High-fidelity additive manufacturing stress-strain simulation method

Abstract

The invention discloses a high-fidelity additive manufacturing stress-strain simulation method, and belongs to the technical field of additive manufacturing simulation, and the method comprises the steps: firstly, loading the temperature calculation result data of CFD into a grid model of an FEM to obtain an FEM temperature field, loading the FEM temperature field into stress analysis in the form of a predefined field to obtain a thermal-mechanical coupling model, and carrying out thermal-mechanical coupling on the thermal-mechanical coupling model to obtain a thermal-mechanical coupling model; calculating a stress-strain process based on a thermal-mechanical coupling model; in the calculation process, model nodes are endowed with solid, liquid or gaseous attributes through a static unit method, and the influence of different attribute nodes on stress and strain is considered, so that the whole calculation process can be close to a real simulation process. According to the method, high-fidelity transmission of laser additive manufacturing molten pool morphology and temperature field data information is achieved, the accuracy of a stress-strain calculation model is effectively improved, meanwhile, the data processing difficulty is reduced, and the calculation efficiency is improved.

Description

Technical Field

[0001] The present invention belongs to the field of additive manufacturing simulation, and in particular to a high-fidelity additive manufacturing stress-strain simulation method. Background Art

[0002] Additive manufacturing technology is an advanced technology for manufacturing parts based on the idea of ​​"discrete-stacking". The technical characteristics of rapid mold-free free forming enable it to manufacture almost any complex geometric structure. However, the technical characteristics of rapid non-equilibrium solidification and repeated thermal cycles bring huge challenges to the high-quality rapid manufacturing of parts, which seriously restricts the development of this technology. "Shape control" and "property control" are two key issues. Shape control refers to obtaining formed parts with high dimensional accuracy, which mainly involves the problem of thermal deformation control during the forming process; while property control refers to obtaining excellent mechanical properties, which mainly involves rapid non-equilibrium solidification and repeated thermal cycle solid-state phase change. Among them, the shape control problem is the main problem that needs to be solved in engineering, that is, how to perform high-precision forming and manufacturing in the face of three-dimensional models with different geometric structures.

[0003] Traditional experimental methods used to verify the planning and optimization of part forming strategies are expensive, but advanced numerical models have been proven to be able to explain various physical phenomena in the additive manufacturing process. These models are widely used to evaluate the quality of additively manufactured parts and reveal the defect formation mechanism and deformation law. Simulation makes up for the inability to conduct expensive and difficult in-situ experimental tests, which is of great significance to saving the production cost of additive manufacturing and reducing repetitive process experiments.

[0004] Laser additive manufacturing involves a series of complex metallurgical and solid-state phase change processes such as melting, rapid solidification and thermal cycling of powder particles. Its typical non-equilibrium and nonlinear characteristics determine that the process stability and forming geometric accuracy control of the process have become research difficulties. By studying the evolution mechanism of thermal stress in the manufacturing process, a theoretical reference can be provided for alleviating residual stress and deformation cracking of parts. The finite element (FEM) simulation method can simulate and calculate the temperature field and stress field of the manufacturing process, and has been widely used in predicting the stress and strain distribution of macroscopic parts, but it cannot accurately reflect the morphology of the molten pool and the heat source information, which greatly reduces the accuracy of the stress and strain predicted by the model calculation. In contrast, the computational fluid dynamics (CFD) simulation method can more accurately reflect the complex morphology, velocity field, and temperature field information of the molten pool evolution process by considering the joint action of multiple physical fields, but its computational complexity limits the ability of this method to calculate larger-scale models. In addition, the Euler grid of fluid mechanics is not suitable for the calculation of structural mechanics, and thus cannot solve the stress and strain of the model. Therefore, CFD-FEM coupling calculation has become a new idea for high-fidelity simulation of the molten pool in additive manufacturing, providing a new method for solving shape control problems.

