints, an intelligent ultra-micro 3D printing repair processing result is generated. This application can improve the accuracy of damage identification, enhance deposition control precision, and improve re-service reliability. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1 This is a flowchart illustrating the artificial intelligence-based ultra-micro 3D printing repair method for metal parts provided in this application embodiment; Figure 2 This is a schematic diagram of the structure of the artificial intelligence-based ultra-micro 3D printing repair system for metal parts provided in this application embodiment; Figure 3 This is a schematic diagram of the structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0011] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0012] In existing technologies, high-temperature and high-pressure metal components are widely used in critical equipment such as aero-engines, energy equipment, petrochemical pipelines, and high-pressure valves. These components are subjected to long-term high-temperature, pressure cycling, corrosive media, and complex load environments, making them prone to damage such as erosion wear, thermal fatigue cracks, creep porosity, oxidation corrosion, localized thinning, and residual stress concentration. Existing metal additive repair methods typically determine the defect area based on geometric scanning results, then generate a repair path according to preset printing parameters, and combine molten pool monitoring or post-repair inspection to judge the repair quality. Although some solutions introduce artificial intelligence models for defect identification, process parameter recommendation, or forming quality prediction, they still mainly focus on geometric repair, local defect detection, or single process optimization.

[0013] However, existing technologies still have significant shortcomings in the repair of high-temperature and high-pressure metal components. On the one hand, existing solutions struggle to uniformly characterize service conditions, geometric defects, material degradation, internal defects, residual stress, and load transfer states, resulting in insufficient correlation between the repair path and the actual damage mechanism of the component. On the other hand, existing printing parameters are typically configured hierarchically or along the entire path, lacking refined path generation and real-time correction mechanisms for ultra-micro deposition units, making it difficult to coordinate interface bonding, thermal impact control, and load recovery. Furthermore, current post-repair quality assessments focus primarily on dimensional recovery, surface morphology, or conventional defect detection, lacking adaptation and verification combined with high-temperature and high-pressure re-service constraints, making it difficult to reliably determine the structural integrity of the repaired area during subsequent service.

[0014] To address the aforementioned issues, this application provides an artificial intelligence-based method for ultra-micro 3D printing repair of metal components. This method first acquires service condition data, inspection data, material state data, and geometric morphology data of the target high-temperature and high-pressure metal component, and maps these data to the same component coordinate system to generate basic data for component repair. Then, based on this basic data, geometric defects, material degradation, internal defects, residual stress, and load transfer status are fused and characterized to generate a high-temperature and high-pressure service damage field. Finally, based on this high-temperature and high-pressure service damage field, the damage mechanism type, repairable boundary, and load-bearing associated region of the target repair area are determined, generating a zoned repair strategy.

[0015] Based on this, this application further divides the target repair area into multiple ultra-micro deposition units according to a zonal repair strategy, and configures the deposition sequence, material constraints, and thermal input constraints for each ultra-micro deposition unit; based on the ultra-micro deposition units, the repair process knowledge graph, and the artificial intelligence repair decision model, an ultra-micro 3D printing repair path and corresponding process parameters are generated; during the ultra-micro 3D printing repair process, the melt pool state data is collected and matched with the predicted forming state to check for deviations, and the printing path and process parameters of subsequent ultra-micro deposition units are updated according to the matching results; finally, based on the post-repair inspection data, printing process records, and re-service constraints, an intelligent ultra-micro 3D printing repair processing result is generated.

[0016] Through the above technical solution, this application can form a continuous data processing closed loop that integrates the service damage mechanism of high-temperature and high-pressure metal components, ultra-micro deposition unit-level printing control, online feedback correction of the molten pool, and post-repair service verification. This improves the accuracy of damage identification, enhances deposition control precision, and strengthens re-service reliability. Compared with existing repair methods that mainly rely on geometric compensation or static process parameters, this application enables the repair path and process parameters to better match the actual damage state of the component and subsequent service constraints, reducing the risk of incomplete fusion, abnormal heat-affected zones, residual stress concentration, and re-service mismatch in the repaired area.

[0017] The technical solution of this application will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0018] Figure 1 This is a flowchart illustrating the artificial intelligence-based ultra-micro 3D printing repair method for metal parts provided in this application embodiment. Figure 1 As shown, the method may specifically include: S101: Acquire the service condition data, inspection data, material state data and geometric morphology data of the target high temperature and high pressure metal component, and map them to the same component coordinate system to generate basic data for component repair. S102, based on the basic data of component repair, integrates and characterizes geometric defects, material degradation, internal defects, residual stress and load transfer status to generate a high temperature and high pressure service damage field. S103, Based on the high temperature and high pressure service damage field, determine the damage mechanism type, repairable boundary and load-bearing related area of ​​the target repair area, and generate a zonal repair strategy; S104, the target repair area is divided into multiple ultra-micro deposition units according to the zoning repair strategy, and deposition sequence, material constraints and heat input constraints are configured for each ultra-micro deposition unit; S105, based on ultra-micro deposition units, repair process knowledge graph and artificial intelligence repair decision model, generates ultra-micro 3D printing repair path and corresponding process parameters; S106, during the ultra-micro 3D printing repair process, the melt pool state data is collected and matched with the predicted forming state to check for deviation. Based on the matching result, the printing path and process parameters of the subsequent ultra-micro deposition units are updated. S107 generates intelligent ultra-micro 3D printing repair results based on post-repair inspection data, printing process records, and re-service constraints.

[0019] In some embodiments, service condition data, inspection data, material condition data, and geometric morphology data of the target high-temperature and high-pressure metal component are acquired and mapped to the same component coordinate system to generate basic data for component repair, including: Based on the structural reference, pressure boundary reference, and repair positioning reference of the target high-temperature and high-pressure metal component, establish the component coordinate system; Acquire multi-source repair data corresponding to the target high-temperature and high-pressure metal component, and perform data cleaning, time alignment, and source identification processing on the multi-source repair data; Based on the component coordinate system, spatial location data, damage location data, and material state location data in multi-source repair data are registered to generate coordinate mapping data. Multi-source repair data are fused based on coordinate mapping data, and data confidence labels are configured for the fused data to generate basic data for component repair.

[0020] Specifically, the target high-temperature and high-pressure metal component can be a high-pressure valve seat, a turbine hot-end component, a supercritical pipeline joint, or a high-temperature reactor sealing component. Before generating the basic data for component repair, the system first determines a unified component coordinate system based on the design model and actual clamping state of the target high-temperature and high-pressure metal component. The component coordinate system is jointly determined by the structural datum, the pressure boundary datum, and the repair positioning datum. Among them, the structural datum is used to define the overall shape and main assembly direction of the component, the pressure boundary datum is used to define the pressure-bearing surface, the sealing surface, and the wall thickness-sensitive area, and the repair positioning datum is used to define the spatial correspondence between subsequent scanning detection, damage localization, and ultra-micro 3D printing equipment.

[0021] When establishing a component coordinate system, the central axis, mounting end face, sealing ring surface, and pressure-bearing inner wall surface can be extracted from the design model of the target high-temperature and high-pressure metal component. Then, the origin and axis directions can be determined by combining the point cloud of the outer contour obtained from the actual scan. For the repair scenario of high-pressure valve seat sealing surface, the valve seat central axis can be used as the axial reference, the plane where the sealing ring surface is located can be used as the height reference, and the preset clamping positioning hole or machining reference groove can be used as the angular reference. In this way, a component coordinate system that can simultaneously express the geometric shape, damage location, and printing trajectory can be established.

