An efficient numerical simulation method and system for dynamic behavior of laser processing molten pool-keyhole

Abstract

The present invention belongs to the field of digital twin technology in laser manufacturing, and discloses a method and system for efficient numerical simulation of the dynamic behavior of a laser-processed molten pool and keyhole. In response to the problem of tight computing resources for large-scale simulation of full welds, the present invention adopts dynamic sliding technology to reduce the large computing domain and reduce the computing resource demand. In response to the information confusion problem caused by high-dimensional data mixed communication of LBM, dimensionality reduction communication technology is adopted to expand the high-dimensional data into one dimension, thereby ensuring efficient and unified communication between parallel processes. In response to the problem of difficulty in tracking the interface position, a serial-parallel hybrid method is adopted in the program architecture to collect the phase field sequence parameters in the sub-process into the main process to achieve accurate tracking of the interface position. Based on the lattice Boltzmann method (LBM), which is easy to parallelize, has a solid physical foundation and higher simulation accuracy, the present invention combines the MPI parallel computing technology to perform numerical simulation on the evolution process of large-scale molten pool and keyhole in laser processing of full welds. Through efficient and accurate calculation and communication, numerical simulation of large-scale laser processing of full welds becomes possible, providing technical support for the prediction of long weld morphology of large components.

Description

Technical Field

[0001] The present invention belongs to the field of laser manufacturing digital twin technology, and in particular relates to an efficient numerical simulation method and system for the dynamic behavior of a laser processing molten pool-keyhole. Background Art

[0002] Laser processing is an advanced manufacturing technology that integrates optics, machines, and electronics. It has the advantages of high energy density, fast processing speed, and good processing flexibility, making it an important means of manufacturing metal components. Laser processing can achieve rapid melting and solidification of metal materials, while ensuring processing quality while also significantly improving the production efficiency of the processing process. It provides effective support for the high-quality and efficient manufacturing of high-performance components of major equipment in the fields of aerospace, rail transportation, and marine ships. However, the interaction process between high-energy laser beams and metal materials is complex, especially in the molten pool-keyhole region where nonlinear thermal-mechanical-fluid interactions exist. The evolution mechanism of its dynamic behavior is unclear, resulting in a certain degree of blindness in process control.

[0003] At present, the research on the dynamic behavior of the molten pool-keyhole in laser processing is mainly divided into experimental observation methods and digital twin numerical simulation methods. The experimental observation method is based on acoustic, optical, and electrical sensors to collect various signals during the processing process. It has the following shortcomings: (1) It is difficult for the experimental device to observe the heat transfer and flow data inside the molten pool-keyhole; (2) It is impossible to achieve data acquisition at the same time with high temporal resolution and high spatial resolution; (3) The collected signals are easily interfered by the environmental noise of the processing process, making effective data extraction and analysis difficult. Therefore, the experimental observation method cannot fully and accurately obtain the highly transient and nonlinear thermal evolution process of the molten pool-keyhole during laser processing.

[0004] Digital twin numerical simulation methods analyze the physical evolution of laser processing and establish conservation equations for mass, momentum, and energy, thereby solving the temperature and flow field information of the melt pool-keyhole during the processing process. This method is the core of digital twin technology for laser manufacturing. With the continuous development of computational fluid dynamics, digital twin numerical simulation technology can achieve high-temporal and spatial resolution and high-fidelity simulation of melt pool-keyhole heat transfer and flow. At the same time, the improvement of computer performance and the promotion and application of parallel computing technology have provided effective means to improve the computational efficiency of numerical simulation of the dynamic behavior of the melt pool-keyhole. However, the shortcomings of existing technologies are: with the increasing demand for digital twin numerical simulation of the full-scale laser processing process of large metal components, the number of mesh values ​​and computing resources required for simulation are increasing rapidly. Due to the limitations of existing computer performance such as memory and processor frequency limits, achieving efficient numerical simulation of the dynamic behavior of the melt pool-keyhole in large-scale laser processing is a huge challenge.

[0005] Therefore, the key to studying the dynamic behavior mechanism of the full-scale laser processing melt pool-keyhole lies in how to achieve efficient calculation based on the existing numerical simulation technology. This mainly includes: (1) how to reasonably plan the computational domain of the numerical simulation based on the full-scale calculation requirements in terms of computational domain size optimization; (2) how to speed up the computational efficiency of the numerical simulation process while ensuring the calculation accuracy.

[0006] Through the above analysis, the problems and defects of the existing technology are as follows:

[0007] With the increasing demand for digital twin numerical simulation of the full-scale laser processing process of large metal components, the number of mesh values ​​and computing resources required for simulation have increased linearly. However, due to the limitations of existing computer performance such as memory and processor maximum processing frequency, it is a huge challenge to achieve efficient numerical simulation of the dynamic behavior of the molten pool-keyhole in large-scale laser processing. Summary of the Invention

[0008] In response to the problems existing in the prior art, the present invention provides a method and system for efficiently simulating the dynamic behavior of a laser-processed molten pool-keyhole.

[0009] The present invention is achieved by providing an efficient numerical simulation method for the dynamic behavior of a laser-processed molten pool-keyhole, comprising:

[0010] S1: Molten pool-keyhole morphology extraction and image processing to determine the initial size of the calculation domain;

[0011] S2: Implementation of 3D dynamic sliding computational domain: Using dynamic sliding technology to construct a dynamic constant computational domain that moves along with the laser beam;

[0012] S3: The dynamic behavior numerical simulation model of the laser processing molten pool-keyhole is constructed using the lattice Boltzmann method and the phase field method. The numerical simulation model is divided into parallel computing modules and serial computing modules based on the MPI parallel computing technology.

[0013] The parallel calculation module includes: velocity particle distribution function calculation module, enthalpy particle distribution function calculation module, phase field particle distribution function calculation module and force calculation module; the serial calculation module includes: interface position recognition module, laser beam multiple reflection heat source module and file output module;

[0014] S4: For the parallel computing module, the particle distribution function introduced by the three-dimensional LBM generates four-dimensional data, while scalar quantities such as flow field, temperature field, and phase field are three-dimensional data. Therefore, the calculation process faces the problem of mixed parallel computing of high-dimensional data;

[0015] S5: Post-processing: Based on the calculation results, the molten pool-keyhole morphology and temperature field, velocity field, and phase field distribution during laser processing are visualized through the Tecplot360 GUI interface.