[0005] Based on the CFD-FEM coupling method, some scholars have carried out some related research work. The publicly published High-fidelity modeling of thermal stress for additive manufacturing by linking thermal-fluid and mechanical models (Materials&Design, 2020, 11) uses the precise temperature field and melt pool morphology extracted from the CFD calculation results to build a FEM stress-strain simulation calculation model with the static element method to solve the stress-strain evolution near the cladding pass. The model accurately reproduces the influence of laser additive manufacturing melt pool morphology and temperature on stress-strain distribution, but the time and space-based interpolation algorithm code written when transferring temperature and melt pool morphology data is complex and consumes a lot of computing resources. The single cladding pass model alone consumed 330 hours (6 cores, 12 threads, 2.66GHz). The publicly published paper Laser direct deposition of AISI H13 tool steelpowder with numerical modeling of solid phase transformation, hardness, and residual stresses (Journal of Materials Processing Technology, 2017, 9) uses the direct energy deposition molten pool contour and temperature distribution information obtained in the experiment, combined with the birth and death unit method to build a FEM residual stress calculation model to predict the evolution of residual stress and phase change driving force during the additive manufacturing process. The model accurately reproduces the dynamic evolution of the molten pool contour during the additive manufacturing process, but the birth and death unit method cannot re-kill the activated units, so it cannot reflect the laser remelting effect and keyhole effect, and cannot perform high-fidelity simulation of the molten pool.The publicly published Thermal Metallurgical Analysis of GMA Welded AH36 Steel Using CFD–FEMframework (Materials&Design, 2016, 2) uses the welding temperature field information calculated by CFD to couple the temperature data to the FEM model by linear interpolation, and uses its high solution efficiency to predict the heat dissipation and thermal stress evolution process of the weld. The model retains the mass transfer and heat transfer temperature field data calculated by CFD, providing high-precision temperature loads for subsequent cooling calculations. However, it only performs average time-space interpolation processing when mapping the temperature data of the Euler grid nodes to the Lagrangian grid nodes. When processing the data near the molten pool, this method will introduce the low-temperature values ​​of the air into the high-temperature area of ​​the liquid molten pool, resulting in a large interface temperature error.

[0006] In summary, the existing CFD-FEM coupling simulation model cannot retain the high-precision melt pool morphology and heat source information of CFD well during the data transmission process. At the same time, the data transmission efficiency is low, the data processing code is lengthy and consumes a lot of computing resources. Summary of the invention

[0007] The purpose of the present invention is to overcome the shortcomings of the above-mentioned prior art and provide a high-fidelity additive manufacturing stress-strain simulation method to solve the problems in the prior art of the CFD-FEM coupling simulation model, in which the processing performed in the process of transmitting data is difficult to retain the high-precision melt pool morphology and heat source information of CFD, and at the same time, the data transmission efficiency is low, the data processing code is lengthy and consumes a large amount of computing resources.

[0008] In order to achieve the above object, the present invention adopts the following technical solutions:

[0009] A high-fidelity additive manufacturing stress-strain simulation method comprises the following steps:

[0010] Step 1, obtaining a CFD calculation result file of the workpiece, wherein the CFD calculation result file includes the node coordinates and node temperatures of the CFD mesh unit, and generating a FEM mesh geometry model based on the node coordinates;

[0011] Step 2, assign thermal analysis related material parameters to the FEM mesh geometry model, set the initial temperature field of the FEM mesh geometry model, perform simulation calculations through FEM without loading heat source loads, and obtain a FEM blank heat transfer analysis result file;

[0012] Step 3, assigning the node temperature of the CFD grid unit to the FEM blank heat transfer analysis result file, sorting the node temperature by time, and obtaining the FEM temperature field file sorted by time;

[0013] Step 4, assigning material parameters related to the thermal-mechanical coupling analysis to the mesh geometry model of the FEM in step 1, creating a static analysis step in the FEM, loading the FEM temperature field file sorted by time into the static analysis step in the form of a predefined field, and obtaining a thermal-mechanical coupling FEM model;

[0014] Step 5, using the temperature field and thermal history of the grid unit in the thermomechanical coupling FEM model as a criterion, judging the properties of the node, wherein the properties of the node include gas material properties, liquid material properties and solid material properties; assigning gaseous parameters, liquid parameters and solid parameters to the nodes of gas material properties, liquid material properties and solid material properties respectively; the solid parameter is greater than the liquid parameter, and the liquid parameter is greater than the gas parameter, thereby obtaining a thermomechanical coupling FEM model with assigned material properties;

[0015] Step 6, performing stress and strain calculation on the thermomechanical coupling FEM model assigned with material properties to obtain the stress and strain field. Each temperature field that changes with time has a corresponding stress and strain field.

