Dual-criteria process window construction and parameter optimization method for laser powder bed fusion of 7075 aluminum alloy
Patent Information
- Application Number
- CN202610605886.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-05-06
- Publication Date
- 2026-09-15
Smart Images
Figure CN122758818A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of laser selective melting and forming technology for aluminum alloys, specifically to a dual-criteria process window construction and parameter optimization method for laser powder bed melting and forming of 7075 aluminum alloys. Background Technology
[0002] 7075 aluminum alloy belongs to the Al-Zn-Mg-Cu series of high-strength aluminum alloys. It has high specific strength, high specific stiffness, good heat treatment strengthening ability and excellent comprehensive mechanical properties. With the continuous improvement of integrated manufacturing of complex components, lightweight structural design and high-performance service requirements, traditional casting, forging and machining methods are gradually becoming unable to meet the needs of high-end manufacturing in terms of material utilization, structural complexity adaptability and manufacturing cycle.
[0003] Selective laser melting (SLM), a key technology in metal additive manufacturing, uses a high-energy laser beam to scan and melt spread metal powder layer by layer, enabling the direct fabrication of three-dimensional solid parts. Compared to traditional subtractive manufacturing methods, SLM offers advantages such as high forming freedom, high material utilization, short manufacturing cycle, and the ability to integrally form complex internal cavities and irregular structures, making it particularly suitable for the fabrication of high-performance, complex metal components. SLM demonstrates significant application potential for lightweight, high-strength, and complex topologically structured components required in aerospace and other fields. Therefore, applying SLM to the fabrication of 7075 aluminum alloy components is considered a crucial direction for expanding the additive manufacturing applications of high-strength aluminum alloys.
[0004] However, due to its high alloy element content, wide solidification temperature range, and significant tendency to hot cracking, 7075 aluminum alloy is highly susceptible to defects such as cracks, porosity, and lack of fusion during the non-equilibrium metallurgical process of laser selective melting (SLM) with rapid melting and solidification. Among these, cracking is particularly prominent, not only weakening the density and mechanical properties of the formed parts but also severely impacting the service reliability and lifespan of components, becoming a key bottleneck restricting the laser selective melting forming of 7075 aluminum alloy. Simultaneously, the large temperature gradient, rapid cooling rate, and complex thermal cycle during laser selective melting, coupled with the interaction of molten pool flow, elemental segregation, residual stress accumulation, and solidification behavior evolution, make 7075 aluminum alloy extremely sensitive to process parameters during forming. Even slight deviations can lead to a decrease in surface forming quality, an increase in internal defects, and enhanced crack susceptibility.
[0005] Existing research indicates that process parameters such as laser power, scanning speed, scanning spacing, and layer thickness collectively determine the heat input level per unit volume, further influencing the molten pool geometry, fusion state, solidification structure, and defect formation mechanism. Specifically, excessively low laser power can lead to incomplete powder melting, resulting in unfused defects and low-density regions; conversely, excessively high laser power or low scanning speed can cause excessive local heat input, leading to enhanced evaporation spatter, increased thermal stress accumulation, and increased cracking tendency. Similarly, variations in scanning speed not only affect molten pool stability and interpass overlap quality but also directly relate to thermal history evolution and solidification behavior regulation. Therefore, laser selective melting forming of 7075 aluminum alloy does not simply depend on a single parameter but exhibits significant multi-parameter coupling characteristics and nonlinear response relationships. Its stable forming window is typically narrow, making process optimization challenging.
