A high-fidelity welding simulation boundary condition correction method and system

Abstract

The application belongs to the technical field of welding simulation, and relates to a high-fidelity welding simulation boundary condition correction method and system. The method comprises the following steps: after the clamping of a welding part is completed and before welding is performed, three-dimensional geometric measured data of the welding part in a clamping constraint state is collected; the measured three-dimensional geometric data is spatially registered with a simulation theoretical model, clamping deformation caused by clamping is extracted, and the clamping deformation is reconstructed into a simulation grid node displacement constraint field; based on the node displacement constraint field, the boundary condition parameters of the welding simulation model are adaptively corrected; welding thermal elastic-plastic finite element calculation is performed by using the corrected simulation model, and a welding simulation result considering clamping deformation is output; the application eliminates systematic errors introduced due to the neglect of clamping deformation in traditional simulation, and significantly improves the accuracy of welding deformation prediction.

Description

Technical Field

[0001] This invention belongs to the field of welding simulation technology, and relates to a high-fidelity welding simulation boundary condition correction method and system. Background Technology

[0002] Welding simulation is an important tool for predicting welding deformation and optimizing welding processes, and its accuracy largely depends on the accuracy of the boundary conditions. Current welding simulations typically set boundary conditions based on an ideal geometric model of the weldment, assuming no initial deformation after clamping. However, in actual production, the weldment inevitably undergoes clamping deformation under the influence of clamping force, its own weight, and assembly stress, leading to deviations between the simulation model and the actual geometric state, thus affecting the accuracy of welding deformation prediction.

[0003] Currently, online monitoring technology for welding processes focuses on real-time trajectory correction during welding, without addressing the pre-welding clamping state; clamping force compensation technology only uses deformation data for measurement correction or fixture adjustment; and simulation methods for preset initial fields rely heavily on empirical estimation, lacking support from measured data. Therefore, existing technologies have not yet achieved a means to effectively convert measured pre-welding clamping deformation into simulation boundary condition correction signals. Some foreign scholars have proposed a method combining deviation simulation with welding simulation, generating non-nominal parts through simulation and representing clamping deformation with translation and rotation matrices before importing it into the welding simulation model. However, the clamping deformation in this method is not the actual measured point cloud data before welding, failing to establish an automatic spatial mapping and constraint conversion mechanism from the measured 3D point cloud to the simulated finite element nodal displacement field, making it difficult to reflect the true initial state of the weldment and hindering further improvement in welding simulation accuracy. To address these technical problems, this invention provides a welding simulation boundary condition correction method based on measured clamping deformation. This method uses a laser 3D scanner to acquire measured 3D point cloud data of the weldment under clamping constraints after clamping and before welding begins. By constructing a spatial mapping and shape function reconstruction mechanism from point cloud data to the simulated mesh node displacement field, the actual clamping deformation before welding is converted into node displacement constraint signals recognizable by the simulation system. Based on this signal, the simulation boundary conditions are adaptively corrected, and high-precision welding simulation calculations considering the initial clamping deformation are performed. This invention realizes the transformation of simulation boundary conditions from "theoretical setting" to "measurement-driven," eliminating the systematic errors introduced by ignoring the initial clamping deformation in traditional simulations, and significantly improving the accuracy of welding deformation prediction. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the prior art and provide a high-fidelity welding simulation boundary condition correction method and system, which aims to solve the technical problem that the prior art cannot effectively convert the measured clamping deformation before welding into a simulated and identifiable boundary condition signal.

[0005] To achieve the objectives of this invention, the following technical solutions will be adopted.

[0006] A high-fidelity welding simulation boundary condition correction method includes:

[0007] S1. Pre-welding measurement data acquisition: After the workpiece is clamped and before welding is carried out, collect the three-dimensional geometric measurement data of the workpiece under clamping and constraint.

[0008] S2. Spatial mapping and function reconstruction: Spatial registration is performed between the measured three-dimensional geometric data and the simulation theoretical model to extract the clamping deformation caused by clamping and reconstruct the clamping deformation into the displacement constraint field of the simulation mesh node;

[0009] S3. Adaptive correction of boundary conditions: Based on the nodal displacement constraint field, the boundary condition parameters of the welding simulation model are adaptively corrected.

