Multi-factor fusion prediction method for stamping fracture risk of integrated door ring laser-welded plate

Abstract

This application provides a multi-factor fusion prediction method for stamping fracture risk of an integrated door ring laser-welded plate, relating to the field of stamping inspection technology. The method includes the following steps: creating a grid pattern covering the weld and base material on both sides of a blank weld seam formed by laser welding at least two steel plates of different thicknesses or materials; stamping the blank; measuring the deformation of each grid on both sides of the weld seam after stamping, obtaining the principal strain value of the grid in the direction perpendicular to the weld seam, and calculating the strain gradient; obtaining the measured hardness value of the weld seam, the elongation of the base material on the thicker side, and the thickness ratio of the base materials on both sides; calculating an equivalent fracture risk index based on the above parameters, comparing it with a critical risk threshold, and determining whether stamping fracture risk exists. This method can accurately predict fracture risk, is compatible with traditional production infrastructure, and reduces trial molding costs.

Description

Technical Field

[0001] This invention belongs to the field of stamping inspection technology, specifically relating to a multi-factor fusion prediction method for stamping fracture risk of integrated door ring laser-welded plate. Background Technology

[0002] The integrated door ring is a key structural component for lightweighting automobile bodies. It is made by using laser welding technology to connect multiple steel plates of different thicknesses or materials into a single blank, which is then stamped into a single piece. Due to the differences in thickness and strength of the base materials on both sides of the weld, the stress distribution in the area near the weld is extremely complex during the stamping process. The heat-affected zone of the weld also experiences localized softening due to the rapid cooling effect of the laser. The coupling effect of multiple factors makes the weld a high-risk location for stamping fracture.

[0003] Currently, stamping workshops mainly rely on two methods to assess the risk of welded plate breakage: one is to predict the forming limit based on finite element simulation. However, the accuracy of the simulation is highly dependent on the material constitutive model and the setting of boundary conditions. For areas such as the heat-affected zone of the weld, where material properties change drastically within a millimeter-level space, the simulation results often deviate significantly from the actual stamping effect. The other method is to measure the deformation distribution of the stamped parts through mesh strain analysis during the trial molding stage. However, this method can only obtain strain values ​​and cannot effectively integrate various factors affecting breakage, such as strain concentration, weld softening, and differences in base material thickness. This results in a one-sided assessment and makes it difficult to directly use it for quantitative determination of breakage risk in production. Summary of the Invention

[0004] In view of the above-mentioned defects or deficiencies in the prior art, a multi-factor fusion prediction method for the stamping fracture risk of an integrated door ring laser-welded plate is provided, including the following steps: A grid pattern is made on both sides of the weld seam of the laser-welded plate blank. The grid pattern covers the weld seam and the base material area on both sides of the weld seam. The laser-welded plate blank is formed by connecting at least two steel plates of different thicknesses or materials through laser welding. The laser-welded blank with the grid pattern is stamped to obtain the stamped part; The deformation of each grid on both sides of the weld on the stamped part is measured, the principal strain value of each grid arranged perpendicular to the weld direction is obtained, and the strain gradient on both sides of the weld is calculated. The strain gradient is the rate of change of the principal strain value of multiple grids extending from the weld position to the base material direction on one side. Obtain the measured hardness value of the weld area and the material elongation of the thicker base material on both sides of the weld. Based on the strain gradient, the measured hardness value, the material elongation of the thicker base material, and the thickness ratio of the base materials on both sides of the weld, the equivalent fracture risk index is calculated. The equivalent fracture risk index is compared with a preset critical risk threshold. If the equivalent fracture risk index exceeds the critical risk threshold, it is determined that there is a risk of stamping fracture.

[0005] According to the technical solution provided in this application, the process of creating a grid pattern on both sides of the weld seam of a laser-welded plate blank includes the following steps: Based on the weld seam orientation of the laser-welded plate blank, the entire weld seam is divided into straight weld seam segments and curved weld seam segments; A grid pattern is made on the straight weld section and the curved weld section respectively. The grid on the straight weld section is a square grid, and the grid on the curved weld section is a radial grid. The grid lines of the radial grid are arranged along the weld tangent direction and the weld normal direction respectively, so as to accommodate the deformation difference of the curved weld section along the weld tangent direction and the weld normal direction during the stamping process.

