Prediction method of melt pool width for 3D printing selective laser melting technology

Abstract

The present invention belongs to the technical field of 3D printing, and provides a method for predicting the molten pool width for 3D printing selective laser melting technology, comprising the steps of: constructing a powder bed model, and obtaining the coordinate information of each particle in the powder bed; creating a geometric mesh model of the powder bed model; inputting the geometric mesh model into computational fluid dynamics software, and determining the properties of the material to be melted used in the geometric mesh model; selecting a physical model in the computational fluid dynamics software; setting each physical model; calculating the equivalent gas flow rate and loading the gas flow field according to the principle that the Reynolds number remains unchanged after the model is reduced and adjusted according to the reduction ratio; performing laser melting simulation after initializing the computational fluid dynamics software and setting parameters, and measuring the molten pool size. The present invention can effectively simulate the complex flow behavior in the molten pool and obtain a more accurate prediction size of the molten pool width.

Description

Technical Field

[0001] The present invention belongs to the technical field of 3D printing, and in particular relates to a method for predicting molten pool width for 3D printing selective laser melting technology. Background Art

[0002] Selective Laser Melting (SLM) is a metal additive manufacturing technology used in 3D printing. Its working principle is to selectively scan the surface of the pre-laid powder bed with a high-energy laser beam to quickly melt and solidify the material, thereby stacking layer by layer to form a three-dimensional entity. SLM technology has a high degree of manufacturing freedom and can be used to produce complex structures, lightweight components and products with excellent performance. This technology has been widely used in high-end manufacturing fields such as aerospace, medical equipment and automobiles.

[0003] In SLM technology, the dynamic behavior of the molten pool is one of the key factors affecting the quality of the final product. The SLM preparation process involves complex multi-physical field interactions, including heat conduction, fluid dynamics, and the interaction between laser and material. The characteristics of the molten pool, such as width, depth, and temperature distribution, directly determine the microstructure, mechanical properties, and dimensional accuracy of the part. Since the SLM process involves a rapid heating and cooling process, the metal flow, heat conduction, and surface tension in the molten pool interact with each other, resulting in a complex and unpredictable morphology of the molten pool. In particular, changes in the molten pool width will directly affect the overlap of the scanning paths, thereby affecting the overall forming quality. For example, a molten pool width that is too small may result in incomplete fusion between the scanning paths, while a width that is too large may result in overburning or porosity defects. Therefore, accurately predicting and controlling the molten pool width is a core issue in improving forming quality and process stability. Summary of the invention

[0004] In view of the shortcomings of the prior art, the present invention provides a method for predicting the molten pool width of selective laser melting technology for 3D printing. Based on computational fluid dynamics software simulation, the molten pool width of selective laser melting technology is predicted by improving the heat source model and considering the influence of gas flow field, thereby improving the accuracy of heat source modeling.

[0005] To achieve the above object, the present invention discloses a method for predicting the molten pool width for 3D printing selective laser melting technology, which comprises the following steps:

[0006] S1. Build a powder bed model and obtain the coordinate information of each particle in the powder bed;

[0007] S2. Creating a geometric mesh model of the powder bed model, where the geometric mesh model includes a substrate, a powder bed and a gas domain;

[0008] S3, inputting the geometric mesh model into computational fluid dynamics software, generating powder bed particles according to the acquired coordinate information of each particle and determining the properties of the material to be melted;

[0009] S4. Select and set a physical model in the computational fluid dynamics software, where the physical model at least includes a heat source model, a heat conduction model, a fluid dynamics model, a phase change model, a surface tension model, and a turbulence model;

[0010] S5. Setting each physical model, including:

[0011] In the heat source model, the heat source is set to a ray tracing heat source. In the ray tracing heat source, the horizontal discrete intensity at any point within the range of the ray tracing heat source is:

[0012] ;

[0013] in, is the intensity distribution of the laser beam, is the laser power, is the laser beam radius, and After the laser beam is dispersed, each ray is and The coordinates of the center origin in the direction, and The laser beam axis before the discrete and The center coordinates in the direction, is the light discrete spacing;

[0014] S6. According to the principle that the Reynolds number remains unchanged after the reduced model and is adjusted according to the reduction ratio, the equivalent gas flow rate is calculated to load the gas flow field in the gas domain;

[0015] S7. After initializing the computational fluid dynamics software and setting parameters, laser melting simulation is performed to measure the molten pool size.