[0010] S4. High-precision welding simulation calculation: The modified simulation model is used to perform welding thermo-elastic-plastic finite element calculation, and the welding simulation results are output considering clamping deformation.

[0011] Furthermore, the three-dimensional geometric measurement data is collected by a non-contact three-dimensional measuring device, including information on the surface deformation of the weldment under the combined action of clamping force, self-weight, and assembly stress.

[0012] Furthermore, the extraction of clamping deformation is achieved by calculating the normal deviation vector of the measured point relative to the theoretical model surface, forming a discrete displacement vector field to characterize the clamping deformation distribution of the weldment.

[0013] Furthermore, the spatial registration involves extracting at least three non-collinear feature points on the weldment and aligning the coordinate system of the measured data with the coordinate system of the simulation model through rigid transformation to obtain the optimal rotation matrix and translation vector.

[0014] Furthermore, the nodal displacement constraint field reconstruction is based on establishing the interpolation relationship between the measured point displacement and the mesh nodal displacement using finite element shape functions, and then solving the overdetermined equations using the least squares method to obtain the equivalent forced displacement field of the mesh nodes.

[0015] Furthermore, the boundary condition parameters include at least one of the constraint node degrees of freedom and the initial displacement of the clamping position.

[0016] Furthermore, the nodal displacement constraint field is compiled into a file format recognizable by the simulation system and automatically written into the simulation model to complete the boundary condition correction.

[0017] Furthermore, the welding simulation results include the welding temperature field, residual stress distribution, and post-weld deformation prediction results.

[0018] A high-fidelity welding simulation boundary condition correction system, characterized in that it includes:

[0019] The three-dimensional geometry acquisition module is used to acquire measured three-dimensional geometric data of the weldment under clamping and constraint conditions after the weldment is clamped and before welding is performed.

[0020] The data conversion module spatially registers the measured three-dimensional geometric data with the simulation theoretical model, extracts the clamping deformation caused by clamping, and reconstructs the clamping deformation into a simulation mesh node displacement constraint field.

[0021] The boundary condition correction module adaptively corrects the boundary condition parameters of the welding simulation model based on the nodal displacement constraint field.

[0022] The simulation calculation module uses the modified simulation model to perform welding thermo-elastic-plastic finite element calculations and outputs welding simulation results that take clamping deformation into account.

[0023] Beneficial effects:

[0024] This invention employs laser 3D scanning to acquire global point cloud data of the weldment under clamping constraints. By constructing a spatial mapping and shape function interpolation reconstruction mechanism, the measured clamping deformation is transformed into nodal displacement constraint signals that drive the simulation model. This method achieves a fundamental shift in welding simulation boundary conditions from "ideal geometric theory setting" to "measured deformation data driving," effectively eliminating systematic initial errors introduced by neglecting clamping force, self-weight, and assembly stress. It significantly improves the prediction accuracy of welding temperature field, residual stress, and post-weld deformation, greatly enhancing simulation fidelity and reducing physical testing costs. Simultaneously, it provides high-precision data support for welding process optimization, fixture layout design, and intelligent prediction of welding quality, powerfully promoting the engineering application of intelligent welding and digital simulation technologies. Attached Figure Description

[0025] Figure 1 The flowchart is as follows: A method for correcting boundary conditions in high-fidelity welding simulation.

[0026] Figure 2 This is a schematic diagram of the registration of the spatial coordinate system between the measured point cloud and the simulation theoretical model;

[0027] Figure 3 A schematic diagram for extracting the deformation vector field and mapping the shape function interpolation. Detailed Implementation

[0028] The present invention will be further described in conjunction with the embodiments and accompanying drawings.

[0029] As an embodiment of the present invention, such as Figures 1 to 3As shown, a high-fidelity welding simulation boundary condition correction method is proposed. The method acquires the global point cloud data of the weldment in the clamping state through laser three-dimensional scanning. After spatial mapping and shape function reconstruction, the measured clamping deformation is converted into the simulation node displacement constraint signal, thereby realizing the measured data-driven adaptive correction of the welding simulation boundary conditions and eliminating the systematic error introduced by ignoring the initial clamping deformation in traditional simulation.