[0022] Multi-source repair data can include service condition data, inspection data, material condition data, and geometric morphology data. Service condition data characterizes the temperature, pressure, media erosion, and load cycle status of the target high-temperature and high-pressure metal component during its historical operation; inspection data characterizes the distribution of surface cracks, internal defects, thinned areas, and abnormal structures; material condition data characterizes hardness, residual stress, oxide layer thickness, and material degradation; and geometric morphology data characterizes the actual shape contour, local defect depth, and boundary of the area to be repaired. When cleaning the multi-source repair data, the system removes duplicate records, abnormal sampling points, and low-quality scan segments; during time alignment, the inspection time, service record time, and material testing time are unified to the same repair evaluation cycle; and during source identification processing, source identifiers are configured for different equipment, different inspection methods, and different data batches to facilitate subsequent reliability differentiation.

[0023] During data registration, the system performs unified mapping of spatial location data, damage location data, and material state location data from multi-source repair data based on the component coordinate system. For 3D point cloud data, rigid registration can be performed based on outer contour feature points and structural references. For damage location data generated by ultrasonic, eddy current, or infrared detection, it can be converted to the component coordinate system based on the detection probe path, scanning start point, and pressure boundary reference. For material state location data generated by hardness, residual stress, and metallographic sampling, corresponding spatial coordinates can be generated based on the sampling location, measurement point spacing, and repair positioning reference. If there are local offsets between different data, the system performs secondary correction through the sealing surface edge, pressure-bearing inner wall contour, and clamping reference points to generate coordinate mapping data.

[0024] Based on coordinate mapping data, the system integrates the geometry, damage detection, material condition, and service conditions within the same region to form basic data for component repair with spatial correspondence. For situations where multiple detection results exist within the same coordinate region, the system configures data confidence labels based on data source, detection resolution, detection time, and spatial consistency. For example, in a nickel-based high-temperature alloy high-pressure valve seat, 3D scanning shows erosion depressions of 0.18 mm to 0.32 mm on the sealing ring surface; eddy current testing shows circumferential microcracks concentrated on the outer side of the sealing line; hardness testing shows a 12% decrease in local hardness compared to the reference area; and service records show that this area has experienced multiple high-temperature pressure shocks. After mapping the above data to the same component coordinate system, the system marks the area near the sealing line as a high-confidence composite damage area, and marks the detection blind zone or single-source inferred area as a low-confidence area requiring verification.

[0025] Through the above processing, repair-related data obtained from different sources, scales, and times can be unified into a spatial representation, providing basic data for subsequent high-temperature and high-pressure service damage field construction, zonal repair strategy generation, and ultra-micro deposition unit division. This embodiment can improve the consistency of damage location, reduce multi-source data fusion errors, and enhance the accuracy of subsequent intelligent ultra-micro 3D printing repair path generation.

[0026] In some embodiments, based on component repair baseline data, geometric defects, material degradation, internal defects, residual stress, and load transfer status are fused and characterized to generate a high-temperature and high-pressure service damage field, including: The basic data for component repair is spatially partitioned according to the component coordinate system to generate damage analysis units corresponding to the target high-temperature and high-pressure metal component. From each damage analysis unit, morphological deviation features, tissue degradation features, defect distribution features, stress state features and load correlation features are extracted to generate multidimensional damage characterization data. Based on multidimensional damage characterization data and high-temperature and high-pressure service constraints, the damage mechanism characterization of each damage analysis unit is determined using a damage identification model that integrates physical constraints. Field fusion is performed based on damage mechanism characterization and spatial adjacency to generate a high-temperature and high-pressure service damage field.

[0027] Specifically, after obtaining the basic data for component repair, the system spatially partitions the target high-temperature and high-pressure metal component according to the component coordinate system, enabling the correlation of geometry, damage detection, material state, and service load at the same spatial granularity. During spatial partitioning, the pressure-bearing surface, sealing surface, transition fillets, thin-walled areas, and historical high-load areas of the target high-temperature and high-pressure metal component can be prioritized for partitioning. The partitioning scale is determined based on changes in component curvature, wall thickness, and detection resolution. For a nickel-based high-temperature alloy high-pressure valve seat sealing surface, a gridded partition can be established along the circumferential and radial directions of the valve seat, dividing the area near the sealing ring surface into multiple damage analysis units. Each damage analysis unit is bound to spatial coordinates, area range, depth range, and relationships with adjacent units.

[0028] After generating damage analysis units, the system extracts multiple damage features from the component repair baseline data corresponding to each damage analysis unit. Morphological deviation features can be determined based on the difference between the actual 3D morphology and the design model, used to characterize erosion depressions, wear steps, and the degree of local thinning; microstructure degradation features can be determined based on hardness distribution, oxide layer thickness, grain anomalies, and the degree of material property deviation, used to characterize the material degradation state caused by high-temperature service; defect distribution features can be determined based on non-destructive testing results, used to characterize the spatial distribution of microcracks, pores, inclusions, and discontinuous regions; stress state features can be determined based on residual stress measurement results and structural load analysis results, used to characterize local tensile stress concentration, compressive stress release, and stress gradient changes; load correlation features can be determined based on historical temperature and pressure cycles, medium flow direction, and load-bearing path, used to characterize the load sensitivity of the corresponding region during high-temperature and high-pressure service.

[0029] In practical applications, for a specific damage analysis unit within the sealing ring surface of a high-pressure valve seat, 3D scanning results show a local depression of 0.26 mm relative to the designed sealing surface. Eddy current testing reveals circumferentially extending microcracks at the edge of the depression. Hardness testing indicates a 10% decrease in hardness compared to the reference area. Residual stress testing shows high tensile stress on the outer side of the sealing line. Service data indicates that the valve seat has undergone multiple high-temperature steam erosion and pressure start-stop cycles. The system integrates these features into multidimensional damage characterization data and assigns corresponding confidence levels to various features, enabling the same damage analysis unit to simultaneously express geometric defects, material degradation, internal defects, residual stress, and load transfer status.

[0030] The system inputs multidimensional damage characterization data into a damage identification model that integrates physical constraints. When determining the damage mechanism, the model not only classifies damage based on a single detection result but also incorporates constraints related to high-temperature and high-pressure service, material degradation patterns, and load transfer relationships. For example, when a damage analysis unit simultaneously exhibits characteristics such as erosion depressions, oxide layer thickening, decreased hardness, and concentrated pressure loads, the model identifies this region as a composite damage resulting from the coupling of erosion wear and high-temperature oxidation. When a damage analysis unit exhibits characteristics such as the aggregation of small pores, decreased hardness, long-term high-temperature recording, and localized tensile stress concentration, the model identifies this region as creep-sensitive damage. When a damage analysis unit exhibits characteristics such as linear cracks, high thermal cycling frequency, and stress concentration at the crack tip, the model identifies this region as thermal fatigue crack damage.

[0031] After determining the damage mechanism characterization of each damage analysis unit, the system further integrates spatial adjacency relationships for field fusion. For regions where the damage mechanism is continuous, the morphology deviation trend is consistent, or the load transfer direction is the same between adjacent damage analysis units, the system merges the relevant units into a continuous damage zone. For adjacent regions where there are abrupt changes in damage mechanism, stress gradient, or significant differences in detection confidence, the system retains boundary transition information and forms a damage field boundary marker. Taking a high-pressure valve seat as an example, the system can merge continuously distributed microcrack units on the outside of the sealing line into a circumferential thermal fatigue damage zone, merge erosion and depression units on the inside of the sealing surface into an erosion thinning damage zone, and merge regions with decreased hardness and dense pores into a material degradation sensitive zone, ultimately generating a high-temperature and high-pressure service damage field that reflects the spatial coupling relationships of multiple types of damage.

[0032] Through the above processing, discrete geometric inspection data, material state data, internal defect data, stress data, and service load data can be transformed into a high-temperature and high-pressure service damage field with spatial continuity and mechanism orientation. This embodiment can improve the accuracy of damage mechanism identification, enhance the ability to express the boundary of the damage region, and provide a reliable data foundation for subsequent generation of partitioned repair strategies, division of ultra-micro deposition units, and configuration of printing process parameters.