[0016] Furthermore, the molten pool-keyhole morphology extraction and image processing are performed to determine the initial size of the calculation domain;

[0017] S1.1: Capture a video file of the dynamic changes in the molten pool-keyhole morphology during laser processing using high-speed video. Select the molten pool-keyhole morphology during the stable processing state as the input for the image processing step.

[0018] S1.2: Import the image of the steady-state melt pool and keyhole into MATLAB. Preprocess the image captured by the high-speed camera using the ImageProcessing Toolbox in MATLAB: read the image in the folder and change the original image format to RGB format to facilitate subsequent image processing.

[0019] S1.3: Use the grayImage function in the toolkit to grayscale the extracted image; use the gaussianBlur function in the toolkit to perform Gaussian filtering on the image to smooth the image, eliminate Gaussian noise and random noise in the image, improve image quality, and facilitate the subsequent extraction of the characteristic line of the processing pool-keyhole boundary;

[0020] S1.4: Use the Sobel operator to sharpen the image and highlight the boundary features between the melt pool and the keyhole. Extract the longest value Smax and the widest value Wmax of the melt pool and the keyhole during the stable processing, as well as the maximum penetration value Hmax, from the boundary feature image.

[0021] S1.5: Determine the length (L), width (W), and height (H) of the optimal calculation domain for numerical simulation of laser processing as 2*Smax, 2*Wmax, and 1.5*Hmax, respectively, to ensure that the molten pool-keyhole is basically in the middle area of ​​the calculation domain during stable processing.

[0022] Furthermore, the three-dimensional dynamic sliding calculation domain is realized by using the dynamic sliding technology to construct a dynamic constant calculation domain that moves along with the laser beam;

[0023] S2.1: Construct a computational domain based on the computational domain dimensions (length (L), width (W), and height (H) obtained in step S1 above, determine the computational domain time step and grid size, and set the slip judgment condition; extract the distance between the position of the laser beam center and the right boundary of the computational domain at each moment, and perform a slip operation on the computational domain when the slip condition is met;

[0024] S2.2: Perform a sliding operation along the direction of beam movement. According to the set number of grids, delete the computational domain on the left boundary of the computational domain and add a blank computational domain of the same size on the right boundary of the computational domain.

[0025] S2.3: After the sliding operation is completed, the coordinate positions of all parameters in the calculation domain are updated, marking the end of a dynamic sliding operation and ensuring that the size of the calculation domain remains constant.

[0026] Furthermore, for the parallel computing module, the particle distribution function introduced by the three-dimensional LBM brings four-dimensional data, while scalars such as flow field, temperature field, and phase field are three-dimensional data. Therefore, the calculation process has the problem of mixed parallel computing of high-dimensional data;

[0027] S4.1: To accurately transfer high-dimensional mixed data between parallel computing nodes, dimensionality reduction processing is performed on the four-dimensional LBM data and three-dimensional data such as temperature and velocity, and they are uniformly converted into one-dimensional data. This enables accurate transfer of high-dimensional data between global processes.

[0028] S4.2: For the serial calculation module, given the difficulty of cross-process calculation in the location of the molten pool-keyhole gas-liquid interface, single-threaded serial calculation is used in the main process for interface position identification and the laser beam multiple reflection heat source model (including beam tracing algorithm and energy distribution calculation); the main process first collects the phase field data of the sub-process through MPIGather and realizes interface position identification, and then calculates the energy distribution on the interface based on the beam tracing algorithm; finally, the main process is transmitted to each sub-process through MPIBcast broadcast, and at the same time, the temperature, speed and other physical quantities of the sub-process are collected through MPIGather and output to the dat file.

[0029] Furthermore, the dynamic slip judgment condition of the three-dimensional computational domain described in S2.1 is as follows: construct the computational domain based on the computational domain dimensions length (L), width (W) and height (H) obtained in the above step S1, determine the computational domain time step Δt and grid size, and determine the number of grids in the X, Y and Z directions to be N respectively. X =L / Δx, Ny=W / Δx, and =H / Δx, set the dynamic slip judgment condition d=0.2*Nx; extract the distance S between the position of the laser beam center and the right boundary of the calculation domain at each moment; when the slip condition S<=d is met, perform a slip operation on the calculation domain;

[0030] The specific implementation of the dynamic sliding operation of the computational domain described in S2.2 is as follows: when the sliding judgment condition of S2.1 is met, the area of ​​the grid in the computational domain with a length of 0 to 0.25*NX along the X direction, a width of Ny and a height of Nz is removed, and at the same time, the right boundary of the computational domain along the X direction is filled with a length of 0.25*N X, a blank computational domain with a width of Ny and a height of Nz, thus ensuring that the computational domain size is always Nx*Ny*Nz;

[0031] The key variable of the dynamic sliding of the three-dimensional computational domain described in S2.3 is the real-time value S of the distance between the center of the beam and the right boundary of the computational domain. The update formula is as follows:

[0032] S=N x -U χ *Step+m*n χ

[0033] Where m is the number of times the beam cycles to the trigger position (the number of times dynamic slip occurs), the initial value is 0, and the value of m increases by 1 each time a dynamic slip operation is performed; n is the number of grids that dynamically slip along the processing direction. χ =0.25*N X , Ux is the moving speed of the laser beam, Step is the number of real-time calculation steps of the current program; the S value changes dynamically in real time, and the update formula is used to update the S value when the judgment condition is met.

[0034] Furthermore, the dimensionality reduction processing formulas for the LBM four-dimensional data and three-dimensional data in S4.1 are as follows:

[0035] Perform dimensionality reduction on the coordinate values ​​of three-dimensional data:

[0036] vector3D(i,j,k)=k+j*Nz+i*Ny*Nz

[0037] For four-dimensional data:

[0038] vector4D(i,j,k,n)=(k+j*Nz+i*Ny*Nz)*Q+n

[0039] Among them, vector3D and vector4D are the coordinates of three-dimensional and four-dimensional data in one-dimensional data, i, j, and k are the current three-dimensional coordinates of the data, n is the directional component of the particle distribution function, and Q is the number of directional components of the particle distribution function; Ny and Nz are the maximum number of grids in the computational domain in the y and z directions, respectively. It can be proved that the above formula can ensure uniqueness when the high-dimensional coordinate data is expanded into one-dimensional coordinates.

