Cross-scale prediction method for high-temperature bending strength of additive manufacturing hard alloy

Abstract

The invention provides a cross-scale prediction method for the high-temperature bending strength of an additive manufacturing hard alloy. The bending strength of an additive manufacturing hard alloy cutter material in a high-temperature state can be accurately predicted. The method comprises the following steps: firstly, determining modeling parameters based on a scanning electron microscope image and a stereology principle; identifying defects in the microstructure of the additive manufacturing hard alloy based on a visual identification method; building a microstructure model of the additive manufacturing hard alloy by utilizing the determined parameters; defects obtained through visual identification are added in the built microscopic structure model through Boolean subtraction, and building of the defect-containing additive manufacturing hard alloy microscopic structure model is completed; embedding a cohesion unit into the built model to simulate the stress state and evolution condition of the microstructure of the material under the action of the thermal load; then, a macroscopic homogenization model of the defect-containing additive manufacturing hard alloy is built, and a three-point bending test model is built; and finally, obtaining the bending strength of the material by using a three-point bending model. The invention belongs to the field of simulation prediction.

Description

Technical Field

[0001] The present invention belongs to the field of simulation and prediction, and particularly relates to a cross-scale prediction method for the high-temperature bending strength of additively manufactured cemented carbide. Background Art

[0002] Due to its excellent mechanical properties and good economy, cemented carbide has become the preferred material in the tool industry. Because of its special principle of layer-by-layer deposition forming, additive manufacturing technology has significant advantages in preparing parts with complex structures. However, due to its unique forming technology and forming method, additively manufactured cemented carbide has its own unique microstructure, among which grain polarization, abnormal growth, and various types of pore defects have become the main factors restricting the development of additively manufactured cemented carbide, having an important impact on its microstructure, and thus reducing the mechanical properties of additively manufactured cemented carbide. Quantitatively establishing the relationship between the microstructure of materials and their mechanical properties is an important direction in materials research, and crystal plasticity finite element technology is one of the most effective research means. The cross-scale analysis method comprehensively analyzes the different characteristics of materials at the micro and macro scales, aims to establish a quantitative relationship between different scales, reveals the influence mechanism of the characteristics of composite materials at different scales on their macroscopic mechanical properties, and provides theoretical support for the comprehensive design of materials.

[0003] Bending strength is an important evaluation index for the mechanical properties of tools. Under thermal load, the microstructure of tool materials changes, resulting in a decrease in bending strength. However, for additively manufactured cemented carbide with specific microstructure and defects, the influence of its microstructure on its macroscopic bending strength at high temperature cannot be achieved by conventional testing methods. Therefore, a cross-scale prediction method for the high-temperature bending strength of additively manufactured cemented carbide is proposed. The cross-scale method applied in the present invention is an innovation point in the field of predicting the bending strength of additively manufactured cemented carbide; the visual recognition technology applied is another innovation point in the field of defect calibration of additively manufactured cemented carbide. Summary of the Invention

[0004] The present invention aims to solve the problem that the specific microstructure of additively manufactured cemented carbide under high-temperature conditions cannot be achieved by conventional testing methods for its macroscopic bending strength, and proposes a cross-scale prediction method for the high-temperature bending strength of additively manufactured cemented carbide, which is implemented according to the following steps:

[0005] Step 1: Determine the microstructure model parameters of additively manufactured cemented carbide;

[0006] Step 2: Identify the defects of additively manufactured cemented carbide based on visual recognition;

[0007] Step 3: Establish a microstructure model of additively manufactured cemented carbide based on the Voronoi algorithm;

[0008] Step 4: Additive manufacturing cemented carbide defect addition method based on Boolean operation;

[0009] Step 5: Determination and establishment of cohesive model parameters;

[0010] Step 6: Macroscopic homogenization modeling of additive manufacturing cemented carbide;

[0011] Step 7: Simulation data processing method.

[0012] Invention effect:

[0013] In order to accurately predict the flexural strength of additive manufacturing cemented carbide at high temperatures, the present invention proposes a cross-scale prediction method for the high-temperature flexural strength of additive manufacturing cemented carbide.