[0033] In some embodiments, the damage mechanism type, repairable boundary, and load-bearing associated region of the target repair area are determined based on the high-temperature and high-pressure service damage field, and a zonal repair strategy is generated, including: Based on the high temperature and high pressure service damage field, the damage mechanism type of each damage analysis unit is identified, and the mechanism continuity relationship between adjacent damage analysis units is determined. The repairable boundary of the target repair area is determined based on the damage mechanism type, material state, and load correlation characteristics. Based on the repairable boundary and load transfer status, determine the load-bearing associated area corresponding to the target repair area; Based on the damage mechanism type, repairable boundary, and load-bearing associated area, the target repair area is functionally partitioned to generate a partitioned repair strategy.

[0034] Specifically, after generating the high-temperature and high-pressure service damage field, the system first reads the damage mechanism characterization, spatial coordinates, material state, load correlation characteristics, and data confidence level identifiers corresponding to each damage analysis unit, and then establishes a damage unit adjacency network according to the component coordinate system. The damage unit adjacency network is used to express the spatial continuity relationship, damage type transition relationship, and load transfer relationship between adjacent damage analysis units. For a nickel-based high-temperature alloy high-pressure valve seat sealing surface, the system can establish an adjacency network along the circumferential and radial directions of the sealing ring surface, and perform correlation analysis on the thermal fatigue crack units located outside the sealing line, the erosion thinning units located inside the sealing surface, and the material degradation units near the pressure-bearing inner wall surface.

[0035] When identifying damage mechanism types, the system determines whether the corresponding region belongs to crack propagation damage, erosion thinning damage, oxidation corrosion damage, creep degradation damage, or a combination of damage based on the mechanism characterization of each damage analysis unit in the high-temperature and high-pressure service damage field. For adjacent damage analysis units, the system further determines the continuity of the mechanism based on whether the damage mechanism type is consistent, whether the damage intensity is continuous, whether the crack direction extends, whether the thinning trend is connected, and whether the load transfer direction is continuous. For example, if eight consecutive damage analysis units on the outer side of a high-pressure valve seat sealing line all have circumferential microcracks, high tensile stress, and thermal cycling sensitivity, the system identifies this area as a continuous thermal fatigue damage zone; if there is a depression area on the inner side of the sealing surface that gradually deepens from the direction of steam scouring, it is identified as a continuous erosion thinning area.

[0036] When determining the repairable boundary, the system comprehensively considers the damage mechanism type, material state, and load correlation characteristics to make boundary judgments. For areas where the crack tip has not yet penetrated the pressure boundary, the matrix hardness and residual stress are within the recoverable range, and the internal void density does not exceed the repair threshold, the system includes them in the repairable range. For areas with severe material degradation, insufficient pressure-bearing wall thickness, or cracks that have penetrated the pressure boundary, the system marks them as areas that cannot be directly printed and outputs prompts for rejection, extended processing, or replacement. For the aforementioned high-pressure valve seat, the system can determine the initial repairable boundary as 0.4 mm outward expansion of the circumferential microcrack and 0.3 mm outward expansion of the erosion depression edge, based on the crack endpoint obtained from eddy current detection, the erosion depression boundary obtained from 3D scanning, the hardness reduction range, and the position of the pressure sealing line. Then, the boundary is corrected according to the distance to the pressure-bearing inner wall and the residual stress gradient to form a target repair area suitable for ultra-micro 3D printing.

[0037] When determining the load-bearing associated region, the system analyzes the correlation between the target repair area and the pressure boundary, sealing contact surface, transition fillet, and wall thickness-sensitive area based on the repairable boundary and load transfer state. For areas directly involved in pressure sealing or bearing the main load, the system defines them as strong load-bearing associated regions; for areas mainly responsible for interface transition, margin shaping, or local surface restoration, the system defines them as weak load-bearing associated regions. Taking a high-pressure valve seat as an example, the area near the sealing line directly affects the high-temperature steam pressure sealing and is therefore classified as a strong load-bearing associated region; the shallow erosion area far from the sealing line is mainly used to restore surface continuity and is therefore classified as a weak load-bearing associated region; the area near the crack tip is classified as a key transition associated region due to stress concentration and propagation risks.

[0038] Subsequently, the system functionally partitions the target repair area according to damage mechanism type, repairable boundary, and load-bearing associated region. For areas containing residual oxide layer, crack tips, and loose tissue, the system designates them as defect removal zones; for areas where a metallurgical transition is needed between the matrix material and the material to be deposited, the system designates them as interface transition zones; for areas requiring restoration of the load-bearing cross-section, sealing profile, and structural continuity, the system designates them as load-bearing recovery zones; for surface areas requiring subsequent finishing, the system designates them as allowance shaping zones; and for areas exceeding the repairable boundary or with excessively high risk, the system designates them as non-printing zones. Different functional zones are respectively bound to boundary coordinates, damage mechanism labels, load-bearing associated levels, and subsequent printing constraints, forming a partitioned repair strategy.

[0039] Through the above processing, the target repair area can be divided into zones based on damage mechanism, material state, and load transfer relationship, rather than solely on the geometric defect range. This embodiment can improve the accuracy of repairable boundary determination, enhance the matching between repair zones and service damage mechanisms, and provide a clear data foundation for subsequent ultra-micro deposition unit division, material constraint configuration, and thermal input control.

[0040] In some embodiments, the target repair area is divided into multiple ultrafine deposition units according to a partitioned repair strategy, and a deposition sequence, material constraints, and thermal input constraints are configured for each ultrafine deposition unit, including: Based on the zoning remediation strategy, determine the deposition accuracy, interface transition requirements, and load-bearing recovery requirements corresponding to different functional zones in the target remediation area; Based on the requirements of deposition accuracy, interface transition and load recovery, the target repair area is spatially discretized at a microscale to generate multiple ultramicro deposition units. Each ultramicro deposition unit is bound with its spatial location, target shape profile, and adjacent connection relationship; Based on the adjacent connection relationship, load transfer state and thermal diffusion state, the deposition sequence, material constraints and thermal input constraints of each ultramicro deposition unit are determined.

[0041] Specifically, after obtaining the zonal repair strategy, the system reads the boundary coordinates, damage mechanism labels, load-bearing correlation levels, and printing constraints of each functional zone within the target repair area, and determines differentiated deposition requirements based on different functional zones. For the area adjacent to the defect removal zone, the system determines the coverage thickness and boundary sealing accuracy after interface cleaning; for the interface transition zone, the system determines the compositional transition gradient between the matrix material and the repair material, the melt depth control range, and interlayer overlap requirements; for the load-bearing recovery zone, the system determines the deposition thickness, cross-sectional compensation amount, and deposition compaction requirements corresponding to the continuity of the pressure boundary; and for the allowance shaping zone, the system determines the finishing allowance and contour approximation accuracy.

[0042] In practical applications, the target high-temperature and high-pressure metal component is a nickel-based high-temperature alloy high-pressure valve seat. The zonal repair strategy indicates that there is a thermal fatigue crack zone on the outer side of the sealing line, an erosion thinning zone on the inner side of the sealing surface, and a metallurgical transition needs to be formed in the middle area. Based on the location of the sealing line, the depth of the erosion pit, and the outward expansion range of the crack tip, the system configures the area near the sealing line as a high-precision deposition zone, the main area of ​​the erosion pit as a load-bearing recovery zone, and the boundary area after crack removal as an interface transition zone. For example, near the sealing line, the single-layer deposition height is required to be controlled between 0.04 mm and 0.08 mm; in the main area of ​​the erosion pit, the deposition height must be continuous with the original sealing surface; and in the interface transition zone, the penetration depth must be less than the preset interface thermal impact threshold and the layers must continuously overlap.