[0006] Currently, the selection of process parameters for laser selective melting (SDM) of high-strength aluminum alloys often employs single-factor experiments, orthogonal experiments, or experience-based point-by-point experiments. These methods involve testing a limited number of parameter combinations and combining observations of surface morphology, relative density, or microstructure to select relatively optimal process conditions. While these methods can reveal the influence of parameter variations on forming quality to some extent, they still have significant shortcomings: Firstly, the limited number of discrete experimental points makes it difficult to accurately reflect the true mapping relationship between process parameters and forming quality within a continuous parameter space. Secondly, traditional trial-and-error selection methods typically involve long experimental cycles, high material consumption, and low optimization efficiency, making it difficult to meet the needs of crack-sensitive high-strength aluminum alloys for rapid identification of precise process windows. For material systems like 7075 aluminum alloy, which have narrow machinability windows and complex defect evolution, relying solely on conventional experimental methods often fails to balance efficiency and accuracy.
[0007] With the development of data-driven methods, machine learning has provided a new technical path for process modeling, performance prediction, and parameter optimization in complex manufacturing processes. Compared with traditional empirical screening and point-by-point experimental methods, machine learning methods have significant advantages in improving parameter optimization efficiency, reducing experimental costs, and enhancing predictive capabilities. However, in the study of laser selective melting forming of 7075 aluminum alloy, existing methods still focus on single-index prediction or single quality target optimization, such as relative density, porosity, or surface forming quality. However, relying on only one index is insufficient to comprehensively reflect the actual forming performance of the material. This is because, under certain parameter conditions, the sample may exhibit high density but still have obvious crack defects, leading to a decline in the overall performance of the component. Furthermore, the systematic construction and crack control research for stable process windows within a continuous parameter space still need further improvement.
[0008] Therefore, in response to the problems existing in the laser selective melting forming of 7075 aluminum alloy, such as high crack sensitivity, narrow process window, low efficiency of traditional parameter screening, and difficulty in simultaneously considering relative density and crack control, how to simultaneously achieve high densification level and low crack sensitivity in the laser selective melting forming process of 7075 aluminum alloy, and how to establish an evaluation system that can reflect dual quality objectives, so as to achieve synergistic control of high relative density and low crack tendency in the laser selective melting forming process of 7075 aluminum alloy, and provide technical support for the stable preparation of complex high-strength aluminum alloy components, is one of the key issues that urgently need to be solved in the current process optimization research. Summary of the Invention
[0009] To address the problems of high crack sensitivity, narrow process window, and low parameter selection efficiency in the laser selective melting forming process of 7075 aluminum alloy, and the difficulty of simultaneously achieving high relative density and low crack area ratio in existing parameter optimization methods, this invention provides a machine learning-based method for the synergistic optimization of dual indicators of 7075 aluminum alloy laser powder bed melting forming quality. By establishing a nonlinear mapping relationship between process parameters and forming quality indicators, the traditional parameter selection method, which relies on experience and point-by-point experiments, is transformed into a prediction and optimization process oriented towards a continuous parameter space. This enables the synergistic control of relative density and crack area ratio during the 7075 aluminum alloy laser powder bed melting forming process.
[0010] The specific technical solution is as follows: This invention provides a method for constructing a dual-criteria process window and optimizing parameters for laser powder bed melting of 7075 aluminum alloy, comprising the following steps: Within the preset process parameter range, multiple sets of different parameter combinations were designed, and laser powder bed melting forming experiments of 7075 aluminum alloy were carried out using multiple sets of different parameter combinations to obtain corresponding forming samples. The relative density of the forming samples and the crack area ratio were quantitatively analyzed to establish a dataset. Based on the dataset, a melt forming prediction model was constructed and trained that can predict the relative density and crack area ratio from laser power and scanning speed. Using the trained melt forming prediction model, the relative density and crack area ratio in the continuous process parameter space are predicted. Based on the preset dual criterion constraints, the prediction results are jointly screened to construct a dual criterion process window that meets the melt forming requirements. Based on the aforementioned dual-criteria process window, the parameter regions suitable for stable laser powder bed melting and forming of 7075 aluminum alloy are identified, and recommended parameter combinations are obtained.
