A method and system for laser directed energy deposition interlayer thermal accumulation and residual stress co-prediction

Abstract

The application discloses a kind of laser directional energy deposition interlayer heat accumulation and residual stress collaborative prediction method and system, comprising: collecting process state data in laser directional energy deposition multilayer multichannel forming process, establish interlayer heat history data set;Interlayer heat accumulation state vector is constructed, interlayer heat accumulation index is calculated and heat accumulation risk grade is divided;Residual stress collaborative prediction model is constructed, and prediction result is obtained, residual stress risk index is calculated, residual tensile stress high-risk area is identified and deformation trend is determined;According to heat accumulation risk grade, residual stress risk evaluation result and deformation trend, interlayer process adjustment strategy is generated;Interlayer temperature deviation data and detection result after forming are obtained, prediction loss function is constructed, heat accumulation evaluation parameter and threshold value are corrected, residual stress risk threshold value, collaborative prediction model parameter and deformation trend judgment parameter are updated.The application can improve laser directional energy deposition component residual stress prediction accuracy and process control reliability.

Description

Technical Field

[0001] This invention relates to the field of metal additive manufacturing and laser processing process control technology, and in particular to a method and system for the synergistic prediction of interlayer thermal accumulation and residual stress in laser-directed energy deposition. Background Technology

[0002] Laser-directed energy deposition (LDED) is a metal additive manufacturing method that uses a high-energy laser beam as a heat source and simultaneously feeds in metal powder or wire to achieve layer-by-layer deposition. This method features high deposition efficiency, high material utilization, and strong adaptability to different forming dimensions, making it suitable for manufacturing large metal components, localized thickening, and damage repair. During multi-layer, multi-pass deposition, the laser heat source repeatedly acts on the deposited metal and the substrate, causing localized areas to undergo rapid heating, rapid cooling, and multiple thermal cycles. As the number of deposition layers increases, the residual heat from the previous layer is not fully dissipated before the next layer continues deposition, easily leading to interlayer heat accumulation in thin-walled areas, areas near free boundaries, corner areas, overlapping areas, arc initiation areas, and arc termination areas. This affects molten pool stability, cooling rate, dimensional accuracy, and residual stress distribution.

[0003] Residual stress is a significant factor affecting the forming accuracy and quality stability of laser-directed energy deposition (LDED) components. Due to the presence of localized high-temperature heat sources, layer-by-layer material addition, and complex geometric constraints in this process, the high-temperature deposition area is restricted during cooling and shrinkage by the surrounding solidified material, the substrate, and the component's own structure. This results in residual tensile stress, warping deformation, and even cracks within the component. Existing methods typically rely on offline finite element simulation, fixed empirical thresholds, or single-point temperature monitoring. However, these methods suffer from low computational efficiency, insufficient adaptability, difficulty in characterizing interlayer thermal history and the effects of local geometric constraints, and are insufficient to meet the rapid prediction requirements of multi-layer, multi-pass forming processes.

[0004] In existing technologies, some methods can monitor and predict molten pool morphology, layer height, porosity, cracks, or final deformation. However, they do not adequately consider the correlation between interlayer thermal accumulation and residual stress response, and lack methods to directly convert the prediction results into strategies for adjusting interlayer waiting time, scanning sequence, local laser power, and scanning speed. For laser-directed energy deposition (LDED), thermal accumulation is not a simple temperature rise phenomenon, but rather the result of the combined effects of interlayer initiation temperature, average temperature, high-temperature duration, linear energy density, unit powder feed energy, local geometric constraints, and cooling capacity. Therefore, it is necessary to propose a method and system for the coordinated prediction of interlayer thermal accumulation and residual stress in LDED to improve the accuracy of residual stress prediction and the reliability of process control. Summary of the Invention

[0005] This invention proposes a method and system for the coordinated prediction of interlayer thermal accumulation and residual stress in laser-directed energy deposition (LDED), which addresses the problems in existing technologies such as the difficulty in accurately characterizing the interlayer thermal accumulation state, insufficient correlation between residual stress prediction and interlayer thermal history, and the difficulty in directly converting prediction results into process adjustment strategies. The aim is to improve the accuracy of residual stress prediction, thermal process control capability, and forming quality stability in the multilayer, multi-pass forming process of LDED.

[0006] On the one hand, to achieve the above objectives, the present invention provides a method for the coordinated prediction of interlayer thermal accumulation and residual stress in laser-directed energy deposition, comprising: The process state data of the multi-layer, multi-pass laser directional energy deposition process were collected, and the geometric data of the components and the thermophysical and mechanical parameters of the materials were obtained. The process state data is preprocessed and encoded in a three-level system: layer-channel-region, based on sedimentary layers, sedimentary channels, and local regions. Layer, First Dao, Di Each local region is used as a prediction unit to establish an interlayer thermal history dataset with spatiotemporal correlation. An interlayer thermal accumulation state vector is constructed based on the interlayer thermal history dataset, and an interlayer thermal accumulation index is calculated based on the interlayer thermal accumulation state vector. Thermal accumulation risk levels are then classified based on a preset first thermal accumulation threshold and a second thermal accumulation threshold. Based on the interlayer thermal accumulation index, geometric constraint factor, and material thermophysical and mechanical parameters, the predicted values ​​of longitudinal residual stress, transverse residual stress, and equivalent residual stress are obtained using a residual stress co-prediction model. A residual stress risk index is calculated based on the interlayer thermal accumulation index, geometric constraint factor, and residual stress prediction results, and high-risk areas of residual tensile stress are identified based on the residual stress risk index. The deformation trend is determined based on the residual stress risk index, geometric constraint factor, and current deposition height. Based on the thermal accumulation risk level, the high-risk area of ​​residual tensile stress, the residual stress risk index, and the deformation trend, an interlayer process adjustment strategy is generated. The interlayer temperature data of each prediction unit during the forming process is acquired and the interlayer temperature deviation is calculated. The residual stress detection results and three-dimensional deformation detection results after forming are collected. A prediction loss function is constructed based on the interlayer temperature deviation, residual stress detection results and three-dimensional deformation detection results. The interlayer thermal accumulation evaluation parameters, the first thermal accumulation threshold and the second thermal accumulation threshold are corrected based on the interlayer temperature deviation. The residual stress risk threshold, residual stress collaborative prediction model parameters and deformation trend judgment parameters are updated based on the prediction loss function.