[0043] During microscale spatial discretization, the system uses the component coordinate system as a reference to segment the target repair area according to functional partition boundaries, local curvature, target forming contour, and the minimum stable deposition scale of the printing equipment, generating multiple ultramicro deposition units. For the area near the sealing line with large curvature changes, a smaller spatial scale is used for discretization, so that a single ultramicro deposition unit covers a shorter arc segment; for the bearing recovery area at the center of the erosion pit, a relatively larger spatial scale is used for discretization to maintain deposition efficiency and cross-sectional continuity; for the interface transition area, a sequence of transition units is generated in a manner that gradually extends from the substrate side to the repair material side, so that adjacent ultramicro deposition units have continuous material constraints and thermal input constraints.

[0044] After each ultramicro deposition unit is generated, the system binds its spatial location, target shaping profile, and adjacent connectivity relationships. The spatial location determines the unit's deposition start point, deposition end point, and deposition layer in the component coordinate system; the target shaping profile defines the required local height, width, curvature, and overlap boundaries of the unit; and the adjacent connectivity relationships characterize the connection methods between this unit and preceding deposition units, subsequent deposition units, and adjacent functional zones. For high-pressure valve seat sealing surfaces, the system binds ultramicro deposition units at the edge of the sealing line to adjacent interface transition units as strong connections, and binds filling units inside erosion pits to residual shaping units as profile continuity relationships.

[0045] When determining the deposition sequence, material constraints, and thermal input constraints, the system comprehensively considers adjacent connectivity, load transfer status, and thermal diffusion status for prioritization. For ultramicro deposition units near the pressure seal line and crack tip, a sequence of low thermal input, dispersed deposition, and intermittent cooling is adopted to avoid localized heat accumulation. For load-bearing recovery areas, the deposition sequence is determined by transitioning from the depression center to the boundary and from strong load-bearing areas to weak load-bearing areas. For interface transition areas, material constraints are configured based on material compatibility and melt depth requirements, gradually changing from matrix-compatible materials to target repair materials, and continuous thermal input variation ranges are configured for adjacent units. Each ultramicro deposition unit forms a data structure containing unit number, spatial coordinates, deposition sequence, material constraints, thermal input constraints, and connectivity relationships.

[0046] Through the above processing, the target repair area can be refined from functional partitions into a set of independently controllable, real-time correctable, and traceable ultra-micro depositional units. This embodiment can enhance the microscale expression capability of the repair path, improve the synergistic control accuracy between interface transition and load recovery, and reduce the risks of local overheating, lack of fusion, and contour mismatch during the repair process.

[0047] In some embodiments, based on ultra-micro deposition units, a repair process knowledge graph, and an artificial intelligence repair decision model, an ultra-micro 3D printing repair path and corresponding process parameters are generated, including: Material damage process association data matching each ultramicro deposition unit is retrieved from the repair process knowledge graph to generate unit process candidate data; The candidate data of unit processes, deposition sequence, material constraints and thermal input constraints are input into the artificial intelligence repair decision model to generate candidate repair paths and candidate process parameters. Based on the high temperature and high pressure service constraints, the forming state prediction and constraint verification of candidate repair paths and candidate process parameters are performed. The ultra-micro 3D printing repair path and corresponding process parameters are determined based on the constraint verification results.

[0048] Specifically, after completing the division of ultrafine deposition units, the system first reads the unit number, spatial coordinates, target forming contour, functional partition label, deposition sequence, material constraints, and thermal input constraints corresponding to each ultrafine deposition unit, and uses the above data as the basic input for path generation. The repair process knowledge graph pre-stores the correlation between material type, damage mechanism, service condition, printing process, forming defects, and repair verification. Based on the matrix material, damage type, load-bearing correlation level, and interface transition requirements corresponding to the ultrafine deposition unit, the system retrieves matching material damage process correlation data in the repair process knowledge graph to form unit process candidate data.

[0049] In specific applications, the target high-temperature and high-pressure metal component is a nickel-based high-temperature alloy high-pressure valve seat. Thermal fatigue cracks exist on the outer side of the sealing line, and erosion thinning zones exist on the inner side of the sealing surface. The interface transition area requires control of matrix dilution and the range of thermally affected components. For ultramicro deposition units near the crack removal boundary, the system retrieves candidate process data related to low-heat-input deposition, interlayer slow cooling, local remelting, and microscale overlap from the repair process knowledge graph. For ultramicro deposition units in the load-bearing recovery area, the system retrieves candidate process data related to deposition compactness, cross-sectional continuity, and pressure boundary repair. For ultramicro deposition units in the margin shaping area, the system retrieves candidate process data related to contour compensation, surface smoothing, and subsequent finishing matching. The resulting unit process candidate data includes at least candidate material ratios, candidate scanning trajectories, candidate energy input ranges, candidate powder delivery ranges, candidate overlap methods, and candidate interlayer control methods.

[0050] Subsequently, the system inputs candidate data for unit processes, deposition sequence, material constraints, and thermal input constraints into the AI-powered remediation decision model. The AI-powered remediation decision model can include a path generation network, a process parameter prediction network, and a constraint scoring network. The path generation network generates candidate remediation paths based on the adjacent connectivity, load transfer state, and thermal diffusion state between each ultramicro deposition unit. The process parameter prediction network outputs corresponding candidate printing parameters based on material constraints and thermal input constraints. The constraint scoring network comprehensively scores the candidate remediation paths and candidate process parameters based on the functional zone to which the unit belongs and historical remediation cases. For example, for ultramicro deposition units near the valve seat sealing line, the model prioritizes generating short arc-shaped paths that transition continuously along the sealing ring surface, and configures smaller layer thickness, lower laser energy, higher monitoring frequency, and interval cooling time. For the central region of the erosion pit, the model generates a filling path that gradually overlaps from the center of the depression to the boundary, and configures a deposition width and powder feed rate that match the restoration of the bearing section.

[0051] After generating candidate repair paths and candidate process parameters, the system performs forming state prediction and constraint verification based on high-temperature and high-pressure service constraints. Forming state prediction can generate temperature distribution, melt pool size, cooling rate, heat-affected zone range, residual stress trend, and interface bonding state for each ultramicro deposition unit. Constraint verification judges the prediction results based on pressure boundary continuity, sealing surface contour consistency, material compatibility, heat-affected zone limitations, and defect risk thresholds. If a candidate path leads to excessively high temperature peaks or residual stress concentration near the sealing line, the system reverts that unit to the AI ​​repair decision model, reducing heat input, adjusting the scanning direction, or increasing interlayer waiting time. If a candidate parameter results in insufficient interface melting depth, the system recombines the number of local remelting cycles, scanning speed, or energy input range until the prediction results meet the constraint verification requirements.

[0052] Based on the constraint verification results, the system determines the candidate repair paths and candidate process parameters that meet the requirements as the ultra-micro 3D printing repair paths and corresponding process parameters, and binds the final path segment, energy input parameters, material supply parameters, motion control parameters, interlayer control parameters, and online monitoring window to each ultra-micro deposition unit. This process ensures that the repair path matches the damage mechanism, material state, and high-temperature, high-pressure re-service constraints, improving the relevance of path generation and the accuracy of process parameter configuration, and reducing the risks of interface non-fusion, localized overheating, abnormal residual stress, and mismatched sealing contours.