[0016] Furthermore, the principle of adjusting by reduction ratio in step S6 is specifically:

[0017] The equivalent gas flow rate is adjusted according to the reduction ratio:

[0018] ;

[0019] in, To adjust the gas flow rate, To adjust the front gas flow rate, is the scale at which the model is reduced.

[0020] Furthermore, in step S5, the coordinates in the horizontal discrete intensity expression of any point within the range of the ray tracing heat source are , and spacing Set up acquisition in the heat source model.

[0021] Furthermore, the ray tracing heat source in step S5 also includes:

[0022] Calculate the direction of reflected light according to the law of mirror reflection:

[0023] ;

[0024] in, is the direction of reflected light, is the incident light direction, is the reflection normal vector.

[0025] Furthermore, the ray tracing heat source in step S5 also includes:

[0026] Reflected light power update:

[0027] ;

[0028] in, is the reflected light power, is the incident light power, is the Fresnel reflectivity, is the angle of incidence.

[0029] Furthermore, the conditions for judging the failure of light in the ray tracing heat source include: when the incident light power drops below a calibrated proportion of the initial power, the light is deemed to be failed and no longer participates in subsequent calculations.

[0030] Furthermore, the calibration ratio is 5%.

[0031] Furthermore, the conditions for judging the failure of light in the ray tracing heat source also include: the path of the light completely leaves the surface area of ​​the material, and there are no more surface particles that can interact with the light, then the light fails.

[0032] Furthermore, the surface tension in the surface tension model in step S4 is:

[0033] ;

[0034] in, is the surface tension, is the surface tension coefficient, is the surface curvature, is the unit vector normal to the free interface, is the average density of each grid, is the average density, , , represent the volume fraction of the metal phase and the volume fraction of the gas phase, respectively. is the metal density, is the gas density, is the modulus of the metal volume fraction.

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] 1. The present invention discretizes the laser beam into multiple light rays, analyzes the propagation, reflection, refraction of the light in space and its interaction with the material surface, captures the surface heterogeneity, multiple reflection effects, and Fresnel reflection characteristics dependent on the incident angle of laser melting, improves the heat source model, and thus improves the accuracy of heat source modeling.

[0037] 2. In the process of multiple iterative calculations of laser light refraction and reflection, the present invention adopts a method of dynamically dividing SPH particles, and merges multiple particles into one particle in low-demand areas, such as stable areas or areas far away from the flow core, thereby reducing calculation overhead and improving calculation efficiency.

[0038] 3. In the simulation process, the present invention loads the gas flow field, and the gas flow field velocity is controlled by keeping the Reynolds number constant and adjusting the reduction ratio to calculate the equivalent gas flow rate, which effectively simulates the influence of the protective gas or process environment on the molten pool in the actual additive manufacturing process. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a schematic flow chart of a method for predicting the molten pool width for 3D printing selective laser melting technology according to the present invention;

[0040] Figure 2 A powder bed model constructed for one embodiment of the present invention;

[0041] Figure 3 A schematic diagram of surface particle identification and normal calculation according to an embodiment of the present invention;

[0042] Figure 4 A schematic diagram of light initialization according to an embodiment of the present invention;

[0043] Figure 5 A schematic diagram of a gas flow field model according to an embodiment of the present invention;

[0044] Figure 6 This is a comparison chart of simulation results and physical printing results of an embodiment of the present invention.

[0045] In the figure:

[0046] 1. Powder particles; 2. Detection ball; 3. Internal particles; 4. Surface particles; 5. Laser beam; 6. Laser beam axis; 7. Separated laser light; 8. Gas inlet; 9. Gas outlet. DETAILED DESCRIPTION

[0047] In order to make the purpose, technical solutions and advantages of the present invention more clear, further description is given below in conjunction with the accompanying drawings and embodiments.