[0040] Another object of the present invention is to provide an efficient numerical simulation system for the dynamic behavior of a laser processing molten pool-keyhole, comprising:

[0041] The image processing module, equipped with MATLAB software and Image Processing Toolbox, is used to import and process laser machining steady-state melt pool-keyhole morphology images, including image format change, grayscale processing, Gaussian filtering, Sobel operator sharpening, and melt pool-keyhole boundary feature extraction;

[0042] The dynamic sliding module uses a decision formula to determine when to perform boundary sliding and uses an update formula to update the S value, thereby achieving dynamic sliding of the three-dimensional calculation domain and ensuring that the size of the calculation domain remains constant during the calculation process;

[0043] The parallel computing module uses LBM simulation technology combined with MPI parallel computing technology to divide a single process into several modules, including: velocity distribution function calculation module, force calculation module, phase field distribution function calculation module, laser energy distribution calculation module, enthalpy distribution function calculation module, and parallel computing adjustment module, thereby completing the numerical realization of the laser processing molten pool-keyhole evolution;

[0044] The efficient communication module uses high-dimensional data dimensionality reduction communication technology to reduce three-dimensional data such as temperature, velocity, and phase field parameters, and four-dimensional data particle distribution functions to one-dimensional data. It uses gradient operators and Laplace operators to conduct three-dimensional data communication tests, and uses particle distribution function migration to conduct four-dimensional data communication tests, ensuring efficient communication of data between different parallel computing processes.

[0045] The serial-parallel hybrid calculation module collects and organizes the phase field parameter data of the sub-process, determines the exact position of the interface in the main process, calculates the energy distribution through the multi-reflection heat source model, and finally broadcasts the energy distribution data of the main process to the sub-process;

[0046] The data output module collects physical quantities such as temperature and speed in the subprocess through the MPI_Gather function and outputs them as dat files;

[0047] The simulation result visualization module uses the visualization function of the Tecplot360 GUI interface to visualize the molten pool-keyhole morphology and temperature field and velocity field distribution during the laser processing process.

[0048] Another object of the present invention is to provide a computer device, which includes a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the efficient numerical simulation method of the dynamic behavior of the laser processing molten pool-keyhole.

[0049] Another object of the present invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, enables the processor to execute the steps of the method for efficient numerical simulation of the dynamic behavior of the laser processing molten pool-keyhole.

[0050] Another object of the present invention is to provide an information data processing terminal, which is used to implement an efficient numerical simulation system for the dynamic behavior of the laser processing molten pool-keyhole.

[0051] In combination with the above technical solutions and the technical problems solved, please analyze the advantages and positive effects of the technical solutions to be protected by the present invention from the following aspects:

[0052] First, in view of the technical problems existing in the above-mentioned prior art and the difficulty of solving these problems, this paper closely combines the technical solutions to be protected by the present invention and the results and data during the research and development process, and analyzes in detail and in depth how the technical solutions of the present invention solve the technical problems and some creative technical effects brought about by solving the problems. The specific description is as follows:

[0053] (1) To address the problem of limited computing resources for large-scale simulation of full welds, dynamic sliding technology is used to reduce the large computational domain and reduce the computing resource requirements. To address the information confusion caused by high-dimensional data hybrid communication in LBM, dimensionality reduction communication technology is used to expand the high-dimensional data into one dimension, ensuring efficient and unified communication between parallel processes. To address the difficulty in tracking interface positions, a serial-parallel hybrid approach is adopted in the program architecture to collect the phase field sequence parameters in the sub-process into the main process, achieving accurate tracking of the interface position.

[0054] (2) The present invention is based on the lattice Boltzmann method (LBM), which is easy to parallelize, has a solid physical foundation and higher simulation accuracy, and combines MPI parallel computing technology to perform numerical simulation of the large-scale molten pool-keyhole evolution process of laser processing of full welds. Through efficient and accurate calculation and communication, the numerical simulation of large-scale laser processing of full welds becomes possible, providing technical support for the prediction of the long weld morphology of large components.

[0055] (3) Compared with the existing experimental methods, the present invention can effectively shorten the research cycle and reduce the research cost; compared with the existing two-dimensional macroscopic molten pool-keyhole evolution simulation and three-dimensional small-scale macroscopic molten pool-keyhole evolution simulation, the present invention can perform large-scale accurate simulation of laser processing of full welds with less computing resources and shorter computing time through dynamic slip, dimensionality reduction communication and serial-parallel mixing. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 This is a flow chart of an efficient numerical simulation method for the dynamic behavior of a laser processing molten pool-keyhole provided by an embodiment of the present invention.

[0057] Figure 2 This is a structural block diagram of an efficient numerical simulation system for the dynamic behavior of a laser processing molten pool-keyhole provided by an embodiment of the present invention.

[0058] Figure 3 This is a diagram of the determination, execution, and update process of dynamic slip in a three-dimensional computational domain for laser processing provided by an embodiment of the present invention.

[0059] Figure 4 Schematic diagram of a parallel computing module provided by an embodiment of the present invention.

[0060] Figure 5 This is a schematic diagram of high-dimensional data dimensionality reduction communication provided by an embodiment of the present invention.

[0061] Figure 6 This is a diagram of a graphical understanding and program architecture using serial-parallel hybrid computing provided by an embodiment of the present invention.

[0062] Figure 7 This is a visualization simulation result diagram of a 3D laser processing molten pool-keyhole provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0063] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0064] like Figure 1 As shown, an embodiment of the present invention provides an efficient numerical simulation method for the dynamic behavior of a laser processing molten pool-keyhole, comprising the following steps:

[0065] S1: Molten pool-keyhole morphology extraction and image processing to determine the initial size of the calculation domain;

[0066] S2: Implementation of 3D dynamic sliding computational domain: Using dynamic sliding technology to construct a dynamic constant computational domain that moves along with the laser beam;

[0067] S3: The dynamic behavior numerical simulation model of the laser processing molten pool-keyhole is constructed using the lattice Boltzmann method and the phase field method. The numerical simulation model is divided into parallel computing modules and serial computing modules based on the MPI parallel computing technology.

[0068] The parallel calculation module includes: velocity particle distribution function calculation module, enthalpy particle distribution function calculation module, phase field particle distribution function calculation module and force calculation module; the serial calculation module includes: interface position recognition module, laser beam multiple reflection heat source module and file output module;

[0069] S4: For the parallel computing module, the particle distribution function introduced by the three-dimensional LBM generates four-dimensional data, while scalar quantities such as flow field, temperature field, and phase field are three-dimensional data. Therefore, the calculation process faces the problem of mixed parallel computing of high-dimensional data;

[0070] S5: Post-processing: Based on the calculation results, the molten pool-keyhole morphology and temperature field, velocity field, and phase field distribution during laser processing are visualized through the Tecplot360 GUI interface.