[0053] In some embodiments, the forming state prediction and constraint verification of candidate repair paths and candidate process parameters are performed based on high temperature and high pressure service constraints. The ultra-micro 3D printing repair path and corresponding process parameters are determined based on the constraint verification results, including: Based on candidate repair paths and candidate process parameters, printing status prediction data corresponding to each ultramicro deposition unit is constructed. Based on the high temperature and high pressure service constraints, the thermoforming state, metallurgical bonding state and structural response state in the printing state prediction data are constrained and verified to generate unit constraint verification results. Based on the unit constraint verification results, the AI ​​repair decision model is called back to correct the path and parameters for ultra-micro deposition units that do not meet the constraints. Candidate repair paths and candidate process parameters that meet the constraints are determined as ultra-micro 3D printing repair paths and corresponding process parameters.

[0054] Specifically, after obtaining candidate repair paths and candidate process parameters, the system constructs printing state prediction data using ultra-micro deposition units as the smallest prediction object. This printing state prediction data is used to simulate and represent the thermoforming state, metallurgical bonding state, and structural response state of each ultra-micro deposition unit during the deposition process before actual ultra-micro 3D printing. For a nickel-based high-temperature alloy high-pressure valve seat sealing surface repair scenario, candidate repair paths include a short arc-segment low-heat input path near the sealing line, a layered filling path at the center of the erosion pit, and a gradient overlapping path in the interface transition region. Candidate process parameters include laser power, scanning speed, powder feed rate, spot diameter, overlap rate, interlayer waiting time, and local remelting parameters. The system binds these paths and parameters according to the ultra-micro deposition unit number to form unit-level prediction input.

[0055] When constructing the printing state prediction data, the system calculates the cumulative heat input, peak temperature, cooling rate, molten pool width, molten pool depth, heat-affected zone range, and interlayer temperature change of the corresponding region based on the spatial location, adjacent connection relationship, and deposition sequence of each ultra-micro deposition unit, thus forming the thermoforming state prediction data; based on the matrix material, repair material, interface transition requirements, and candidate process parameters, it predicts the dilution range, fusion depth, material transition continuity, and potential metallurgical defect risk at the unit interface, thus forming the metallurgical bonding state prediction data; based on the pressure boundary location, sealing surface profile, initial state of residual stress, and candidate deposition sequence, it predicts the local deformation, residual stress redistribution, pressure-bearing section continuity, and sealing profile deviation of the repair area, thus forming the structural response state prediction data.

[0056] The system performs constraint verification on the predicted printing status data based on high-temperature and high-pressure service constraints. These constraints can include the target component's re-service temperature range, pressure rating, allowable deviation of the sealing surface profile, pressure boundary continuity requirements, interface fusion depth range, upper limit of the heat-affected zone width, and residual stress control range. For ultra-micro deposition units near the high-pressure valve seat sealing line, if the predicted temperature peak exceeds the upper limit of the interface heat-affected zone, or if residual tensile stress is concentrated outside the sealing line, the system marks the unit as failing the thermoforming constraint; if the predicted fusion depth is below the lower limit of interface bonding, or if there is a dilution abrupt change in the material transition region, the system marks the unit as failing the metallurgical bonding constraint; if the predicted curvature of the sealing surface after deposition deviates from the design profile, or if there is a local discontinuity in the pressure-bearing section, the system marks the unit as failing the structural response constraint. Each verification result is linked to the unit number, risk type, and deviation location to generate unit constraint verification results.

[0057] For ultra-micro deposition units that do not meet the constraints, the system redirects the unit constraint verification results to the AI ​​repair decision model and performs path and parameter corrections based on the risk type. If the unit has a risk of local overheating, the model reduces energy input, increases scanning speed, increases interlayer waiting time, or adjusts the deposition sequence to allow heat to diffuse to adjacent low-temperature regions. If the unit has a risk of insufficient fusion, the model adds local remelting paths, adjusts the spot diameter, or reduces scanning speed to bring the interface fusion depth into the target range. If the unit has a risk of residual stress concentration, the model changes the continuous deposition path to a segmented staggered path and configures slow cooling or micro-shaping parameters before and after high-stress regions. If the unit has a risk of deviating from the sealing profile, the model corrects the path height compensation and overlap rate to ensure that the predicted forming profile is consistent with the designed sealing surface.

[0058] After callback correction, the system regenerates the printing status prediction data for the corresponding ultra-micro deposition unit and repeats the constraint verification until the thermoforming state, metallurgical bonding state, and structural response state of each ultra-micro deposition unit meet the high-temperature and high-pressure service constraints. The system determines the candidate repair paths and candidate process parameters that meet the constraints as the final ultra-micro 3D printing repair paths and corresponding process parameters, and generates printing instruction data containing unit number, path segment, process parameters, prediction status, and verification results.

[0059] The above embodiments can improve the matching degree between process parameters and service constraints by using pre-printing unit-level prediction and constraint callback correction, reduce the risk of local overheating, insufficient fusion, residual stress concentration and contour deviation, and improve the forming stability and re-service reliability of high-temperature and high-pressure metal parts after repair.

[0060] In some embodiments, during the ultra-micro 3D printing repair process, molten pool state data is collected and matched with the predicted forming state to identify deviations. Based on the matching results, the printing path and process parameters of subsequent ultra-micro deposition units are updated, including: According to the deposition sequence of the ultramicro deposition units, multimodal online monitoring data corresponding to the current ultramicro deposition unit are collected, and the melt pool state characterization is extracted. The molten pool state characterization is matched with the predicted forming state corresponding to the current ultrafine deposition unit to generate a unit forming deviation characterization. The type of forming risk and the scope of influence of the deviation in the current ultrafine deposition unit are determined based on the characterization of the unit forming deviation. Based on the type of forming risk and the range of influence of deviations, the printing path and process parameters of subsequent ultra-micro deposition units are updated.

[0061] Specifically, before performing ultra-micro 3D printing repair, the system loads the final printing instruction data into the printing control unit according to the deposition sequence of the ultra-micro deposition units, and simultaneously writes the predicted forming state, online monitoring window, and deviation judgment threshold corresponding to each ultra-micro deposition unit into the monitoring control unit. For the repair scenario of the high-pressure valve seat sealing surface of nickel-based high-temperature alloy, the printing area includes low-heat-input short arc segment units near the sealing line, layered filling units in the center of the erosion pit, and gradient overlapping units in the interface transition area. The system performs deposition sequentially according to the unit number, and starts multi-modal online monitoring when each current ultra-micro deposition unit enters the printing window.

[0062] Multimodal online monitoring data can include molten pool images, infrared temperature, spectral intensity, acoustic emission signals, spatter status, powder feeding status, motion attitude, and protective gas status. The system performs time synchronization and spatial calibration on the above data, ensuring that each frame of monitoring data corresponds to the specific location of the current ultramicro deposition unit in the component coordinate system. Subsequently, the system extracts molten pool state characteristics from the multimodal online monitoring data, which can include molten pool length, molten pool width, molten pool brightness distribution, temperature peak, cooling slope, spectral fluctuations, acoustic emission abrupt changes, spatter density, and powder feeding stability. For ultramicro deposition units near the high-pressure valve seat sealing line, if the infrared temperature curve shows a temperature peak lower than the predicted range, and the molten pool image shows a narrow molten pool width, the system records the molten pool state characteristics of this unit as a low heat input deviation; if the spatter density suddenly increases and the spectral intensity fluctuates significantly, the system records this unit as an overheating disturbance deviation.

[0063] The system performs deviation matching between the current molten pool state characterization of the ultrafine deposition unit and the corresponding predicted forming state. The predicted forming state includes the molten pool size, peak temperature, cooling rate, heat-affected zone range, and forming profile predicted before printing. During deviation matching, the system calculates the difference between the actual molten pool width and the predicted molten pool width, the difference between the actual peak temperature and the predicted peak temperature, and the difference between the actual cooling slope and the predicted cooling slope, and combines these with acoustic emission, sputtering, and powder feeding stability to generate a unit forming deviation characterization. This deviation characterization records not only the magnitude of the deviation but also its direction, duration, spatial location, and confidence level.