[0014] Under the action of severe thermal loads during high-speed cutting, the microstructure of cemented carbide tools changes, affecting the tool life. Flexural strength is an important evaluation index for the mechanical properties of tools. Therefore, establishing the correlation between the microstructure of cemented carbide tool materials and macroscopic flexural strength has become an urgent problem to be solved. To solve this problem, the present invention proposes a cross-scale prediction method for the high-temperature flexural strength of additive manufacturing cemented carbide. First, the microstructure parameters of cemented carbide materials are determined using stereology principles and a scanning electron microscope; secondly, based on the determined parameters, a microstructure model of additive manufacturing cemented carbide is established using the Voronoi algorithm; the unique microstructure defects of additive manufacturing cemented carbide are determined using visual recognition technology, and defect addition is performed using Boolean operations to construct a microstructure model of additive manufacturing cemented carbide with defects; the Cohesive damage constitutive model is used to insert cohesive elements at grain boundaries and within grains to complete the construction of the damage model; then, through a homogenization model, a cross-scale model of additive manufacturing cemented carbide materials is constructed, and its flexural strength at high temperatures is predicted; finally, the simulation data is processed through a calculation formula to obtain the flexural strength of the additive manufacturing cemented carbide tool material at high temperatures. By comparing the simulation results with the experimental results, the accuracy of the model is verified.

[0015] 1. The microstructure parameters of cemented carbide materials are determined by using stereology principles in combination with a scanning electron microscope

[0016] In this study, aiming at the problem that the additive manufacturing cemented carbide tool material is an opaque cube and the scanning electron microscope cannot judge the parameters in the microstructure, it is proposed to use stereology principles to determine the microstructure parameters of the additive manufacturing cemented carbide tool material.

[0017] 2. The unique microstructure and defects of additive manufacturing cemented carbide are determined by using visual recognition methods, and defect addition is completed through Boolean operations to construct a microstructure model of additive manufacturing cemented carbide with defects.

[0018] 3. A cross-scale prediction method for the high-temperature flexural strength of additive manufacturing cemented carbide can accurately predict the flexural strength of cemented carbide at high temperatures.

[0019] In this study, aiming at the modeling problem of additive manufacturing cemented carbide at high temperatures with specific microstructures and defects, a method is proposed to calculate the microstructure parameters of additive manufacturing cemented carbide by using stereology principles, determine the specific microstructures and defects by using visual recognition methods, and complete the model establishment between the microstructure-macroscopic models by using cross-scale methods, thereby effectively reflecting the flexural strength of additive manufacturing cemented carbide at the macroscopic scale and realizing the establishment of a cross-scale model of additive manufacturing cemented carbide at high temperatures. Brief Description of the Drawings

[0020] Figure 1 It is a schematic diagram of a three-point bending model and boundary conditions;

[0021] Figure 2 It is the flow chart of the cross-scale analysis method. Detailed Implementation Modes

[0022] Detailed Implementation Mode 1: A cross-scale prediction method for the high-temperature flexural strength of additive manufacturing cemented carbide includes the following steps:

[0023] Step 1: Determine the parameters of the microstructure model of additive manufacturing cemented carbide;

[0024] Step 2: Identify the defects of additive manufacturing cemented carbide based on visual recognition;

[0025] Step 3: Establish a microstructure model of additive manufacturing cemented carbide based on the Voronoi algorithm;

[0026] Step 4: A method for adding defects to additive manufacturing cemented carbide based on Boolean operations;

[0027] Step 5: Determine and establish the parameters of the cohesive force model;

[0028] Step 6: Macroscopic homogenization modeling of additive manufacturing cemented carbide;

[0029] Step 7: Simulation data processing method.