[0071] The molten pool-keyhole morphology extraction and image processing provided by the embodiment of the present invention determine the initial size of the calculation domain;

[0072] S1.1: Capture a video file of the dynamic changes in the molten pool-keyhole morphology during laser processing using high-speed video. Select the molten pool-keyhole morphology during the stable processing state as the input for the image processing step.

[0073] S1.2: Import the image of the steady-state melt pool and keyhole into MATLAB. Preprocess the image captured by the high-speed camera using the ImageProcessing Toolbox in MATLAB: read the image in the folder and change the original image format to RGB format to facilitate subsequent image processing.

[0074] S1.3: Use the grayImage function in the toolkit to grayscale the extracted image; use the gaussianBlur function in the toolkit to perform Gaussian filtering on the image to smooth the image, eliminate Gaussian noise and random noise in the image, improve image quality, and facilitate the subsequent extraction of the characteristic line of the processing pool-keyhole boundary;

[0075] S1.4: Use the Sobel operator to sharpen the image and highlight the boundary features between the melt pool and the keyhole. Extract the longest value Smax and the widest value Wmax of the melt pool and the keyhole during the stable processing, as well as the maximum penetration value Hmax, from the boundary feature image.

[0076] S1.5: Determine the length (L), width (W), and height (H) of the optimal calculation domain for numerical simulation of laser processing as 2*Smax, 2*Wmax, and 1.5*Hmax, respectively, to ensure that the molten pool-keyhole is basically in the middle area of ​​the calculation domain during stable processing.

[0077] The implementation of the three-dimensional dynamic sliding calculation domain provided by the embodiment of the present invention: using dynamic sliding technology to construct a dynamic constant calculation domain that moves with the laser beam;

[0078] S2.1: Construct a computational domain based on the computational domain dimensions (length (L), width (W), and height (H) obtained in step S1 above, determine the computational domain time step and grid size, and set the slip judgment condition; extract the distance between the position of the laser beam center and the right boundary of the computational domain at each moment, and perform a slip operation on the computational domain when the slip condition is met;

[0079] S2.2: Perform a sliding operation along the direction of beam movement. According to the set number of grids, delete the computational domain on the left boundary of the computational domain and add a blank computational domain of the same size on the right boundary of the computational domain.

[0080] S2.3: After the sliding operation is completed, the coordinate positions of all parameters in the calculation domain are updated, marking the end of a dynamic sliding operation and ensuring that the size of the calculation domain remains constant.

[0081] In the parallel computing module provided by the embodiment of the present invention, the particle distribution function introduced by the three-dimensional LBM generates four-dimensional data, while scalar quantities such as flow field, temperature field, and phase field are three-dimensional data. Therefore, the computing process faces the problem of mixed parallel computing of high-dimensional data.

[0082] S4.1: To accurately transfer high-dimensional mixed data between parallel computing nodes, dimensionality reduction processing is performed on the four-dimensional LBM data and three-dimensional data such as temperature and velocity, and they are uniformly converted into one-dimensional data. This enables accurate transfer of high-dimensional data between global processes.

[0083] S4.2: For the serial calculation module, given the difficulty of cross-process calculation in the location of the molten pool-keyhole gas-liquid interface, single-threaded serial calculation is used in the main process for interface position identification and the laser beam multiple reflection heat source model (including beam tracing algorithm and energy distribution calculation); the main process first collects the phase field data of the sub-process through MPIGather and realizes interface position identification, and then calculates the energy distribution on the interface based on the beam tracing algorithm; finally, the main process is transmitted to each sub-process through MPIBcast broadcast, and at the same time, the temperature, speed and other physical quantities of the sub-process are collected through MPIGather and output to the dat file.

[0084] The dynamic slip judgment condition of the three-dimensional computational domain described in S2.1 provided in the embodiment of the present invention is as follows: construct the computational domain based on the computational domain dimensions length (L), width (W) and height (H) obtained in the above step S1, determine the computational domain time step Δt and grid size, and determine the number of grids in the X, Y and Z directions to be N respectively. X =L / Δx, Ny=W / Δx, and =H / Δx, set the dynamic slip judgment condition d=0.2*Nx; extract the distance S between the position of the laser beam center and the right boundary of the calculation domain at each moment; when the slip condition S<=d is met, perform a slip operation on the calculation domain;

[0085] The specific implementation of the dynamic sliding operation of the computational domain described in S2.2 is: when the sliding judgment condition of S2.1 is met, the grid in the computational domain is moved from 0 to 0.25*N along the X direction. X The area with length, width Ny and height Nz is removed, and the right boundary along the X direction in the calculation domain is filled with a length of 0.25*N X , a blank computational domain with a width of Ny and a height of Nz, thus ensuring that the computational domain size is always Nx*Ny*Nz;

[0086] The key variable of the dynamic sliding of the three-dimensional computational domain described in S2.3 is the real-time value S of the distance between the center of the beam and the right boundary of the computational domain. The update formula is as follows:

[0087] S=N χ -U χ *Step+m*n x

[0088] Where m is the number of times the beam cycles to the trigger position (the number of times dynamic slip occurs), the initial value is 0, and the value of m increases by 1 each time a dynamic slip operation is performed; n is the number of grids that dynamically slip along the processing direction. χ =0.25*N X , Ux is the moving speed of the laser beam, Step is the number of real-time calculation steps of the current program; the S value changes dynamically in real time, and the update formula is used to update the S value when the judgment condition is met.

[0089] The dimensionality reduction processing formulas for LBM four-dimensional data and three-dimensional data in S4.1 provided in the embodiment of the present invention are as follows:

[0090] Perform dimensionality reduction on the coordinate values ​​of three-dimensional data:

[0091] vector3D(i,j,k)=k+j*Nz+i*Ny*Nz

[0092] For four-dimensional data:

[0093] vector4D(i,j,k,n)=(k+j*Nz+i*Ny*Nz)*Q+n

[0094] Among them, vector3D and vector4D are the coordinates of three-dimensional and four-dimensional data in one-dimensional data, i, j, and k are the current three-dimensional coordinates of the data, n is the directional component of the particle distribution function, and Q is the number of directional components of the particle distribution function; Ny and Nz are the maximum number of grids in the computational domain in the y and z directions, respectively. It can be proved that the above formula can ensure uniqueness when the high-dimensional coordinate data is expanded into one-dimensional coordinates.