[0006] According to the technical solution provided in this application, the calculation of the strain gradient on both sides of the weld includes the following steps: In a series of grids arranged perpendicular to the weld direction on one side of the weld, starting from the grid closest to the weld centerline, the difference between the principal strain values ​​of two adjacent grids is taken in sequence in the direction away from the weld to obtain a set of principal strain changes piecewise. Calculate the maximum value among the absolute values ​​of the segmental changes in the principal strain, and take it as the maximum inter-segment change on that side; Divide the maximum inter-segment variation by the distance between the center points of two adjacent grids to obtain the strain gradient on that side; Perform the above steps on both sides of the weld, and take the larger absolute value of the strain gradient on both sides as the strain gradient on both sides of the weld.

[0007] According to the technical solution provided in this application, taking the larger absolute value of the strain gradients on both sides as the strain gradient on both sides of the weld includes the following steps: Obtain the elongation of the base materials on both sides of the weld; Divide the absolute value of the strain gradient on one side of the weld by the elongation of the base material on that side to obtain the normalized strain gradient on that side. Perform the same operation on the other side of the weld to obtain the normalized strain gradient on the other side. The original strain gradient corresponding to the larger of the normalized strain gradients on both sides is taken as the strain gradient on both sides of the weld.

[0008] According to the technical solution provided in this application, before calculating the equivalent fracture risk index, the following steps are also included: The strain gradient is compared with a preset gradient activation threshold; If the strain gradient is less than the gradient activation threshold, it is determined that the weld area has not yet entered a high deformation gradient state, which is used to indicate that the measured hardness value is included in the calculation with the first participation weight when calculating the equivalent fracture risk index. If the strain gradient is greater than or equal to the gradient activation threshold, it is determined that the weld area has entered a high deformation gradient state, which is used to indicate that the measured hardness value is included in the calculation with a second participation weight when calculating the equivalent fracture risk index. The second participation weight is greater than the first participation weight.

[0009] According to the technical solution provided in this application, the gradient activation threshold is determined in the following way: From the stamped part, select a region on the thinner side of the base material on both sides of the weld, and whose distance from the weld is greater than a preset distance threshold, as the reference base material region; Within the reference base material region, principal strain values ​​of multiple grids are extracted along the direction perpendicular to the weld. The average value of the principal strain values ​​of the multiple grids within the reference base material region is calculated and used as the base material reference strain value. The reference strain value of the parent material is used as the gradient activation threshold.

[0010] According to the technical solution provided in this application, the first participation weight is a preset benchmark value; the second participation weight is determined by calculating the ratio of the strain gradient to the benchmark strain value of the base material as the weld deformation concentration multiple; the second participation weight is equal to the product of the first participation weight and the weld deformation concentration multiple.

[0011] According to the technical solution provided in this application, the critical risk threshold takes different values ​​on the straight weld segment and the curved weld segment, wherein the critical risk threshold on the curved weld segment is lower than the critical risk threshold on the straight weld segment.

[0012] According to the technical solution provided in this application, obtaining the measured hardness value of the weld area includes the following steps: On the stamped part, the initial hardness values ​​of multiple measuring points are obtained sequentially along the weld seam. The multiple measuring points are distributed on the weld seam centerline and at positions offset by a preset distance on both sides of the weld seam centerline. The minimum value among the initial hardness values ​​from multiple measuring points is selected as the measured hardness value of the weld area.

[0013] According to the technical solution provided in this application, before stamping the laser-welded plate blank with the grid pattern, the following steps are also included: From the laser-welded plate blank that has completed the mesh pattern production but has not yet been stamped, select multiple meshes located on both sides of the weld and adjacent to the weld as initial reference meshes. Measure the initial deformation of the initial reference mesh; If the initial deformation exceeds the preset initial deformation threshold, the deformation offset of the initial reference mesh is recorded. Measuring the deformation of each grid on both sides of the weld seam on the stamped part includes the following steps: Subtract the deformation offset corresponding to the grid from the deformation measured after stamping to obtain the net deformation of the grid caused by stamping.