[0007] Preferably, the interlayer thermal accumulation state vector is: ; In the formula, For the first Layer, First Dao, Di Interlayer thermal accumulation state vector of a local region; This is the interlayer initiation temperature; Average temperature; Duration of high temperature; Linear energy density; Energy for powder delivery per unit; Geometric constraint factor; Cooling factor; This is the state mapping function.

[0008] Preferably, the geometric constraint factor is used to comprehensively characterize the influence of the local wall thickness, boundary distance, corner radius or path curvature, overlapping area location and degree of overlapping, arc initiation position and arc termination position of the prediction unit on local heat dissipation, repeated heating and stress release, specifically as follows: ; In the formula, This is the evaluation function for geometric constraints; It is a wall thickness sensitive factor used to characterize the influence of the local wall thickness of the prediction unit on heat conduction and structural constraints; This is a boundary sensitivity factor used to characterize the effect of the distance between the prediction unit and the free boundary on local heat dissipation and stress release; The path curvature factor is used to characterize the effect of the turning radius or path curvature on the degree of local heat input concentration. The overlap factor is used to characterize the effect of overlapping of adjacent sedimentary channels on local reheating. The arc initiation and termination factor is used to characterize the effect of the distance between the prediction unit and the arc initiation and termination positions on local repetitive heating.

[0009] Preferably, the interlayer heat accumulation index is: ; In the formula, For the first Layer, First Dao, Di Interlayer heat accumulation index of a local region; This is the thermal accumulation evaluation function; For the first Layer, First Dao, Di Interlayer thermal accumulation state vector of a local region; Based on the interlayer heat accumulation index and the preset first heat accumulation threshold Second thermal accumulation threshold Based on the comparison results, the prediction unit is divided into low heat accumulation risk level, medium heat accumulation risk level, or high heat accumulation risk level, wherein... < .

[0010] Preferably, the residual stress collaborative prediction model is as follows: ; In the formula, For the first Layer, First Dao, Di Vector of residual stress prediction results for each local region; This is a function for predicting residual stress. For the first Layer, First Dao, Di Interlayer heat accumulation index of a local region; Geometric constraint factor; It is a collection of thermophysical and mechanical parameters of materials; The residual stress prediction results include at least the longitudinal residual stress prediction value, the transverse residual stress prediction value, and the equivalent residual stress prediction value; The set of thermophysical and mechanical parameters of the material It includes at least one of thermal conductivity, specific heat capacity, coefficient of thermal expansion, elastic modulus, Poisson's ratio, and yield strength.

[0011] Preferably, the high-risk area of ​​residual tensile stress is identified by a residual stress risk index, which is specifically: ; In the formula, For the first Layer, First Dao, Di Residual stress risk index for a local area; This is a function for assessing residual stress risk. For the first Layer, First Dao, Di Interlayer heat accumulation index of a local region; For the first Layer, First Dao, Di Predicted equivalent residual stress values ​​for a local region; For the first Layer, First Dao, Di Geometric constraint factors for a local region; When the residual stress risk index Greater than or equal to the preset residual stress risk threshold When this happens, the corresponding prediction unit is identified as a high-risk area for residual tensile stress.

[0012] Preferably, the maximum tensile stress component is: ; In the formula, For the first Layer, First Dao, Di Predicted longitudinal residual stress values ​​for a local region; For the first Layer, First Dao, Di Predicted values ​​of transverse residual stress in a local area.

[0013] Preferably, the interlayer process adjustment strategy is as follows: ; In the formula, For the first Layer, First Dao, Di Interlayer process adjustment strategies for specific local areas; The risk level is determined by heat accumulation. The residual stress risk index; This indicates a trend towards deformation. Generate functions for process strategies; When the prediction unit is determined to be at a high heat accumulation risk level, the interlayer process adjustment strategy includes at least one of the following: extending the interlayer waiting time, reducing local laser power, increasing scanning speed, and adjusting the scanning sequence; when the prediction unit is determined to be a high-risk area of ​​residual tensile stress, the interlayer process adjustment strategy includes at least one of the following: path partitioning adjustment, scanning direction adjustment, local heat input restriction, and arc start and end position offset.

[0014] Preferably, the prediction loss function is: ; In the formula, This refers to the interlayer temperature deviation. This represents the error in residual stress prediction. Error in judging deformation trend; , , All are non-negative weighting coefficients, and ; For prediction loss function; The interlayer thermal accumulation evaluation parameters, the first thermal accumulation threshold, and the second thermal accumulation threshold are corrected based on the interlayer temperature deviation, and the residual stress risk threshold, residual stress collaborative prediction model parameters, and deformation trend judgment parameters are updated based on the prediction loss function.

[0015] On the other hand, to achieve the above objectives, the present invention also provides a system for co-predicting interlayer thermal accumulation and residual stress in laser-directed energy deposition, comprising: The process data acquisition unit is used to acquire process state data during the laser-directed energy deposition multilayer multi-pass forming process, and to obtain component geometric data and material thermophysical and mechanical parameters. The data preprocessing unit is used to preprocess the process state data and establish an interlayer thermal history dataset according to the deposition layer, deposition channel and local area. The inter-layer heat accumulation evaluation unit is used to construct an inter-layer heat accumulation state vector based on the inter-layer heat history dataset, calculate the inter-layer heat accumulation index, and classify the heat accumulation risk level. The residual stress collaborative prediction unit is used to obtain the prediction results of longitudinal residual stress, transverse residual stress and equivalent residual stress based on the interlayer thermal accumulation index, geometric constraint factor and material thermophysical and mechanical parameters, calculate the residual stress risk index, identify high-risk areas of residual tensile stress based on the residual stress risk index, and determine the deformation trend based on the residual stress risk index, geometric constraint factor and current deposition height. The interlayer process strategy generation unit is used to generate an interlayer process adjustment strategy based on the thermal accumulation risk level, the high-risk area of ​​residual tensile stress, the residual stress risk index, and the deformation trend. The feedback update unit is used to acquire interlayer temperature data of each prediction unit and calculate interlayer temperature deviation, collect residual stress detection results and three-dimensional deformation detection results, construct a prediction loss function based on the interlayer temperature deviation, residual stress detection results and three-dimensional deformation detection results, correct the interlayer thermal accumulation evaluation parameters, the first thermal accumulation threshold and the second thermal accumulation threshold based on the interlayer temperature deviation, and update the residual stress risk threshold, residual stress collaborative prediction model parameters and deformation trend judgment parameters based on the prediction loss function.