[0095] like Figure 2As shown, an embodiment of the present invention provides an efficient numerical simulation system for the dynamic behavior of a laser processing molten pool-keyhole, including:

[0096] The image processing module, equipped with MATLAB software and Image Processing Toolbox, is used to import and process laser machining steady-state melt pool-keyhole morphology images, including image format change, grayscale processing, Gaussian filtering, Sobel operator sharpening, and melt pool-keyhole boundary feature extraction;

[0097] The dynamic sliding module uses a decision formula to determine when to perform boundary sliding and uses an update formula to update the S value, thereby achieving dynamic sliding of the three-dimensional calculation domain and ensuring that the size of the calculation domain remains constant during the calculation process;

[0098] The parallel computing module uses LBM simulation technology combined with MPI parallel computing technology to divide a single process into several modules, including: velocity distribution function calculation module, force calculation module, phase field distribution function calculation module, laser energy distribution calculation module, enthalpy distribution function calculation module, and parallel computing adjustment module, thereby completing the numerical realization of the laser processing molten pool-keyhole evolution;

[0099] The efficient communication module uses high-dimensional data dimensionality reduction communication technology to reduce three-dimensional data such as temperature, velocity, and phase field parameters, and four-dimensional data particle distribution functions to one-dimensional data. It uses gradient operators and Laplace operators to conduct three-dimensional data communication tests, and uses particle distribution function migration to conduct four-dimensional data communication tests, ensuring efficient communication of data between different parallel computing processes.

[0100] The serial-parallel hybrid calculation module collects and organizes the phase field parameter data of the sub-process, determines the exact position of the interface in the main process, calculates the energy distribution through the multi-reflection heat source model, and finally broadcasts the energy distribution data of the main process to the sub-process;

[0101] The data output module collects physical quantities such as temperature and speed in the subprocess through the MPI_Gather function and outputs them as dat files;

[0102] The simulation result visualization module uses the visualization function of the Tecplot360 GUI interface to visualize the molten pool-keyhole morphology and temperature field and velocity field distribution during the laser processing process.

[0103] Another object of the present invention is to provide a computer device, which includes a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the efficient numerical simulation method of the dynamic behavior of the laser processing molten pool-keyhole.

[0104] Another object of the present invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, enables the processor to execute the steps of the method for efficient numerical simulation of the dynamic behavior of the laser processing molten pool-keyhole.

[0105] Another object of the present invention is to provide an information data processing terminal, which is used to implement an efficient numerical simulation system for the dynamic behavior of the laser processing molten pool-keyhole.

[0106] The present invention is specifically implemented:

[0107] The present invention provides an efficient calculation method and system for numerical simulation of laser processing molten pool-keyhole evolution. The calculation steps are specifically implemented according to the following steps:

[0108] The implementation of the dynamic slip calculation domain of the present invention includes the following steps:

[0109] S1: Molten pool-keyhole morphology extraction and image processing to determine the initial size of the calculation domain.

[0110] S1.1: Use high-speed video to capture a video file of the dynamic changes in the molten pool-keyhole morphology during laser processing. Select the molten pool-keyhole morphology during the stable processing state as the input for the image processing step.

[0111] S1.2: Import the image of the steady-state melt pool-keyhole morphology into MATLAB, and preprocess the image captured by the high-speed camera based on the ImageProcessing Toolbox in MATLAB: read the image in the folder and change the original image format to RGB format to facilitate subsequent image processing programs.

[0112] S1.3: Use the grayImage function in the toolkit to grayscale the extracted image. Use the gaussianBlur function in the toolkit to perform Gaussian filtering on the image to smooth it, eliminate Gaussian noise and random noise in the image, and improve image quality to facilitate the subsequent extraction of feature lines at the boundary between the molten pool and the keyhole.

[0113] S1.4: Use the Sobel operator to sharpen the image and highlight the boundary features between the melt pool and the keyhole. Extract the longest and widest values ​​(Smax) and (Wmax) of the melt pool and keyhole during the stable machining process, as well as the highest penetration depth (Hmax).

[0114] S1.5: Determine the length (L), width (W), and height (H) of the optimal calculation domain for numerical simulation of laser processing as 2*Smax, 2*Wmax, and 1.5*Hmax, respectively, to ensure that the molten pool-keyhole is basically in the middle area of ​​the calculation domain during stable processing.

[0115] S2: Implementation of three-dimensional dynamic sliding computational domain: Using dynamic sliding technology to construct a dynamic constant computational domain that moves along with the laser beam, such as Figure 3 shown.

[0116] S2.1: Construct the computational domain based on the length (L), width (W), and height (H) dimensions obtained in step S1. Determine the computational domain time step and grid size, and set the slip criteria. Extract the distance between the laser beam center and the right boundary of the computational domain at each moment. Slide the computational domain if the slip criteria are met.

[0117] Furthermore, the dynamic slip judgment condition of the three-dimensional computational domain described in S2.1 is as follows: construct the computational domain based on the computational domain dimensions length (L), width (W) and height (H) obtained in the above step S1, determine the computational domain time step Δt and grid size, and determine the number of grids in the X, Y and Z directions to be N respectively. X = L / Δx, Ny = W / Δx, and = H / Δx, set the dynamic slip judgment condition d = 0.2*Nx. Extract the distance S between the laser beam center and the right edge of the computational domain at each moment. When the slip condition S <= d is met, perform a slip operation on the computational domain.

[0118] S2.2: Perform a sliding operation along the direction of beam movement. According to the set number of grids, delete part of the computational domain on the left boundary and add a blank computational domain of the same size on the right boundary.

[0119] Furthermore, the specific implementation of the dynamic sliding operation of the computational domain described in S2.2 is: when the sliding judgment condition of S2.1 is met, the grid in the computational domain is moved from 0 to 0.25*N along the X direction. X The area with length, Ny and Nz as width and height is removed, and the right boundary of the computational domain along the X direction is filled with 0.25*N. X A blank computational domain with Nx as length, Ny as width, and Nz as height, ensuring that the computational domain size is always Nx*Ny*Nz.