[0030] Detailed Implementation Mode 2: The difference between this implementation mode and Detailed Implementation Mode 1 is that the specific operation for determining the parameters of the microstructure model of additive manufacturing cemented carbide in Step 1 is as follows:

[0031] Step 1-1: First, determine the volume fraction of the main components of additive manufacturing cemented carbide. The area fraction of the two-dimensional microstructure of the cross-section of additive manufacturing cemented carbide is used to replace the volume fraction for modeling, and the cross-section of additive manufacturing cemented carbide is characterized by a scanning electron microscope. According to stereology principles, the volume fraction can be calculated by the following formula:

[0032] V V = A A (11)

[0033] Where: V V is the measured volume ratio of the tissue in the three-dimensional space; A A is the measured area ratio of the tissue in the two-dimensional image.

[0034] The volume fractions of the phases in the additive manufactured cemented carbide were measured using ImageJ software and the area method, and binarization and denoising were performed on them. Since the additive manufactured cemented carbide only contains WC and Co phases, the volume fractions of WC and Co phases in the additive manufactured cemented carbide can be obtained by measuring the average gray level:

[0035]

[0036] Step 1-2: Calculate the grain size using the equivalent diameter of a circle. Separate and classify the grains through ImageJ software, measure the grain area and perimeter, and calculate the equivalent circle diameter and shape factor of the grains based on the obtained area and perimeter. The calculation formulas are as follows:

[0037] Equivalent circle diameter D:

[0038]

[0039] Shape influence factor S:

[0040]

[0041] Where: A is the area occupied by the particle; L is the particle perimeter

[0042] Calculate the equivalent circle diameter of each grain using the above formula, and then calculate its average value as the average grain size for subsequent modeling.

[0043] Step 1-3: Calculate the grain adjacency. The additive manufactured cemented carbide material only contains two phases: one is the a phase of WC grains, and the other is the b phase of Co grains. Based on SEM and ImageJ software, the grain adjacency can be calculated according to Equation (7).

[0044]

[0045] Where: L aa is the phase boundary length between the a phase and the a phase in the two-dimensional image; L ab is the phase boundary length between the a phase and the b phase in the two-dimensional image.

[0046] Other steps and parameters are the same as those in the specific implementation method 1

[0047] Specific Embodiment 3: The difference between this embodiment and Specific Embodiment 1 or 2 is that the specific operation of defect recognition of additive manufacturing cemented carbide based on visual recognition in Step 2 is as follows:

[0048] Perform visual recognition using MATLAB software. Store the SEM end face of the additive manufacturing cemented carbide obtained in Step 1 in a specified path. Use MATLAB language to denoise the image to highlight the defect position. Then, use binary code for visual recognition to obtain the size, shape, and position information of the special tissue defects of the additive manufacturing cemented carbide, so as to provide data support for subsequent model establishment.

[0049] Specific Embodiment 4: The difference between this embodiment and the previous three specific embodiments is that the specific operation of establishing the microstructure model of additive manufacturing cemented carbide based on the Voronoi algorithm in Step 3 is as follows:

[0050] Use Matlab software to generate Voronoi polycrystal geometric figures. Combine the average grain size obtained in Step 1. Characterize the microstructure characteristics of cemented carbide by controlling geometric parameters to adjust the geometric shape of polygons; use the elastic modulus of the model as the mechanical evaluation index of the RVE unit, and the loading method is compressive load; to ensure that its deformation is still within the elastic deformation range, the applied displacement load is selected as 0.1% of the material length, and the mesh type is C3D10M. After the test, use Hooke's law to determine the elastic modulus of the material. When the calculated elastic modulus of the material in different directions tends to be stable, this size is the RVE size of the material. The RVE model size determined according to the above method is 60μm×60μm. Import it into Abaqus finite element software in the form of a script through Python language. Based on the volume fraction obtained in Step 1, use Python language to create a set of grains in the microstructure model of the cemented carbide material to control the phase ratio in the model and complete the model establishment.

[0051] Other steps and parameters are the same as the previous three specific embodiments.

[0052] Specific Embodiment 5: The difference between this embodiment and the previous four specific embodiments is that the specific operation of the method for adding defects to additive manufacturing cemented carbide based on Boolean operation in Step 4 is as follows:

[0053] Based on the defect position, shape, and size of the additive manufacturing cemented carbide obtained by visual recognition, use Python language to map it into the finite element simulation software, and use Boolean subtraction operation to add defects to the Voronoi model established in Step 3 to complete the construction of the microstructure model of the additive manufacturing cemented carbide with defects.