[0030] This embodiment uses the butt welding of two 6mm thick 6061-T6 aluminum alloy sheets as an example to illustrate the specific implementation process of the method of the present invention. Welding simulation is performed using general simulation software for thermo-elastic-plastic finite element analysis.

[0031] To comprehensively verify the technical effectiveness of the method of this invention, multiple sets of comparative verification experiments were systematically conducted, focusing on typical thin-plate butt welding. The experiments covered three typical engineering materials (6061-T6 aluminum alloy, Q345 low-carbon steel, and 304 stainless steel), with plate thicknesses of 2mm, 4mm, and 6mm. Three typical boundary constraint forms were used for the fixture layout: two-end clamping, evenly distributed clamping on all four sides, and cantilever beam clamping. Each set of experiments was independently repeated five times to evaluate repeatability. The welding method was TIG flat plate butt welding. Before welding, a laser 3D scanner was used to acquire measured point cloud data of each test piece under fixture constraints. Welding deformation and thermo-mechanical coupling simulation calculations were performed using both the traditional nominal geometric boundary condition simulation method and the measured driven boundary condition correction method of this invention, and the results were compared with the measured values ​​obtained from the 3D scanning after welding.

[0032] S1: Acquisition of pre-welding measured data

[0033] First, the aluminum alloy sheet to be welded is clamped onto a specialized welding fixture, and a preset clamping force is applied to simulate actual production conditions. After the workpiece is clamped but before welding begins, a handheld blue light 3D laser scanner is used to perform a non-contact, full-area scan of the workpiece under clamping constraints. During the scan, the scanner collects data around the workpiece at multiple angles, acquiring measured 3D point cloud data of the workpiece under the combined action of clamping force, its own weight, and assembly stress. This point cloud data includes information on clamping deformations such as minor indentations and warping on the workpiece surface caused by clamping. The obtained point cloud data is exported and stored in a format recognizable by the simulation software.

[0034] S2: Spatial Mapping and Function Reconstruction

[0035] This step establishes a spatial mapping and shape function reconstruction mechanism from laser 3D scanning point cloud data to the simulated mesh node displacement field, converting the actual pre-welding clamping deformation into node displacement constraint signals recognizable by the simulation system. The specific implementation steps are as follows:

[0036] (1) Spatial coordinate system registration: Import the measured point cloud obtained from the scan into the supporting preprocessing software. Extract the centers of the positioning pin holes at the four corners of the weldment as non-collinear feature points (the number should be greater than or equal to 3), such as Figure 2 As shown. The rotation matrix R and translation vector T are calculated using the least squares method, such that the registration error E satisfies:

[0037] ;

[0038] In the formula, These are the coordinates of the corresponding feature points on the simulation theoretical model. R represents the coordinates of the feature points extracted from the measured point cloud, R is the rotation matrix, and T is the translation vector.

[0039] The rotation matrix R and translation vector T are calculated as follows:

[0040] Suppose the coordinates of the n (n≥3) non-collinear feature points selected in the theoretical model are... The coordinates of the corresponding feature points in the measured point cloud are: Calculate the centroid coordinates of the two sets of points respectively:

[0041] ;

[0042] ;

[0043] In the formula, Non-collinear feature points are The coordinates of the centroid, The corresponding feature points in the measured point cloud are The coordinates of the centroid.

[0044] Translate the two sets of points to a local coordinate system with their respective centroids as the origin to obtain the decentralized coordinates:

[0045] ;

[0046] ;

[0047] In the formula, The coordinates of non-collinear feature points after decentering. These are the centered coordinates of the corresponding feature points in the measured point cloud.

[0048] Construct the covariance matrix by using the decentralized point set to build a 3×3 covariance matrix H:

[0049] ;

[0050] Perform singular value decomposition on the covariance matrix H:

[0051] ;

[0052] In the formula, U and V are 3×3 orthogonal matrices. It is a diagonal matrix whose diagonal elements are singular values ​​of H.