[0048] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0049] As an embodiment of the present invention, Figure 1 As shown, the present invention provides a method for predicting the molten pool width for 3D printing selective laser melting technology, which comprises the following steps:

[0050] S1. Build a powder bed model and obtain the coordinate information of each particle in the powder bed.

[0051] First, build a powder bed model: First, use particle simulation software to generate powder particles with a particle size of 15-53 microns, and export the coordinate information of each particle; the constructed powder bed model is as follows Figure 2 shown.

[0052] Particle simulation software is a software used to simulate and analyze particulate matter. After generating powder particles 1 in the particle simulation software, a powder bed model can be obtained. The powder bed model contains a large amount of geometric information about the particles, such as the shape, size, arrangement and distribution of the particles (referring to the coordinate information of each particle). The powder bed model can also be used to simulate and analyze the stacking behavior, flow characteristics, compaction process, etc. of particles in a container. Particle simulation software provides intuitive 3D visualization functions to observe the movement of particles and the behavior of the powder bed.

[0053] S2. Create a geometric mesh model of the powder bed model, where the geometric mesh model includes a substrate, a powder bed and a gas domain.

[0054] The powder bed model is created through geometric modeling software. The geometric modeling software can be considered to use open source software such as FreeCAD. The model size is created to be 1x0.2x0.25mm 3, where the substrate size is 1x0.2x0.1mm 3 , powder bed size is 1x0.2x0.03mm 3 , gas domain size 1x0.2x0.12mm 3 The present invention uses computational fluid dynamics (CFD) mesh generation tools to mesh the geometric model, with a minimum mesh size of 3 microns.

[0055] Importing the powder bed model from the particle simulation software into the CFD mesh generation tool for model creation will result in a CFD mesh model suitable for fluid flow and heat transfer simulation. This mesh model is the basis for subsequent CFD simulations.

[0056] S3. Import the grid model file into the computational fluid dynamics (CFD) software, and generate the particles of the powder bed model in the above step S1 and determine the properties of the material to be melted in the computational fluid dynamics (CFD) software according to the coordinate information of each particle exported by the UDF code.

[0057] At present, the traditional method generally involves testing with large parameters and physical materials, and then narrowing the scope to obtain relatively good results. Computational fluid dynamics (CFD) software can be used to simulate and analyze fluid flow and heat transfer phenomena, discretize physical quantities into computational grids, and iteratively solve them through conservation equations (mass, momentum, and energy). This method has a high degree of stability and accuracy, and is particularly suitable for dealing with heat conduction and fluid dynamics coupling problems. It can effectively simulate the complex flow behavior in the molten pool, including the convection and heat transfer processes caused by laser heating. Its functions mainly include: Fluid flow simulation: Computational fluid dynamics CFD software can simulate the flow behavior of various fluids, including gases and liquids, in complex geometric structures; Heat transfer and heat transfer analysis: The software can analyze the heat exchange between the fluid and the solid surface, as well as the temperature distribution inside the fluid; Multi-physics field coupling: In addition to fluid flow and heat transfer, computational fluid dynamics CFD software can also simulate complex physical processes such as chemical reactions, multiphase flow, and particle tracking; Powerful pre-processing and post-processing functions: Provide grid generation, result visualization and data analysis tools to help users prepare simulations and interpret results; Customization and extensibility: Users can expand the functionality of the software by writing UDF (User-Defined Functions); High-efficiency calculation: Supports parallel computing and GPU acceleration, and can handle large-scale and complex simulation problems. By using computational fluid dynamics CFD software, users can better understand and predict fluid behavior, thereby optimizing product performance in the design stage, reducing experimental costs, and accelerating the R&D cycle. Computational fluid dynamics CFD software can be ANSYS Fluent, for example.

[0058] After the grid model file is imported into the computational fluid dynamics CFD software, the particles in the above powder bed model are generated by the UDF code. It should be noted that the particles generated in the particle simulation software cannot be used directly, so the shape, size, arrangement and distribution of the particles (referring to the coordinate information of each particle) are required. These coordinate information are used to regenerate these particles in the computational fluid dynamics CFD software through the UDF code, and these generated particles interact with the light. The present invention selects 316L stainless steel as the material to be melted in the computational fluid dynamics CFD software, and determines the material properties of 316L stainless steel, which specifically include thermal conductivity, specific heat capacity, material density, viscosity and gas-solid surface tension; the material properties can be obtained through public software.

