Data correction method and system combined with lattice structure additive manufacturing process characteristics

Abstract

The invention provides a data correction method and system combined with lattice structure additive manufacturing process characteristics, and the method comprises the steps: firstly employing a Newton iteration method, taking the radius of a pillar as a variable, optimizing the relative density of an iteration target and generating a lattice structure geometric model under the condition that a process constraint condition is satisfied, and extracting actual geometric parameters after forming through CT scanning, the elastic modulus is corrected through the pillar diameter deviation and the defect volume fraction, a compression failure mechanism is combined, a density-related failure criterion is introduced, a nonlinear relation with the relative density is constructed through specific energy absorption data, material plasticity parameters are inversely optimized, and a defect coupling evaluation model is constructed according to the surface powder sticking rate and the pillar diameter deviation. And establishing a Gaussian mixture model based on a stress-strain curve to screen out abnormal data, and finally fusing the parameters to construct a data correction model. The dot matrix structure design precision can be improved, and then a high-quality data basis is provided for performance prediction-structure design two-way feedback.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of dot matrix structure data optimization, and more particularly to a data correction method and system combining dot matrix structure additive manufacturing process features. BACKGROUND

[0002] With the continuous pursuit of equipment performance limits in cutting-edge fields such as aerospace and power energy, structural lightweighting has become a key bottleneck for improving the thrust-to-weight ratio, load efficiency and endurance capability. The traditional lightweighting design method is limited by manufacturability and structural topology, and its weight reduction potential has approached the theoretical limit. In this context, dot matrix structures with theoretical specific strength / specific stiffness close to the material upper limit and based on the classic porous material theory model of Gibson-Ashby show great advantages, have extreme lightweighting potential, high design freedom, and their open pore characteristics are more convenient for the integration of functions such as thermal management and energy absorption. Additive manufacturing (AM) technology provides a unique and almost unique technical approach for the integrated forming of such complex embedded dot matrix structures.

[0003] However, due to the complex three-dimensional cell geometry, powder accumulation and high surface roughness often occur in the overhanging part, which leads to significant deviations in the geometric shape and size of the manufactured component from the design values. For example, experimental observations have found that thin-walled rods often exhibit elliptical cross-sections rather than ideal cylindrical shapes, and these geometric shape deviations have a significant impact on the structural mechanical properties. Current research uses CT technology to reconstruct and numerically simulate typical TPMS dot matrix structures, emphasizing the key role of process defects on the relationship between structural properties, and proposing corresponding numerical methods to capture their influence. However, most current simulation analyses are still based on ideal geometric models, and do not fully consider actual process defects, so the prediction accuracy is low, and there is a large deviation from the actual mechanical properties. SUMMARY

[0004] The present application provides a data correction method and system combining dot matrix structure additive manufacturing process features, which can quantify the evolution law of key defects such as powder sticking rate and rod diameter deviation based on experiments and CT analysis, establish a data correction model of actual bearing density based on material sensitivity coefficients, significantly improve data quality and simulation accuracy, and data efficiency is improved to more than 90%.

[0005] In the first aspect, the present application provides a data correction method combined with the additive manufacturing process characteristics of a lattice structure, the method comprising: using the Newton iteration method to iteratively optimize the target relative density with the lattice pillar radius as a variable and combining preset process constraints to limit the minimum pillar diameter and pillar spacing to generate a geometric model of the lattice structure; obtaining the actual geometric parameters of the formed lattice structure through CT scanning and defining a corrected elastic modulus based on the defect volume fraction and pillar diameter deviation between the actual geometric parameters and the geometric model; introducing a density-dependent failure criterion based on the compression failure mechanism and establishing a nonlinear relationship between the specific energy absorption data and the relative density based on the specific energy absorption data, and inversely optimizing and correcting the material plastic parameters; establishing a defect coupling evaluation model for screening out defect data based on the surface powder adhesion rate of the lattice structure and the pillar diameter deviation parameters and constructing a Gaussian mixture model for screening out abnormal data based on the compression stress-strain curve characteristics; constructing a data correction model for two-way feedback of performance prediction and structural design based on the defect coupling evaluation model, Gaussian mixture model, corrected elastic modulus, failure criterion and corrected material plastic parameters.

[0006] In an optional solution of the first aspect, when performing relative density iteration using the Newton iteration method with the lattice pillar radius as a variable, the method includes: setting a relative density iteration formula: Among them, r k is the radius of the lattice structure pillar of the kth iteration, ρ k is the relative density value of the kth iteration, ρ t is the expected value of relative density; is the derivative of the relative density with respect to the pillar radius obtained by approximate calculation using the finite difference method, where where r i For r k Subtract the preset step size, ρ i For r i Corresponding relative density; when the relative density difference between two consecutive iterations is less than the preset relative density difference, the step size is reduced to the preset relative density difference; for the preset density range, the initial pillar radius r0 is based on the empirical formula: r0 = 0.7ρ t +0.1 settings.

[0007] In an optional solution of the first aspect, when generating the geometric model of the lattice structure, the minimum forming diameter of the pillars and the distance between adjacent pillars are set according to the additive manufacturing process characteristics of the lattice structure.

[0008] In an optional solution of the first aspect, in defining the corrected elastic modulus, the method comprises: obtaining the actual strut forming diameter, the surface powder sticking rate and the porosity of the formed lattice structure by CT scanning; obtaining the actual relative density according to the average strut forming diameter; obtaining the actual bearing density according to the measured relative density and the surface powder sticking rate; generating the diameter deviation according to the actual strut diameter and the designed strut diameter of the geometric model and generating the relative density deviation according to the actual bearing density and the designed relative density of the geometric model; Δρ = ρ actual -ρ design = -(αV d + βΔd), wherein ρ actual is the actual bearing density, ρ design is the designed relative density, α and β are material sensitivity coefficients, V d is the defect volume fraction, and Δd is the diameter deviation; and defining the corrected elastic modulus according to the defect volume fraction and the diameter deviation: wherein E corr is the corrected elastic modulus, E design is the uncorrected elastic modulus, and d design is the designed strut diameter.

[0009] In an optional solution of the first aspect, in introducing the density-dependent failure criterion, the density-dependent failure criterion is introduced according to elastic buckling, shear band expansion and rapid collapse: for the lattice structure below a first preset relative density, a Timoshenko beam model is used to simulate buckling-dominant failure; and for the lattice structure above a second preset relative density, a Gurson-Tvergaard-Needleman damage model is used to represent crack expansion.

[0010] In an optional solution of the first aspect, in inversely optimizing the material plasticity parameters, the method comprises: obtaining actual stress-strain curves of the compression test of the lattice structure at different relative densities and calculating actual specific energy absorption data according to the actual stress-strain curves; establishing a nonlinear relationship between the actual specific energy absorption data and the relative density data; obtaining simulation stress-strain curves of the simulation compression test of the geometric model at different relative densities and calculating simulation specific energy absorption data according to the simulation stress-strain curves; taking the square relative error between the actual specific energy absorption data and the simulation specific energy absorption data as an error objective function and iteratively optimizing the material plasticity parameters according to the error objective function by using a preset optimization method.

[0011] In an optional solution of the first aspect, in establishing the defect coupling evaluation model, the method comprises:

[0012] a volume sum, a voxel sum, a surface sum, a defect sum, and a projection area sum of the three-dimensional image data after CT scanning are added, and the volume sum of the surface sticking powder is divided by the volume sum of the lattice structure to obtain a surface sticking powder rate; a forming strut diameter data of the lattice structure is obtained by adding the three-dimensional image data after CT scanning, and a strut diameter deviation is generated according to the forming strut diameter and the designed strut diameter; and a defect coupling evaluation model is established according to the surface sticking powder rate and the strut diameter deviation: wherein Radhesion is the surface sticking powder rate, p is the designed relative density of the lattice structure, Ad is the strut diameter deviation, d is the designed strut diameter of the lattice structure, and y R is a weighting coefficient for adjusting the weight of the surface sticking powder rate in the total score, and y is a weighting coefficient for adjusting the weight of the strut diameter deviation in the total score.