[0014] Compared with the prior art, the beneficial effects of this application are as follows: I. Effective Integration of Multiple Fracture Risk Factors. This method integrates four factors—strain gradient on both sides of the weld, measured hardness value of the weld region, material elongation of the thicker base material, and thickness ratio of the base materials on both sides of the weld—into a unified equivalent fracture risk index. Specifically, the strain gradient reflects the degree of deformation concentration near the weld, the measured hardness value reflects the degree of local softening in the heat-affected zone of the weld, the material elongation of the thicker base material reflects the plastic deformation capacity of the base material on the side primarily bearing the deformation, and the thickness ratio reflects the degree of uneven stress distribution between the base materials on both sides of the unequal-thickness welded plate. These four factors characterize fracture risk from four dimensions: deformation concentration, material weakening, intrinsic plasticity, and structural differences. The integrated index can more comprehensively reflect the true risk level of stamping fracture in welded plates.

[0015] Second, all input data are derived from actual production process measurements and do not rely on finite element simulation. Strain gradients are directly obtained from mesh deformation measurements of the stamped parts, hardness values ​​are directly obtained from hardness testing of the weld area, and material elongation and thickness ratios are directly obtained from the material specifications of the base material. This avoids calculation errors caused by the difficulty in accurately setting material parameters in the weld heat-affected zone, and the evaluation results directly reflect the actual fracture risk of the welded plate under actual stamping conditions.

[0016] Third, the evaluation process has clear operational steps and quantitative criteria, and can be implemented directly on the stamping production site. Mesh fabrication adopts conventional electrochemical etching or laser etching processes, mesh deformation measurement adopts a general mesh strain measurement system, and hardness testing adopts a portable hardness tester. All operations can be completed under the existing equipment conditions of traditional mold manufacturing enterprises, without the need to purchase additional expensive testing equipment. Attached Figure Description

[0017] Other features, objects, and advantages of this application will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1 A flowchart illustrating the steps of the multi-factor fusion prediction method for the stamping fracture risk of the integrated door ring laser-welded plate provided in this application. Detailed Implementation

[0018] The present application will now be described in further detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and not intended to limit it. Furthermore, it should be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings.

[0019] It should be noted that, unless otherwise specified, the embodiments and features described in this application can be combined with each other. This application will now be described in detail with reference to the accompanying drawings and embodiments.

[0020] As mentioned in the background section, this application proposes a multi-factor fusion prediction method for the stamping fracture risk of an integrated door ring laser-welded plate, such as... Figure 1 As shown, it includes the following steps: S1. A grid pattern is made on both sides of the weld seam of the laser-welded plate blank. The grid pattern covers the weld seam and the base material area on both sides of the weld seam. The laser-welded plate blank is formed by connecting at least two steel plates of different thicknesses or materials through laser welding. S2. The laser-welded blank with the grid pattern is stamped to obtain the stamped part; S3. Measure the deformation of each grid on both sides of the weld on the stamped part, obtain the principal strain value of each grid arranged perpendicular to the weld direction, and calculate the strain gradient on both sides of the weld. The strain gradient is the rate of change of the principal strain value of multiple grids extending from the weld position to the base material direction on one side. S4. Obtain the measured hardness value of the weld area and the material elongation of the thicker base material on both sides of the weld. S5. Based on the strain gradient, the measured hardness value, the material elongation of the thicker base material, and the thickness ratio of the base materials on both sides of the weld, calculate the equivalent fracture risk index. S6. Compare the equivalent fracture risk index with a preset critical risk threshold. If the equivalent fracture risk index exceeds the critical risk threshold, it is determined that there is a risk of stamping fracture.

[0021] Specifically, the first step is to create a grid pattern on both sides of the weld seam of the laser-welded plate blank. The laser-welded plate blank is made of at least two steel plates of different thicknesses or materials, such as 22MnB5 steel plates with thicknesses of 1.2 mm and 1.8 mm. Before creating the grid pattern, the surface of the blank is cleaned to remove oil and oxide layers. The grid pattern is created using an electrochemical etching method. The specific operation is as follows: a wire mesh template with a grid pattern is placed on the surface of the blank, and the mesh holes on the template form the grid lines to be etched; an electrode dipped in electrolyte is moved against the template, and after energizing, the exposed metal surface is electrolytically etched to form grid grooves with a depth of 0.01 to 0.03 mm. The grid is square, with a side length of 2 mm and a line width of 0.2 mm. The grid coverage extends at least 15 mm to each side from the weld seam to ensure that the strain distribution near the weld seam can be completely captured after stamping. After etching, the blank is cleaned and dried, and the grid pattern is completed.