[0059] S4. Select the physical model in the computational fluid dynamics (CFD) software;

[0060] In computational fluid dynamics (CFD) software, different physical phenomena and effects can correspond to different model types. The physical model selected in the embodiment of the present invention at least includes a heat source model, a heat conduction model, a fluid dynamics model, a phase change model, a surface tension model, and a turbulence model.

[0061] S5. Setting each physical model in computational fluid dynamics (CFD) software;

[0062] Corresponding to the heat source model: the present invention improves the traditional heat source model, and sets the heat source as a ray tracing heat source in the heat source model to simulate the heat input generated by the laser light source on the surface or inside of the material.

[0063] The traditional volumetric heat source model ignores the Fresnel reflection effect, that is, the change of laser absorption rate with the incident angle, and cannot capture physical properties such as surface roughness and local reflectivity changes. For concave surfaces with complex geometric shapes, the energy absorption changes caused by multiple reflections cannot be simulated, which is quite different from the actual results. The ray tracing RT method is based on the theory of geometric optics and discretizes the laser beam into multiple rays. These rays propagate, reflect and refract in space and interact with the surface of the material. Compared with the traditional volumetric heat source model, ray tracing can capture surface heterogeneity, multiple reflection effects, and Fresnel reflection characteristics that depend on the incident angle, thereby improving the accuracy of heat source modeling. The specific implementation steps are as follows:

[0064] P1, initialize the light and determine the horizontal discrete intensity of any point within the range of the ray tracing heat source;

[0065] Initializing the ray requires Figure 4As shown in the figure on the right, the XOY plane light in the figure is discretized with equal spacing according to Gaussian distribution, and the laser beam 5 is discretized into multiple light rays, each of which has a central origin, direction and energy. The initial direction of the light is set to the direction of the laser beam, vertically downward along the z axis. The laser beam is arranged (vertically passing through) in a spaced On the 2D horizontal uniform grid with a spacing, the axis of the laser beam axis 6 is at the center of the 2D horizontal uniform grid, and the center origin of each separated laser light 7 is arranged at the center of each small grid.

[0066] Calculate the coordinates of each discrete light ray, and update the initial light power according to the above Gaussian beam intensity distribution formula. The horizontal discrete intensity of any point within the heat source range is traced by the ray, that is:

[0067] ;

[0068] in, is the intensity distribution of the laser beam, is the laser power, is the laser beam radius, and After discretization, each ray is and The coordinates of the center origin in the direction, and The laser beam axis before the discrete and The center coordinates in the direction, is the discrete spacing of the light, and the coordinate , and spacing Set up acquisition in the heat source model.

[0069] P2. Calculate the direction of reflected light according to the law of mirror reflection.

[0070] Smooth particle hydrodynamics (SPH) is used to reconstruct the surface particles of the material, and the radius of the detection ball 2 is used to control the surface area represented by the particles. The surface particles are identified by calculating physical effects such as surface tension and boundary heat loss (such as radiation and convection). When the laser moves in the negative direction of the Z axis, the distance between the laser and the SPH particle is less than or equal to the radius of the detection ball and the product of the incident laser vector and the SPH particle reflection normal is negative, it is determined that the laser interacts with the surface area of ​​the SPH particle. The general principle is: the detection ball determines the interaction and energy absorption, and then redistributes the energy through the absorption ball. The closer the place is, the higher the energy is, and the farther the place is, the lower the energy is.

[0071] like Figure 3As shown, surface particles 4 are identified in the laser action area, and the reflection normal vector is calculated. The present invention sets the absorption sphere radius to 8 microns, uses the absorption sphere as a surface characterization tool, and calculates the reflection normal vector :

[0072] ;

[0073] ;

[0074] in: For the surface particles in the particle The index of the absorption sphere, that is, is the collection of all SPH particles within the absorbing sphere centered at is the corresponding average weight, refers to Each particle in the SPH particle set The distance For particles The normal vector of is the radius of the absorbing sphere.