[0011] This invention addresses the complex nonlinear relationship between process parameters (laser power, scanning speed) and two key quality indicators (relative density, crack area ratio) in the laser powder bed melting process of 7075 aluminum alloy. It employs a supervised learning regression model for high-precision fitting, enabling quantitative prediction and joint constraint screening of both indicators within a continuous parameter space. This avoids the pitfalls of single-indicator optimization (e.g., pursuing only high density) leading to high density but severe cracking, and single-indicator optimization (e.g., pursuing only low cracking) leading to crack-free but porous structures. By simultaneously predicting both indicators and performing joint constraint screening, the invention quantitatively identifies the equilibrium region where both indicators simultaneously meet thresholds within the parameter space. This region corresponds to a specific energy input window, where the thermodynamic conditions of the molten pool ensure sufficient melting and gas escape while controlling thermal stress above the solidus line of the 7075 alloy and below the crack initiation threshold. This is impossible to achieve with single-indicator methods. The final output dual-criteria process window is a continuous region, significantly improving process robustness.
[0012] Preferably, the preset process parameters include: laser power and scanning speed; wherein, the preset laser power range is 200-350 W, and the preset scanning speed range is 600-2100 mm / s.
[0013] Preferably, the crack area ratio is obtained by performing the following image processing on the microscopic image of the formed sample cross section: grayscale processing of the microscopic image, threshold segmentation to obtain a binarized image, identification of the crack target area and removal of noise and false defects, and statistical analysis of the proportion of the crack area to the total analysis area. The relative density is obtained by Archimedes' method or cross-sectional image method.
[0014] Alternatively, the crack area percentage is calculated using the following formula: , in, This represents the percentage of the crack area. The area of the cracked region. This represents the total area of the image analysis region.
[0015] Preferably, the construction and training of the melt forming prediction model includes: The model is selected from one or more of the following: Gaussian process regression model, random forest regression model, support vector regression model, or extreme gradient boosting regression model. With multiple sets of different parameter combinations as input, the relative density and crack area ratio are predicted by the basic model to complete the construction of the melt forming prediction model. Based on the constructed molten forming prediction model, the molten forming prediction model with the best prediction accuracy is selected through cross-validation to obtain the trained molten forming prediction model.
[0016] In a further preferred embodiment, the basic model is a Gaussian process regression model Preferably, the preset double-criterion constraint conditions comprise: the relative density is greater than or equal to 98.5%, and the proportion of crack area is less than or equal to 0.20.
[0017] In laser powder bed fusion forming of 7075 aluminum alloy, increasing laser energy input (increasing power or reducing scanning speed) usually increases relative density (promoting fusion and reducing air pores), but at the same time increases thermal stress and solidification shrinkage, and intensifies crack initiation and propagation; reducing laser energy input can reduce thermal stress and inhibit cracks, but easily leads to incomplete powder melting, poor interlayer bonding and formation of unfused pores. Therefore, the establishment of double-criterion constraint conditions in the present invention essentially screens two non-cooperative or even contradictory quality objectives, and only when both indicators reach an acceptable level at the same time is the parameter combination determined to be qualified. This avoids the pseudo-optimization problem caused by one-sided indicator selection in traditional methods.
[0018] Preferably, the combined screening refers to: at each parameter point in the continuous process parameter space, simultaneously judging whether the predicted relative density value corresponding to the point meets the preset density requirement and whether the predicted crack area proportion value meets the preset crack requirement, and only when the predicted values of both indicators meet the respective threshold conditions is the parameter point included in the double-criterion process window.
[0019] In another aspect, the present invention provides a laser powder bed fusion apparatus, comprising a control unit, wherein the control unit has built-in double-criterion process window data constructed by the double-criterion process window construction and parameter optimization method described above, and is configured to: allowing the forming operation to be performed when the received process parameter combination falls within the double-criterion process window; automatically adjusting to a recommended parameter combination within the window when the process parameter combination does not fall within the window.