[0022] The second step involves stamping the blank with the grid pattern. The blank, heated or at room temperature, is placed into a mold. After the upper and lower molds close, the sheet metal flows within the mold cavity to form a one-piece door ring part. During stamping, due to differences in thickness and material, the plastic flow of the base material on both sides of the weld is asynchronous, resulting in strain concentration near the weld. After stamping, the mold is opened and the part is removed. At this point, the grid on the part's surface is distorted due to material deformation, and this distorted grid records this deformation process.

[0023] The third step involves measuring the deformation of each grid on both sides of the weld on the stamped part, obtaining the principal strain values, and calculating the strain gradient. A grid strain measurement system, including a high-resolution industrial camera and a computer, is used. During operation, the part is fixed, and the camera is perpendicularly aimed at the grid area to capture clear images. The computer software identifies the node positions of the distorted grid and compares them with the node positions of the standard grid before deformation. Specifically, the strain of a single grid element is calculated as follows: four corner points of the grid are taken, and their original coordinates on the blank and their current coordinates on the image after stamping are recorded. Using the displacements of the four corner points, the deformation gradient tensor of the grid is calculated using finite strain theory. Then, the right Cauchy-Green tensor is solved from the deformation gradient tensor, and the square root of its eigenvalue minus 1 is the principal strain value of the grid. After calculating the principal strain values ​​of all grids, they are sequentially arranged along the direction perpendicular to the weld, starting from the weld centerline and moving towards one side of the base material, to extract the principal strain values ​​of each grid, forming a sequence. The strain gradient is defined as the rate of change of the principal strain values ​​in this sequence; a larger rate of change indicates that the deformation is more concentrated near the weld.

[0024] The fourth step is to obtain the measured hardness value of the weld area and the elongation of the thicker base material on both sides of the weld. Using a portable Vickers hardness tester, three measuring points are selected on the weld surface of the stamped part, with a spacing of 1 mm between the measuring points and the center of the measuring points located on the center line of the weld. After applying a test force to each measuring point, the hardness value is read, and the average of the three values ​​is taken as the measured hardness value of the weld area. The elongation of the material is obtained directly from the material certificate or incoming inspection report of the base material, selecting the data from the thicker base material.

[0025] The fifth step is to calculate the equivalent fracture risk index. This involves calculating the risk index using a pre-defined fusion formula, which combines four factors: strain gradient, measured hardness value, elongation of the thicker base material, and the thickness ratio of the base materials on both sides of the weld. An example fusion formula is: the risk index equals the strain gradient multiplied by the hardness softening ratio multiplied by the thickness ratio raised to a power, then divided by the elongation. The hardness softening ratio is the original hardness of the softer base material divided by the measured hardness value of the weld. This formula results in a higher index value for more concentrated deformation, a softer weld, greater thickness differences, and poorer base material plasticity.

[0026] The sixth step is to compare the equivalent fracture risk index with the preset critical risk threshold. The critical risk threshold is determined based on the statistical analysis of risk indices of multiple parts from historically qualified batches of the mold, specifically the average of the historical qualified indices plus three times the standard deviation. If the current index exceeds the threshold, a stamping fracture risk is identified; otherwise, the risk is considered controllable.

[0027] The technical advantage of this implementation method is that by extracting the degree of deformation concentration through mesh-based strain measurement, and combining the weld hardness softening, base material plasticity and thickness differences, multiple independent risk factors are integrated into a quantifiable index, thereby achieving an objective judgment of fracture risk based on actual production data and overcoming the problem that simulation methods are difficult to accurately set the material parameters of the weld area.

[0028] In a preferred embodiment, creating a grid pattern on both sides of the weld seam of the laser-welded plate blank includes the following steps: Based on the weld seam orientation of the laser-welded plate blank, the entire weld seam is divided into straight weld seam segments and curved weld seam segments; A grid pattern is made on the straight weld section and the curved weld section respectively. The grid on the straight weld section is a square grid, and the grid on the curved weld section is a radial grid. The grid lines of the radial grid are arranged along the weld tangent direction and the weld normal direction respectively, so as to accommodate the deformation difference of the curved weld section along the weld tangent direction and the weld normal direction during the stamping process.