[0120] S2.3: After the sliding operation is completed, the coordinate positions of all parameters in the calculation domain are updated, marking the end of a dynamic sliding operation and ensuring that the size of the calculation domain remains constant.

[0121] Furthermore, the key variable of the dynamic sliding of the three-dimensional computational domain described in S2.3 is the real-time value S of the distance between the center of the beam and the right boundary of the computational domain. The update formula is as follows:

[0122] S=N χ -U χ *Step+m*n χ

[0123] Where m is the number of times the beam cycles to the trigger position (the number of times dynamic slip occurs), the initial value is 0, and the value of m increases by 1 each time a dynamic slip operation is performed; n is the number of grids that dynamically slip along the processing direction. χ =0.25*N X , Ux is the moving speed of the laser beam, and Step is the number of real-time calculation steps of the current program. The S value changes dynamically in real time, and the update formula is used to update the S value when the judgment condition is met. The graphical understanding of the above formula is as follows Figure 3 shown.

[0124] S3: Model calculation process: This patent establishes a numerical model of the dynamic behavior of the laser processing molten pool-keyhole based on LBM, and combines MPI parallel computing to achieve efficient solution of the model. Parallel computing module and serial computing module are as follows Figure 4 The parallel calculation module includes: velocity particle distribution function calculation module, enthalpy particle distribution function calculation module, phase field particle distribution function calculation module and force calculation module; the serial calculation module includes: interface position recognition module, laser beam multiple reflection heat source module and file output module.

[0125] The velocity distribution function calculation module and the force calculation module, that is, the flow field distribution calculation formula during laser processing, are as follows:

[0126]

[0127] Among them, g i (r+e i δt,t+δt) is the particle velocity distribution function of the next time step, g i (r,t) is the current particle velocity distribution function, is the particle equilibrium velocity distribution function, G i (r,t) is the forced convection phase, τ g For relaxation time.

[0128] The laser energy distribution calculation module includes three components: the laser beam multiple reflection model, laser beam energy distribution, and Fresnel energy absorption. The phase field distribution function calculation module calculates the distribution of the current phase field parameters and uses the interface tracking formula to accurately track the melt pool and keyhole. The calculation formula is detailed in the subsequent serial and parallel hybrid calculation section.

[0129] The enthalpy distribution function calculation module, that is, the temperature field distribution calculation formula during laser processing, is as follows:

[0130]

[0131] Among them, h i (r+e i δt,t+δt) is the enthalpy distribution function of the next time step, h i (r,t) is the current enthalpy distribution function, is the equilibrium enthalpy distribution function, ω i Qδt is the energy input item affected by the above calculation module, ω i is the weight coefficient of the particle distribution function in each direction, τ h is the relaxation time Q in is the laser energy input.

[0132] The parallel computing adjustment module uses the MPI control function to ensure that the computing progress of each process is consistent, maintain the order of the computing process, and ensure its efficiency and uniformity.

[0133] S4: Computational process processing: The following solutions are proposed to solve the problems of chaotic high-dimensional data communication in the computational process and low position tracking accuracy of the parallel computing interface.

[0134] Furthermore, the dimensionality reduction processing formulas for the four-dimensional data of the LBM particle distribution function and the three-dimensional data of the temperature field, flow field, and phase field in S4.1 are as follows:

[0135] Perform dimensionality reduction on the coordinate values ​​of three-dimensional data:

[0136] vector3D(i,j,k)=k+j*Nz+i*Ny*Nz

[0137] For four-dimensional data:

[0138] vector4D(i,j,k,n)=(k+j*Nz+i*Ny*Nz)*Q+n

[0139] Where vector3D and vector4D represent the coordinates of three-dimensional and four-dimensional data in one-dimensional data, i, j, and k represent the current three-dimensional coordinates of the data, n represents the directional component of the particle distribution function, and Q represents the number of directional components of the particle distribution function. Ny and Nz represent the maximum number of grid cells in the computational domain in the y and z directions, respectively. It has been proven that the above formula guarantees uniqueness when high-dimensional coordinate data is expanded into one-dimensional coordinates.

[0140] Furthermore, the program uses gradient operators and Laplace operators to test three-dimensional data communication, and uses particle distribution function migration to test four-dimensional data communication, to ensure accurate communication of data between different parallel computing processes. Graphical understanding of high-level data dimensionality reduction communication is as follows: Figure 5 shown.

[0141] S4.2: To address the low accuracy of interface position tracking in parallel computing, serial and parallel hybrid computing is performed on the data in the computational domain. Interface position identification, beam tracing algorithm, and energy distribution calculation are performed using single-threaded serial computing in the main process. The main process transmits the beam tracing results and energy distribution data to each sub-process via MPIBcast broadcast. At the same time, the temperature, speed, and other physical quantities of the sub-process are collected via MPIGather and output to a dat file. The program architecture of serial and parallel hybrid computing is as follows: Figure 6 shown.

[0142] The interface tracking formula is as follows:

[0143]

[0144] in, is the phase field order parameter, u is the velocity vector, M is the interface diffusion coefficient, n is the normal vector of the current interface, and W is the interface thickness. A grid size of [3, 5] is generally used. The phase field order parameter is set to a constant in the range [0, 1]. This patent sets the material phase order parameter to 1 and the gas phase to 0. is the order parameter constant, set to 0.5.

[0145] The formula for calculating the beam reflection direction vector using the beam multiple reflection heat source model is as follows:

[0146]

[0147] in, is the beam reflection direction vector, N i is the direction vector of the incident beam, and n is the normal vector of the reflective interface. The model assumes that the laser beam reflects 10 times or stops reflecting when it reaches the boundary of the computational domain. The energy of the laser beam is assumed to have a Gaussian distribution.

[0148] Then the intensity distribution function of each laser beam is as follows:

[0149]

[0150] Where I0(r) is the intensity of the laser beam at the reflection point, in W / mm 2 , P is the laser processing power, is the radius constant of the laser beam focal plane, and r is the distance between the reflection point and the origin of the laser beam plane.

[0151] The Fresnel energy absorption rate is used to calculate the energy absorbed by the material when it is irradiated by the laser beam. The calculation formula considering only reflection is as follows:

[0152]

[0153] Among them, α Fis the laser beam energy absorption rate of the material at the reflection point, ε is the Fresnel absorption coefficient of the material, is the angle between the incident light beam and the normal of the reflecting interface. The energy at the reflection point can be calculated using the following formula:

[0154] Q=I0(r)*α F

[0155] S5: Post-processing: The numerical simulation is exported as a dat file containing the calculation results. The morphology of the molten pool-keyhole and the distribution of temperature and velocity fields during laser processing are visualized through the Tecplot360 GUI interface. The results of the efficient numerical simulation of the dynamic behavior of the molten pool-keyhole during laser processing are as follows: Figure 7 shown.