[0020] In another aspect, the present invention provides a 7075 aluminum alloy formed member prepared by laser powder bed fusion, which is prepared by laser powder bed fusion forming using the recommended parameter combination obtained by the double-criterion process window construction and parameter optimization method described above.
[0021] Preferably, the preparation method of the 7075 aluminum alloy laser powder bed fusion formed member comprises: using 7075 aluminum alloy powder as a raw material, in laser powder bed fusion equipment, setting laser power and scanning speed based on the obtained recommended parameter combination, performing layer-by-layer powder spreading and selective melting to obtain the formed member.
[0022] More preferably, the laser powder bed melting equipment adopts a Gaussian spot mode with a laser spot diameter of 80 μm.
[0023] Compared with the prior art, the present invention has the following advantages: (1) This invention addresses the problems of high crack sensitivity, narrow process window, and complex parameter coupling in the laser powder bed melting process of 7075 aluminum alloy. It proposes a dual-index collaborative optimization method based on machine learning, which takes into account both high densification level and low crack sensitivity, and can more accurately characterize the comprehensive influence of process parameters on forming quality. Using relative density and crack area ratio as joint quality evaluation indicators, it breaks through the limitation of existing technologies that mainly rely on single relative density or surface forming quality for parameter optimization, and can more comprehensively reflect the quality of 7075 aluminum alloy laser powder bed melting.
[0024] (2) The present invention can establish a nonlinear mapping relationship between process parameters and forming quality indicators based on limited experimental data, and realize quality prediction and process window construction in continuous parameter space, thereby improving parameter screening efficiency and reducing experimental costs.
[0025] (3) The dual-criteria process window constructed in this invention can achieve synergistic constraints on high relative density and low crack area ratio, providing parameter basis and technical support for the stable forming of 7075 aluminum alloy laser powder bed melting and the preparation of complex components. Attached Figure Description
[0026] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below.
[0027] Figure 1 This is a flowchart illustrating the dual-criteria process window construction and parameter optimization method for laser powder bed melting forming of 7075 aluminum alloy.
[0028] Figure 2 Typical defects and local unfused areas are shown in the morphology diagrams of 7075 aluminum alloy forming samples.
[0029] Figure 3 The image shows the distribution morphology of cracks and pores in a 7075 aluminum alloy forming sample.
[0030] Figure 4 The graph shows the variation of crack area ratio with scanning speed under different laser power conditions.
[0031] Figure 5 The graph shows the variation of relative density with scanning speed under different laser power conditions.
[0032] Figure 6This is a predicted distribution of relative density in the laser power-scanning speed parameter space.
[0033] Figure 7 This is a predicted distribution of the crack area ratio in the parameter space.
[0034] Figure 8 This is the optimal process window diagram based on the dual criteria of relative density and crack area ratio. Detailed Implementation
[0035] The present invention will be further described below with reference to specific embodiments and accompanying drawings. It should be understood that these embodiments are for illustrative purposes only and are not intended to limit the scope of the invention. Experimental methods in the following embodiments that do not specify specific conditions are generally performed under conventional experimental conditions in the art or according to the operating conditions recommended by the equipment manufacturer.
[0036] The technical concept of this invention lies in addressing the problems of high crack sensitivity, narrow process window, and low parameter selection efficiency in the laser selective melting forming process of 7075 aluminum alloy, and the difficulty of simultaneously achieving high relative density and low crack area ratio in existing parameter optimization methods. This invention provides a machine learning-based method for the synergistic optimization of dual indicators for the quality of 7075 aluminum alloy laser powder bed melting forming. By establishing a nonlinear mapping relationship between process parameters and forming quality indicators, the traditional parameter selection method, which relies on experience and point-by-point experiments, is transformed into a prediction and optimization process oriented towards a continuous parameter space. This achieves synergistic control of relative density and crack area ratio during the 7075 aluminum alloy laser powder bed melting forming process.