[0016] A further improvement of the present invention is:

[0017] Preferably, in step 1, the process of generating the mesh geometry model of FEM based on the node coordinates is: constructing the FEM Lagrangian mesh unit node coordinates corresponding to the CFD Euler mesh unit node coordinates, constructing hexahedral mesh units through adjacent nodes, and obtaining the mesh geometry model of FEM.

[0018] Preferably, in step 2, the mesh geometry model of the FEM is transformed into an assembly model, and thermal analysis related material parameters are assigned to the assembly model; the thermal analysis related material parameters include density, specific heat capacity and thermal conductivity.

[0019] Preferably, in step 2, the temperature of the FEM mesh geometry model is set to room temperature.

[0020] Preferably, after step 3, temperature data irrelevant to stress-strain calculation is eliminated from the FEM temperature field file sorted by time; the temperature data irrelevant to stress-strain includes node temperature data exceeding material vaporization and abnormal temperature data.

[0021] Preferably, in step 4, the mesh geometry model of the FEM is transformed into an assembly model, and the assembly model is assigned material parameters related to the thermal-mechanical coupling analysis; the material parameters related to the thermal-mechanical coupling analysis include thermal expansion coefficient, Young's modulus and Poisson's ratio.

[0022] Preferably, in step 4, in the thermal-mechanically coupled FEM, completely fixed constraints are applied to the unit nodes on the bottom surface of the substrate.

[0023] Preferably, in step 5, the solid-state parameter is an order of magnitude multiple of the liquid-state parameter, and the liquid-state parameter is an order of magnitude multiple of the gas-state parameter.

[0024] Preferably, in step 5, the process of determining the attributes of the node is:

[0025] S1, using the coordinates of the nodes as a criterion, obtaining the substrate in the thermal-mechanical coupling FEM model, and assigning solid material properties to the substrate in the FEM model;

[0026] In S2, the node temperature is used as the criterion to select nodes that are always at room temperature from the thermal-mechanical coupling FEM model and assign gas material properties; nodes that exceed the melting point of the powder are selected and assigned liquid material properties;

[0027] In S3, the thermal history of the node is used as a criterion. In the thermal-mechanical coupling FEM model, the melted powder particle nodes are assigned solid material properties, and the unmelted powder particle nodes are assigned gas material properties.

[0028] Preferably, in S3, the process of determining the attributes of the node is completed by a subroutine, and the subroutine is written by FEM simulation software.

[0029] Compared with the prior art, the present invention has the following beneficial effects:

[0030] The present invention discloses a high-fidelity additive manufacturing stress and strain simulation method. The simulation method first directly applies the CFD temperature field calculation results to the FEM mesh geometry model. Compared with the traditional FEM laser additive manufacturing thermal stress and strain simulation calculation model, the CFD temperature field calculation results provide irregular and time-varying heat source models and molten pool geometry, and the boundary conditions of stress and strain calculation are more realistic. Then, in view of the problem that the birth and death unit method used in the laser additive manufacturing stress and strain calculation cannot kill the activated unit, so that the remelting and keyhole effect under the action of the laser cannot be simulated, the present invention introduces the static element method in the thermal-mechanical coupling FEM model. The static element method distinguishes the properties of the unit in the stress and strain calculation process, and assigns solid parameters of large order of magnitude to the solid unit that has a large influence on the stress and strain calculation results, and assigns gas-liquid parameters with small influence to the gas-liquid unit that has little or no influence on the stress and strain calculation results, thereby avoiding the problem of numerical calculation convergence difficulties caused by repeated activation and killing of units. The present invention realizes high-fidelity transmission of laser additive manufacturing molten pool morphology and temperature field data information, effectively improves the accuracy of the stress-strain calculation model, and reduces the difficulty of data processing.

[0031] Furthermore, compared with the traditional static element method, which does not distinguish the role of unmelted powder particles in the stress and strain calculation process of laser additive manufacturing and directly participates in the calculation, this model screens them out and assigns gas material stiffness parameters based on the criterion of node temperature history, so that they will not have a major impact on the calculation results of the thermal stress and strain of the model, and the model building logic is more perfect.