[0156] Example 1

[0157] In this example, it is assumed that the processing material is TC4 titanium alloy, the Fresnel absorption coefficient is 0.2, the heat source model adopts the Gaussian heat source model, and the laser beam is incident perpendicularly on the material surface. According to the modeling steps of the efficient numerical simulation system for the dynamic behavior of the laser processing molten pool-keyhole, the evolution behavior of the molten pool-keyhole during the laser processing is numerically simulated, and the simulation results of the large-scale molten pool-keyhole of the entire weld are obtained as follows: Figure 7 shown. Figure 7 (a)-(c) are pictures of the temperature field and flow field distribution at different times in the central longitudinal section of the calculation domain, showing the changes of the molten pool-keyhole in the central section. Figure 7 (d)-(f) are the morphological evolution diagrams of the molten pool-keyhole at different moments in a three-dimensional perspective. In this case, the dynamic sliding technology is used to complete the large-scale weld simulation of the entire weld with a relatively small amount of computing resources, proving the feasibility and superiority of this technology. In this case, high-dimensional data dimensionality reduction communication technology is used to ensure efficient and unified communication of data between different processes, and a serial-parallel hybrid computing method is used to ensure the accuracy of interface position tracking and energy calculation. The gas-liquid interface and the solid-liquid interface are clear and smooth, achieving the simulation effect. Therefore, the efficient numerical simulation method and system for the dynamic behavior of the laser processing molten pool-keyhole proposed in the present invention effectively solve the problems of large computing resources for full weld simulation, chaotic high-dimensional data communication, and low accuracy of parallel solution of interface position.

[0158] It should be noted that the embodiments of the present invention can be implemented by hardware, software, or a combination of software and hardware. The hardware portion can be implemented using dedicated logic; the software portion can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. Those skilled in the art will appreciate that the above-mentioned devices and methods can be implemented using computer-executable instructions and / or contained in processor control code, for example, such as a carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The devices and modules of the present invention can be implemented by hardware circuits such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., can also be implemented by software executed by various types of processors, or can be implemented by a combination of the above-mentioned hardware circuits and software, such as firmware.

[0159] The above description is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with this technical field within the technical scope disclosed by the present invention and within the spirit and principles of the present invention should be covered by the scope of protection of the present invention.

Claims

1. An efficient numerical simulation method for the dynamic behavior of laser processing molten pool-keyhole, characterized by: The method for efficiently simulating the dynamic behavior of a laser processing molten pool and a keyhole comprises the following steps: S1: Molten pool-keyhole morphology extraction and image processing to determine the initial size of the calculation domain; S2: Implementation of 3D dynamic sliding computational domain: Using dynamic sliding technology to construct a dynamic constant computational domain that moves along with the laser beam; S3: The dynamic behavior numerical simulation model of the laser processing molten pool-keyhole is constructed using the lattice Boltzmann method and the phase field method. The numerical simulation model is divided into parallel computing modules and serial computing modules based on the MPI parallel computing technology. The parallel calculation module includes: velocity particle distribution function calculation module, enthalpy particle distribution function calculation module, phase field particle distribution function calculation module and force calculation module; the serial calculation module includes: interface position recognition module, laser beam multiple reflection heat source module and file output module; S4: For the parallel computing module, the particle distribution function introduced by the three-dimensional LBM generates four-dimensional data, while the flow field, temperature field, and phase field scalars are three-dimensional data. Therefore, the calculation process faces the problem of mixed parallel computing of high-dimensional data; S5: Post-processing: Based on the calculation results, the molten pool-keyhole morphology and temperature field, velocity field and phase field distribution during laser processing are visualized through the Tecplot360 GUI interface; The molten pool-keyhole morphology extraction and image processing are used to determine the initial size of the calculation domain; S1.1: Capture a video file of the dynamic changes in the molten pool-keyhole morphology during laser processing using high-speed video. Select the molten pool-keyhole morphology during the stable processing state as the input for the image processing step. S1.2: Import the image of the steady-state melt pool and keyhole into MATLAB. Preprocess the image captured by the high-speed camera using the ImageProcessing Toolbox in MATLAB: read the image in the folder and change the original image format to RGB format to facilitate subsequent image processing. S1.3: Use the grayImage function in the toolkit to grayscale the extracted image; use the gaussianBlur function in the toolkit to perform Gaussian filtering on the image to smooth the image, eliminate Gaussian noise and random noise in the image, improve image quality, and facilitate the subsequent extraction of the characteristic line of the processing pool-keyhole boundary; S1.4: Use the Sobel operator to sharpen the image and highlight the boundary features between the melt pool and the keyhole. Extract the longest value Smax and the widest value Wmax of the melt pool and the keyhole during the stable processing, as well as the maximum penetration value Hmax, from the boundary feature image. S1.5: Determine the optimal computational domain length (L), width (W), and height (H) for numerical simulation of laser processing to be 2*Smax, 2*Wmax, and 1.5*Hmax, respectively, to ensure that the molten pool-keyhole is basically in the middle area of ​​the computational domain during stable processing.

2. The efficient numerical simulation method for dynamic behavior of laser processing molten pool-keyhole according to claim 1, characterized in that: Implementation of the three-dimensional dynamic sliding calculation domain: using dynamic sliding technology to construct a dynamic constant calculation domain that moves along with the laser beam; S2.1: Construct a computational domain based on the length (L), width (W), and height (H) of the computational domain obtained in step S1 above. Determine the computational domain time step and grid size, and set the slip judgment condition. Extract the distance between the laser beam center and the right boundary of the computational domain at each moment. When the slip condition is met, perform a slip operation on the computational domain. S2.2: Perform a sliding operation along the direction of beam movement. According to the set number of grids, delete the computational domain on the left boundary of the computational domain and add a blank computational domain of the same size on the right boundary of the computational domain. S2.3: After the sliding operation is completed, the coordinate positions of all parameters in the calculation domain are updated, marking the end of a dynamic sliding operation and ensuring that the size of the calculation domain remains constant.