[0064] Based on the characterization of unit forming deviations, the system determines the forming risk type and deviation impact range of the current ultrafine deposition unit. If the melt pool width is too small, the peak temperature is too low, and the cooling rate is too fast, the system judges that the current unit has a risk of insufficient fusion and includes adjacent overlapping units and subsequent interface transition units in the deviation impact range; if the peak temperature is too high, the spatter density increases, and the melt pool boundary fluctuates, the system judges that the current unit has a risk of local overheating or porosity and includes subsequent units in the heat diffusion direction in the deviation impact range; if the acoustic emission signal changes abruptly and is accompanied by an abnormal cooling slope, the system judges that the current unit has a risk of microcracks and includes subsequent units on the same load transfer path in the deviation impact range.

[0065] After determining the type of forming risk and the scope of the deviation, the system updates the printing path and process parameters of subsequent ultra-micro deposition units. For the risk of insufficient fusion, the system can reduce the scanning speed of subsequent adjacent units, increase local energy input, increase the overlap rate, or add a local remelting path; for the risk of local overheating, the system can reduce energy input, increase interlayer waiting time, change the deposition sequence, or adjust the continuous path to a segmented path; for the risk of microcracks, the system can add slow cooling control, adjust the scanning direction, or reduce the thermal gradient of adjacent units. Taking a high-pressure valve seat as an example, when a low heat input deviation occurs in a short arc segment unit outside the sealing line, the system reduces the scanning speed of the next two adjacent units by 8% from the original set value, adds a local remelting path, and marks the area as a key re-inspection location after repair.

[0066] Through the above processing, the real-time state of the molten pool can be matched with the pre-printing prediction results at the unit level during the ultra-micro 3D printing repair process, and the forming deviation can be promptly transmitted to the subsequent ultra-micro deposition units. This embodiment can improve the dynamic correction capability of the printing process for local insufficient fusion, overheating disturbance and microcrack risk, improve the consistency of ultra-micro deposition forming, and enhance the interface bonding quality and re-service reliability of the repair area of ​​high-temperature and high-pressure metal parts.

[0067] In some embodiments, intelligent ultra-micro 3D printing repair results are generated based on post-repair inspection data, printing process records, and re-service constraints, including: The repaired inspection data and the printing process records are correlated according to the component coordinate system to generate repair evidence data; Based on the repair evidence data, the forming consistency, interface bonding state and defect residual state of the target repair area are evaluated to generate a repair quality characterization. Based on the repair quality characterization and reservice constraints, the service adaptation verification of the target repair area is performed, and the reservice verification results are generated. Based on the re-service verification results, intelligent ultra-micro 3D printing repair processing results are generated.

[0068] Specifically, after completing the ultra-micro 3D printing repair, the system acquires post-repair inspection data of the target repair area and uses the printing process records and re-service constraints for correlation evaluation. The post-repair inspection data can be derived from 3D contour re-measurement, ultrasonic testing, eddy current testing, infrared thermography, surface roughness testing, hardness testing, residual stress testing, and local microstructure testing. This data characterizes the actual contour, internal continuity, interface state, surface quality, mechanical state, and material microstructure of the repair area. The printing process records include the deposition sequence of each ultra-micro deposition unit, printing path, process parameters, molten pool state characterization, unit forming deviation characterization, path correction records, and key re-inspection markers. Re-service constraints include the temperature range, pressure rating, allowable deviation of the sealing contour, pressure continuity requirements, interface bonding requirements, and defect residue limitations corresponding to the target high-temperature, high-pressure metal component.

[0069] For a nickel-based high-temperature alloy high-pressure valve seat sealing surface repair scenario, the system uses the component coordinate system as a unified reference to spatially correlate the post-repair inspection data with the printing process records. For example, 3D contour re-measurement shows a local height deviation of 0.03 mm on the sealing ring surface, ultrasonic testing shows no continuous defects inside the erosion pit repair area, eddy current testing shows no abnormal signals in the original circumferential microcrack area, and residual stress testing shows that the tensile stress on the outer side of the sealing line is lower than a preset threshold. The system maps the above inspection results to the corresponding ultra-micro deposition units and binds them with the low heat input deviation, local remelting path, scanning speed correction, and key re-inspection markers recorded during the printing process to generate repair evidence data. For short arc segments where the melt pool width was too small during the printing process, the system increases the weight of the inspection results in that area and incorporates the contour re-measurement results, interface inspection results, and residual stress results after local remelting into the evidence chain.

[0070] After generating repair evidence data, the system evaluates the shape consistency, interface bonding state, and defect residue state of the target repair area. Shape consistency evaluation compares the deviation between the actual repaired contour and the target shape contour, and combines the height continuity, curvature continuity, and overlap smoothness between adjacent ultrafine deposition units to form a contour consistency evaluation result. Interface bonding state evaluation determines the bonding continuity between the matrix material and the deposited material based on non-destructive testing, hardness transition, residual stress distribution, and local microstructure data. Defect residue state evaluation identifies residual risks such as incomplete fusion, pores, cracks, inclusions, and abnormal heat-affected zones. The system summarizes the above evaluation results according to functional zones and ultrafine deposition units to generate a repair quality characterization.

[0071] Subsequently, the system performs service adaptation verification between the repair quality characterization and reservice constraints. For the area near the high-pressure valve seat sealing line, the system focuses on verifying whether the sealing profile deviation, interface bonding continuity, and residual stress distribution meet the high-temperature steam pressure sealing requirements; for the load recovery area, the system focuses on verifying whether the repair section continuity, internal defect residue, and pressure load transfer path meet the pressure boundary requirements; for the interface transition area, the system focuses on verifying whether the hardness transition, material microstructure continuity, and heat-affected zone meet the high-temperature cycling requirements. When a certain area fully meets the reservice constraints, the system marks the area as a reserviceable area; when a local area has profile deviations, insufficient detection confidence, or defect residue risks, the system generates a secondary inspection area, a secondary repair area, or a reduced-load use identifier.

[0072] Based on the re-service verification results, the system generates intelligent ultra-micro 3D printing repair processing results. These results include repair quality level, re-service verification status, key re-inspection locations, secondary repair prompts, permissible service boundaries, and process traceability information. Through this processing, post-repair inspection results, printing process data, and high-temperature, high-pressure re-service requirements can be integrated into a closed-loop evaluation. This embodiment improves the traceability of repair quality evaluation, enhances the accuracy of service compatibility assessment of the repaired area, and reduces the risk of sealing failure, pressure mismatch, and defect propagation in repaired components under high-temperature, high-pressure environments.

[0073] The following are system embodiments of this application, which can be used to execute the method embodiments of this application. For details not disclosed in the system embodiments of this application, please refer to the method embodiments of this application.

[0074] Figure 2 This is a schematic diagram of the structure of the artificial intelligence-based ultra-micro 3D printing repair system for metal parts provided in an embodiment of this application. Figure 2 As shown, the system includes: The acquisition module 201 is used to acquire the service condition data, inspection data, material state data and geometric morphology data of the target high temperature and high pressure metal component, and map them to the same component coordinate system to generate basic data for component repair. The fusion module 202 is used to fuse and characterize geometric defects, material degradation, internal defects, residual stress and load transfer status based on the component repair basic data, and generate a high temperature and high pressure service damage field. The determination module 203 is used to determine the damage mechanism type, repairable boundary and load-bearing associated area of ​​the target repair area based on the high temperature and high pressure service damage field, and generate a zoning repair strategy. The partitioning module 204 is used to divide the target repair area into multiple ultra-micro deposition units according to the partition repair strategy, and to configure the deposition sequence, material constraints and thermal input constraints for each ultra-micro deposition unit. Repair module 205 is used to generate ultra-micro 3D printing repair paths and corresponding process parameters based on ultra-micro deposition units, repair process knowledge graphs and artificial intelligence repair decision models; The matching module 206 is used to collect molten pool state data and match it with the predicted forming state during the ultra-micro 3D printing repair process, and update the printing path and process parameters of subsequent ultra-micro deposition units according to the matching results. The generation module 207 is used to generate intelligent ultra-micro 3D printing repair results based on the post-repair inspection data, printing process records, and re-service constraints.