[0032] Furthermore, the present invention uses the software provided by FEM for programming, which further lowers the threshold for users to master programming technology compared to traditional CFD-FEM, and reduces the possibility of introducing unnecessary errors when writing time and space interpolation algorithms. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is a flow chart of the high-fidelity additive manufacturing stress-strain simulation method of the present invention;

[0034] Figure 2 The FEM simulation calculation grid model of the present invention;

[0035] Figure 3 This is the model diagram of FEM temperature field;

[0036] Among them, (a) is the morphology of the molten pool; (b) is the temperature distribution diagram in the laser scanning direction;

[0037] Figure 4 It is the model diagram of FEM stress field;

[0038] Among them, (a) shows the morphology of the cladding track; (b) shows the stress distribution in the laser scanning direction. DETAILED DESCRIPTION

[0039] The present invention is further described in detail below in conjunction with the accompanying drawings:

[0040] In the description of the present invention, it should be noted that the orientations or positional relationships indicated by the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inside", "outside", etc. are based on the orientations or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention; the terms "first", "second", and "third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance; in addition, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection or a detachable connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the internal communication of two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0041] See also Figure 1 The first aspect of the present invention discloses a high-fidelity additive manufacturing stress-strain simulation method, which specifically comprises the following steps:

[0042] Step 1: perform CFD calculation on the workpiece, analyze the CFD calculation result file, extract the node coordinate data and temperature data therein, and generate the FEM model file using the node coordinate data.

[0043] The specific process is as follows: construct the FEM model file according to the node coordinates and temperature data result files calculated by CFD, build the FEM Lagrangian mesh unit node coordinates corresponding to the CFD Euler mesh unit node coordinates, and number the nodes. Use the search algorithm to find each node coordinate, build a hexahedral mesh unit according to the position relationship of the adjacent node coordinates, create a hexahedral mesh unit for every six adjacent unit nodes, specify the unit number, and establish the index relationship between the unit number and the node number to obtain the geometric model; output the constructed geometric model in a file format that can be read by the FEM software to realize the automated modeling of the FEM mesh geometric model.

[0044] Step 2: Create a heat transfer analysis job based on the FEM model file and obtain a blank result file.

[0045] The specific process is: input the geometric model file containing mesh units into the FEM simulation software, set the temperature field analysis unit type for the geometric model containing mesh units, create an assembly model, and assign FEM thermal analysis related material parameters to the assembly model, including density, specific heat capacity, and thermal conductivity. Create a heat conduction analysis step, set a predefined temperature field for the entire assembly model assigned with material parameters, set an arbitrary temperature value for the predefined temperature field, do not apply any heat source load to the model, set conventional convection heat transfer conditions, create an analysis job and submit the solution calculation, thereby obtaining a result file without valid temperature data, which is a FEM blank heat transfer analysis result file.

[0046] As a preferred solution, the temperature set in the predefined temperature field is room temperature, because room temperature is the temperature closest to the ambient temperature. When the temperature field is initially established, the model can be free of heat transfer, thus saving calculation time.

[0047] Step 3, using the CFD temperature data to modify the FEM blank heat transfer analysis result file, to obtain a result file of the FEM temperature field calculation data sorted by time.

[0048] Modify the temperature data calculation result file. The specific process is: assign the temperature data based on the node coordinates in the CFD calculation results to the node coordinates in the corresponding FEM temperature calculation result file, so as to construct a result file containing valid FEM temperature field calculation data, and sort the constructed FEM result temperature field data at different times according to the CFD calculation temperature field time and insert them into the corresponding time to obtain the FEM temperature field file sorted by time.

[0049] Step 4: Create FEM static analysis work, couple the temperature field calculation results, and create different field variable unit classifications.

[0050] The geometric model file containing the mesh unit is input into the FEM simulation software, the stress analysis unit type is set for the mesh model, the assembly model is created, and the thermal coupling analysis related material parameters required for the FEM thermal coupling analysis are assigned to the model, including the thermal expansion coefficient, Young's modulus and Poisson's ratio. A general static analysis step is created in the FEM software, and the constructed time-ordered FEM temperature field file is loaded into the stress analysis step in the form of a predefined field, in preparation for the subsequent solution of the model stress-strain results using temperature as the load, and a completely fixed constraint is imposed on the unit nodes corresponding to the bottom surface of the substrate in the model.