3. The efficient numerical simulation method for dynamic behavior of laser processing molten pool-keyhole according to claim 1, characterized in that: For the parallel computing module, the particle distribution function introduced by the three-dimensional LBM brings four-dimensional data, while the flow field, temperature field, and phase field scalars are three-dimensional data. Therefore, the calculation process has the problem of mixed parallel computing of high-dimensional data; S4.1: To accurately transfer high-dimensional mixed data between parallel computing nodes, dimensionality reduction processing is performed on the four-dimensional LBM data and the three-dimensional temperature and velocity data, uniformly converting them into one-dimensional data. This enables accurate transfer of high-dimensional data between global processes. S4.2: For the serial calculation module, given the difficulty of cross-process calculation for the location of the molten pool-keyhole gas-liquid interface, single-threaded serial calculation is used in the main process for interface position identification, the laser beam multiple reflection heat source model including the beam tracing algorithm and energy distribution calculation. The main process first collects the phase field data of the sub-process through MPIGather and realizes interface position identification, and then calculates the energy distribution on the interface based on the beam tracing algorithm. Finally, the main process transmits the data to each sub-process through MPIBcast broadcasting, and at the same time, the temperature and velocity physical quantities of the sub-process are collected through MPIGather and output to the dat file.

4. The method for efficient numerical simulation of dynamic behavior of laser processing molten pool-keyhole according to claim 2, characterized in that: The dynamic slip judgment condition of the three-dimensional calculation domain in step S2.1 is as follows: construct the calculation domain based on the length (L), width (W) and height (H) of the calculation domain obtained in step S1 above, and determine the time step of the calculation domain. and grid size, determine the number of grids in the X, Y, and Z directions as N X =L / x、Ny=W / x and =H / x sets the dynamic slip judgment condition d=0.2*Nx; extract the distance S between the position of the laser beam center and the right boundary of the calculation domain at each moment; when the slip condition S<=d is met, the calculation domain is slipped; The specific implementation of the dynamic sliding operation of the computational domain described in S2.2 is: when the sliding judgment condition of S2.1 is met, the grid in the computational domain is moved from 0 to 0.25*N along the X direction. X The area with length, width Ny and height Nz is removed, and the right boundary along the X direction in the calculation domain is filled with a length of 0.25*N X , a blank computational domain with a width of Ny and a height of Nz, thus ensuring that the computational domain size is always Nx*Ny*Nz; The key variable of the dynamic sliding of the three-dimensional computational domain described in S2.3 is the real-time value S of the distance between the center of the beam and the right boundary of the computational domain. The update formula is as follows: ； Where m is the number of times the beam cycles to the trigger position, that is, the number of times dynamic slip occurs. The initial value is 0. Each time a dynamic slip operation is performed, the value of m increases by 1. n is the number of grids that dynamically slip along the processing direction. =0.25*N X , Ux is the moving speed of the laser beam, Step is the number of real-time calculation steps of the current program; the S value changes dynamically in real time, and the update formula is used to update the S value when the judgment condition is met.

5. The efficient numerical simulation method for dynamic behavior of laser processing molten pool-keyhole according to claim 3, characterized in that: The dimensionality reduction processing formulas for LBM four-dimensional data and three-dimensional data in S4.1 are as follows: Perform dimensionality reduction on the coordinate values ​​of three-dimensional data: ； For four-dimensional data: ； Among them, vector3D and vector4D are the coordinates of three-dimensional and four-dimensional data in one-dimensional data, i, j, and k are the current three-dimensional coordinates of the data, n is the directional component of the particle distribution function, and Q is the number of directional components of the particle distribution function; Ny and Nz are the maximum number of grids in the computational domain in the y and z directions, respectively. It can be proved that the above formula can ensure uniqueness when the high-dimensional coordinate data is expanded into one-dimensional coordinates.

6. A system for efficiently simulating the dynamic behavior of a molten pool and a keyhole in laser processing, which implements the method for efficiently simulating the dynamic behavior of a molten pool and a keyhole in laser processing according to any one of claims 1 to 5, characterized in that: The laser processing molten pool-keyhole dynamic behavior efficient numerical simulation system includes: The image processing module, equipped with MATLAB software and Image Processing Toolbox, is used to import and process laser machining steady-state melt pool-keyhole morphology images, including image format change, grayscale processing, Gaussian filtering, Sobel operator sharpening, and melt pool-keyhole boundary feature extraction; The dynamic sliding module uses a decision formula to determine when to perform boundary sliding and uses an update formula to update the S value, thereby achieving dynamic sliding of the three-dimensional calculation domain and ensuring that the size of the calculation domain remains constant during the calculation process; The parallel computing module uses LBM simulation technology combined with MPI parallel computing technology to divide a single process into several modules, including: velocity distribution function calculation module, force calculation module, phase field distribution function calculation module, laser energy distribution calculation module, enthalpy distribution function calculation module, and parallel computing adjustment module, thereby completing the numerical realization of the laser processing molten pool-keyhole evolution; The efficient communication module uses high-dimensional data dimensionality reduction communication technology to reduce the three-dimensional data of temperature, velocity, and phase field parameters, and the four-dimensional data particle distribution function to one-dimensional data. It uses gradient operators and Laplace operators to conduct three-dimensional data communication tests, and uses particle distribution function migration to conduct four-dimensional data communication tests, ensuring efficient communication of data between different parallel computing processes. The serial-parallel hybrid calculation module collects and organizes the phase field parameter data of the sub-process, determines the exact position of the interface in the main process, calculates the energy distribution through the multi-reflection heat source model, and finally broadcasts the energy distribution data of the main process to the sub-process; The data output module collects the temperature and speed physical quantities in the subprocess through the MPI_Gather function and outputs them as dat files; The simulation result visualization module uses the visualization function of the Tecplot360 GUI interface to visualize the molten pool-keyhole morphology and temperature field and velocity field distribution during the laser processing process.

7. A computer device, characterized in that: The computer device includes a memory and a processor, the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the efficient numerical simulation method of the dynamic behavior of the laser processing molten pool-keyhole as described in any one of claims 1 to 5.

8. A computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor executes the steps of the method for efficient numerical simulation of dynamic behavior of a laser processing molten pool-keyhole as claimed in any one of claims 1 to 5.

9. An information data processing terminal, characterized in that: The information data processing terminal is used to implement the efficient numerical simulation system for the dynamic behavior of the laser processing molten pool-keyhole as described in claim 6.

Images