[0013] In an optional solution of the first aspect, when the Gaussian mixture model is constructed, the method comprises: obtaining actual stress-strain curves of the lattice structure under compression tests at different relative densities; extracting the elastic modulus, the platform stress, the peak stress, and the energy absorption rate contained in the actual stress-strain curves and constructing the Gaussian mixture model.

[0014] In an optional solution of the first aspect, when the defect coupling evaluation model and the Gaussian mixture model are used to screen out defects and abnormal data, the method comprises: judging whether the defect score of the defect coupling evaluation model is greater than a preset defect threshold, if yes, determining that the data has geometric defects and removing it, and if not, determining that the data is qualified and retaining it; for each data, judging whether the performance index deviates from a preset range under the Gaussian mixture model, if yes, determining that the data has performance abnormalities and removing it, and if not, determining that the data is qualified and retaining it.

[0015] In a second aspect, the application provides a data correction system for using the binding dot array structure additive manufacturing process feature according to the above method, comprising: a simulation design module for adopting Newton iteration method to iteratively optimize the target relative density with the dot array pillar radius as the variable and combining the preset process constraint to limit the pillar minimum diameter and the pillar spacing, and generating a geometric model of the dot array structure; a modulus correction module for obtaining the actual geometric parameters of the formed dot array structure through CT scanning and defining the corrected elastic modulus according to the defect volume fraction and the pillar diameter deviation between the actual geometric parameters and the geometric model; a material correction module for introducing a density-dependent failure criterion according to the compression failure mechanism and establishing a nonlinear relationship between the specific energy absorption data and the relative density according to the specific energy absorption data, and inversely optimizing the corrected material plasticity parameters; a data screening module for establishing a defect coupling evaluation model for screening defect data according to the surface powder sticking rate and the pillar diameter deviation parameters of the dot array structure, and constructing a Gaussian mixture model for screening abnormal data according to the compression stress-strain curve characteristics; a model construction module for constructing a data correction model for bidirectional feedback of performance prediction and structure design according to the defect coupling evaluation model, the Gaussian mixture model, the corrected elastic modulus, the failure criterion and the corrected material plasticity parameters.

[0016] It should be understood that the general description above and the following detailed description are only exemplary and do not limit the application. BRIEF DESCRIPTION OF DRAWINGS

[0017] The accompanying drawings, which are incorporated herein and form part of the specification, illustrate one or more embodiments of the application and, together with the description, explain the principles of the application and enable a person skilled in the relevant art to make and use the application.

[0018] Figure 1 is a structural schematic diagram of an exemplary typical dot array structure cell according to some embodiments of the application.

[0019] Figure 2 is a schematic diagram of an iterative process of a BCC dot array structure with a relative density of 60% according to some embodiments of the application.

[0020] Figure 3 is a schematic diagram of a three-dimensional reconstruction of the surface powder of BCC dot array structures with different relative densities according to some embodiments of the application; wherein (a) is a 20% relative density sample, (b) is a 40% relative density sample, (c) is a 60% relative density sample, and (d) is an 80% relative density sample.

[0021] Figure 4 is a schematic diagram of the surface powder rate of BCC dot array structures with different relative densities according to some embodiments of the application.

[0022] Figure 5 is a schematic diagram of three-dimensional reconstruction of different relative density BCC lattice structure pillars according to some embodiments of the present application; wherein (a) is a 20% relative density sample, (b) is a 40% relative density sample, (c) is a 60% relative density sample, and (d) is an 80% relative density sample.

[0023] Figure 6 is a schematic diagram of powder sticking on the surface of pillars according to some embodiments of the present application.

[0024] Figure 7 is a schematic diagram of engineering stress-strain curves of four relative densities of BCC structure according to some embodiments of the present application; wherein (a) is a 20% relative density sample, (b) is a 40% relative density sample, (c) is a 60% relative density sample, and (d) is an 80% relative density sample.

[0025] Figure 8 is a schematic diagram of compression stress-strain curves of different relative density samples according to some embodiments of the present application.

[0026] Figure 9 is a schematic diagram of fracture scanning electron microscope (SEM) of different relative density samples according to some embodiments of the present application; wherein (a) is a 20% relative density sample, (b) is a 40% relative density sample, (c) is a 60% relative density sample, and (d) is an 80% relative density sample.

[0027] Figure 10 is a schematic diagram of porosity of different relative density BCC lattice structure according to some embodiments of the present application.

[0028] Figure 11 is a schematic diagram of LOF porosity and spherical pore ratio of different relative density BCC lattice structure according to some embodiments of the present application.

[0029] Figure 12 is a schematic diagram of powder sticking rate on the surface of different relative density BCC lattice structure according to some embodiments of the present application.

[0030] Figure 13 is a schematic diagram of average rod diameter of different relative density BCC lattice structure according to some embodiments of the present application.

[0031] Figure 14 is a schematic diagram of a flow of a data correction method according to some embodiments of the present application.

[0032] Figure 15is an exemplary experimental data and uncorrected simulation result fitting error diagram according to some embodiments of the present application.

[0033] Figure 16 is an exemplary screening experimental abnormal data diagram according to some embodiments of the present application.

[0034] Figure 17 is an exemplary terminal device connection diagram according to some embodiments of the present application. DETAILED DESCRIPTION

[0035] Example implementations will now be described more fully with reference to the accompanying drawings. Example implementations can be implemented in any of various forms, and should not be limited to the specific examples described herein; rather, these described examples are intended to provide specific implementations thereof. Like reference numerals can be used to denote like elements throughout the accompanying drawings. The illustrated implementation can be used to enable a user to interact with a computing device in a variety of ways. The described features, structures, or characteristics can be combined in any suitable manner in one or more implementations. In the following description, numerous specific details are provided to give a thorough understanding of implementations of the application. Implementations can be used in a variety of different applications.

[0036] With the rapid development of aerospace technology, the demand for lightweight structure of aerospace equipment is becoming more and more urgent. Because the structure design is limited by the processing technology, the traditional lightweight design method has reached the limit and it is difficult to meet the weight reduction requirement of aerospace equipment. In the late 20th century, some scholars proposed three-dimensional lattice structure and theoretically proved that it is an effective technical path to greatly reduce the relative density of the structure while ensuring reliable mechanical properties. In the early 21st century, some scholars used methods such as investment casting, melt gas injection, and metal superplastic forming to prepare metal porous structures, but these methods are costly and the shape and size of the formed structure are limited, making it difficult to prepare complex three-dimensional lattice structures.

[0037] In recent years, additive manufacturing technology with free manufacturing capability has gradually matured, making it possible to manufacture complex lightweight structures such as lattice structures. Compared with traditional processing methods, additive manufacturing breaks the original design limitations, enabling the production of complex parts while reducing production cycle and processing costs. Laser powder bed fusion (LPBF) is one of the most promising methods for manufacturing complex metal parts using additive manufacturing technology. The density of parts manufactured by LPBF process can reach 99.9%, and the mechanical properties are closer to those of parts manufactured by traditional methods. The main advantage of LPBF process for lattice structure forming is the release of design freedom for lattice structure, which can design and manufacture lattice structures with multiple functions such as bearing, energy absorption, vibration reduction, and damping. It is an important embodiment of current material-structure-process integrated design and manufacturing. The main principle of LPBF process is to use a high-energy laser beam to scan and melt the metal powder pre-coated on the powder bed according to the predetermined path, realizing the layer-by-layer fusion and accumulation of materials.

[0038] A three-dimensional lattice structure is a topologically ordered three-dimensional structure composed of a plurality of repeating unit cells arranged periodically in space. The high strength, lightweight, and excellent thermal control characteristics of the lattice structure, as well as the maturity of additive manufacturing technology, make it a new idea for lightweight structural design. The combination of three-dimensional lattice structures and additive manufacturing technology has become a new method for high-performance lightweight multifunctional structural design and manufacturing. Many researchers have designed various lattice structure cells such as cubes, octahedrons, honeycombs, and grids, as shown in Figure 1 . Figure 1 The structural schematic diagrams of typical lattice structure cells are shown. These structures exhibit unique mechanical and functional properties when responding to different stress conditions.