[0053] Based on the singular value decomposition results, the rotation matrix R is calculated using the following formula:

[0054] ;

[0055] After obtaining the rotation matrix R, the translation vector T is determined by the centroid coordinates of the two sets of points:

[0056] ;

[0057] Through the above calculation process, the optimal rotation matrix R and translation vector T for rigidly transforming the measured point cloud coordinate system to the simulation model coordinate system can be obtained.

[0058] (2) Extraction of the deformation vector field of the clamping: For the transformed measured point cloud, the nearest neighbor search algorithm is used to traverse each measured point. Find its nearest neighbor point on the surface of the simulation theoretical model. Calculate the normal deviation vector using the following formula, and use it as the clamping deformation at that point, such as... Figure 3 As shown:

[0059] ;

[0060] In the formula, These are the coordinates of the measured point. The coordinates of the nearest point on the theoretical model surface. For theoretical surfaces in The unit normal vector at that location.

[0061] (3) Shape function interpolation mapping: Determine the finite element e in which each measured point is located and its natural coordinates within that element. Based on the shape function interpolation relationship:

[0062] ;

[0063] In the formula, n is the number of unit nodes, N i (e) Let U be the shape function of element node i. i (e) Let be the displacement vector of the mesh node to be determined;

[0064] Let the measured deformation displacement equal to interpolation displacement Establish an overdetermined linear system of equations for the displacement U of all grid nodes by combining all measured points:

[0065] ;

[0066] Here, matrix A is composed of the shape function values ​​corresponding to each point, and vector b is composed of the measured deformation displacement components. The system of equations is solved using the least squares method:

[0067] ;

[0068] In this system of equations, matrix A is composed of shape functions, and vector b consists of measured deformation displacement components. The system of equations is solved using the least squares method, and the resulting mesh node displacement vector U is the reconstructed equivalent forced displacement field of the mesh nodes. This displacement field data is then output as a boundary condition adjustment signal.

[0069] S3: Adaptive Boundary Condition Correction

[0070] The mesh node displacement vector U obtained in step two is compiled into a boundary condition input file that the simulation system can recognize, in a format recognizable by the simulation software. A script is written to read this displacement field file and automatically write keywords into the input file of the welding simulation model. Specifically, the reconstructed node displacement values ​​are assigned to the nodes at the corresponding clamping positions as their initial displacement constraints. This drives the simulation system to automatically correct the constraint node degrees of freedom, the initial displacement of the clamping position, and the initial deformation of the weldment, generating a corrected welding simulation model.

[0071] Step 4: High-precision welding simulation calculation

[0072] A modified simulation model was generated based on the corrected boundary conditions, and welding thermo-elastic-plastic finite element calculations were performed. A double ellipsoidal heat source model was used to simulate the welding heat input, and the thermophysical and mechanical property parameters of the material as a function of temperature were defined. Geometric and material nonlinearities were considered during the calculation. After the calculation was completed, the welding temperature field distribution, welding residual stress distribution, and welding deformation distribution data were extracted from the output results database (.odb file).

[0073] Taking the clamping condition of a 4mm thick Q345 low-carbon steel plate at both ends as an example, the prediction accuracy of key field variables during welding was evaluated. Regarding the temperature field, using the peak temperature at the weld center as the evaluation index, the measured peak temperature by thermocouple was 1276℃, while the traditional method's simulation prediction was 1093℃, with an error of approximately 14.3%; the method of this invention predicted 1234℃, reducing the error to approximately 3.3%. Regarding the stress field, using the longitudinal residual tensile stress at the weld center as the evaluation index, the measured value by X-ray diffraction was 312MPa, while the traditional method's simulation prediction was 384MPa, with an error of approximately 23.1%; the method of this invention predicted 325MPa, reducing the error to approximately 4.2%. The simultaneous improvement in the prediction accuracy of the temperature field and residual stress further verifies that the method of this invention, by correcting the initial geometric boundary conditions before welding, not only improves deformation prediction but also fundamentally improves the overall accuracy of the thermo-mechanical coupling simulation of the welding process.