[0016] Compared with the prior art, the present invention has the following advantages and technical effects: (1) This invention divides the multi-layer, multi-channel deposition process into specific deposition layers, deposition channels, and local areas, so that locations prone to heat accumulation, such as thin-walled areas, areas near free boundaries, corner areas, overlapping areas, arc initiation areas, and arc termination areas, are no longer obscured by the overall temperature results. As a result, the system can determine the interlayer heat accumulation state for specific areas, avoiding rough judgments based solely on single-point temperatures or the overall temperature field, making the heat accumulation identification results more consistent with the actual forming process.

[0017] (2) The interlayer initiation temperature, average temperature, high temperature duration, linear energy density, unit powder feeding energy, geometric constraints and cooling conditions are organized into an interlayer thermal accumulation state vector, and the local thermal history is characterized by the thermal accumulation index. This links the prediction of residual stress with the interlayer thermal state in the actual forming process, reduces the judgment bias caused by relying solely on offline simulation or empirical thresholds, and is conducive to identifying high-risk areas of residual tensile stress and deformation trends.

[0018] (3) After obtaining the thermal accumulation risk and residual stress risk, the present invention transforms the prediction results into process adjustment strategies such as interlayer waiting time, scanning sequence, local laser power, scanning speed and path partitioning, so that the prediction results can directly serve the subsequent deposition process. The interlayer temperature deviation data recorded during the forming process is used to correct the thermal accumulation evaluation parameters, the first thermal accumulation threshold and the second thermal accumulation threshold. The residual stress detection results and three-dimensional deformation detection results after the manufacturing is completed are used to update the residual stress risk threshold, residual stress co-prediction model parameters and deformation trend judgment parameters, so that the prediction model can be continuously corrected according to the actual manufacturing results of similar components, and provide a more stable process basis for the forming process under similar material, geometric structure and process parameter conditions. Attached Figure Description

[0019] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments and descriptions of this application are used to explain this application and do not constitute an undue limitation of this application. In the drawings: Figure 1 This is a flowchart of a method for co-predicting interlayer thermal accumulation and residual stress in laser-directed energy deposition, according to an embodiment of the present invention. Figure 2 This is a schematic diagram of a laser-directed energy deposition interlayer thermal accumulation and residual stress co-prediction system according to an embodiment of the present invention. Detailed Implementation

[0020] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0021] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0022] This embodiment describes a multi-layer, multi-pass forming process using powder-fed laser-directed energy deposition (LDED). The forming material can be titanium alloy, high-temperature alloy, stainless steel, mold steel, or high-strength marine steel. The formed component can be a thin-walled part, cylindrical part, curved-wall part, corner structure, or locally thickened structure. The LED equipment includes a laser, a powder feeding device, a motion control system, an infrared thermal imager, a molten pool temperature measuring device, a data acquisition module, and a host computer processing module. The infrared thermal imager is used to acquire the surface temperature field of the formed component, the molten pool temperature measuring device is used to acquire the molten pool temperature, and the motion control system is used to record the scanning path, laser power, scanning speed, powder feeding rate, overlap rate, layer number, and inter-layer waiting time.

[0023] Specifically, such as Figure 1 As shown in this embodiment, a method for co-predicting interlayer thermal accumulation and residual stress in laser-directed energy deposition includes the following steps: S1. Collect process status data during the laser-directed energy deposition multilayer multi-pass forming process.

[0024] During the forming process, the process data acquisition unit collects data in real time, including the infrared temperature field, melt pool temperature, scanning path, interlayer waiting time, laser power, scanning speed, powder feeding rate, overlap rate, and layer number. Simultaneously, it reads the material's thermophysical and mechanical parameters and the component's geometric data.

[0025] The material's thermophysical and mechanical parameters include at least one of thermal conductivity, specific heat capacity, coefficient of thermal expansion, elastic modulus, Poisson's ratio, and yield strength. The component's geometric data includes at least one of local wall thickness, boundary distance, corner radius or path curvature, overlap area location and overlap degree, arc initiation location, and arc termination location.

[0026] In this embodiment, the first Layer, First The laser power, scanning speed, and powder feed rate of the channel are denoted as follows: , and ,in, Number the sedimentary layers. Number the sedimentary channels. For the first... Layer The linear energy density of the Tao is expressed as: ; In the formula, Linear energy density is used to characterize the laser energy input intensity per unit scan length.

[0027] When it is necessary to further consider the impact of powder feed rate on the thermal process, the unit powder feed energy should be established as follows: ; In the formula, For the first Layer, First The unit energy of powder delivery in the channel; For the first Layer, First The laser power of the channel; This refers to the powder delivery mass flow rate converted to a unified time unit. Unit powder delivery energy is used to characterize the laser energy input corresponding to a unit mass of powder.

[0028] It should be noted that linear energy density and unit powder delivery energy are both parameters related to heat input and are not directly equivalent to the total heat actually absorbed by the component. Actual heat absorption is also affected by material absorption rate, powder utilization rate, protective airflow, and local heat dissipation conditions.

[0029] S2. Preprocess the process status data and establish an interlayer thermal history dataset.

[0030] The data preprocessing unit uses the timestamp of the motion control system as a unified time reference to synchronize the infrared temperature field, molten pool temperature, scanning path, and process parameters. For data with different sampling frequencies, interpolation or resampling methods are used to unify them into the same time series.

[0031] For abnormal temperature points in the infrared temperature field, if they exceed the preset upper temperature limit, are below the lower limit of the ambient temperature, or have a difference from adjacent sampling points exceeding a preset fluctuation threshold, they are identified as abnormal points and removed. For short-term missing data, data from adjacent time points or historical data from the same area are used for compensation.

[0032] After time synchronization is completed, the infrared temperature field, molten pool temperature, and scanning path are mapped to the same component coordinate system. The component coordinate system is established with the upper surface of the substrate as the reference plane and the starting direction of the scanning path as the reference direction. For each pixel in the temperature field, it is converted into a temperature sampling point in the component coordinate system according to the camera calibration parameters and the spatial position of the component.

[0033] Subsequently, the formed components were divided according to sedimentary layers, sedimentary channels, and local areas, and then assigned layer numbers. , road number and area code Number them, and put the first Layer, First Dao, Di Each local region is treated as a prediction unit. Each prediction unit corresponds to at least one set of temperature time series, path coordinates, process parameters, and geometric parameters, thereby establishing an interlayer thermal history dataset.

[0034] For each prediction unit, extract its temperature time series: ; In the formula, For the first Layer, First Dao, Di Temperature time series of a local region For the first Layer, First Dao, Di The local region in the first Temperature values ​​at each sampling time.