[0039] However, in the laser melting process of the LPBF process, the heat transfer from the molten pool to the powder bed is unstable due to factors such as material, laser spot diameter, scanning speed, and laser energy density. There is an error between the molten pool width and the size of the rod, especially near the intersection nodes of the lattice structure. The main performance is that part of the powder is attached around the rod, the node is severely caked, and the forming quality deteriorates as the rod angle decreases.

[0040] In the design and optimization of lattice structures, although there are software that can generate complex geometries and perform topology optimization, most existing simulation software uses idealized simulation models for simulation and does not consider the performance impact of additive manufacturing process characteristics. ANSYS developed a lattice structure-specific auxiliary design software (LatticeSimulation plug-in), but its functionality is limited to regular lattices and is difficult to handle multi-objective optimization under complex loads. nTopology enables rapid modeling and filling of lattice structures, but it still only solves the modeling process, which is a gradual design and testing process from existing structures to performance. It uses a topology optimization algorithm to design a three-dimensional lattice support, achieving a 50% weight reduction for a cube satellite, but lacks systematic correction of additive manufacturing process defects. In addition, these tools lack a feedback mechanism between real processes and performance, often relying on designer experience to iteratively adjust parameters, and lack an automated closed-loop optimization process.

[0041] In the optimization design of three-dimensional lattice structures, relative density is a key design parameter that directly determines the balance between the mechanical properties and lightweight effect of the structure. To achieve precise relative density control, the Newton-Raphson method is used as the core optimization algorithm, which gradually adjusts the lattice strut radius to approach the target density value, thereby constructing a three-dimensional model that meets the design requirements. The Newton-Raphson method has significant advantages in solving nonlinear equations due to its quadratic convergence characteristics, making it particularly suitable for high-degree-of-freedom design problems such as lattice structures.

[0042] Newton iteration is an efficient numerical solution method, which is often used to solve nonlinear equations. For the calculation of relative density, the goal is to find a suitable lattice structure strut radius, so that the relative density of the structure reaches the expected value. The core idea of Newton iteration is to linearize the objective function through Taylor expansion, and gradually approach the root of the equation. For the relative density optimization problem, the objective function can be defined as the difference between the current relative density p and the target value p t : f(r) = p(r) - p t , where the relative density refers to the ratio of the actual material volume occupied by the structure to the total volume, and r is the strut radius of the lattice structure.

[0043] The first-order finite difference method is used for approximate calculation in this application. The step size is selected as Dr = 0.1 mm. The specific calculation steps are as follows: first, set the initial radius r0 = 0.5. According to the initial radius, a three-dimensional model is established, and the relative density (volume fraction) under this radius is calculated through the three-dimensional model volume. According to the difference between the current relative density and the target value, it is judged whether it is within the set tolerance range. If it is, the iteration is stopped, otherwise the new radius is calculated according to the iteration formula, and the above steps are repeated. In this application, the set tolerance is 0.002, that is, when the target relative density is 0.2, the current relative density is within the range of 0.198-0.202, and the iteration can be stopped.

[0044] Therefore, the relative density iteration formula is: where r k is the strut radius of the lattice structure in the kth iteration, p k is the relative density value in the kth iteration, p t is the expected value of the relative density; is the derivative of the relative density with respect to the strut radius calculated by the finite difference method. It is relatively complex to directly calculate, so the finite difference method is used for approximate calculation in this application, that is, the current radius r k is selected, and the preset step size is 0.1, so that r i = r k - 0.1, and the corresponding relative densities are p k and p i , respectively. The derivative approximation formula of r k point is:

[0045] All calculations of relative density in this application are implemented in Python. This iterative method can theoretically satisfy the relative density iteration of all truss-type lattice structures. The effectiveness of this method has been verified in the optimization of BCC (body-centered cubic) and FCC (face-centered cubic) lattice structures. Its average convergence iteration number is 5-8 times, which is significantly better than traditional numerical methods such as the bisection method. In practical applications, the Newton iteration method may diverge due to improper selection of initial values ​​or excessive nonlinearity of the objective function. To this end, this application introduces an adaptive step size adjustment mechanism:

[0046] 1. Convergence criterion: Set tolerance ∈ = 0.002, when ρ k -ρ t The iteration is terminated when ≤∈.

[0047] 2. Dynamic adjustment of step size: If the density difference between two consecutive iterations is less than 0.8, the step size will be reduced to 0.05mm to avoid overshoot.

[0048] 3. For high density interval (ρ t >0.8), the initial radius r0 is from the empirical formula r0=0.7ρ t +0.1 to improve the convergence speed.

[0049] The actual experiments show that the adaptive strategy can reduce the number of iterations in the high-density interval from 12 to 7, while avoiding the oscillation problem in the low-density interval. Figure 2 As shown, Figure 2 The schematic diagram of the iteration process of the BCC lattice structure with 60% relative density in some embodiments of the present application is shown. In addition, to address the multi-cell coupling effect, the topology sensitivity factor (TSF) is introduced to modify the iteration direction so that the density distribution is more in line with local mechanical requirements.

[0050] The geometric model generated by the Newton iteration method must meet the additive manufacturing process constraints. That is, according to the characteristics of the laser selective melting process, the minimum formable diameter of the pillar dmin = 0.3mm, and the distance between adjacent pillars must be greater than 2dmin to avoid thermal accumulation defects. During the iteration process, if the calculated result of the pillar radius exceeds the process window, the constraint adjustment algorithm is automatically triggered to redistribute the density gradient. Through integrated process-design collaborative optimization, the method adopted in this application increases the manufacturability qualification rate to more than 95% while ensuring relative density accuracy.

[0051] Specifically, due to the limitations of the metal lattice structure additive manufacturing powder material characteristics and powder bed fusion additive manufacturing process, there is a geometric deviation between the additive manufacturing metal lattice structure component and the design model, such as surface powder sticking, pillar diameter deviation, actual bearing density deviation and other geometric defects. When the pillar is inclined, the downward surface is formed above the powder, and because the thermal conductivity of the powder is much lower than that of the solid, the material at the suspended position will become overheated, resulting in insufficient melting of the powder below the pillar, and the unmelting powder is seriously adhered. The causes of these deviations can be attributed to the following three aspects: 1) melt pool dynamics and thermal conductivity limitations: during the forming process of the inclined pillar (inclination < 45°), the powder below the suspended area is low in thermal conductivity (the thermal conductivity of the powder is only 5%-10% of that of the dense material), resulting in excessive accumulation of laser energy and increase of the melt pool temperature gradient. At this time, the un-melted powder is adsorbed to the pillar surface by the surface tension of the melt pool, forming a powder sticking defect; step effect: due to the limitation of the layered manufacturing principle, the inclined pillar surface presents a stepped geometric feature, which aggravates the unevenness of local energy input. Actual experiments show that when the inclination angle of the pillar is 30°, the actual diameter deviation can be as high as 18%-22% of the designed value; powder-melt interaction: the random accumulation characteristics of metal powder cause the melt pool boundary to fluctuate, especially when forming small features (rod diameter < 0.5mm), the powder agglomeration effect is significant, causing the rod to be locally too thick or broken.

[0052] In actual forming process, due to the existence of step effect in PBF-L process, some pillars are over-thick or over-thin, the pillar diameter deviates from the designed model, and the relative density of the actual component bearing load deviates from the designed model. These defects cause geometric mismatch between the actual component and the design model, affecting the mechanical properties of the actual formed component.

[0053] Existing researches mainly focus on the influence mechanism of process parameters on the forming metallurgical defects of metal lattice structure, and lack of research on the evolution law of geometric forming quality under different relative densities, which is also valuable for guiding actual engineering application. This application will explain the evolution law of geometric forming quality of lattice structure under different relative densities.