[0054] Other steps and parameters are the same as the previous four specific embodiments.

[0055] Embodiment Six: The difference between this embodiment and the previous five embodiments is that the specific operation for determining and establishing the cohesive model parameters in Step Five is as follows:

[0056] Step 5-1: The failure cracks of the additively manufactured cemented carbide cutting tools may propagate along the grain boundaries, i.e., intergranular fracture, or may also propagate inside the grains, i.e., transgranular fracture. Therefore, the quadrilateral cohesive elements containing damage judgment criteria are embedded in the grain boundaries and inside the grains. The microstructure model is divided by linear triangular elements, and the Cohesive elements are inserted into the microstructure model that has been divided by linear triangular elements before solving.

[0057] The Cohesive elements between the triangular elements are mainly completed through the splitting and recombination of shared nodes, and at the same time, the direction of the viscous elements needs to be ensured. First, based on the crystal structure, the triangular mesh is divided; then, the numbering information of the elements and their nodes in the crystal structure is obtained; finally, node splitting and recombination occur at the shared nodes to complete the creation of the new element COH2D4, thereby realizing the insertion of the 0-thickness Cohesive element.

[0058] The most commonly used constitutive model for the Traction-Separation Criterion of the Cohesive element in the cohesive model is the bilinear constitutive model. In the figure, the abscissa is the separation displacement (δ n / mm) between the two interfaces, and the ordinate is the traction stress (T / MPa) between the two interfaces. The slope K n of the curve in the elastic stage represents the unit penalty stiffness, which can be calculated by Equation (17). When the interface traction stress reaches the peak value T max , that is, the damage initiation displacement is δ n init , damage begins to occur at the interface and then immediately decays; when the separation displacement reaches the damage failure displacement δ n fail , the interface completely opens or shear fractures; the area enclosed by the curve and the horizontal axis is the fracture energy G TC , which represents the energy required for the unit to completely fail and can be obtained by Equation (18).

[0059]

[0060] Here, the damage initiation criterion of the Cohesive element adopts the maximum nominal stress criterion (Maxs DamageCriterion), which means that the element is damaged when the ratio of the maximum nominal stress reaches 1. This criterion can be expressed as:

[0061]

[0062] In the formula: are the maximum normal stress, the first shear stress, and the second shear stress, respectively.

[0063] To describe the damage evolution process of the Cohesive element after damage initiation, a damage variable d is determined, which increases monotonically in the interval [0, 1]. When d is 0, it means no damage; when d is 1, it means complete damage. Then the damage evolution can be expressed as:

[0064]

[0065] The damage evolution of the Cohesive element adopts the BK mixed mode. To simplify the solution, the stress-strain relationship and tensile properties of the Cohesive element under shear behavior are set the same, that is, isotropic behavior is assumed. Abaqus mainly controls the penalty stiffness K of the cohesive element n the maximum traction stress T max and the fracture energy G TC to describe the properties of the Cohesive element in the model.

[0066] Take T of the Cohesive element max as 0.002 times of the elastic modulus. At the same time, to successfully simulate the crack effect caused by the damage failure of the Cohesive element, the fracture energy G required for the complete failure of the cohesive element needs to be calculated TC , which can be obtained by calculating from equations (17) and (18). Here, we use the concept of the damage initiation ratio δ ratio , that is, 0 < δ ratio < 1, which can be expressed as:

[0067]

[0068] To ensure that the influence of the Cohesive element on the overall model is small, a relatively large penalty stiffness of the Cohesive element should be used as much as possible. Here, the damage initiation ratio is assumed to be 0.001. According to equations (17), (18) and (21), the property parameters of the Cohesive element at different positions in the model can be calculated.

[0069] Step 52: Based on the principle of embedding 0-thickness Cohesive elements, the construction of the cohesive model for the microstructure of cemented carbide materials is realized by running Python programming scripts in Abaqus. When performing linear triangular element meshing, the mesh size should be as small as possible to clearly describe the initiation and propagation of cracks in the subsequent process. This method realizes the simultaneous distribution of Cohesive with traction-separation criteria in the grains and grain boundaries of the microscopic finite element model, providing conditions for the intergranular and transgranular propagation of cracks.