[0075] Calculate the direction of reflected light according to the law of mirror reflection :

[0076] ;

[0077] in: is the incident light direction, is the direction of reflected light, is the reflection normal vector.

[0078] P3. Calculate the material absorption efficiency and update the reflected light power according to the Fresnel reflection law. The specific formula is as follows:

[0079] The expression for updating the reflected light power is as follows:

[0080] ;

[0081] ;

[0082] in: is the reflected light power, is the incident light power, is the Fresnel reflectivity; is the incident angle, is the actual refractive index, is the relative magnetic permeability.

[0083] In the process of multiple iterative calculations, the method of dynamically dividing SPH internal particles 3 is adopted to optimize the calculation efficiency by dynamically adjusting the particle distribution density while ensuring the physical accuracy. In low-demand areas, such as stable areas or areas far away from the flow core, multiple particles are merged into one particle to reduce the calculation overhead.

[0084] This model uses two conditions to determine the failure of light. First, if the power of the incident light drops below 5% of the initial power, the light is considered to be failed and no longer participates in subsequent calculations. Second, if the path of the light completely leaves the surface area of ​​the material and there are no more surface particles that can interact with the light, the light is considered to be failed.

[0085] Corresponding heat conduction model: In the heat conduction model, heat transfer is simulated by heat transfer radiation. Heat transfer radiation can be simulated by heat transfer radiation function. The heat transfer radiation function includes:

[0086] ;

[0087] ;

[0088] in, is the convection loss, is the radiation loss, is the convective heat transfer coefficient, To simulate the current temperature, is the ambient temperature, is the Stefan-Boltzmann constant, is the emissivity, is the norm of the metal volume fraction, is the average density of each grid, is the average density, , , represent the volume fraction of the metal phase and the volume fraction of the gas phase, respectively. is the metal density, is the gas density.

[0089] Corresponding surface tension model: In the surface tension model, surface tension is used to simulate interfacial tension and its effect on fluid flow. Surface tension can be simulated by the surface tension function, which is: ;

[0090] in, is the surface tension, is the surface tension coefficient, is the surface curvature, is the norm of the metal volume fraction, is the unit vector normal to the free interface, is the average density of each grid.

[0091] Corresponding fluid dynamics model: In the fluid dynamics model, the momentum equation of the fluid is affected by considering the thermal buoyancy in the gravity source term. The thermal buoyancy can be simulated by the thermal buoyancy function, which is:

[0092] ;

[0093] in, is the acceleration due to gravity, is the norm of the metal volume fraction, is the thermal expansion coefficient of the material.

[0094] Corresponding phase change models include:

[0095] 1) In the evaporation-condensation model, the recoil pressure is added as a source term to the momentum equation. The recoil pressure can be simulated by the recoil pressure function, which is:

[0096] .

[0097] in, is the atmospheric pressure, To simulate the current temperature, is the ideal gas constant, is the latent heat of vaporization, is the molar mass of the metal, is the latent heat of vaporization, is the modulus of the metal volume fraction.

[0098] 2) In the evaporation-condensation model, the mass transfer from the liquid phase to the gas phase is simulated by evaporation loss. The evaporation loss can be simulated by the evaporation loss function, which is:

[0099] ;

[0100] in, To simulate the current temperature, is the ideal gas constant, is the latent heat of vaporization, is the molar mass of the metal, is the latent heat of vaporization, is the norm of the metal volume fraction, is the average density, , , represent the volume fraction of the metal phase and the volume fraction of the gas phase, respectively. , represents the heat capacity of the metal, is the heat capacity of the gas, is the average density of each grid.

[0101] Corresponding turbulence model: In the turbulence model, the Marangoni effect is added as an additional source term to the momentum equation to take into account the flow caused by temperature or concentration gradients. The Marangoni effect can be simulated by the Marangoni effect function, which is:

[0102] ;

[0103] in, is the equivalent volume force of the Marangoni effect, is the effect of temperature on surface tension, is the average density of each grid, is the temperature gradient.

[0104] S6. Consider the influence of gas flow field: Based on the principle that the Reynolds number remains unchanged after the model is reduced and adjusted according to the reduction ratio, the equivalent gas flow rate is calculated and the gas flow field in the gas domain is loaded.