[0054] I. Evolution law of surface powder sticking under different relative densities: in low relative density (ρ < 0.3) lattice structure, the pillar spacing is large and the surface area ratio is high, the sticking probability of un-melted powder increases significantly. In order to determine whether the process parameters selected in the forming experiment are reasonable, the densification of different relative density samples is measured by Archimedes method, and the results are shown in the following table:

[0055] Table: Densification of lattice structure samples with different relative densities prepared by PBF-L process

[0056]

[0057] The relative density of each lattice structure sample is more than 98%, which indicates that the internal forming quality of the sample is stable, and can ensure the uniformity and stability of the sample deformation under stress. The optimal parameters of the BLT-S400 device for forming Ti-6Al-4V solid structure are used for the forming experiment of the lattice structure, which can achieve the same density level as the solid structure (100% relative density in the table is the solid structure). This shows that the process parameters selected for the forming experiment are reasonable, and the density of the lattice structure at different relative densities does not change significantly.

[0058] The three-dimensional reconstruction diagram of the surface powder adhesion of the BCC lattice structure with different relative densities is shown in Figure 3 Figure 3 The three-dimensional reconstruction diagram of the surface powder adhesion of the BCC lattice structure with different relative densities is shown in some embodiments of the application, wherein Figure 3 (a) is a 20% relative density sample, Figure 3 (b) is a 40% relative density sample, Figure 3 (c) is a 60% relative density sample, and Figure 3 (d) is an 80% relative density sample; Figure 3 The color band indicates the volume of the adhered powder. The surface powder is mainly distributed inside the lower surface of the inclined strut, and the lattice structure with a 20% relative density has the most serious surface powder adhesion. The characterization results show that cracks usually occur at the thinnest part of the strut, indicating that the additional material volume caused by surface powder adhesion (mainly caused by Z-direction growth, which increases the strut size in the building direction) does not effectively improve the load sharing capacity of the structure. Therefore, when analyzing the influence of surface powder adhesion, the influence on the geometric volume should be considered, rather than the direct influence on the mechanical properties.

[0059] The three-dimensional image data after CT scanning is processed by using the inclusion analysis function of the VGStudioMAX software. The total volume, total voxel, total surface, total defect, and total projection area are obtained by adding all the test data, and then the total volume of the surface powder is divided by the volume of the lattice structure material to obtain the surface powder rate.

[0060] The surface powder rate of the BCC lattice structure with different densities is shown in Figure 4 Figure 4 The surface powder rate of the BCC lattice structure with different densities is shown in some embodiments of the application. By analyzing the surface powder rate of the BCC lattice structure with different densities, it can be found that the surface powder rate decreases exponentially with the relative density, and the equation obtained by fitting the data is y=3.2716e-4.987x, R 2 is 0.9972.

[0061] ​​The reason for the exponential decrease of the surface powder sticking rate with the relative density is that at a low relative density, the diameter of the struts is small, the internal space of the lattice is large, and the surface of the struts is easy to adsorb and adhere to the unmelted powder particles, so the surface powder sticking rate is high. With the increase of the relative density, the diameter of the lattice structure struts increases, the overall surface area decreases, and the adhesion of the unmelted powder decreases significantly, resulting in a rapid decrease of the surface powder sticking rate.

[0062] II. Evolution law of strut diameter at different relative densities: The three-dimensional reconstruction diagram of the strut diameter of the BCC lattice structure at different relative densities is shown in FIG. 3. Figure 5 Figure 5 FIG. 3 shows the three-dimensional reconstruction diagram of the strut diameter of the BCC lattice structure at different relative densities according to some embodiments of the present application, wherein Figure 5 (a) of FIG. 3 is a 20% relative density sample, Figure 5 (b) of FIG. 3 is a 40% relative density sample, Figure 5 (c) of FIG. 3 is a 60% relative density sample, and Figure 5 (d) of FIG. 3 is an 80% relative density sample; Figure 5 The color band in the figure represents the strut diameter.

[0063] In the process of forming the lattice structure by additive manufacturing, on the one hand, it is limited by the process conditions of the existing PBF-L forming equipment, and on the other hand, due to the complex internal space of the lattice structure, the finally formed lattice component will deviate from the original design model in the strut diameter.

[0064] The three-dimensional image data after CT scanning is processed by using the wall thickness analysis function of the VGStudioMAX software. The minimum strut diameter, the maximum strut diameter, the average strut diameter and the standard deviation are obtained by adding all the test data. The average strut diameter of the BCC lattice structure at different densities is shown in the following table:

[0065] Table: Formed diameter and designed diameter of the strut of the BCC lattice structure at different relative densities

[0066]

[0067] By analyzing the average strut diameter of the BCC lattice structure at different relative densities, it can be found that the strut diameter increases linearly with the relative density, and the deviation between the formed diameter and the designed diameter decreases with the decrease of the relative density. This is because the volume of the strut is in a quadratic relationship with the diameter, while the surface area is in a linear relationship with the diameter. The unmelted powder adheres to the surface of the strut, and as the diameter of the strut increases, the ratio of the volume of the unmelted powder on the strut to the volume of the strut becomes smaller and smaller, so the deviation between the formed diameter and the designed diameter becomes smaller and smaller. These formed diameter values provide data support for improving the accuracy of subsequent lattice structure simulation.

[0068] ​III. Evolution of actual bearing density under different relative densities:

[0069] Actual bearing density refers to the proportion of the material actually participating in bearing external load in the overall structure in the point array structure, reflecting the proportion of the effective bearing part in the structure. In order to analyze the main reasons for the change of the overall pillar diameter and its influence, and provide accurate data guidance for subsequent finite element modeling, this section selects local pillars with obvious diameter changes for analysis and comparison.

[0070] For BCC structures of different densities, through analysis and comparison of the selected local pillar CT images, it is found that there are two reasons for the change of the pillar diameter: one is the change of the pillar diameter caused by process conditions such as laser spot size and powder particle size; the second is that in the forming process, the powder near the laser scanning area is partially melted, and the surface powder is adsorbed on the surface of the pillar to form a surface powder sticking diagram, as shown in Figure 6 , Figure 6 shows a schematic diagram of the surface powder sticking rate of the BCC point array structure of different relative densities according to some embodiments of the application. Since the surface powder is not completely melted during the forming process, and there is no effective contact between the unmelted powder, the mechanical performance contribution under actual working conditions can be ignored.

[0071] The measured relative density is calculated and analyzed based on the average pillar diameter results, and the actual bearing density is obtained through the measured relative density-surface powder sticking rate. The actual bearing density of BCC point array structures of different relative densities is shown in the following table:

[0072] Table: Actual bearing density values of point array structures of different relative densities

[0073]

[0074] By analyzing the actual bearing density of BCC point array structures of different densities, it can be found that the deviation degree of the actual bearing density from the designed relative density decreases with the increase of the relative density. The deviation of the actual bearing density (ρ actual ) from the designed value (ρ design ) is caused by the cumulative effect of geometric defects. Through CT scanning and mechanical test correlation analysis, it is found that:

[0075] Δρ=ρ actual -ρ design =-(αV d +βΔd), where ρ actual is the actual bearing density, ρ design is the designed relative density, α and β are the material sensitivity coefficients, V d is the defect volume fraction, and Δd is the diameter deviation; where α is 0.15 and β is 0.08.

[0076] For example, the sample with p = 0.2 has a Δρ of 26.5%, mainly due to the excessive invalid volume ratio caused by powder sticking. The actual bearing density value can provide a reference for the optimization of the three-dimensional model used for simulation to improve the simulation accuracy.

[0077] Specifically, in the study of the mechanical properties of lattice structures, relative density, as an important design parameter, directly affects the compression performance in practical applications. Lattice structures exhibit mechanical properties that are quite different from traditional solid materials due to their unique geometric morphology and internal void structure. For example, solid materials only have fixed mechanical response characteristics, while honeycomb lattice structures can provide more obvious energy absorption and dispersion effects than other structures, providing good cushioning performance; octahedral lattice structures exhibit more excellent stiffness and stability through the geometric stability of triangular cells. At the same time, by adjusting the relative density of the lattice structure, its mechanical behavior can be significantly changed, thereby controlling its performance in various engineering applications, such as lightweight structures, high-energy absorption materials, and high-strength supports. The lower the relative density, the higher the void ratio in the lattice structure, and the overall stiffness and compressive strength generally decrease, but the energy absorption characteristics and lightweight effect will be improved. Therefore, systematically studying the compression performance characteristics at different relative densities is crucial for optimizing the design of lattice structures and improving their application effects in practical engineering.