[0035] In this embodiment, local areas can be divided according to path length, geometric features, or thermally sensitive regions. For ordinary straight-line deposition areas, they can be divided at equal intervals according to a fixed path length; for thin-walled areas, areas adjacent to free boundaries, corner areas, overlapping areas, arc initiation areas, and arc termination areas, they are divided into independent prediction units. This division method can prevent areas prone to heat accumulation from being masked by the overall average temperature.

[0036] S3. Construct an interlayer thermal accumulation state vector based on the interlayer thermal history dataset and calculate the interlayer thermal accumulation index.

[0037] The interlayer heat accumulation evaluation unit extracts the interlayer initiation temperature, average temperature, high temperature duration, linear energy density, unit powder delivery energy, geometric constraint factor, and cooling factor based on the temperature time series, process parameters, and geometric parameters of each prediction unit.

[0038] Among them, interlayer initiation temperature Indicates the first Before the first layer begins to deposit, the first Dao, Di The initial temperature of a local area is used to reflect the residual thermal state of that area after the previous layer was deposited.

[0039] For the Layer, First Dao, Di A temperature statistics time window is constructed for a local region, with the moment when the center of the laser heat source enters the prediction unit as the starting moment of temperature statistics and the moment after a preset cooling time after the laser heat source leaves the prediction unit as the ending moment of temperature statistics. The temperature statistics time window is used to characterize the temperature change characteristics of the prediction unit under laser heat input and subsequent cooling process.

[0040] Average temperature Obtained through the following methods: ; In the formula, The number of sampling points within the temperature statistics time window. For the first The temperature value corresponding to each sampling time.

[0041] Duration of high temperature Obtained through the following methods: ; In the formula, The time interval between adjacent sampling points. To preset the high temperature threshold, This is a decision function; it takes the value 1 if the condition within the parentheses is true, and 0 otherwise.

[0042] The duration of high temperature is calculated using the same temperature statistical time window as the average temperature. The duration of high temperature is used to characterize the degree to which the prediction unit remains in a high-temperature state within this temperature statistical time window.

[0043] Cooling factor The cooling rate is determined based on the cooling rate and the interlayer waiting time. In this embodiment, the cooling rate can be expressed as: ; In the formula, and These are two sampling times during the cooling phase, and ( > ); and These are the temperature values ​​at the corresponding sampling times; during the cooling phase, there are typically... ,therefore The larger the value, the faster the local area cools down.

[0044] The cooling factor can be expressed in the following form: ; In the formula, The function for calculating the cooling factor. For the first Inter-layer waiting time.

[0045] In one specific embodiment, the cooling factor Calculated using a dimensionless weighted method based on reference values: ; In the formula, This is a reference value for the cooling rate; This serves as a reference value for inter-floor waiting time. and It can be determined based on historical forming samples, forming data of similar materials, or the preset process allowable range; and These are non-negative weighting coefficients, and .

[0046] Cooling factor The larger the value, the more efficient the heat dissipation in the corresponding local area; cooling factor The smaller the value, the more significant the residual heat between layers in that region. Under the same conditions of other state parameters, the cooling factor... Interlayer heat accumulation index There is a negative correlation, that is, the higher the cooling factor, the lower the interlayer heat accumulation index.

[0047] Geometric constraint factor The expression used to comprehensively characterize the effects of local wall thickness, boundary distance, corner radius or path curvature, overlap area location and degree of overlap, arc initiation position and arc termination position on local heat dissipation, repeated heating and stress release of the prediction unit can be written as: ; In the formula, This is the evaluation function for geometric constraints; It is a wall thickness sensitive factor used to characterize the influence of the local wall thickness of the prediction unit on heat conduction and structural constraints; This is a boundary sensitivity factor used to characterize the effect of the distance between the prediction unit and the free boundary on local heat dissipation and stress release; The path curvature factor is used to characterize the effect of the turning radius or path curvature on the degree of local heat input concentration. The overlap factor is used to characterize the effect of overlapping of adjacent sedimentary channels on local reheating. The arc initiation and termination factor is used to characterize the effect of the distance between the prediction unit and the arc initiation and termination positions on local repetitive heating.

[0048] In this embodiment, geometric constraint factor Calculated using a normalized weighted method: ; In the formula, , , , and All are non-negative weighting coefficients, and .

[0049] Wall thickness sensitive factor It can be represented as: ; In the formula, For the first Layer, First Dao, Di Local wall thickness in a specific region; This is a reference value for wall thickness, which can be determined based on the typical wall thickness of similar components or a preset wall thickness range. The smaller the local wall thickness, the higher the wall thickness sensitivity factor. The larger the value, the more significant the impact of local wall thickness on heat conduction and structural constraints.

[0050] Boundary Sensitive Factor It can be represented as: ; In the formula, This is the distance from the predicted cell to the free boundary; The boundary affects the attenuation distance.

[0051] The closer the prediction unit is to the free boundary, The larger the value, the more significant the effect of the free boundary on local heat dissipation and stress release.

[0052] Path curvature factor It can be represented as: ; In the formula, The local curvature of the scan path corresponding to the prediction unit; This serves as a reference value for path curvature normalization.

[0053] The greater the path curvature, The larger the value, the higher the risk of localized heat input concentration caused by corners or curved paths.

[0054] Overlap factor It can be represented as: ; In the formula, The area of ​​the repeated heating zone between adjacent sedimentary channels within the predicted unit; This represents the area of ​​the sedimentary channel influence zone corresponding to the predicted unit.

[0055] The larger the overlap area, The larger the value, the more significant the effect of repeated heating from adjacent sedimentary channels on the predicted unit.

[0056] Arc initiation and termination factors It can be represented as: ; In the formula, This is the distance from the predicted cell to the arc initiation position; This is the distance from the predicted cell to the arc termination position; and These represent the attenuation distances caused by the arc initiation and arc termination positions, respectively.

[0057] The closer the prediction unit is to the arc start or end position, The larger the value, the more pronounced the repeated heating effect caused by the arc initiation or termination.