[0105] In the computational fluid dynamics CFD software simulation process, such as Figure 5 As shown, the gas flow field including the gas inlet 8 and the gas outlet 9 is to simulate the effect of the protective gas or process environment on the molten pool in the actual additive manufacturing process. This factor is very important in laser additive manufacturing. The wind field takes away the heat from the surface of the molten pool and reduces the local temperature gradient, thereby affecting the cooling rate and solidification behavior of the molten pool. At the same time, the shear force generated by the gas flow will change the flow pattern on the surface of the molten pool, thereby affecting the shape and depth of the molten pool.

[0106] At the same time, in order to reduce computing resources when the simulation model is reduced, the velocity and height of the wind field need to be adjusted to ensure that the simulation results have physical significance. This scaling usually needs to be based on similarity theory to ensure that key dimensionless parameters remain unchanged, such as the Reynolds number. :

[0107] ;

[0108] in, is the gas density, is the gas flow rate, is the characteristic length, i.e. the height or width of the model, is the dynamic viscosity of gas.

[0109] To ensure that the Reynolds number remains unchanged after reducing the model, the flow rate should be adjusted according to the reduction ratio:

[0110] ;

[0111] in, is the scale at which the model is reduced.

[0112] The gas flow field is loaded, and the gas flow field velocity is adjusted by keeping the Reynolds number unchanged and reducing the ratio. The equivalent gas flow velocity is calculated to be 0.043m / s. After the gas flow field is loaded, the subsequent calculation can be started.

[0113] S7. After initializing the computational fluid dynamics (CFD) software and setting parameters, laser melting simulation is performed to measure the molten pool size.

[0114] Initialize the computational fluid dynamics (CFD) software parameters, and set the model area, material, fluid domain, boundary conditions and solution method in the functional area of ​​the computational fluid dynamics (CFD) software according to the material parameters recorded above. Select the PISO solver as the solution method, set the solution time step to 10000 and the time step to 10. -9 to 10 -8 Second.

[0115] In simple terms, boundary conditions are to define each surface so that it can realize its function. Figure 5 Many surfaces in the system are regarded as walls, and then set to have convection and radiation with the air. There is a gas inlet on one wall, which is also a velocity inlet, and there is a pressure outlet on the opposite wall, which is mainly used to set the wind field.

[0116] Determine the parameters during the simulation in the computational fluid dynamics (CFD) software.

[0117] In the present invention, the parameters in the simulation process include: ideal gas constant, latent heat of melting of metal materials, latent heat of evaporation of metal materials, relative molecular mass of metal materials, surface emissivity, Boltzmann constant, solid phase line of metal materials, liquid phase line of metal materials, gas phase line of metal materials, melting point of metal materials, solid-liquid phase line difference, ambient pressure, thermal expansion coefficient of metal materials, surface tension coefficient, thermal dependence coefficient of surface tension, Boltzmann constant. All of the above parameters can be obtained by querying in public documents.

[0118] In addition, in the embodiment of the present invention, the laser moving speed is 1400 mm / s, the laser power is 300 W, the laser radius is 20 μm, the ambient temperature is 300 K, the mushy zone constant is 10 6 .

[0119] After the computational fluid dynamics (CFD) software is finished, the simulation results can be Figure 6 As shown, the molten pool size is measured. The simulation results of one embodiment of the present invention are shown in Figure 6 The right picture shows the actual printed result. Figure 6 As shown in the left picture.

[0120] Furthermore, when predicting the molten pool width, it is necessary to ensure that all relevant physical processes are simulated correctly, and the above parameters can also be adjusted according to the specific simulation goals and the characteristics of the molten pool. Usually, this process requires iteration and verification to ensure that the simulation results are consistent with experimental data or theoretical predictions. By inputting material parameters and process parameters, the simulation can calculate the temperature field and flow field of the molten pool and further extract its geometric characteristics.