[0078] To further explore the specific influence of relative density on the compression performance of lattice structures, the embodiments conduct compression experiments on four different relative density lattice structures and solid materials (reference, i.e., 100% relative density) to analyze the stress-strain curve characteristics. This part of the research can help reveal the compression performance evolution law of BCC lattice structures at different relative densities. For the design of lightweight structural materials, this analysis can clearly define the relationship between the strength and mass of the lattice structure, thereby providing theoretical support for the aerospace, automotive, and other industries.

[0079] Reference Figure 7 As shown, Figure 7 The compression stress-strain curves of the BCC structure with 20%, 40%, 60%, and 80% relative density of one embodiment of the present application are shown, wherein, Figure 7 (a) of FIG. 1 is a 20% relative density sample, Figure 7 (b) of FIG. 1 is a 40% relative density sample, Figure 7 (c) of FIG. 1 is a 60% relative density sample, and Figure 7(d) is 80% relative density specimen. Each relative density is repeated three times. It can be seen from the figure that the compression test has good stability and repeatability, and the stress-strain curve shows that the BCC lattice structure of the four relative densities has obvious elastic stage, yield stage, hardening stage and fracture stage. When the relative density is low, the stress-strain curve of the lattice structure during the compression process usually presents a relatively wide plastic collapse platform. Due to the thin cell wall and the easy bending or local buckling, the structure will not collapse completely immediately after entering the yield stage, but gradually form multiple local failure zones to release energy in a long strain interval. Macroscopically, the curve maintains a relatively flat “platform section” at a low stress level, and gradually enters the densification stage and significantly increases the stress as the strain increases.

[0080] When the relative density is at a medium level, the wall thickness and the cell geometry stiffness are both improved. The initial elastic stage slope increases significantly, and the stress level of the collapse platform also increases accordingly. At this time, the lattice can withstand higher load before starting to fail, but once the local area reaches the critical stress, the collapse process will spread to the surrounding cells more quickly than at low relative density. The stress-strain curve often shows a certain amplitude of decline after reaching the peak, and then a certain range of collapse platform can still be seen, and the strain interval of the platform is usually shortened, and the steep drop of the curve is more obvious.

[0081] When the relative density further increases to a high level, the wall thickness and the support capacity are greatly enhanced, and the structure can accumulate higher stress in the elastic stage. However, with the increase of carrying capacity, once the crack or failed cell is generated, the surrounding area will quickly take on the load, leading to more severe chain collapse. At this time, the stress-strain curve can be observed to have a prominent peak and a more sudden stress drop. The collapse stage is relatively shorter than that at medium and low relative densities, and the whole presents the characteristics of rapid destruction. Although it has strong resistance to local damage before the crack is generated, once the failure threshold is crossed, the whole structure collapses rapidly and the structure fails.

[0082] Reference Figure 8 As shown in the figure, Figure 8 The stress-strain curves of the 20%, 40%, 60%, 80% relative density and the same size solid structure of an embodiment of the application are shown. From the initial elastic stage, as the relative density gradually increases from 20% to 100%, the curve slope (i.e. the overall stiffness of the material in the elastic stage) increases significantly. The elastic stage stress-strain curve slope of the lattice structure with the lowest relative density is the smallest, indicating that the stiffness is low in a small strain interval; while the overall stiffness of the structure with a relative density of 80% has been significantly improved, and the initial elastic slope is much larger than that of the low density group. The curve of the solid structure is the highest and steepest, indicating that the dense material has the strongest resistance to deformation in the elastic deformation stage.

[0083] From the yield and plastic deformation stage, with the increase of relative density, the average load that the material can withstand before reaching the peak stress is significantly improved. The lattice structure with 20% and 40% relative density enters a relatively flat plastic collapse platform at a small strain, indicating that the low relative density structure is prone to bending or buckling, and after local failure begins, it does not immediately cause overall collapse, but gradually expands to form a longer energy absorption interval. Although the lattice structure with 60% and 80% relative density also shows a certain degree of plastic collapse, the stress level corresponding to the stress platform is already quite high, and the platform interval tends to be shortened, indicating that the cell wall strength and local stability are enhanced, and the failure process is more concentrated and tends to appear rapid stress drop.

[0084] From the maximum bearing capacity, the lattice structures with different relative densities each show a peak stress on the stress-strain curve. With the increase of relative density, this peak stress rapidly increases, which reflects that the lattice is closer to the strength level of solid materials in the high density state. The peak stress of the lattice structure with 80% relative density is already very considerable, and although it still lags behind the solid sample, the improvement amplitude relative to the low density group is extremely significant. The solid material as a dense structure has the longest linear interval before plastic deformation and the largest slope, and after entering the plastic stage, it also maintains a high stress level until a large strain appears flow deformation or local damage, so its overall strength and stiffness exceed the lattice structures with various relative densities.

[0085] Reference Figure 9 As shown in the drawings, Figure 9 The SEM scanning images of the compression fracture of the four relative density samples of an embodiment of the present application are shown, wherein, Figure 9 (a) of FIG. 1 is a 20% relative density sample, Figure 9 (b) of FIG. 1 is a 40% relative density sample, Figure 9 (c) of FIG. 1 is a 60% relative density sample, and Figure 9 (d) of FIG. 1 is an 80% relative density sample. It can be seen that the four relative density samples are all ductile fractures, mainly showing that there are a large number of dimples in the fracture, showing the typical micropore aggregation fracture morphology characteristics. With the increase of relative density, the dimples gradually deepen, which indicates that the material absorbs a high plastic deformation energy during the fracture process.

[0086] In terms of failure mode or collapse characteristics, the low relative density lattice structure exhibits a longer collapse platform, meaning a slower and more dispersed cell wall failure characteristic at lower strength levels. As the relative density increases, the lattice is more likely to collapse rapidly after reaching peak strength, with a steeper decline in the curve, reflecting a kind of "macroscopic brittleness" characteristic: the thicker the supporting wall, the higher the load it can resist in the early stage, but once failure occurs, the load redistribution speed of the adjacent area is fast, and the overall structure tends to be quickly destroyed. Solid materials have the strongest integrity before failure, but when yield or crack propagation occurs, they also show a rapid decline in strength.

[0087] The current ICT test results show that the pores randomly distributed in the lattice structure test cells mainly include LOF pores and spherical pores. The presence of pores can lead to a decrease in the cooling rate of the area around the defect, an interruption in heat distribution, and an increase in dislocation density. Smaller β grains and thicker α lathes can be formed near the pores. Irregularly shaped and slender LOF pores can induce severe strain accumulation near sharp edges, ultimately leading to poor tensile mechanical properties; the strain accumulation caused by spherical pores can be negligible, but the oxidation often found in pores can lead to oxide-induced microcracks. However, the relatively small strain accumulation of the pores can prevent these induced microcracks from further propagating from the brittle oxides during the tensile loading process. This application explores the defect distribution obtained from non-destructive testing and its distribution mechanism.

[0088] Pore rate of BCC lattice structure of different densities Figure 10 As shown, Figure 10 The schematic diagram of the pore rate of the BCC lattice structure of different relative densities of some embodiments of the application is shown. By analyzing the pore rate of the BCC lattice structure of different densities, it can be found that the pore rate decreases according to the binomial law with the relative density, and the equation obtained by fitting the data is shown as y = B1x 2 +B2x+C, where x is the relative density, y is the pore rate, B1, B2 and C are coefficients obtained by fitting experiments; the values of each term are shown in the following table:

[0089] Table: Pore rate fitting polynomial coefficient values

[0090]

[0091] The residual sum of squares is 0.00238, and R 2 = 0.98378.