[0051] Step 5, using the static element method to solve the stress and strain values, the specific process is to use the subroutine of the FEM simulation software, divide the grid units into three categories based on the temperature field data and thermal history of the unit, assign different field variables to the nodes in the model, and assign different material properties according to the unit field variable values. Among them, the thermal history of the unit is the temperature change process of the unit node, which is used to determine whether the node has melted.

[0052] Specifically, the first category corresponds to the air and unmelted powder particle nodes in the CFD calculation model. They have no effect on the stress calculation of the cladding track in the stress-strain calculation, and need to be assigned gas material properties. The corresponding setting field variable is 1, and the gaseous parameters are assigned at the same time; the second category is the liquid nodes inside the molten pool. They have little effect on the stress-strain calculation of the cladding track, and need to be assigned liquid material properties and liquid parameters. The field variable of the unit is set to 2; the third category is the nodes located on the substrate and the powder particle units in the cladding track that have been melted and deposited on the substrate. As the main participating nodes in the stress-strain calculation of the cladding track, they need to be assigned solid material properties, correspondingly assigned solid parameters, and the field variable is set to 3. Among them, the solid parameters are greater than the liquid parameters, and the liquid parameters are greater than the gas parameters. The liquid and gas nodes are in a secondary position in the stress calculation. Therefore, the gas parameters and liquid parameters only need to meet the model calculation convergence requirements, and should have a clear order of magnitude difference from the solid parameters of the solid nodes that mainly participate in the stress calculation. There is also a clear order of magnitude difference between the liquid parameters and the gas parameters, so that the gas nodes and liquid nodes can ensure the overall convergence of the model when calculating the stress and strain of the solid unit without having a significant impact on the stress and strain calculation results.

[0053] The subroutine containing the field variable control criteria is associated with the interface reserved for user subroutines in the FEM solver to perform analysis work.

[0054] Specifically, based on the temperature field data and thermal history of the node, the process of node division is as follows:

[0055] (1) Using the node Z coordinate as the criterion, the unit nodes with Z < the set value are screened out. The units in this part correspond to the substrate in the model and always participate in the stress calculation. Solid material properties are given to such nodes, and field variables 3 are assigned to them.

[0056] (2) Using temperature as the criterion, select nodes that are always at room temperature, assign gas material properties to such nodes, and allocate field variable 1 to them; using temperature as the criterion, select nodes that exceed the melting point of the powder, assign liquid material properties to such nodes, and allocate field variable 2 to them.

[0057] (3) Based on the thermal history, the melted nodes in the powder particle nodes are recorded and their Boolean values ​​are marked as 1, and the unmelted powder particle nodes are marked as 0, so as to distinguish the unit properties; for the melted powder particle nodes, when the temperature drops below the melting point of the material, their solid field variable 3 is assigned; on the contrary, for the unmelted powder particle nodes, their gas field variable 1 is assigned.

[0058] (4) Finally, the various conditional criteria for node classification are combined and written into a code that can be run by FEM simulation software using a subroutine.

[0059] Step 6: Submit the solution work to obtain the influence of the molten pool morphology on the thermal stress distribution. During the solution process, the solution is performed using hexahedral grid units. After the calculation is completed, a high-fidelity FEM stress-strain calculation result based on the CFD temperature field calculation result is obtained.

[0060] In some implementation modes of the present invention, a filtering process is further provided after step 3.

[0061] Specifically, the temperature analysis results of the structure are viewed in the FEM analysis software, and the temperature data transmitted by CFD is further filtered to eliminate temperature data that is meaningless for stress-strain calculations, such as node temperature data that exceeds the material vaporization point, and abnormal temperature values ​​in the CFD temperature field calculation results caused by unstable calculation values. Specific abnormal temperature values ​​are data points that far exceed reasonable temperature values.

[0062] Furthermore, the temperature data can be filtered repeatedly to generate temperature loads suitable for thermal-mechanical coupling analysis.

[0063] The implementation of the technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments:

[0064] Example 1

[0065] Step 1, see Figure 2 , automatically generates FEM analysis grid model based on CFD grid model node coordinate data.