[0074] Comparative analysis revealed that after correcting the boundary conditions using the method of this invention, the prediction error of post-weld angular deformation under different materials, plate thicknesses, and fixture layouts was generally between 33% and 40%, while the prediction errors of temperature field peak value and residual stress were reduced to below 5%. This invention's method, by introducing pre-weld measured point cloud data to automatically correct the simulation boundary conditions, significantly reduced the prediction error to 7.5%–9.1%, and the standard deviation of five repeated experiments was controlled within 1.1%, indicating that the method has good repeatability and robustness. These results confirm that the method of this invention can effectively eliminate the systematic errors introduced by neglecting clamping deformation in traditional simulations, realizing the transformation of simulation boundary conditions from "theoretical setting" to "measurement-driven."

Claims

1. A method for correcting boundary conditions in high-fidelity welding simulation, characterized in that, include: S1. Pre-welding measurement data acquisition: After the workpiece is clamped and before welding is carried out, collect the three-dimensional geometric measurement data of the workpiece under clamping and constraint. S2. Spatial mapping and function reconstruction: Spatial registration is performed between the measured three-dimensional geometric data and the simulation theoretical model to extract the clamping deformation caused by clamping and reconstruct the clamping deformation into the displacement constraint field of the simulation mesh node; S3. Adaptive correction of boundary conditions: Based on the nodal displacement constraint field, the boundary condition parameters of the welding simulation model are adaptively corrected. S4. High-precision welding simulation calculation: The modified simulation model is used to perform welding thermo-elastic-plastic finite element calculation, and the welding simulation results are output considering clamping deformation.

2. The high-fidelity welding simulation boundary condition correction method according to claim 1, characterized in that, The three-dimensional geometric measurement data in the clamping deformation conversion is collected by a non-contact three-dimensional measuring device, which includes information on the clamping deformation of the weldment surface under the combined action of clamping force, self-weight, and assembly stress.

3. The high-fidelity welding simulation boundary condition correction method according to claim 1, characterized in that, The extraction of clamping deformation is achieved by calculating the normal deviation vector of the measured point relative to the theoretical model surface, forming a discrete displacement vector field to characterize the clamping deformation distribution of the weldment.

4. The high-fidelity welding simulation boundary condition correction method according to claim 1, characterized in that, The spatial registration involves extracting at least three non-collinear feature points on the weldment and aligning the coordinate system of the measured data with the coordinate system of the simulation model through rigid transformation to obtain the optimal rotation matrix and translation vector.

5. The high-fidelity welding simulation boundary condition correction method according to claim 1, characterized in that, The nodal displacement constraint field reconstruction is based on establishing the interpolation relationship between the measured point displacement and the mesh nodal displacement using finite element shape functions. The overdetermined equations consisting of the optimal rotation matrix and translation vector are solved by the least squares method to obtain the equivalent forced displacement field of the mesh nodal.

6. The high-fidelity welding simulation boundary condition correction method according to claim 1, characterized in that, The boundary condition parameters include at least one of the following: constrained node degrees of freedom and initial displacement of the clamping position.

7. The high-fidelity welding simulation boundary condition correction method according to claim 1, characterized in that, The nodal displacement constraint field is compiled into a file format recognizable by the simulation system and automatically written into the simulation model to complete the boundary condition correction.

8. The high-fidelity welding simulation boundary condition correction method according to claim 1, characterized in that, The welding simulation results include the welding temperature field, residual stress distribution, and post-weld deformation prediction results.

9. A high-fidelity welding simulation boundary condition correction system, characterized in that, include: The three-dimensional geometry acquisition module is used to acquire measured three-dimensional geometric data of the weldment under clamping and constraint conditions after the weldment is clamped and before welding is performed. The data conversion module spatially registers the measured three-dimensional geometric data with the simulation theoretical model, extracts the clamping deformation caused by clamping, and reconstructs the clamping deformation into a simulation mesh node displacement constraint field. The boundary condition correction module adaptively corrects the boundary condition parameters of the welding simulation model based on the nodal displacement constraint field. The simulation calculation module uses the modified simulation model to perform welding thermo-elastic-plastic finite element calculations and outputs welding simulation results that take clamping deformation into account.

Images