[0029] Specifically, the laser-welded seam of the integrated door ring is not entirely straight; it has curved sections at the corners. During stamping, the straight sections primarily bear tension perpendicular to the weld, with the principal strain direction aligned with the weld normal. However, the curved sections bear both normal tension and tangential shear during stamping, with the principal strain direction continuously changing along the weld bend. If the curved sections also use the same square grid as the straight sections, the square grid will be distorted after stamping, and its principal strain direction will no longer align with the original grid line direction. Extracting the principal strain value along the original normal direction at this point will result in errors due to direction mismatch, leading to distorted strain gradient calculations. This implementation uses a differentiated grid design to make the grid line direction adapt to the weld direction, ensuring that the extracted principal strain direction after deformation is consistent with the actual deformation direction.

[0030] In practice, the first step is to divide the weld seam into straight and curved segments based on the geometry of the weld centerline. From the weld geometry data of the laser-welded blank, samples are taken at equal intervals along the weld centerline, and the radius of curvature at each sampling point is calculated. A radius of curvature threshold is set, for example, 300 mm. Continuous segments with a radius of curvature greater than this threshold are designated as straight weld seams, and segments with a radius of curvature less than or equal to this threshold are designated as curved weld seams. In actual door rings, generally the four corners are curved segments, and the rest are straight segments.

[0031] The second step involves creating a square grid on the straight weld section. The square grid has a side length of 2 mm, with one set of grid lines parallel to the weld centerline and the other set perpendicular to it. The fabrication method is the same as described previously, employing electrochemical etching. The grid orientation is controlled by a wire mesh template, ensuring precise alignment of the grid lines with the weld direction. This ensures that when extracting the principal strain values ​​along the direction perpendicular to the weld after stamping, the grid line direction is highly consistent with the principal deformation direction, resulting in reliable strain measurements.

[0032] The third step is to create a radially radiating grid on the curved weld segment. The configuration of the radially radiating grid is as follows: first, determine the center of curvature of the curved weld segment; then, using this point as the center, draw a set of radial lines, the direction of which is the normal direction at each point of the weld; next, draw a family of arcs orthogonal to the radial lines, the direction of which is the tangent direction at each point of the weld. The grid cells formed by the interweaving of radial and arc lines are approximately fan-shaped at the weld, with a grid spacing of 2 mm along both the normal and tangential directions. During fabrication, the pattern of the screen template is designed based on parameters such as the arc length and radius of curvature of the curved weld segment. Computer-aided drawing is used to generate the radial pattern, which is then transferred to the sheet metal surface through electrochemical etching. For curved segments with different curvatures, corresponding templates need to be fabricated separately to ensure the accuracy of the grid line direction.

[0033] After stamping, in the curved weld section, the grid lines in the radial direction are arranged along the normal direction. The principal strain values ​​of each grid can be directly extracted along this direction for calculating the strain gradient. Since the grid line direction is adaptively consistent with the principal direction of stamping deformation, the extracted principal strain values ​​do not require direction correction, thus ensuring the accuracy of the strain gradient calculation for the curved weld section.

[0034] The technical advantage of this embodiment is that by introducing a differentiated grid that matches the geometry of the weld, the measurement error caused by grid orientation mismatch in the curved section is eliminated, and the reliability of the risk assessment of the entire weld fracture is improved. In particular, it achieves accurate deformation capture in high-risk areas such as door ring corners.

[0035] In a preferred embodiment, calculating the strain gradient across the weld includes the following steps: In a series of grids arranged perpendicular to the weld direction on one side of the weld, starting from the grid closest to the weld centerline, the difference between the principal strain values ​​of two adjacent grids is taken in sequence in the direction away from the weld to obtain a set of principal strain changes piecewise. Calculate the maximum value among the absolute values ​​of the segmental changes in the principal strain, and take it as the maximum inter-segment change on that side; Divide the maximum inter-segment variation by the distance between the center points of two adjacent grids to obtain the strain gradient on that side; Perform the above steps on both sides of the weld, and take the larger absolute value of the strain gradient on both sides as the strain gradient on both sides of the weld.