[0066] First, analyze the CFD result data, write a search algorithm, record the maximum and minimum values ​​of the CFD grid node coordinate data x, y, z, and use the CFD hexahedral grid with a side length of 5μm as the seed length within this range to divide the FEM grid. Based on the node coordinate data provided by CFD, establish 800,000 FEM grid nodes with corresponding coordinates. Then, number each node, establish a table of node coordinates indexed by the node number, and output it to the node part in the data file. Next, based on the node coordinate position relationship, query the numbers of the 8 nodes that can form a hexahedral unit, build an index relationship between the node numbers of the hexahedral units, and build 770,329 grid units based on 800,000 nodes, such as Figure 1 Finally, the node numbers in the grid are stored in the counterclockwise direction. The node coordinates are stored in sequence. The unit number is followed by the node numbers of the constituent units stored according to the right-hand rule. A table is established with the unit number as the index to associate the node number. The table is then output to the unit part of the data file, thus completing the automated modeling of the entire FEM grid model.

[0067] Step 2: Based on the constructed mesh model, use FEM simulation software to generate a blank temperature field data result file.

[0068] First, the geometric model file containing the mesh unit is input into the FEM simulation analysis software, the mesh unit is assigned the heat transfer analysis unit attribute, and the assembly model is generated based on the mesh unit component model; then, the material parameters required for heat transfer analysis are assigned to the assembly model, including: material density, specific heat capacity and thermal conductivity. Here, the material for analysis is set to AlSi10Mg; then, a heat conduction analysis step is created, and an analysis step is set. The analysis step and incremental step are both 1μs. In the predefined temperature field, all unit nodes are assigned an initial temperature value of 26℃ at room temperature. No heat load is applied to the model, and conventional convection heat transfer conditions and an ambient temperature of 26℃ are set; finally, a heat conduction analysis job is created and submitted to the FEM solver for solution, and a result file without valid temperature field data is obtained.

[0069] Step 3: Build the FEM temperature field calculation result file based on the CFD calculation temperature field result data.

[0070] First, the temperature values ​​in the fourth column of the CFD temperature field calculation result file from 0 to 60 μs are read in sequence according to the coordinate information and stored in the computer memory; then, the 2nd to 13th frames of result data are constructed in the previously generated empty FEM temperature field data result file in the order of outputting one frame of result data every 5 μs; then, the node temperature field data in the memory are assigned to the FEM temperature field result grid unit nodes of the corresponding time and coordinates according to the CFD calculation time and the node coordinate index, and the newly generated temperature field data is saved to the result file; finally, the temperature distribution cloud map is drawn using the grid interpolation algorithm provided by the FEM software, and the animation of the temperature field evolution process after the imported temperature field data is visually rendered is viewed in the software.

[0071] Step 4: Analyze the constructed FEM temperature field result data and filter the node temperature data that is not suitable for temperature field calculation.

[0072] First, observe the visualized temperature field data cloud map frame by frame to observe the distribution position and changes of the unit nodes that exceed the material vaporization temperature data. Then, the node temperature value that exceeds the material vaporization temperature is modified to 1107°C, that is, the vaporization temperature, based on the temperature as the criterion. At the same time, the splash caused by the laser scanning process, that is, the high-temperature unit node temperature in the air is modified to room temperature 26°C based on the temperature node coordinates. Next, eliminate the unit node temperature values ​​below room temperature in the calculation caused by the numerical instability of the CFD calculation process, that is, modify the node temperature data below room temperature in the model temperature field to 26°C. Finally, load the filtered temperature field result file again, view the processed temperature field cloud map in the FEM software visualization module, and check the filtering quality.

[0073] like Figure 3 As shown in the figure, the filtered FEM temperature field data retains the high-fidelity morphology of the molten pool while eliminating the temperature load that is irrelevant to the stress-strain calculation, which not only improves the convergence of the model calculation, but also more accurately reflects the physical phenomena such as the molten pool keyhole and remelting.

[0074] Step 5: Set up stress-strain solution in FEM simulation analysis software.