[0070] Through the above operations, sets of Cohesive elements at different grains and grain boundaries can be obtained, specifically defined as: the Set-WC-coh, Set-TiC-coh, and Set-Co-coh sets inside the grains, and the Set-WC / WC-coh, Set-WC / TiC-coh, Set-TiC / TiC-coh, Set-WC / Co-coh, and Set-TiC / Co-coh sets at the grain boundaries.

[0071] Other steps and parameters are the same as those in the first five specific embodiments.

[0072] Specific Embodiment 7: The difference between this embodiment and the first six specific embodiments is that the specific operation of additive manufacturing cemented carbide macro-homogenization modeling in Step 6 is as follows:

[0073] The macroscopic mechanical properties of additive manufacturing cemented carbide materials are the comprehensive manifestation of the properties of their internal microstructures. Due to the inhomogeneity of the internal microstructural characteristics, the macroscopic mechanical properties of the materials show great dispersion and difference. Different microstructures in different regions, combined with different grain orientations and the presence of defects inside the grains and at the grain boundaries, will affect the stress distribution and strength of the cemented carbide materials, thus showing the instability of the macroscopic mechanical properties of the materials. Therefore, the method of homogenization modeling is used to establish the connection between the mechanical properties of the microstructure and the macroscopic properties of the cemented carbide materials to achieve cross-scale analysis.

[0074] First, a two-dimensional finite element model is established through finite element simulation software. Figure 2The finite element simulation model and boundary conditions for the three-point bending test. To achieve the transfer of mechanical properties between the micro and macro scales and establish the connection between the material microstructure and its macro mechanical properties, the size of the material microstructure model is used as the size of the macro model mesh element. Therefore, when dividing the model into mesh elements, a global seed of 0.06 mm is selected for this model, and a total of 30,000 mesh elements and 30,651 nodes are divided. The element type is a four-node plane stress element (CPS4R). To simulate the stress of the additively manufactured cemented carbide microstructure under thermal load, geometric nonlinearity is turned on, and the temperature is increased using a linear loading method, where the initial temperature is 25 °C and the final temperature is 800 °C; to ensure that the average strain rate of the bending finite element model is the same as the loading strain rate of the three-point bending test, a longitudinal velocity load is applied to the indenter of the model, where the peak value of the velocity load is 0.0083 mm / s, which is obtained by multiplying the average strain rate of the model by the longitudinal initial thickness of the finite element model.

[0075] Then, the mechanical properties of the additively manufactured cemented carbide microstructure with defects are randomly assigned to each element of the bending specimen model through the Python language to achieve the homogenization of different microstructural properties in the macro finite element model.

[0076] Finally, the extended finite element method is used to simulate the failure process of the additively manufactured cemented carbide material and predict the flexural strength. In the whole process, there is no need to prefabricate the initial crack position. The extended finite element mainly judges the failure position through the stress state and strength of the elements in the material model, so as to realize the simulation of crack propagation. The maximum principal stress criterion (MaxpsDamage Criterion) is used as the damage initiation criterion, and the maximum principal stress criterion can be expressed as:

[0077]

[0078] In the formula: is the maximum allowable principal stress; when the maximum principal stress of the material reaches the maximum allowable principal stress, it indicates that damage begins and crack initiation and propagation occur.

[0079] Other steps and parameters are the same as those in the first six specific embodiments.

[0080] Specific Embodiment 8: The difference between this embodiment and the first seven specific embodiments is that the specific operation of the simulation data processing method in Step 7 is as follows:

[0081] The flexural strength, maximum shear modulus, and flexural modulus of the additively manufactured cemented carbide material can be calculated by the following formula:

[0082]

[0083]

[0084] Where: R bb is the flexural strength, unit: MPa; F max is the maximum load required for the specimen to fracture, unit: N; L s is the distance between two support points, i.e., the span, unit: mm; b is the width of the bending specimen, unit: mm; h is the thickness of the bending specimen, unit: mm; τ max is the maximum shear strength of the bending specimen, unit: MPa; E f is the bending modulus of the specimen, unit: MPa; ΔF is the load increment in the test, unit: N; Δf is the deflection increment in the test, unit: mm.