[0037] like Figure 1 As shown in the figure, this embodiment provides a method for constructing a dual-criteria process window and optimizing parameters for laser powder bed melting of 7075 aluminum alloy, which includes the following steps: S1. Design multiple sets of different parameter combinations within the preset process parameter range, and conduct laser powder bed melting forming experiments on 7075 aluminum alloy using multiple sets of different parameter combinations to obtain corresponding forming samples. Perform relative density testing and quantitative analysis of crack area ratio on the forming samples to establish a dataset.
[0038] In this embodiment, 7075 aluminum alloy powder was used as the forming raw material, and forming experiments were conducted using a laser powder bed melting device. The 7075 aluminum alloy powder is generally spherical or nearly spherical, has good flowability and powder spreading performance, and can meet the requirements of laser powder bed melting forming.
[0039] Within the preset process parameters, a combination of laser power of 200-350 W and scanning speed of 600-2100 mm / s was designed to layer-by-layer powder spreading and selective melting of 7075 aluminum alloy powder, resulting in multiple sets of shaped samples under different process conditions. After forming, the samples underwent cross-sectional preparation, polishing, and microscopic observation. Figure 2 and Figure 3 Typical defect morphologies in the formed samples are shown, including localized lack of fusion, cracks, and porosity. 7075 aluminum alloy exhibits significant crack sensitivity during laser powder bed fusion forming. Quantitative evaluation of forming quality cannot be achieved solely based on a single appearance or localized microstructure. Therefore, this embodiment further introduces relative density and crack area ratio as joint evaluation indicators.
[0040] In this embodiment, the crack area ratio is obtained through digital processing of microscopic images. Specifically, the metallographic microscopic image of the formed sample cross-section is imported into image processing software to identify crack regions and perform area statistics on the microscopic image.
[0041] First, the original microscopic image is converted to grayscale to reduce the influence of color information on defect identification and enhance the grayscale difference between the crack region and the substrate region. Then, a segmentation threshold is set based on the grayscale characteristics of the crack region and the substrate region, and the image is segmented to obtain a binarized image. Based on the binarized result, the target crack region is identified, and noise, small-area pseudo-defects, and non-crack interference areas are removed to obtain the final crack region. Finally, the crack region area and the total analysis area of the image are statistically analyzed, and the proportion of the crack region area to the total analysis area is calculated as the specific value of the crack area ratio under the given process parameters. Through these image processing steps, the crack morphology information in the microscopic image can be transformed into quantifiable data for subsequent model construction and process window analysis.
[0042] The variation of crack area ratio with scanning speed under different process parameters is as follows: Figure 4 As shown. The relative density was measured using the Archimedes method. The variation of relative density with scanning speed under different process parameters is shown below. Figure 5 As shown, the relative density data and crack area ratio data corresponding to different laser powers and scanning speeds are processed to construct a supervised learning dataset.
[0043] S2. Based on the dataset, construct and train a melt forming prediction model that can predict the relative density and crack area ratio from laser power and scanning speed.
[0044] In this embodiment, based on the established supervised learning dataset, laser power and scanning speed are used as input variables, and relative density and crack area ratio are used as output variables. In this embodiment, Gaussian process regression, random forest, support vector regression, and gradient boosting regression models are established respectively, and trained and cross-validated on the aforementioned supervised learning dataset. The model evaluation metrics include the coefficient of determination (R²), mean absolute error (MAE), and root mean square error (RMSE). R² is used to evaluate the model's fit to the experimental data's variation patterns, while MAE and RMSE are used to evaluate the magnitude of the model's prediction error. Based on the evaluation results of different models in the relative density prediction task and the crack area ratio prediction task, as shown in Tables 1 and 2, the model with better overall prediction ability is selected as the subsequent continuous parameter space prediction model.