[0058] For ordinary straight-line depositional regions, the geometric constraint factor is set to a relatively low value; for thin-walled regions, regions adjacent to free boundaries, corner regions, overlapping regions, arc initiation regions, and arc termination regions, the geometric constraint factor is set to a relatively high value. The value of the geometric constraint factor is used to characterize the comprehensive influence of local geometry on heat dissipation, repeated heating, and stress release. An increase in its value only indicates an enhanced degree of local geometric influence and does not directly indicate an increase in interlayer heat accumulation or residual stress. The specific direction of influence is jointly determined by the heat accumulation evaluation function, the residual stress co-prediction function, and the residual stress risk evaluation function. In this way, the influence of local geometry on heat accumulation and residual stress can be incorporated into the evaluation and prediction process.

[0059] Furthermore, the above variables are constructed into an interlayer thermal accumulation state vector: ; In the formula, For the first Layer, First Dao, Di Interlayer thermal accumulation state vector of a local region; This is the interlayer initiation temperature; Average temperature; Duration of high temperature; Linear energy density; Energy for powder delivery per unit; Geometric constraint factor; Cooling factor; It is a state mapping function used to normalize and combine features of temperature, time, energy input, geometric parameters, and cooling parameters with different dimensions.

[0060] Furthermore, the interlayer heat accumulation index is: ; In the formula, For the first Layer, First Dao, Di Interlayer heat accumulation index of a local region; This is the thermal accumulation evaluation function; For the first Layer, First Dao, Di The interlayer thermal accumulation state vector represents a local region. A larger thermal accumulation index indicates a higher degree of interlayer thermal accumulation in that region.

[0061] In this embodiment, the thermal accumulation evaluation function A positive mapping is used for the interlayer initiation temperature, average temperature, high-temperature duration, linear energy density, and unit powder feeding energy, while a negative mapping is used for the cooling factor. The mapping relationship of the geometric constraint factor and various evaluation parameters are determined based on historical forming samples or calibration test results. The interlayer thermal accumulation evaluation parameters include at least one of the following: state mapping function parameters, thermal accumulation evaluation function parameters, and cooling factor calculation weights.

[0062] Furthermore, based on the interlayer heat accumulation index and the preset first heat accumulation threshold... Second thermal accumulation threshold Based on the comparison results, the prediction unit is divided into low heat accumulation risk level, medium heat accumulation risk level, or high heat accumulation risk level, wherein... < .

[0063] Specifically, in this embodiment, the heat accumulation risk level is classified according to the interlayer heat accumulation index: ; In the formula, For the first Layer, First Dao, Di The heat accumulation risk level of a local area; This is the first heat accumulation threshold; The second heat accumulation threshold is given, and ( < ).

[0064] S4. Construct a residual stress collaborative prediction model based on the interlayer thermal accumulation index, and predict residual stress risk and deformation trend.

[0065] The residual stress co-prediction unit inputs the interlayer thermal accumulation index, geometric constraint factor, and the set of material thermophysical and mechanical parameters into the residual stress co-prediction model. The set of material thermophysical and mechanical parameters is denoted as... It includes at least one of thermal conductivity, specific heat capacity, coefficient of thermal expansion, elastic modulus, Poisson's ratio, and yield strength.

[0066] The residual stress collaborative prediction model is as follows: ; In the formula, For the first Layer, First Dao, Di Vector of residual stress prediction results for each local region; This is a function for predicting residual stress. For the first Layer, First Dao, Di Interlayer heat accumulation index of a local region; Geometric constraint factor; It is a set of thermophysical and mechanical parameters of materials.

[0067] In practice, The model can be a fitting model based on finite element calculation results, a regression prediction model, a neural network model, a lookup table correction model, or a combination of the above models. The fitting model based on finite element calculation results provides the physical constraint basis for residual stress; the regression prediction model or neural network model is used to correct prediction deviations under different materials, paths, and geometries based on process data; the lookup table correction model is used to retrieve residual stress results from historical process cases under similar materials, similar geometries, and similar process parameters.

[0068] In this embodiment, the residual stress co-prediction function adopts a multivariate regression model, with interlayer thermal accumulation index, geometric constraint factor and material thermophysical and mechanical parameters as inputs, and longitudinal residual stress, transverse residual stress and equivalent residual stress as outputs. The model coefficients are obtained by fitting finite element calculation samples and historical forming test samples.

[0069] In this embodiment, the residual stress prediction results include longitudinal residual stress prediction values, transverse residual stress prediction values, and equivalent residual stress prediction values, which together constitute a residual stress prediction result vector: ; In the formula, For the first Layer, First Dao, Di Predicted longitudinal residual stress values ​​for a local region For the first Layer, First Dao, Di Predicted values ​​of transverse residual stress in a local area For the first Layer, First Dao, Di Predicted equivalent residual stress values ​​for a local area.

[0070] The longitudinal residual stress direction is the local tangential direction of the deposition channel corresponding to the prediction unit; the transverse residual stress direction is the direction perpendicular to the local tangential direction within the tangential plane of the local surface where the prediction unit is located.

[0071] To identify high-risk areas of residual tensile stress, a residual stress risk index is constructed: ; In the formula, For the first Layer, First Dao, Di Residual stress risk index for a local area; This is a function for assessing residual stress risk. For the first Layer, First Dao, Di Interlayer heat accumulation index of a local region; For the first Layer, First Dao, Di Predicted equivalent residual stress values ​​for a local region; For the first Layer, First Dao, Di Geometric constraint factors for a local region.

[0072] The maximum tensile stress component is calculated using the following formula: ; In the formula, For the first Layer, First Dao, Di Predicted longitudinal residual stress values ​​for a local region For the first Layer, First Dao, Di Predicted transverse residual stress values ​​for a local region. When both the predicted longitudinal and transverse residual stress values ​​are compressive stresses... The value is 0; when at least one direction is tensile stress, Take the larger tensile stress value between the predicted longitudinal residual stress and the predicted transverse residual stress.

[0073] In this embodiment, the residual stress risk assessment function The interlaminar thermal accumulation index, the predicted value of equivalent residual stress, the maximum tensile stress component, and the geometric constraint factor are obtained by dimensionless and weighted combination. The weights of each item are determined based on the material yield strength, historical forming samples, or calibration test results.

[0074] When the residual stress risk index Greater than or equal to the preset residual stress risk threshold When this occurs, the corresponding prediction unit is identified as a high-risk area for residual tensile stress; a preset residual stress risk threshold is set. It is determined based on at least one of the following: material yield strength, allowable residual tensile stress level, and statistical value of residual stress risk index of historical qualified process samples.

[0075] In practice, for thin-walled regions, regions adjacent to free boundaries, corner regions, overlapping regions, arc initiation regions, and arc termination regions, geometric constraint factors are used to correct the residual stress risk assessment results of the corresponding prediction units.