[0092] Therefore, the present application regards the irregular pores with sphericity less than 0.6 as LOF pores, and regards the spherical pores with sphericity greater than 0.6 as spherical pores. According to the above standard, the pore data is divided, the sum of the LOF pore volume and the spherical pore volume is divided by the volume of the lattice structure material to obtain the LOF porosity and the spherical porosity, and the LOF porosity and the spherical porosity of the BCC lattice structure with different densities are shown in the following table. Figure 11 Figure 11 The schematic diagram of the LOF porosity and the spherical porosity of the BCC lattice structure with different relative densities is shown.

[0093] Through the analysis of the LOF porosity and the spherical porosity of the BCC lattice structure with different densities, it can be found that the LOF porosity and the spherical porosity and the relative density all decrease according to the binomial law. The equation fitted by the LOF pore data is shown by the formula y=B1x 2 +B2x+C, and the values of each term are shown in the following table.

[0094] Table: LOF pore fitting polynomial coefficient values

[0095]

[0096] The residual sum of squares is 0.00172, R 2 =0.974, and it can be found that the LOF pores are more inclined to exponential change. The equation fitted by the spherical pore data is shown by the formula y=B1x 2 +B2x+C, and the values of each term are shown in the following table.

[0097] Table: spherical pore fitting polynomial coefficient values

[0098]

[0099] The residual sum of squares is 1.50413E-4, R 2 =0.99081. The present application compares the LOF porosity and the spherical porosity under the same density, and finds that the LOF porosity decreases faster than the spherical porosity with the increase of the relative density.

[0100] In summary, it can be found that the porosity of the BCC lattice structure with 20% relative density is the highest, and the spherical pores are dominant, and the porosity of the BCC lattice structure with 80% relative density is the lowest, and the LOF pores are dominant.

[0101] ​After comparing the parameter settings in the forming test, the present application considers that the rod of the BCC structure is smaller in diameter at low relative density. Under the parameter setting of the laser power of the LPBF process, the molten pool size of the thin rod is small and unstable during melting, which easily leads to incomplete melting of the powder or gas retention, forming a ball hole. At the same time, due to the larger specific surface area of the thin rod, the cooling rate is fast, and the gas in the thin rod cannot escape in time, which will further aggravate the formation of the ball hole. On the contrary, the rod of the BCC structure is thicker or more dense at high relative density. Higher energy input and more complex scanning strategy are required during manufacturing. If the interlayer fusion parameters, such as laser power and scanning speed, are not matched, it is more likely to cause LOF porosity. This phenomenon reflects the coupling effect of process parameters and structure design in additive manufacturing: thin rods tend to have gas hole defects, and thick rods are more susceptible to fusion problems.

[0102] The experimental results show that cracks usually occur at the thinnest part of the strut, indicating that the additional volume of material caused by surface powder sticking increases the strut size in the building direction, but does not effectively contribute to the load sharing capacity of the structure, so when analyzing the influence of surface powder sticking, mainly consider its influence on geometric volume rather than its influence on mechanical properties. The surface powder sticking rate of BCC lattice structures with different densities is shown in the following table. Figure 12 Figure 12 The schematic diagram of the surface powder sticking rate of BCC lattice structures with different relative densities according to some embodiments of the present application is shown.

[0103] The present application analyzes the surface powder sticking rate of BCC lattice structures with different densities and finds that the surface powder sticking rate decreases exponentially with the relative density. The equation obtained by fitting the data is consistent with the formula: y = a - b x C x As shown in the following table, the values of each item are as follows:

[0104] Table: Fitting exponential term coefficient values of surface powder sticking rate

[0105]

[0106] After comparing the surface powder ICT detection results and the parameter settings of the air inlet speed during the forming process, the present application considers that the surface powder is the powder that is not completely removed and is attached to the surface or pores of the BCC lattice structure due to mechanical jamming or electrostatic adsorption, and the severity mainly depends on three aspects: pore size, structure surface area and powder removal efficiency. The pores of the BCC lattice structure with low relative density are more, and the powder is free to enter at the same time, and it is difficult to be completely removed by the airflow in the forming chamber. At the same time, more ball holes are easy to form "traps", and the powder is jammed at the intersection of the rod due to gravity and friction. The pores of the BCC lattice structure with high relative density are significantly reduced, and the fusion channel overlap is more obvious due to the existence of large area scanning, and the rod structure surface is smoother, the powder adhesion decreases, and at the same time, due to the uniformity of the pore channel, the local flow rate increases under the Bernoulli effect, and the powder is more easily blown away.​

[0107] In the additive manufacturing lattice structure forming process, on the one hand, the shrinkage of the material itself during the forming process leads to the shrinkage of the rod diameter due to the process condition limitation of the existing LPBF forming equipment, and on the other hand, due to the complex internal space of the lattice structure, it is easy to produce pores due to powder LOF and spherical holes, which affects the nondestructive testing results, and finally the formed lattice component will deviate from the theoretical model in the rod diameter. The average rod diameter of the BCC lattice structure with different densities is shown in Table 1. Figure 13 Figure 13 The schematic diagram of the average rod diameter of the BCC lattice structure with different relative densities is shown. By analyzing the average rod diameter of the BCC lattice structure with different densities, it can be found that the average rod diameter and the relative density increase according to the binomial law, and the equation obtained by fitting the data satisfies the formula y = B1x 2 +B2x+C. The values of each term are shown in the following table:

[0108] Table: Fitting polynomial coefficient values of rod diameter

[0109]

[0110] The residual sum of squares is 4.7045E-4, and R 2 = 0.99967.

[0111] In summary, the present application believes that as the relative density increases, the deviation of the rod diameter gradually decreases, which is determined by the coupling effect of pore distribution and surface powder adhesion. The deviation of the rod diameter is greatly affected by the surface powder adhesion, and there is also an error in the printing material shrinkage rate in the additive manufacturing process. The deviation of the rod diameter measured by ICT and the theoretical model decreases with the increase of the relative density.

[0112] Therefore, in some embodiments of the present application, the data correction method is shown in Figure 14 Figure 14 The flowchart of the data correction method of some embodiments of the present application is shown. In order to solve the above problems, the present application designs a data correction method combined with the process characteristics of the lattice structure additive manufacturing, which comprises the following steps:

[0113] S1: Using Newton iteration method to iteratively optimize the target relative density with lattice strut radius as variable and combining with the preset process constraint to limit the minimum diameter of the strut and the distance between the struts, a geometric model of the lattice structure is generated.

[0114] S2: The actual geometric parameters of the formed lattice structure are obtained by CT scanning, and the correction elastic modulus is defined according to the defect volume fraction and the deviation of the strut diameter between the actual geometric parameters and the geometric model.

[0115] ​​Wherein, when defining the modified elastic modulus, the method includes:

[0116] The actual pillar diameter, surface powder adhesion rate and porosity of the formed lattice structure are obtained through CT scanning;

[0117] The actual relative density is obtained based on the average pillar forming diameter;

[0118] The actual bearing density is obtained based on the measured relative density and surface powder adhesion rate:

[0119] Generate a diameter deviation based on the actual pillar diameter and the designed pillar diameter of the geometric model, and generate a relative density deviation based on the actual load-bearing density and the designed relative density of the geometric model:

[0120] Δρ=ρ actual -ρ design =-(αV d +βΔd), where ρ actual is the actual bearing density, ρ design is the design relative density, α and β are the material sensitivity coefficients, V d is the defect volume fraction, Δd is the diameter deviation;

[0121] According to the defect volume fraction and diameter deviation, the modified elastic modulus is defined as:

[0122] Among them E corr is the modified elastic modulus, E design is the uncorrected elastic modulus, d design is the design pillar diameter.

[0123] S3: introducing a density-dependent failure criterion according to the compression failure mechanism and establishing a nonlinear relationship between the specific energy absorption data and the relative density according to the specific energy absorption data, and inversely optimizing and correcting the material plasticity parameters.

[0124] Among them, according to the compression failure mechanism (elastic buckling, shear band extension, and rapid collapse), a density-dependent failure criterion is introduced: for low-density lattice structures (ρ<0.3), the Timoshenko beam model is used to simulate buckling-dominated failure; for high-density lattice structures (ρ>0.6), the GTN (GursonTvergaard-Needleman) damage model is embedded to characterize crack propagation.