[0085] The flexural strength, shear strength and bending modulus of the additive manufacturing cemented carbide are calculated according to the results obtained from the finite element simulation.

[0086] Other steps and parameters are the same as those in the first seven specific embodiments.

Claims

1. A cross-scale prediction method for the high-temperature flexural strength of additively manufactured cemented carbides, characterized in that The cross-scale prediction method for the high-temperature bending strength of additive manufacturing cemented carbide includes the following steps: Step 1: Determination of the parameters of the microstructure model of additive manufacturing cemented carbide; Step 2: Defect identification of additive manufacturing cemented carbide based on visual recognition; Step 3: Establishment of the microstructure model of additive manufacturing cemented carbide based on the Voronoi algorithm; Step 4: Method for adding defects to additive manufacturing cemented carbide based on Boolean operations; Step 5: Determination and establishment of the parameters of the cohesive force model; Step 6: Macroscopic homogenization modeling of additive manufacturing cemented carbide; Step 7: Simulation data processing method.

2. The cross-scale prediction method for the high-temperature bending strength of additively manufactured cemented carbide according to claim 1, characterized in that, The specific operation for determining the parameters of the microstructure model of additive manufacturing cemented carbide in Step 1 is as follows: Step 1-1: Determination of the volume fraction of the main components. The cross-section of the additive manufacturing cemented carbide is characterized by a scanning electron microscope (SEM). Based on the stereology principle, its volume fraction can be calculated by Equation (1): （1） Where: V V is the measured volume ratio of the tissue in the three-dimensional space; A A is the measured area ratio of the tissue in the two-dimensional image. The volume fraction of each phase in the cemented carbide is measured using ImageJ software and the area method. Then, its microstructure is binarized and denoised, and the gray value of the binarized image is measured. The volume fractions of the WC and Co phases in the additive manufacturing cemented carbide material can be obtained from the measured average gray value, as shown in Equations (2) and (3): （2） （3） Step 1-2: Calculation of the average grain size and shape factor of WC. The grain size is calculated using the equivalent diameter of a circle, and the average grain size is taken as the total grain size. ImageJ is used to separate and classify the grains, and the average grain area and perimeter of WC are measured, as shown in Equations (5) and (6). Equivalent circle diameter D: （4） Shape influence factor S: （5） In the formula: A is the area occupied by the particle; L is the particle perimeter The equivalent circle diameter of each grain is calculated using the above formula, and then the average value is calculated as the average grain size. Step 1-3: Calculation of grain adjacency. The additive manufacturing cemented carbide material only contains two phases: one is the a phase of WC grains, and the other is the b phase of Co grains. Based on the SEM and ImageJ software, the grain adjacency can be calculated according to Equation (6). （6） Where: is the phase boundary length between phase a and phase a in the two-dimensional image; is the phase boundary length between phase a and phase b in the two-dimensional image.

3. The cross-scale prediction method for the high-temperature flexural strength of additively manufactured cemented carbide according to claim 2, wherein The specific operation for defect identification of additive manufacturing cemented carbide based on visual recognition in Step 2 is as follows: MATLAB software is used for visual recognition. The SEM end face of the additive manufacturing cemented carbide obtained in Step 1 is stored in the specified path. The MATLAB language is used to denoise the image to highlight the defect position. Then, visual recognition is performed using the binarization code to obtain the size, morphology, and position information of the special tissue defects of the additive manufacturing cemented carbide, providing data support for subsequent model establishment.