[0076] The deformation trend is determined based on the residual stress risk index, geometric constraint factor, and current deposition height, and is expressed as follows: ; In the formula, For the first Layer, First Dao, Di Deformation trends in local areas; This is a function for judging deformation trends; The residual stress risk index; ; This represents the current deposition height.

[0077] Current deposition height Determined based on the sedimentary layer number and the set thickness of each sedimentary layer; when the... The layer thickness is set to At that time, the current deposition height is expressed as: ; The deformation trend includes three states: insignificant, relatively significant, and significant, coded as 0, 1, and 2, respectively. The deformation trend is predicted by a deformation trend judgment function. Output: The actual deformation trend is determined based on the comparison between the 3D deformation detection result and the preset first deformation threshold and second deformation threshold, and adopts the same coding rule as the predicted deformation trend. The first deformation threshold is less than the second deformation threshold; when the measured 3D deformation is less than the first deformation threshold, the actual deformation trend is coded as 0; when the measured 3D deformation is greater than or equal to the first deformation threshold and less than the second deformation threshold, the actual deformation trend is coded as 1; when the measured 3D deformation is greater than or equal to the second deformation threshold, the actual deformation trend is coded as 2.

[0078] When the residual stress risk index is greater than or equal to the preset residual stress risk threshold, and the geometric constraint factor indicates that the corresponding prediction unit is located in a thin-walled region, a region adjacent to a free boundary, a corner region, an overlapping region, an arc initiation region, or an arc termination region, the deformation trend of the prediction unit is determined to be at least relatively obvious.

[0079] S5. Generate interlayer process adjustment strategies based on thermal accumulation risk and residual stress risk.

[0080] The interlayer process strategy generation unit generates a process adjustment strategy based on the thermal accumulation risk level, residual stress risk assessment results, and deformation trends. The residual stress risk assessment results include a residual stress risk index and high-risk areas for residual tensile stress. The interlayer process adjustment strategy is expressed by the following formula: ; in, For the first Layer, First Dao, Di Interlayer process adjustment strategies for specific local areas; The risk level is determined by heat accumulation. The residual stress risk index; This indicates a trend towards deformation; Generate functions for process strategies.

[0081] The interlayer process adjustment strategy is used in areas where the current layer has not yet been deposited, or in local areas in subsequent deposited layers corresponding to the prediction unit that triggers the interlayer process adjustment strategy.

[0082] In this embodiment, the process strategy generation function The corresponding control action is output according to different trigger conditions.

[0083] when When the corresponding prediction unit is determined to be at a high heat accumulation risk level, strategies such as extending the interlayer waiting time, reducing local laser power, increasing scanning speed, or adjusting the scanning sequence are prioritized.

[0084] The adjustment method can be expressed as: ; ; ; In the formula, This refers to the adjusted inter-floor waiting time. This refers to the waiting time between floors before the adjustment. The adjusted laser power, The laser power before adjustment. The adjusted scan speed, The original scan speed. , , This is an adjustment coefficient. All the above adjustment values ​​must meet the allowable range of the equipment and the stable forming range of the material.

[0085] when When the corresponding prediction unit is identified as a high-risk area for residual tensile stress, priority is given to generating strategies such as path partitioning adjustment, scanning direction adjustment, local heat input restriction, or arc initiation and termination position offset. For thin-walled areas and areas adjacent to free boundaries, local heat input is restricted first; for corner areas, scanning order is adjusted first or path transition is added first; for arc initiation and termination areas of cylindrical parts, the method of staggering the arc initiation positions of adjacent layers is preferred to reduce repeated heating in the same area.

[0086] when and During execution, a combined control strategy is employed, including extending the inter-layer waiting time, reducing local laser power, adjusting the scanning sequence, and changing the local scanning direction. After the strategy is generated, the corresponding layer number, track number, area number, adjustment parameters, and triggering reason are written into the process record table.

[0087] S6. Acquire process and post-forming inspection data, and update the data accordingly.

[0088] During the forming process, the feedback update unit acquires the interlayer temperature data of each prediction unit and calculates the interlayer temperature deviation between the measured temperature characteristic value and the target temperature characteristic value. After the component is manufactured, residual stress detection results and three-dimensional deformation detection results are collected. The residual stress detection results are obtained through at least one of X-ray diffraction, blind hole method, contour method, or neutron diffraction method; the three-dimensional deformation detection results are obtained through at least one of three-dimensional scanning, structured light measurement, or laser displacement sensor.

[0089] The spatial coordinates of residual stress detection points and three-dimensional deformation detection points are mapped to the component coordinate system and associated with the corresponding prediction units according to the layer number, channel number, and area number. For prediction units with corresponding detection results, the residual stress prediction error and deformation trend judgment error are calculated respectively.

[0090] Interlayer temperature deviation, residual stress prediction error, and deformation trend judgment error are respectively expressed as: ; ; ; in, For the first Layer, First Dao, Di Measured temperature characteristic values ​​of a local area For the target temperature feature value of the corresponding prediction unit, This serves as a normalized reference value for temperature deviation. To predict the actual residual stress of the corresponding prediction unit in a preset direction or a preset equivalent form, These are the residual stress prediction values ​​selected from the residual stress prediction result vector, representing the values ​​in the same direction or with the same equivalent form. This is a normalized reference value for residual stress; To determine the actual deformation trend based on the measured three-dimensional deformation results, The output of the deformation trend judgment function is the predicted deformation trend. This serves as a normalized reference value for the error in judging the deformation trend.

[0091] The measured temperature characteristic value and the target temperature characteristic value adopt the same characteristic form, which is the interlayer initiation temperature or average temperature, or a weighted combination of the interlayer initiation temperature and average temperature; the target temperature characteristic value is determined based on the temperature characteristic value of the corresponding prediction unit in the historical qualified process sample, the stable forming temperature range of similar materials, or the preset process temperature range.

[0092] The prediction loss function is calculated using the following formula: ; In the formula, This refers to the interlayer temperature deviation. This represents the error in residual stress prediction. Error in judging deformation trend; , , All are non-negative weighting coefficients, and ; The loss function is used for prediction; the weighting coefficients are preset based on the importance of the three types of errors in the feedback update or determined based on historical samples.

[0093] The prediction loss function When updating a single prediction unit, the value is the unit loss of that prediction unit; when evaluating or updating the entire process sample, the value is the average or weighted average of the losses of each prediction unit with the corresponding detection results.