[0125] Wherein, when inverse optimization is performed to correct the plastic parameters of the material, the method includes:

[0126] The actual stress-strain curves of the compression test of the lattice structure at different relative densities are obtained and the actual specific energy absorption data are calculated based on the actual stress-strain curves.

[0127] establishing a nonlinear relationship of the actual specific energy absorption data and the relative density data: SEA = 25.4p 1.8 (J / g).

[0128] obtaining a simulation stress-strain curve of the geometric model under different relative densities in a simulated compression test and calculating simulation specific energy absorption data according to the simulation stress-strain curve.

[0129] taking the square relative error between the actual specific energy absorption data and the simulation specific energy absorption data as an error objective function and iteratively optimizing the material plasticity parameters according to the error objective function by using a preset optimization method.

[0130] The preset optimization method includes gradient-based methods such as L-BFGS-B, Newton method, genetic algorithm (GA), particle swarm optimization (PSO), Bayesian optimization, etc.

[0131] S4: According to the surface powder adhesion rate and the strut diameter deviation parameters of the lattice structure, a defect coupling evaluation model for screening out defect data is established, and a Gaussian mixture model for screening out abnormal data is constructed according to the compression stress-strain curve characteristics.

[0132] The method comprises the following steps when establishing the defect coupling evaluation model:

[0133] The three-dimensional image data after CT scanning is added to obtain volume sum, voxel sum, surface sum, defect sum, and projection area sum data, and the volume sum of the surface powder is divided by the total volume of the lattice structure to obtain the surface powder adhesion rate.

[0134] The three-dimensional image data after CT scanning is added to obtain the forming strut diameter data of the lattice structure, and the forming strut diameter and the designed strut diameter are used to generate the strut diameter deviation.

[0135] The defect coupling evaluation model is established according to the surface powder adhesion rate and the strut diameter deviation:

[0136] Wherein Radhesion is the surface powder adhesion rate, p is the designed relative density of the lattice structure, Ad is the strut diameter deviation, d is the designed strut diameter of the lattice structure, γ R is a weighting coefficient for adjusting the weight of the surface powder adhesion rate in the total score, and γ is a weighting coefficient for adjusting the weight of the strut diameter deviation in the total score.

[0137] For example, the defect coupling evaluation model is established as follows:

[0138] The method comprises the following steps when constructing the Gaussian mixture model:

[0139] Obtaining the actual stress-strain curves of the lattice structure under compression test at different relative densities.

[0140] Extracting the elastic modulus, plateau stress, peak stress and energy absorption rate contained in the actual stress-strain curve and constructing a Gaussian mixture model.

[0141] Among them, in the defect coupling evaluation model and the Gaussian mixture model, when the defects and abnormal data are screened out, the method comprises:

[0142] Determine whether the defect score of the defect coupling evaluation model is greater than the preset defect threshold (for example, 1.0). If yes, it is determined that the data has geometric defects and is rejected. If not, it is determined that the data is qualified and is retained.

[0143] For each data, determine whether its performance index deviates from the preset range under the Gaussian mixture model. If yes, it is determined that the data has performance abnormalities and is rejected. If not, it is determined that the data is qualified and is retained.

[0144] For example, the energy absorption efficiency (EA) of a certain ρ = 0.4 sample is 68%, which is significantly lower than the mean value of 75% ± 5% of the same density group, and is marked as abnormal and rejected.

[0145] S5: Construct a data correction model for bidirectional feedback of performance prediction and structure design according to the defect coupling evaluation model, the Gaussian mixture model, the modified elastic modulus, the failure criterion and the modified material plasticity parameters.

[0146] Among them, according to the ISO / ASTM52902:2019 standard, set threshold values such as laser power fluctuation tolerance (±5%), scanning speed deviation (±3%), etc. Automatically mark the experimental batches that exceed the range. For example, the compression strength of a certain group of ρ = 0.6 samples deviates from the mean value by 3σ, and tracing back finds that its laser power fluctuation reaches 7%, triggering the data rejection mechanism. Combined with the above iterative optimization, the optimization failure cases that do not meet the tolerance are screened out to avoid invalid design parameters entering the database.

[0147] In summary, through comprehensive statistics of the forming and actual mechanical properties of the lattice structure, data support is provided for the subsequent lattice structure-mechanical property intelligent feedback system. In the mechanical property and geometric parameter part, the mechanical property is normalized; for defects and organizational characteristics, they are summarized according to their distribution characteristics in different regions of the lattice structure. The team designs and develops a metal lattice structure nondestructive testing and mechanical property evaluation database to realize fast query of simulation calculation and performance detection data, and through the establishment of a correction function, the data is screened to obtain effective and accurate experimental data, such as Figure 15It is shown that the mechanical property data obtained after the simulation model is corrected are more consistent with the actual performance, such as Figure 16 It is shown that the outliers of the experiment are proposed after screening. High-quality data basis is provided for the two-way feedback of performance prediction-structure design.

[0148] Therefore, in some embodiments of the present application, to solve the above problems, the present application also designs a data correction system using the binding point lattice structure additive manufacturing process characteristics according to the above method, comprising: a simulation design module for adopting Newton iteration method to iteratively optimize the target relative density with the lattice strut radius as the variable and combining the preset process constraint to limit the minimum diameter of the strut and the strut spacing, to generate a geometric model of the lattice structure; a modulus correction module for obtaining the actual geometric parameters of the formed lattice structure by CT scanning and defining the corrected elastic modulus according to the defect volume fraction and strut diameter deviation between the actual geometric parameters and the geometric model; a material correction module for introducing a density-dependent failure criterion according to the compression failure mechanism and establishing a nonlinear relationship between the specific energy absorption data and the relative density according to the specific energy absorption data to inversely optimize the corrected material plasticity parameters; a data screening module for establishing a defect coupling evaluation model for screening defect data according to the surface powder adhesion rate and strut diameter deviation parameters of the lattice structure and constructing a Gaussian mixture model for screening abnormal data according to the compression stress-strain curve characteristics; a model construction module for constructing a data correction model for performance prediction and structure design two-way feedback according to the defect coupling evaluation model, the Gaussian mixture model, the corrected elastic modulus, the failure criterion and the corrected material plasticity parameters.

[0149] In some embodiments, referring to Figure 17 It is shown that Figure 17 A connection diagram of a terminal device for implementing the embodiments of the present application is shown. The terminal device 3 includes a memory 301 and a processor 302, and the memory 301 stores a computer program executable on the processor 302. The processor 302 implements the method in the above embodiments when executing the computer program. The number of the memory 301 and the processor 302 can be one or more.

[0150] The terminal device 3 further includes:

[0151] A communication interface 303 for communicating with external devices and performing data interaction transmission.

[0152] If the memory 301, the processor 302 and the communication interface 303 are independently implemented, the memory 301, the processor 302 and the communication interface 303 can be connected to each other through a bus and complete communication among them.

[0153] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 17 Only one thick line is used in the figure to represent the bus, but it does not mean that there is only one bus or only one type of bus.

[0154] Optionally, in a specific implementation, if the memory 301, the processor 302, and the communication interface 303 are integrated on a chip, the memory 301, the processor 302, and the communication interface 303 can complete communication with each other through an internal interface.

[0155] The embodiment of the present application provides a computer readable storage medium, which stores a computer program, and the program is executed by the processor 302 to implement the method provided in the embodiment of the present application.

[0156] The embodiment of the present application further provides a chip, which includes a processor 302, and the processor 302 is used to call and run instructions stored in the memory 301, so that a communication device installed with the chip executes the method provided in the embodiment of the present application.

[0157] The embodiment of the present application further provides a chip, which includes an input interface, an output interface, a processor 302, and a memory 301, and the input interface, the output interface, the processor 302, and the memory 301 are connected through an internal connection path. The processor 302 is used to execute code in the memory 301, and when the code is executed, the processor 302 is used to execute the method provided in the embodiment of the present application.