[0121] The present invention uses computational fluid dynamics (CFD) software to simulate the behavior inside the molten pool, and can provide detailed data from aspects such as fluid flow, temperature field, stress field, and material transfer, helping to optimize process parameters, improve product quality, reduce defects, save time and cost, and promote the application and research of new technologies. Especially in areas involving complex flows and high-precision requirements, CFD simulation can provide unparalleled support and is an indispensable tool in modern manufacturing processes. The simulation results of the present invention can not only intuitively display the morphology of the molten pool, but also provide a basis for optimizing process parameters.

[0122] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

[0123] The present invention and its embodiments are described above, and such description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. In short, if ordinary technicians in the field are inspired by it, without departing from the purpose of the invention, they can design a structure and embodiment similar to the technical solution without creativity, which should belong to the protection scope of the present invention.

Claims

1. A method for predicting the molten pool width for 3D printing selective laser melting technology, characterized in that: It includes the steps of: S1. Build a powder bed model and obtain the coordinate information of each particle in the powder bed; S2. Create a geometric mesh model of the powder bed model, where the geometric mesh model includes a substrate, a powder bed and a gas domain; S3, inputting the geometric mesh model into computational fluid dynamics software, generating powder bed particles according to the acquired coordinate information of each particle and determining the properties of the material to be melted; S4. Select and set a physical model in the computational fluid dynamics software, where the physical model at least includes a heat source model, a heat conduction model, a fluid dynamics model, a phase change model, a surface tension model, and a turbulence model; S5. Setting each physical model, including: In the heat source model, the heat source is set to a ray tracing heat source. In the ray tracing heat source, the horizontal discrete intensity at any point within the range of the ray tracing heat source is: ； in, is the intensity distribution of the laser beam, is the laser power, is the laser beam radius, and After the laser beam is dispersed, each ray is and The coordinates of the center origin in the direction, and The laser beam axis before the discrete and The center coordinates in the direction, is the light discrete spacing; Ray tracing heat sources also include: Calculate the direction of reflected light according to the law of mirror reflection: ； in, is the direction of reflected light, is the reflection normal vector; S6. According to the principle that the Reynolds number remains unchanged after the reduced model and is adjusted according to the reduction ratio, the equivalent gas flow rate is calculated to load the gas flow field in the gas domain; S7. After initializing the computational fluid dynamics software and setting parameters, laser melting simulation is performed to measure the molten pool size.

2. The method for predicting the molten pool width for 3D printing selective laser melting technology according to claim 1, characterized in that: The principle of adjusting by reduction ratio in step S6 is specifically: The equivalent gas flow rate is adjusted according to the reduction ratio: ； in, To adjust the gas flow rate, To adjust the front gas flow rate, is the scale at which the model is reduced.

3. The method for predicting the molten pool width for 3D printing selective laser melting technology according to claim 1, characterized in that: The coordinates in the horizontal discrete intensity expression of any point within the range of the ray tracing heat source in step S5 , and spacing Set up acquisition in the heat source model.

4. The method for predicting the molten pool width for 3D printing selective laser melting technology according to claim 1, characterized in that: The ray tracing heat source in step S5 also includes: Reflected light power update: ； in, is the reflected light power, is the incident light power, is the Fresnel reflectivity, is the angle of incidence.

5. The method for predicting the molten pool width for 3D printing selective laser melting technology according to claim 4, characterized in that: The conditions for determining ray failure in ray tracing heat sources include: If the incident light power drops below a calibrated proportion of the initial power, the light is considered to have failed and no longer participates in subsequent calculations.

6. The method for predicting the molten pool width for 3D printing selective laser melting technology according to claim 5, characterized in that: The nominal ratio is 5%.

7. The method for predicting the molten pool width for 3D printing selective laser melting technology according to claim 4, characterized in that: The conditions for judging the failure of light in the ray tracing heat source also include: the path of the light completely leaves the surface area of ​​the material, and there are no more surface particles that can interact with the light, then the light fails.

8. The method for predicting the molten pool width for 3D printing selective laser melting technology according to claim 2, characterized in that: The surface tension in the surface tension model in step S4 is: ； in, is the surface tension, is the surface tension coefficient, is the surface curvature, is the unit vector normal to the free interface, is the average density of each grid, is the average density, , , represent the volume fraction of the metal phase and the volume fraction of the gas phase, respectively. is the metal density, is the gas density, is the modulus of the metal volume fraction.

Images