[0075] First, the geometric model file containing the mesh unit is input into the FEM simulation analysis software, the mesh unit is assigned the static analysis unit attributes, and the assembly model is generated based on the mesh unit component model. Then, the material parameters required for static analysis are assigned to the assembly model, including: material thermal expansion coefficient, Young's modulus, Poisson's ratio, and the material analyzed here is set to AlSi10Mg. Next, a general static analysis step is created, an analysis step is set, the analysis step is 60μs, the incremental step is a variable incremental step, the minimum value is 1μs, and the maximum value is 2μs. The temperature field calculation result file is coupled to the predefined field of the static analysis step, and the 2nd to 13th frame data of the temperature field calculation result is imported. The model stress and strain are solved with temperature as the load. Finally, a completely fixed constraint is applied to the unit node corresponding to the bottom surface of the substrate in the model.

[0076] Step 6: Use subroutines to classify different units in the model so that the model can adapt to the requirements of solving stress and strain using the static element method.

[0077] First, the units in the model are divided into three categories according to their material stiffness properties, namely gas parameters, liquid parameters and solid parameters. Different field variables 1, 2 and 3 are assigned to the units in the three states. The units with solid parameter field variables are mainly involved in stress-strain calculations, and the material parameters of other units only need to meet the calculation convergence requirements. Then, the Z coordinate of the unit node is used as the criterion to screen out the units with Z < -4.25×10 -5 The unit nodes of this part correspond to the substrate in the model and are always involved in the stress calculation. The field variable 3 is assigned to them. Taking temperature as the criterion, the gas units that are always at room temperature are screened out and the field variable 1 is assigned to them. Taking temperature as the criterion, the unit nodes that exceed the powder melting point of 580°C are screened out and the field variable 2 is assigned to them. Then, taking thermal history as the criterion, the unit nodes that have exceeded the melting point in the powder particle units are recorded and their Boolean values ​​are marked as 1. The powder particle unit nodes that have not melted are marked as 0. When the melted powder particle units are cooled to below the material melting point, their solid-state field variable 3 is assigned. On the contrary, the unmelted powder particle units are assigned their gaseous field variable 1. Finally, the various conditional criteria for unit classification are combined and written into a code that can be run by FEM simulation calculation software using a subroutine.

[0078] Step 7, associate the subroutine code with the FEM static analysis main program, create an analysis job, and obtain high-fidelity stress and strain calculation results.

[0079] First, a general statics analysis task is created, and the written subroutine is associated with the interface reserved for user subroutines in the FEM solver; then, the Intel(R) Xeon(R) W-2245CPU (4 cores, 3.9GHz) is used to calculate for 3.5 hours to obtain high-fidelity stress and strain calculation results of the model; then, the calculation results are viewed in the FEM simulation software view module, and in the post-processing module, a unit grouping is created with the field variable unit result value interval of 2 to 3, and the model stress and strain calculation results are viewed under this unit grouping condition; finally, a segmented view is created to view the evolution of the cladding path morphology with the movement of the laser heat source, as well as the distribution state of stress and strain.

[0080] like Figure 4 As shown in the figure, the stress result cloud map calculated by the static element method retains the edge morphology of the irregular cladding channel intact, the stress calculation result is reasonable, and the dynamic evolution process of the stress distribution of the laser additive manufacturing molten pool with the movement of the heat source is predicted more accurately.

[0081] This method is based on the CFD-FEM coupling method, extracts the CFD calculation of the high-fidelity molten pool morphology and heat source temperature field distribution of laser additive manufacturing, and builds a FEM thermal stress and strain high-fidelity calculation model. Combined with the static element method, the influence of complex physical phenomena such as remelting and keyhole in the laser additive manufacturing process on the stress and strain calculation of the cladding channel is introduced. At the same time, the time-space interpolation algorithm provided by the FEM simulation software is used to greatly simplify the workload of CFD model temperature field data processing, and ingeniously realize the thermal-mechanical coupling calculation of laser additive manufacturing. Therefore, this method effectively improves the accuracy of stress and strain simulation calculation of laser additive manufacturing, and opens up a new path for high-fidelity simulation modeling of laser additive manufacturing.