[0075] In some embodiments, Figure 2 The acquisition module 201 establishes a component coordinate system based on the structural reference, pressure boundary reference, and repair positioning reference of the target high-temperature and high-pressure metal component; acquires multi-source repair data corresponding to the target high-temperature and high-pressure metal component, and performs data cleaning, time alignment, and source identification processing on the multi-source repair data; based on the component coordinate system, it registers the spatial location data, damage location data, and material state location data in the multi-source repair data to generate coordinate mapping data; it fuses the multi-source repair data according to the coordinate mapping data, and configures data confidence labels for the fused data to generate basic data for component repair.

[0076] In some embodiments, Figure 2 The fusion module 202 spatially partitions the basic data for component repair according to the component coordinate system, generating damage analysis units corresponding to the target high-temperature and high-pressure metal component; it extracts morphological deviation features, microstructure degradation features, defect distribution features, stress state features, and load correlation features from each damage analysis unit to generate multidimensional damage characterization data; based on the multidimensional damage characterization data and high-temperature and high-pressure service constraints, it uses a damage identification model that integrates physical constraints to determine the damage mechanism characterization of each damage analysis unit; and it performs field fusion based on the damage mechanism characterization and spatial adjacency relationship to generate a high-temperature and high-pressure service damage field.

[0077] In some embodiments, Figure 2 The determination module 203 identifies the damage mechanism type of each damage analysis unit based on the high temperature and high pressure service damage field and determines the mechanism continuity relationship between adjacent damage analysis units; it determines the repairable boundary of the target repair area based on the damage mechanism type, material state and load correlation characteristics; it determines the load-bearing correlation area corresponding to the target repair area based on the repairable boundary and load transfer state; and it performs functional partitioning of the target repair area according to the damage mechanism type, repairable boundary and load-bearing correlation area to generate a partitioned repair strategy.

[0078] In some embodiments, Figure 2The partitioning module 204 determines the deposition accuracy, interface transition requirements, and load recovery requirements corresponding to different functional zones in the target repair area according to the partition repair strategy; based on the deposition accuracy, interface transition requirements, and load recovery requirements, the target repair area is spatially discretized at a microscale to generate multiple ultra-micro deposition units; the spatial position, target forming contour, and adjacent connection relationship are bound to each ultra-micro deposition unit; the deposition sequence, material constraints, and thermal input constraints of each ultra-micro deposition unit are determined according to the adjacent connection relationship, load transfer state, and thermal diffusion state.

[0079] In some embodiments, Figure 2 The repair module 205 retrieves material damage process association data matching each ultramicro deposition unit from the repair process knowledge graph to generate unit process candidate data; it inputs the unit process candidate data, deposition sequence, material constraints, and thermal input constraints into the artificial intelligence repair decision model to generate candidate repair paths and candidate process parameters; it performs forming state prediction and constraint verification on the candidate repair paths and candidate process parameters based on high temperature and high pressure service constraints; and it determines the ultramicro 3D printing repair path and corresponding process parameters based on the constraint verification results.

[0080] In some embodiments, Figure 2 The repair module 205 constructs printing state prediction data corresponding to each ultra-micro deposition unit based on candidate repair paths and candidate process parameters; based on high temperature and high pressure service constraints, it performs constraint verification on the thermoforming state, metallurgical bonding state, and structural response state in the printing state prediction data, and generates unit constraint verification results; based on the unit constraint verification results, it calls back the artificial intelligence repair decision model to correct the path and parameters for ultra-micro deposition units that do not meet the constraints; and determines the candidate repair paths and candidate process parameters that meet the constraints as ultra-micro 3D printing repair paths and corresponding process parameters.

[0081] In some embodiments, Figure 2 The matching module 206 collects multimodal online monitoring data corresponding to the current ultramicro deposition unit according to the deposition sequence of the ultramicro deposition unit, and extracts the melt pool state characterization; it matches the melt pool state characterization with the predicted forming state corresponding to the current ultramicro deposition unit to generate a unit forming deviation characterization; it determines the forming risk type and deviation influence range of the current ultramicro deposition unit based on the unit forming deviation characterization; and it updates the printing path and process parameters of subsequent ultramicro deposition units based on the forming risk type and deviation influence range.

[0082] In some embodiments, Figure 2The generation module 207 associates the post-repair inspection data with the printing process record according to the component coordinate system to generate repair evidence data; based on the repair evidence data, it evaluates the forming consistency, interface bonding state and defect residue state of the target repair area to generate repair quality characterization; based on the repair quality characterization and re-service constraints, it performs service adaptation verification on the target repair area to generate re-service verification results; based on the re-service verification results, it generates intelligent ultra-micro 3D printing repair processing results.

[0083] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0084] Figure 3 This is a schematic diagram of the electronic device 3 provided in an embodiment of this application. Figure 3 As shown, the electronic device 3 of this embodiment includes: a processor 301, a memory 302, and a computer program 303 stored in the memory 302 and executable on the processor 301. When the processor 301 executes the computer program 303, it implements the steps in the various method embodiments described above. Alternatively, when the processor 301 executes the computer program 303, it implements the functions of each module / unit in the various system embodiments described above.

[0085] Electronic device 3 can be a desktop computer, laptop, handheld computer, cloud server, or other electronic device. Electronic device 3 may include, but is not limited to, processor 301 and memory 302. Those skilled in the art will understand that... Figure 3 This is merely an example of electronic device 3 and does not constitute a limitation on electronic device 3. It may include more or fewer components than shown, or different components.

[0086] The processor 301 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0087] The memory 302 can be an internal storage unit of the electronic device 3, such as a hard disk or memory of the electronic device 3. The memory 302 can also be an external storage device of the electronic device 3, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the electronic device 3. The memory 302 can also include both internal and external storage units of the electronic device 3. The memory 302 is used to store computer programs and other programs and data required by the electronic device.

[0088] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0089] If integrated modules / units are implemented as software functional units and sold or used as independent products, they can be stored in a readable storage medium (e.g., a computer-readable storage medium). Based on this understanding, all or part of the processes in the methods of the above embodiments can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program may include computer program code, which may be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable storage medium may include: any entity or device capable of carrying computer program code, recording media, USB flash drives, portable hard drives, magnetic disks, optical disks, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signals, telecommunication signals, and software distribution media, etc.

[0090] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. An artificial intelligence-based ultra-micro three-dimensional printing repair method for metal parts, characterized in that, include: Acquire service condition data, inspection data, material condition data, and geometric morphology data of the target high-temperature and high-pressure metal component, and map them to the same component coordinate system to generate basic data for component repair; Based on the basic data of component repair, geometric defects, material degradation, internal defects, residual stress and load transfer status are fused and characterized to generate a high temperature and high pressure service damage field. Based on the high temperature and high pressure service damage field, the damage mechanism type, repairable boundary and load-bearing related area of ​​the target repair area are determined, and a zonal repair strategy is generated. According to the zoning repair strategy, the target repair area is divided into multiple ultra-micro deposition units, and deposition sequence, material constraints and heat input constraints are configured for each ultra-micro deposition unit. Based on ultra-micro deposition units, a knowledge graph of repair processes, and an artificial intelligence repair decision-making model, an ultra-micro 3D printing repair path and corresponding process parameters are generated. During the ultra-micro 3D printing repair process, the molten pool state data is collected and matched with the predicted forming state to check for deviation. Based on the matching results, the printing path and process parameters of the subsequent ultra-micro deposition units are updated. Based on the post-repair inspection data, printing process records, and re-service constraints, intelligent ultra-micro 3D printing repair results are generated.