[0094] When performing parameter feedback updates, for the residual stress co-prediction model and deformation trend judgment function using a differentiable model, the model parameter update is expressed as: ; In the formula, This is the updated set of model parameters; This is the set of model parameters before the update. To update the step size; For prediction loss function For the model parameter set The gradient. The model parameter set includes residual stress co-prediction model parameters and deformation trend judgment parameters.

[0095] When using regression fitting models, finite element result fitting models, or lookup table correction models that do not apply gradient updates, updates are performed through refitting, error compensation, or parameter correction.

[0096] When used for the prediction loss function of the entire process sample evaluation When the error is less than a preset error threshold, the corresponding process sample is marked as a valid prediction sample and written into the process case library; when the prediction loss function used for the evaluation of the entire process sample... When the error is greater than or equal to a preset error threshold, the interlayer thermal accumulation evaluation parameters, the first thermal accumulation threshold, and the second thermal accumulation threshold are corrected based on the interlayer temperature deviation. The residual stress risk threshold, residual stress co-prediction model parameters, and deformation trend judgment parameters are also updated. The updated parameters and model are used for subsequent laser-directed energy deposition forming processes under similar materials, similar geometries, and similar process parameter conditions.

[0097] Through the above steps, this embodiment forms a complete process from process status data acquisition, interlayer thermal history dataset establishment, interlayer thermal accumulation index calculation, residual stress co-prediction, interlayer process strategy generation to feedback update. This method can identify local thermal accumulation areas and high-risk areas of residual tensile stress in the multilayer multi-pass laser directional energy deposition forming process, and convert the prediction results into executable interlayer process adjustment strategies.

[0098] This embodiment also provides a system for the coordinated prediction of interlayer thermal accumulation and residual stress in laser-directed energy deposition, such as... Figure 2 ,include: The process data acquisition unit is used to acquire process state data during the laser-directed energy deposition multilayer multi-pass forming process, and to obtain component geometric data and material thermophysical and mechanical parameters. The data preprocessing unit is used to preprocess the process state data and establish an interlayer thermal history dataset according to the deposition layer, deposition channel and local area. The inter-layer heat accumulation evaluation unit is used to construct an inter-layer heat accumulation state vector based on the inter-layer heat history dataset, calculate the inter-layer heat accumulation index, and classify the heat accumulation risk level. The residual stress collaborative prediction unit is used to obtain the prediction results of longitudinal residual stress, transverse residual stress and equivalent residual stress based on the interlayer thermal accumulation index, geometric constraint factor and material thermophysical and mechanical parameters, calculate the residual stress risk index, identify high-risk areas of residual tensile stress based on the residual stress risk index, and determine the deformation trend based on the residual stress risk index, geometric constraint factor and current deposition height. The interlayer process strategy generation unit is used to generate an interlayer process adjustment strategy based on the thermal accumulation risk level, the high-risk area of ​​residual tensile stress, the residual stress risk index, and the deformation trend. The feedback update unit is used to acquire interlayer temperature data of each prediction unit and calculate interlayer temperature deviation, collect residual stress detection results and three-dimensional deformation detection results, construct a prediction loss function based on the interlayer temperature deviation, residual stress detection results and three-dimensional deformation detection results, correct the interlayer thermal accumulation evaluation parameters, the first thermal accumulation threshold and the second thermal accumulation threshold based on the interlayer temperature deviation, and update the residual stress risk threshold, residual stress collaborative prediction model parameters and deformation trend judgment parameters based on the prediction loss function.

[0099] The above are merely preferred embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for jointly predicting interlayer thermal accumulation and residual stress in laser-directed energy deposition, characterized in that, include: The process state data of the multi-layer, multi-pass laser directional energy deposition process were collected, and the geometric data of the components and the thermophysical and mechanical parameters of the materials were obtained. The process state data is preprocessed and encoded in a three-level system: layer-channel-region, based on sedimentary layers, sedimentary channels, and local regions. Layer, First Dao, Di Each local region is used as a prediction unit to establish an interlayer thermal history dataset with spatiotemporal correlation. An interlayer thermal accumulation state vector is constructed based on the interlayer thermal history dataset, and an interlayer thermal accumulation index is calculated based on the interlayer thermal accumulation state vector. Thermal accumulation risk levels are then classified based on a preset first thermal accumulation threshold and a second thermal accumulation threshold. Based on the interlayer thermal accumulation index, geometric constraint factor, and material thermophysical and mechanical parameters, the longitudinal residual stress prediction value, transverse residual stress prediction value, and equivalent residual stress prediction value are obtained using the residual stress collaborative prediction model; the residual stress risk index is calculated based on the interlayer thermal accumulation index, geometric constraint factor, and residual stress prediction results, and high-risk areas of residual tensile stress are identified based on the residual stress risk index. The deformation trend is determined based on the residual stress risk index, geometric constraint factor, and current deposition height. Based on the thermal accumulation risk level, the high-risk area of ​​residual tensile stress, the residual stress risk index, and the deformation trend, an interlayer process adjustment strategy is generated. The interlayer temperature data of each prediction unit during the forming process is acquired and the interlayer temperature deviation is calculated. The residual stress detection results and three-dimensional deformation detection results after forming are collected. A prediction loss function is constructed based on the interlayer temperature deviation, residual stress detection results and three-dimensional deformation detection results. The interlayer thermal accumulation evaluation parameters, the first thermal accumulation threshold and the second thermal accumulation threshold are corrected based on the interlayer temperature deviation. The residual stress risk threshold, residual stress collaborative prediction model parameters and deformation trend judgment parameters are updated based on the prediction loss function.

2. The method for co-predicting interlayer thermal accumulation and residual stress in laser-directed energy deposition according to claim 1, characterized in that, The interlayer thermal accumulation state vector is: ； In the formula, For the first Layer, First Dao, Di Interlayer thermal accumulation state vector of a local region; This is the interlayer initiation temperature; Average temperature; Duration of high temperature; Linear energy density; Energy for powder delivery per unit; Geometric constraint factor; Cooling factor; This is the state mapping function.