[0045] Table 1. Performance Comparison Results of Relative Density Prediction Models GPR 0.969 0.686 0.996 RF 0.963 0.821 1.102 SVR 0.927 1.037 1.493 XGB -0.118 5.381 6.044 Table 2. Performance Comparison Results of Crack Area Proportion Prediction Models GPR 0.951 0.038 0.046 RF 0.880 0.058 0.071 SVR 0.854 0.067 0.078 XGB -0.373 0.214 0.243 Among them, GPR is Gaussian Process Regression, RF is Random Forest, SVR is Support Vector Regression, and XGB is Extreme Gradient Boosting.
[0046] Based on the optimized Gaussian process regression model, the continuous parameter space of laser power and scanning speed is predicted, and the predicted distribution of relative density in the parameter space is obtained, such as... Figure 6 As shown; the predicted distribution of the crack area ratio in the parameter space is obtained, as follows. Figure 7 As shown. Based on this, dual-criteria constraints are set to jointly screen the relative density prediction results and the crack area ratio prediction results, thereby obtaining the parameter region that simultaneously meets the requirements of high density and low crack tendency, forming a process window, such as... Figure 8 As shown.
[0047] In this embodiment, the dual-criteria constraint conditions are set as follows: relative density greater than or equal to 98.5%, and crack area ratio less than or equal to 0.20. By jointly constraining and screening the prediction results, the problems of "high relative density but obvious cracks" or "few cracks but insufficient density" that occur in traditional single-index screening methods can be effectively avoided, thereby achieving synergistic optimization of the laser powder bed melting forming quality of 7075 aluminum alloy.
[0048] Through the above embodiments, the present invention realizes the joint prediction and evaluation of relative density and crack area ratio during the laser powder bed melting process of 7075 aluminum alloy, and constructs a dual-criteria process window based on a continuous parameter space, which can be used for process parameter screening and optimization.
[0049] In summary, this invention establishes a nonlinear mapping relationship between process parameters and forming quality by jointly modeling and predicting the relative density and crack area ratio during the laser powder bed melting process of 7075 aluminum alloy, and constructs a dual-criteria process window based on a continuous parameter space. Compared with traditional parameter screening methods that rely on experience or point-by-point experiments, this invention can effectively characterize the influence of process parameters under limited experimental data conditions, improve parameter optimization efficiency, and reduce experimental costs. Furthermore, by introducing the crack area ratio, a quantitative indicator reflecting crack sensitivity, and combining it with relative density for synergistic constraints, the limitations of single-indicator evaluation can be avoided, achieving unified control over high density and low crack tendency. Therefore, the method proposed in this invention is applicable to the process optimization of crack-sensitive high-strength aluminum alloy laser powder bed melting process, and has good application value and promotion prospects.
[0050] Based on the same inventive concept, an embodiment provides a laser powder bed melting device, including a control unit. The control unit has built-in dual-criteria process window data constructed by the aforementioned dual-criteria process window construction and parameter optimization method, and is configured as follows: When the received combination of process parameters falls within the dual-criteria process window, the forming operation is allowed to be performed; If the combination of process parameters does not fall within the window, it will automatically adjust to the recommended combination within the window.
[0051] On the other hand, based on the same inventive concept, the embodiment also provides a 7075 aluminum alloy laser powder bed melting forming component, which is prepared by laser powder bed melting forming using the recommended parameter combination obtained by the aforementioned dual-criteria process window construction and parameter optimization method. Specifically, it includes: using 7075 aluminum alloy powder as raw material, in a laser powder bed melting device, based on the obtained recommended parameter combination, setting the laser power and scanning speed, performing layer-by-layer powder spreading and selective melting to obtain the formed component.
[0052] The laser powder bed melting equipment adopts a Gaussian spot mode with a laser spot diameter of 80 μm.
[0053] Furthermore, it should be understood that after reading the above description of the present invention, those skilled in the art can make various alterations or modifications to the present invention, and these equivalent forms also fall within the scope defined by the appended claims.