2. The method according to claim 1, characterized in that, The process involves acquiring service condition data, inspection data, material condition data, and geometric morphology data of the target high-temperature and high-pressure metal component, mapping them to the same component coordinate system, and generating basic data for component repair, including: Based on the structural reference, pressure boundary reference, and repair positioning reference of the target high-temperature and high-pressure metal component, establish the component coordinate system; Acquire multi-source repair data corresponding to the target high-temperature and high-pressure metal component, and perform data cleaning, time alignment, and source identification processing on the multi-source repair data; Based on the component coordinate system, spatial location data, damage location data, and material state location data in multi-source repair data are registered to generate coordinate mapping data. Multi-source repair data are fused based on coordinate mapping data, and data confidence labels are configured for the fused data to generate basic data for component repair.

3. The method according to claim 1, characterized in that, Based on the component repair baseline data, geometric defects, material degradation, internal defects, residual stress, and load transfer status are fused and characterized to generate a high-temperature and high-pressure service damage field, including: The basic data for component repair is spatially partitioned according to the component coordinate system to generate damage analysis units corresponding to the target high-temperature and high-pressure metal component. From each damage analysis unit, morphological deviation features, tissue degradation features, defect distribution features, stress state features and load correlation features are extracted to generate multidimensional damage characterization data. Based on multidimensional damage characterization data and high-temperature and high-pressure service constraints, the damage mechanism characterization of each damage analysis unit is determined using a damage identification model that integrates physical constraints. Field fusion is performed based on damage mechanism characterization and spatial adjacency to generate a high-temperature and high-pressure service damage field.

4. The method according to claim 1, characterized in that, The step of determining the damage mechanism type, repairable boundary, and load-bearing associated region of the target repair area based on the high-temperature and high-pressure service damage field, and generating a zonal repair strategy, includes: Based on the high temperature and high pressure service damage field, the damage mechanism type of each damage analysis unit is identified, and the mechanism continuity relationship between adjacent damage analysis units is determined. The repairable boundary of the target repair area is determined based on the damage mechanism type, material state, and load correlation characteristics. Based on the repairable boundary and load transfer status, determine the load-bearing associated area corresponding to the target repair area; Based on the damage mechanism type, repairable boundary, and load-bearing associated area, the target repair area is functionally partitioned to generate a partitioned repair strategy.

5. The method according to claim 1, characterized in that, The process involves dividing the target repair area into multiple ultrafine deposition units according to a zonal repair strategy, and configuring deposition sequence, material constraints, and thermal input constraints for each ultrafine deposition unit, including: Based on the zoning remediation strategy, determine the deposition accuracy, interface transition requirements, and load-bearing recovery requirements corresponding to different functional zones in the target remediation area; Based on the requirements of deposition accuracy, interface transition and load recovery, the target repair area is spatially discretized at a microscale to generate multiple ultramicro deposition units. Each ultramicro deposition unit is bound with its spatial location, target shape profile, and adjacent connection relationship; Based on the adjacent connection relationship, load transfer state and thermal diffusion state, the deposition sequence, material constraints and thermal input constraints of each ultramicro deposition unit are determined.

6. The method according to claim 1, characterized in that, The process, based on ultra-micro deposition units, a knowledge graph of repair processes, and an artificial intelligence repair decision model, generates an ultra-micro 3D printing repair path and corresponding process parameters, including: Material damage process association data matching each ultramicro deposition unit is retrieved from the repair process knowledge graph to generate unit process candidate data; The candidate data of unit processes, deposition sequence, material constraints and thermal input constraints are input into the artificial intelligence repair decision model to generate candidate repair paths and candidate process parameters. Based on the high temperature and high pressure service constraints, the forming state prediction and constraint verification of candidate repair paths and candidate process parameters are performed. The ultra-micro 3D printing repair path and corresponding process parameters are determined based on the constraint verification results.

7. The method according to claim 6, characterized in that, The process involves predicting the forming state and verifying the constraints of candidate repair paths and process parameters based on high-temperature and high-pressure service constraints. The ultra-micro 3D printing repair path and corresponding process parameters are then determined based on the constraint verification results, including: Based on candidate repair paths and candidate process parameters, printing status prediction data corresponding to each ultramicro deposition unit is constructed. Based on the high temperature and high pressure service constraints, the thermoforming state, metallurgical bonding state and structural response state in the printing state prediction data are constrained and verified to generate unit constraint verification results. Based on the unit constraint verification results, the AI ​​repair decision model is called back to correct the path and parameters for ultra-micro deposition units that do not meet the constraints. Candidate repair paths and candidate process parameters that meet the constraints are determined as ultra-micro 3D printing repair paths and corresponding process parameters.

8. The method according to claim 1, characterized in that, During the ultra-micro 3D printing repair process, the molten pool state data is collected and matched with the predicted forming state to identify deviations. Based on the matching results, the printing path and process parameters of subsequent ultra-micro deposition units are updated, including: According to the deposition sequence of the ultramicro deposition units, multimodal online monitoring data corresponding to the current ultramicro deposition unit are collected, and the melt pool state characterization is extracted. The molten pool state characterization is matched with the predicted forming state corresponding to the current ultrafine deposition unit to generate a unit forming deviation characterization. The type of forming risk and the scope of influence of the deviation in the current ultrafine deposition unit are determined based on the characterization of the unit forming deviation. Based on the type of forming risk and the range of influence of deviations, the printing path and process parameters of subsequent ultra-micro deposition units are updated.

9. The method according to claim 1, characterized in that, The intelligent ultra-micro 3D printing repair process results are generated based on post-repair inspection data, printing process records, and re-service constraints, including: The repaired inspection data and the printing process records are correlated according to the component coordinate system to generate repair evidence data; Based on the repair evidence data, the forming consistency, interface bonding state and defect residual state of the target repair area are evaluated to generate a repair quality characterization. Based on the repair quality characterization and reservice constraints, the service adaptation verification of the target repair area is performed, and the reservice verification results are generated. Based on the re-service verification results, intelligent ultra-micro 3D printing repair processing results are generated.

10. A micro-3D printing repair system for metal parts based on artificial intelligence, characterized in that, include: The acquisition module is used to acquire service condition data, inspection data, material state data and geometric morphology data of the target high temperature and high pressure metal component, and map them to the same component coordinate system to generate basic data for component repair. The fusion module is used to fuse and characterize geometric defects, material degradation, internal defects, residual stress and load transfer status based on the basic data of component repair, and generate a high-temperature and high-pressure service damage field. The determination module is used to determine the damage mechanism type, repairable boundary and load-bearing associated area of ​​the target repair area based on the high temperature and high pressure service damage field, and generate a zonal repair strategy. The partitioning module is used to divide the target repair area into multiple ultra-micro deposition units according to the partitioning repair strategy, and to configure the deposition sequence, material constraints and thermal input constraints for each ultra-micro deposition unit. The repair module is used to generate ultra-micro 3D printing repair paths and corresponding process parameters based on ultra-micro deposition units, repair process knowledge graphs, and artificial intelligence repair decision models. The matching module is used to collect molten pool state data and match it with the predicted forming state during the ultra-micro 3D printing repair process, and update the printing path and process parameters of subsequent ultra-micro deposition units based on the matching results. The generation module is used to generate intelligent ultra-micro 3D printing repair results based on post-repair inspection data, printing process records, and re-service constraints.

Images