[0082] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A high-fidelity additive manufacturing stress-strain simulation method, characterized in that: The following steps are involved: Step 1, obtaining a CFD calculation result file of the workpiece, wherein the CFD calculation result file includes the node coordinates and node temperatures of the CFD mesh unit, and generating a FEM mesh geometry model based on the node coordinates; Step 2, assign thermal analysis related material parameters to the FEM mesh geometry model, set the initial temperature field of the FEM mesh geometry model, perform simulation calculations through FEM without loading heat source loads, and obtain a FEM blank heat transfer analysis result file; Step 3, assigning the node temperature of the CFD grid unit to the FEM blank heat transfer analysis result file, sorting the node temperature by time, and obtaining the FEM temperature field file sorted by time; Step 4, assigning material parameters related to the thermal-mechanical coupling analysis to the mesh geometry model of the FEM in step 1, creating a static analysis step in the FEM, loading the FEM temperature field file sorted by time into the static analysis step in the form of a predefined field, and obtaining a thermal-mechanical coupling FEM model; Step 5, using the temperature field and thermal history of the grid unit in the thermomechanical coupling FEM model as a criterion, judging the properties of the node, wherein the properties of the node include gas material properties, liquid material properties and solid material properties; assigning gaseous parameters, liquid parameters and solid parameters to the nodes of gas material properties, liquid material properties and solid material properties respectively; the solid parameter is greater than the liquid parameter, and the liquid parameter is greater than the gas parameter, thereby obtaining a thermomechanical coupling FEM model with assigned material properties; Step 6, performing stress and strain calculation on the thermomechanical coupling FEM model assigned with material properties to obtain the stress and strain field. Each temperature field that changes with time has a corresponding stress and strain field.

2. A high-fidelity additive manufacturing stress-strain simulation method according to claim 1, characterized in that: In step 1, the process of generating the mesh geometry model of FEM based on the node coordinates is as follows: constructing the FEM Lagrangian mesh unit node coordinates corresponding to the CFD Euler mesh unit node coordinates, constructing hexahedral mesh units through adjacent nodes, and obtaining the FEM mesh geometry model.

3. A high-fidelity additive manufacturing stress-strain simulation method according to claim 1, characterized in that: In step 2, the mesh geometry model of the FEM is transformed into an assembly model, and thermal analysis related material parameters are assigned to the assembly model; the thermal analysis related material parameters include density, specific heat capacity and thermal conductivity.

4. A high-fidelity additive manufacturing stress-strain simulation method according to claim 1, characterized in that: In step 2, the temperature of the FEM mesh geometry model is set to room temperature.

5. The high-fidelity additive manufacturing stress-strain simulation method according to claim 1, characterized in that: After step 3, the temperature data irrelevant to the stress-strain calculation in the FEM temperature field file sorted by time is eliminated; the temperature data irrelevant to the stress-strain includes the node temperature data exceeding the material vaporization and the abnormal temperature data.

6. A high-fidelity additive manufacturing stress-strain simulation method according to claim 1, characterized in that: In step 4, the mesh geometry model of the FEM is transformed into an assembly model, and the assembly model is assigned material parameters related to the thermal-mechanical coupling analysis; the material parameters related to the thermal-mechanical coupling analysis include thermal expansion coefficient, Young's modulus and Poisson's ratio.

7. A high-fidelity additive manufacturing stress-strain simulation method according to claim 1, characterized in that: In step 4, in the thermal-mechanical coupled FEM, fully fixed constraints are applied to the element nodes on the bottom surface of the substrate.

8. The high-fidelity additive manufacturing stress-strain simulation method according to claim 1, characterized in that: In step 5, the solid state parameter is an order of magnitude multiple of the liquid state parameter, and the liquid state parameter is an order of magnitude multiple of the gas state parameter.

9. A high-fidelity additive manufacturing stress-strain simulation method according to claim 1, characterized in that: In step 5, the process of determining the attributes of the node is: S1, using the coordinates of the nodes as a criterion, obtaining the substrate in the thermal-mechanical coupling FEM model, and assigning solid material properties to the substrate in the FEM model; In S2, the node temperature is used as the criterion to select nodes that are always at room temperature from the thermal-mechanical coupling FEM model and assign gas material properties; nodes that exceed the melting point of the powder are selected and assigned liquid material properties; In S3, the thermal history of the node is used as a criterion. In the thermal-mechanical coupling FEM model, the melted powder particle nodes are assigned solid material properties, and the unmelted powder particle nodes are assigned gas material properties.

10. The high-fidelity additive manufacturing stress-strain simulation method according to any one of claims 1 to 9, characterized in that: In S3, the process of determining the attributes of the node is completed through a subroutine, and the subroutine is written through FEM simulation software.

Images