Claims
1. A dual-criteria process window construction and parameter optimization method for laser powder bed fusion of 7075 aluminum alloy, characterized in that, Includes the following steps: Within the preset process parameter range, multiple sets of different parameter combinations were designed, and laser powder bed melting forming experiments of 7075 aluminum alloy were carried out using multiple sets of different parameter combinations to obtain corresponding forming samples. The relative density of the forming samples and the crack area ratio were quantitatively analyzed to establish a dataset. Based on the dataset, a melt forming prediction model was constructed and trained that can predict the relative density and crack area ratio from laser power and scanning speed. Using the trained melt forming prediction model, the relative density and crack area ratio in the continuous process parameter space are predicted. Based on the preset dual criterion constraints, the prediction results are jointly screened to construct a dual criterion process window that meets the melt forming requirements. Based on the aforementioned dual-criteria process window, the parameter regions suitable for stable laser powder bed melting and forming of 7075 aluminum alloy are identified, and recommended parameter combinations are obtained.
2. The dual-criteria process window construction and parameter optimization method of claim 1, wherein, The preset process parameters include: laser power and scanning speed; the preset laser power range is 200-350 W; the preset scanning speed range is 600-2100 mm / s.
3. The dual-criteria process window construction and parameter optimization method of claim 1, wherein, The crack area ratio was obtained by performing the following image processing on the microscopic image of the formed specimen cross section: grayscale processing of the microscopic image, threshold segmentation to obtain a binary image, identification of the crack target area and removal of noise and false defects, and statistical analysis of the proportion of the crack area to the total analysis area. The relative density is obtained by Archimedes' method or cross-sectional image method.
4. The dual-criteria process window construction and parameter optimization method of claim 1, wherein, The construction and training of the melt forming prediction model includes: One or more of the following models are selected as the base model: Gaussian process regression model, random forest regression model, support vector regression model, extreme gradient boosting regression model. With multiple sets of different parameter combinations as input, the relative density and crack area ratio are predicted through the base model to complete the construction of the melt forming prediction model. Based on the constructed molten forming prediction model, the molten forming prediction model with the best prediction accuracy is selected through cross-validation to obtain the trained molten forming prediction model.
5. The dual-criteria process window construction and parameter optimization method of claim 1, wherein, The preset dual-criteria constraints include: relative density greater than or equal to 98.5%, and crack area ratio less than or equal to 0.
20.
6. The method for constructing a dual-criteria process window and optimizing parameters according to claim 5, characterized in that, The joint screening refers to simultaneously determining whether the predicted relative density value and the predicted crack area ratio at each parameter point in the continuous process parameter space meet the preset density requirements and crack area ratio requirements. Only when the predicted values of both indicators meet their respective threshold conditions is this parameter point included in the dual-criteria process window.
7. A laser powder bed melting device, characterized in that, The device includes a control unit, which has built-in dual-criteria process window data constructed using the dual-criteria process window construction and parameter optimization method according to any one of claims 1-6, and is configured as follows: When the received combination of process parameters falls within the dual-criteria process window, the forming operation is allowed to be performed; If the combination of process parameters does not fall within the window, it will automatically adjust to the recommended combination within the window.
8. A 7075 aluminum alloy laser powder bed fusion forming component, characterized in that, The formed component is prepared by laser powder bed melting forming using the recommended parameter combination obtained by the dual-criteria process window construction and parameter optimization method described in any one of claims 1-6.
9. The method for preparing 7075 aluminum alloy laser powder bed fusion forming components according to claim 8, characterized in that, include: Using 7075 aluminum alloy powder as raw material, in a laser powder bed melting device, based on the obtained recommended parameter combination, the laser power and scanning speed are set to carry out layer-by-layer powder spreading and selective melting to obtain shaped components.
10. The method for preparing 7075 aluminum alloy laser powder bed fusion forming components according to claim 9, characterized in that, The laser powder bed melting equipment adopts a Gaussian spot mode with a laser spot diameter of 80 μm.