4. The cross-scale prediction method for the high-temperature flexural strength of additively manufactured cemented carbide according to claim 3, wherein, The specific operation for establishing the microstructure model of additive manufacturing cemented carbide based on the Voronoi algorithm in Step 3 is as follows: Generate Voronoi polycrystalline geometries using Matlab software. Combine with the average grain size obtained in Step 1, and characterize the microstructural features of cemented carbide by adjusting the geometric morphology of polygons through controlling geometric parameters. Determine its RVE unit according to the elastic modulus, and finally determine the model size as 60μm×60μm, and import it into Abaqus finite element software in the form of a script through Python language. Based on the volume fraction obtained in Step 1, use Python language to create a set of grains in the microstructure model of cemented carbide material, realize the control of the phase ratio in the model, and complete the model establishment.

5. The cross-scale prediction method for the high-temperature bending strength of additively manufactured cemented carbide according to claim 4, wherein, The specific operation of the method for adding defects to additively manufactured cemented carbide based on Boolean operation in Step 4 is as follows: Based on the positions, morphologies, and sizes of the defects of the additively manufactured cemented carbide obtained by visual recognition, use Python language to map them into the finite element simulation software, and use Boolean subtraction operation to add defects to the Voronoi model established in Step 3 to complete the construction of the microstructure model of additively manufactured cemented carbide with defects.

6. The cross-scale prediction method for the high-temperature flexural strength of additively manufactured cemented carbide according to claim 5, characterized in that, The specific operation of determining and establishing the cohesive model parameters in Step 5 is as follows: Step 5-1: Divide the microstructure model using linear triangular elements, and insert Cohesive elements to divide the microstructure model. Embed quadrilateral cohesive elements containing damage judgment criteria inside the grains and at the grain boundaries. Use a bilinear constitutive model to describe the traction-separation criterion of the Cohesive element, and the damage initiation criterion uses the maximum nominal stress criterion, which can be expressed as: （7） Wherein: , , are the maximum normal stress, the first shear stress direction and the second shear stress direction, respectively. The damage evolution of the Cohesive element adopts the BK mixed mode, and the penalty stiffness K of the cohesive element is controlled n , the maximum traction force T max and the fracture energy G TC to describe the properties of the Cohesive element in the model by three parameters. Step 5-2: Use Python programming script to establish the cohesive model of the microstructure of cemented carbide material. When performing linear triangular element division, the mesh size should be as small as possible to clearly describe the initiation and propagation of cracks subsequently.

7. The cross-scale prediction method for the high-temperature bending strength of additively manufactured cemented carbide according to claim 6, characterized in that, The specific operation of the macroscopic homogenization modeling of additively manufactured cemented carbide in Step 6 is as follows: Establish a two-dimensional finite element model through Abaqus software. The model and boundary conditions are shown in Figure 2, and the cross-scale analysis method flow is shown in Figure 3. Use the microstructure size of the material as the mesh element size of the macroscopic model, select a 0.06mm size for global seeding of the model, and the mesh type is selected as four-node plane stress element (CPS4R). Apply a longitudinal velocity load to the indenter, and its peak value is calculated by multiplying the average strain rate of the model by the longitudinal initial thickness of the finite element model. At the same time, turn on geometric nonlinearity, and use linear loading to apply thermal load, with the starting temperature of 25°C and the ending temperature of 800°C. Then use Python language to randomly assign the mechanical properties of the microstructure with microvoid defects obtained in Step 4 to each element of the bending specimen model to realize the homogenization of different microstructural properties in the macroscopic finite element model.

8. A cross-scale prediction method for the high-temperature flexural strength of additively manufactured cemented carbides according to claim 7, characterized in that The specific operation of the simulation data processing method in Step 7 is as follows: Referring to the calculation formulas in national standards and industry standards, the flexural strength, maximum shear modulus, and bending modulus of the material can be calculated by Equation (5), Equation (6), and Equation (7) respectively: （8） （9） （10） Where: R bb is the flexural strength, unit: MPa; F max is the maximum load required for specimen fracture, unit: N; L s is the distance between two support points, i.e., the span, unit: mm; b is the width of the bending specimen, unit: mm; h is the thickness of the bending specimen, unit: mm; τ max is the maximum shear strength of the bending specimen, unit: MPa; E f is the flexural modulus of the specimen, unit: MPa; ΔF is the load increment during the test, unit: N; Δf is the deflection increment during the test, unit: mm.

Images