[0158] It is to be understood that the above-described processor 302 can be a central processing unit (CPU), but can also be other general purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs) or other programmable logic devices, discrete gates or transistor logic components, discrete hardware components, or the like. The general purpose processor can be a microprocessor or any conventional processor, or the like. It is to be noted that the processor 302 can be a processor supporting an advanced RISC machine (ARM) architecture.

[0159] Further, the above-described memory 301 can include read-only memory and random access memory, and can also include non-volatile random access memory. The memory 301 can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories. Among them, the non-volatile memory can include read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM) or flash memory. The volatile memory can include random access memory (RAM) used as an external cache. By way of example but not limitation, many forms of RAM are available. For example, static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM) and direct memory bus random access memory (Direct Rambus RAM, DRRAM).

[0160] The above merely provides the specific implementation of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of the changes or replacements within the technical range disclosed by the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A data correction method combining the characteristics of lattice structure additive manufacturing process, characterized in that: The method comprises: The Newton iteration method is used to iteratively optimize the target relative density with the lattice pillar radius as the variable. The minimum pillar diameter and pillar spacing are limited by the preset process constraints to generate the geometric model of the lattice structure. obtaining actual geometric parameters of the formed lattice structure through CT scanning and defining a modified elastic modulus based on the deviation of the defect volume fraction and the strut diameter between the actual geometric parameters and the geometric model; A density-dependent failure criterion is introduced according to the compression failure mechanism, and a nonlinear relationship between the specific energy absorption data and the relative density is established according to the specific energy absorption data, and the plastic parameters of the material are modified by inverse optimization; Based on the surface powder adhesion rate of the lattice structure and the pillar diameter deviation parameters, a defect coupling evaluation model for screening defect data was established, and a Gaussian mixture model for screening abnormal data was constructed based on the characteristics of the compressive stress-strain curve. A data correction model for bidirectional feedback of performance prediction and structural design is constructed based on the defect coupling evaluation model, Gaussian mixture model, modified elastic modulus, failure criterion and modified material plasticity parameter.

2. The method according to claim 1, characterized in that When the Newton iteration method is used to perform relative density iteration with the lattice pillar radius as a variable, the method includes: Set the relative density iteration formula: Among them, r k is the radius of the lattice structure pillar of the kth iteration, ρ k is the relative density value of the kth iteration, ρ t is the expected value of relative density; is the derivative of the relative density with respect to the pillar radius obtained by approximate calculation using the finite difference method, where where r i For r k Subtract the preset step size, ρ i For r i The corresponding relative density; When the relative density difference between two consecutive iterations is less than the preset relative density difference, the step size is reduced to the preset relative density difference; For the preset density range, the initial pillar radius r0 is based on the empirical formula: r0 = 0.7ρ t +0.1 settings.

3. The method according to claim 1 or 2, characterized in that When generating the geometric model of the lattice structure, the minimum forming diameter of the pillars and the distance between adjacent pillars are set according to the additive manufacturing process characteristics of the lattice structure.

4. The method according to claim 3, characterized in that In defining the modified elastic modulus, the method includes: The actual pillar diameter, surface powder adhesion rate and porosity of the formed lattice structure are obtained through CT scanning; The actual relative density is obtained based on the average pillar forming diameter; The actual bearing density is obtained based on the measured relative density and surface powder adhesion rate: Generate a diameter deviation based on the actual pillar diameter and the designed pillar diameter of the geometric model, and generate a relative density deviation based on the actual load-bearing density and the designed relative density of the geometric model: Δρ=ρ actual -ρ design =-(αV d +βΔd), where ρ actual is the actual bearing density, ρ design is the design relative density, α and β are the material sensitivity coefficients, V d is the defect volume fraction, Δd is the diameter deviation; According to the defect volume fraction and diameter deviation, the modified elastic modulus is defined as: Among them E corr is the modified elastic modulus, E design is the uncorrected elastic modulus, d design is the design pillar diameter.

5. The method according to claim 4, characterized in that When introducing the density-dependent failure criterion, based on elastic buckling, shear band expansion, and rapid collapse, the density-dependent failure criterion is introduced: For lattice structures with relative density lower than the first preset value, the Timoshenko beam model is used to simulate buckling-dominated failure; For lattice structures with a relative density higher than the second preset value, the GTN damage model is used to characterize crack propagation.

6. The method according to claim 5, characterized in that When inverse optimization is performed to correct the plastic parameters of the material, the method includes: Obtain the actual stress-strain curves of the lattice structure under compression test at different relative densities and calculate the actual specific energy absorption data based on the actual stress-strain curves; Establishing a nonlinear relationship between the actual specific energy absorption data and the relative density data; Obtaining simulated stress-strain curves of the geometric model under different relative densities for simulating compression tests and calculating simulated specific energy absorption data based on the simulated stress-strain curves; The square relative error between the actual specific energy absorption data and the simulated specific energy absorption data is used as an error objective function and a preset optimization method is used to iteratively optimize the material plasticity parameters according to the error objective function.

7. The method according to claim 4 or 6, characterized in that When establishing the defect coupling assessment model, the method includes: The three-dimensional image data after CT scanning are summed to obtain the volume sum, voxel sum, surface sum, defect sum, and projection area sum data, and the surface powder adhesion rate is obtained by dividing the surface powder adhesion volume sum by the lattice structure volume sum. The three-dimensional image data after CT scanning is summed to obtain the diameter data of the formed struts of the lattice structure and the strut diameter deviation is generated according to the formed strut diameter and the designed strut diameter; A defect coupling evaluation model is established based on the surface powder adhesion rate and pillar diameter deviation: Where Radhesion is the surface powder adhesion rate, ρ is the design relative density of the lattice structure, Δd is the pillar diameter deviation, d is the design pillar diameter of the lattice structure, γ R is the weighting coefficient used to adjust the weight of the surface powder adhesion rate in the total score, and γ is the weighting coefficient used to adjust the weight of the pillar diameter deviation in the total score.

8. The method according to claim 7, characterized in that When constructing a Gaussian mixture model, the method includes: Obtain the actual stress-strain curves of the compression test of the lattice structure at different relative densities; The elastic modulus, platform stress, peak stress and energy absorption rate contained in the actual stress-strain curve are extracted and a Gaussian mixture model is constructed.

9. The method according to claim 8, characterized in that When screening out defects and abnormal data from the multi-source data set according to the defect coupling assessment model and the Gaussian mixture model, the method includes: Determine whether the defect score of the defect coupling assessment model is greater than a preset defect threshold. If so, the data is determined to have geometric defects and is discarded. If not, the data is determined to be qualified and is retained. For each data, determine whether its performance index deviates from the preset range under the Gaussian mixture model. If so, the data is determined to have performance anomalies and is eliminated. If not, the data is determined to be qualified and is retained.

10. A data correction system using the method according to any one of claims 1 to 9 in combination with lattice structure additive manufacturing process features, characterized in that: include: A simulation design module is used to iteratively optimize the target relative density using the Newton iteration method with the lattice pillar radius as a variable, and to generate a geometric model of the lattice structure by limiting the minimum pillar diameter and pillar spacing in combination with preset process constraints; a modulus correction module, configured to obtain actual geometric parameters of the formed lattice structure through CT scanning and define a corrected elastic modulus based on the deviation of the defect volume fraction and the strut diameter between the actual geometric parameters and the geometric model; A material correction module is used to introduce a density-dependent failure criterion based on the compression failure mechanism and establish a nonlinear relationship between the specific energy absorption data and the relative density based on the specific energy absorption data, and inversely optimize and correct the material plasticity parameters; The data screening module is used to establish a defect coupling assessment model for screening defect data based on the surface powder adhesion rate of the lattice structure and the pillar diameter deviation parameters, and to construct a Gaussian mixture model for screening abnormal data based on the characteristics of the compressive stress-strain curve; The model construction module is used to construct a data correction model for bidirectional feedback of performance prediction and structural design based on the defect coupling evaluation model, Gaussian mixture model, modified elastic modulus, failure criterion and modified material plasticity parameter.

Images