3. The method for synergistic prediction of interlayer thermal accumulation and residual stress in laser-directed energy deposition according to claim 2, characterized in that, The geometric constraint factors are used to comprehensively characterize the effects of the local wall thickness, boundary distance, corner radius or path curvature, overlapping area location and degree of overlapping, arc initiation position and arc termination position of the prediction unit on local heat dissipation, repeated heating and stress release, specifically: ； In the formula, This is the evaluation function for geometric constraints; It is a wall thickness sensitive factor used to characterize the influence of the local wall thickness of the prediction unit on heat conduction and structural constraints; This is a boundary sensitivity factor used to characterize the effect of the distance between the prediction unit and the free boundary on local heat dissipation and stress release; The path curvature factor is used to characterize the effect of the turning radius or path curvature on the degree of local heat input concentration. The overlap factor is used to characterize the effect of overlapping of adjacent sedimentary channels on local reheating. The arc initiation and termination factor is used to characterize the effect of the distance between the prediction unit and the arc initiation and termination positions on local repetitive heating.

4. The method for synergistic prediction of interlayer thermal accumulation and residual stress in laser-directed energy deposition according to claim 2, characterized in that, The interlayer heat accumulation index is: ； In the formula, For the first Layer, First Dao, Di Interlayer heat accumulation index of a local region; This is the thermal accumulation evaluation function; For the first Layer, First Dao, Di Interlayer thermal accumulation state vector of a local region; Based on the interlayer heat accumulation index and the preset first heat accumulation threshold Second thermal accumulation threshold Based on the comparison results, the prediction unit is divided into low heat accumulation risk level, medium heat accumulation risk level, or high heat accumulation risk level, wherein... < .

5. The method for co-predicting interlayer thermal accumulation and residual stress in laser-directed energy deposition according to claim 1, characterized in that, The residual stress collaborative prediction model is as follows: ； In the formula, For the first Layer, First Dao, Di Vector of residual stress prediction results for each local region; This is a function for predicting residual stress. For the first Layer, First Dao, Di Interlayer heat accumulation index of a local region; Geometric constraint factor; It is a collection of thermophysical and mechanical parameters of materials; The residual stress prediction results include at least the longitudinal residual stress prediction value, the transverse residual stress prediction value, and the equivalent residual stress prediction value; The set of thermophysical and mechanical parameters of the material It includes at least one of thermal conductivity, specific heat capacity, coefficient of thermal expansion, elastic modulus, Poisson's ratio, and yield strength.

6. The method for co-predicting interlayer thermal accumulation and residual stress in laser-directed energy deposition according to claim 5, characterized in that, The high-risk areas of residual tensile stress are identified by a residual stress risk index, which is specifically: ； In the formula, For the first Layer, First Dao, Di Residual stress risk index for a local area; This is a function for assessing residual stress risk. For the first Layer, First Dao, Di Interlayer heat accumulation index of a local region; For the first Layer, First Dao, Di Predicted equivalent residual stress values ​​for a local region; For the first Layer, First Dao, Di Geometric constraint factors for a local region; When the residual stress risk index Greater than or equal to the preset residual stress risk threshold When this happens, the corresponding prediction unit is identified as a high-risk area for residual tensile stress.

7. The method for synergistic prediction of interlayer thermal accumulation and residual stress in laser-directed energy deposition according to claim 6, characterized in that, The maximum tensile stress component is: ； In the formula, For the first Layer, First Dao, Di Predicted longitudinal residual stress values ​​for a local region; For the first Layer, First Dao, Di Predicted values ​​of transverse residual stress in a local area.

8. The method for co-predicting interlayer thermal accumulation and residual stress in laser-directed energy deposition according to claim 1, characterized in that, The interlayer process adjustment strategy is as follows: ； In the formula, For the first Layer, First Dao, Di Interlayer process adjustment strategies for specific local areas; The risk level is determined by heat accumulation. The residual stress risk index; This indicates a trend towards deformation. Generate functions for process strategies; When the prediction unit is determined to be at a high heat accumulation risk level, the interlayer process adjustment strategy includes at least one of the following: extending the interlayer waiting time, reducing local laser power, increasing scanning speed, and adjusting the scanning sequence; when the prediction unit is determined to be a high-risk area of ​​residual tensile stress, the interlayer process adjustment strategy includes at least one of the following: path partitioning adjustment, scanning direction adjustment, local heat input restriction, and arc start and end position offset.

9. The method for co-predicting interlayer thermal accumulation and residual stress in laser-directed energy deposition according to claim 1, characterized in that, The prediction loss function is: ； In the formula, This refers to the interlayer temperature deviation. This represents the error in residual stress prediction. Error in judging deformation trend; , , All are non-negative weighting coefficients, and ; For prediction loss function; The interlayer thermal accumulation evaluation parameters, the first thermal accumulation threshold, and the second thermal accumulation threshold are corrected based on the interlayer temperature deviation, and the residual stress risk threshold, residual stress collaborative prediction model parameters, and deformation trend judgment parameters are updated based on the prediction loss function.

10. A system for co-predicting interlayer thermal accumulation and residual stress in laser-directed energy deposition, characterized in that, The system for implementing the method for co-predicting interlayer thermal accumulation and residual stress in laser-directed energy deposition as described in any one of claims 1-9 comprises: The process data acquisition unit is used to acquire process state data during the laser-directed energy deposition multilayer multi-pass forming process, and to obtain component geometric data and material thermophysical and mechanical parameters. The data preprocessing unit is used to preprocess the process state data and establish an interlayer thermal history dataset according to the deposition layer, deposition channel and local area. The inter-layer heat accumulation evaluation unit is used to construct an inter-layer heat accumulation state vector based on the inter-layer heat history dataset, calculate the inter-layer heat accumulation index, and classify the heat accumulation risk level. The residual stress collaborative prediction unit is used to obtain the prediction results of longitudinal residual stress, transverse residual stress and equivalent residual stress based on the interlayer thermal accumulation index, geometric constraint factor and material thermophysical and mechanical parameters, calculate the residual stress risk index, identify high-risk areas of residual tensile stress based on the residual stress risk index, and determine the deformation trend based on the residual stress risk index, geometric constraint factor and current deposition height. The interlayer process strategy generation unit is used to generate an interlayer process adjustment strategy based on the thermal accumulation risk level, the high-risk area of ​​residual tensile stress, the residual stress risk index, and the deformation trend. The feedback update unit is used to acquire interlayer temperature data of each prediction unit and calculate interlayer temperature deviation, collect residual stress detection results and three-dimensional deformation detection results, construct a prediction loss function based on the interlayer temperature deviation, residual stress detection results and three-dimensional deformation detection results, correct the interlayer thermal accumulation evaluation parameters, the first thermal accumulation threshold and the second thermal accumulation threshold based on the interlayer temperature deviation, and update the residual stress risk threshold, residual stress collaborative prediction model parameters and deformation trend judgment parameters based on the prediction loss function.

Images