[0036] Specifically, in grid strain measurement, due to factors such as camera resolution and grid line recognition errors, the principal strain values ​​of each grid contain a certain degree of random noise. If a linear fitting of the entire segment is used to calculate the slope, local noise will be averaged out, potentially missing the true local strain concentration peaks. Furthermore, the stamping fracture of welded plates is usually triggered by a severe deformation gradient in a very narrow region near the weld, not by the average gradient along the entire measurement line. Therefore, this implementation method uses a segmented differential method, taking the maximum change and then dividing by the interval, to directly lock onto the local segment with the most concentrated deformation, as the strain gradient of the entire measurement line.

[0037] In practical implementation, taking one side of the weld as an example, the first step is to select n grids sequentially outward from the weld centerline on the image of the stamped part, perpendicular to the weld direction, with n being at least 5. The center points of these grids are roughly arranged along a normal line. The principal strain values ​​of each grid are extracted and denoted as the sequence ε1, ε2, ..., ε n The subscript 1 corresponds to the grid closest to the weld.

[0038] The second step involves calculating the difference between the principal strain values ​​of two adjacent grids, starting from ε1, to obtain a set of segment-by-segment changes in principal strain Δε. i = |εi+1 -ε i | (i=1 to n-1). The absolute value is taken here because the risk of rupture is related to the magnitude of change, and not to the sign of the direction of change.

[0039] The third step is to start from this set of Δε i Find the maximum value Δε_max. The adjacent mesh pair corresponding to this maximum value is the local location where the strain change is most drastic on that side of the parent material, which is often the crack initiation point.

[0040] The fourth step is to read the distance d between the center points of the two meshes in the direction perpendicular to the weld. This distance is determined during mesh creation; for example, when the side length of a square mesh is 2 mm and there are no gaps, the center-to-center distance between adjacent meshes is 2 mm. If there are gaps between meshes, calculate based on the actual center-to-center distance. Divide Δε_max by d to obtain the strain gradient G1 on that side.

[0041] Fifth, perform the same steps on the base material on the other side of the weld to obtain the strain gradient G2 on the other side.

[0042] Step 6: Compare |G1| and |G2|, and take the larger one as the strain gradient G at that weld location. The reason for taking the larger one is that fracture may occur on either side, and the deformation capacity of the base materials on both sides may be different. Choosing the side with more concentrated deformation as the risk assessment input can avoid missing risks by only taking one side.

[0043] The technical advantage of this implementation method is that it replaces the complex full-segment fitting with simple difference and maximum value operations. The calculation process is clear, it has strong resistance to local noise interference, and it can sensitively capture the extremely narrow high gradient region that may exist near the weld, thereby improving the accuracy of strain gradient calculation and the sensitivity of early warning of rupture risk.

[0044] In a preferred embodiment, taking the larger absolute value of the strain gradients on both sides as the strain gradient on both sides of the weld includes the following steps: Obtain the elongation of the base materials on both sides of the weld; Divide the absolute value of the strain gradient on one side of the weld by the elongation of the base material on that side to obtain the normalized strain gradient on that side. Perform the same operation on the other side of the weld to obtain the normalized strain gradient on the other side. The original strain gradient corresponding to the larger of the normalized strain gradients on both sides is taken as the strain gradient on both sides of the weld.

[0045] Specifically, in previous technical solutions, strain gradients were calculated on both sides of the weld, and the absolute values ​​of the two gradients were directly compared, with the larger one being taken as the strain gradient at that weld location. This approach implicitly assumes that the plastic deformation tolerance of the base materials on both sides of the weld is the same, and the side with a larger gradient value has a higher risk of fracture. However, in integrated door ring welded plates, the thickness and material of the base materials on both sides of the weld are often different, and their elongation rates are also different. Base materials with higher elongation rates can withstand greater local deformation without necking fracture, while base materials with lower elongation rates, even with a smaller absolute gradient value, may already be close to their plastic limit. If the difference in elongation rates between the two base materials is ignored and the absolute values ​​of the gradients are simply compared, two misjudgments can occur: First, the gradient on the side with lower elongation rates may be slightly lower, but it is actually on the verge of fracture, which is masked by the larger gradient on the opposite side with higher elongation rates, leading to underreporting of risk; second, the gradient on the side with higher elongation rates may be large, but it is far from reaching its fracture limit, yet it is judged as the dangerous side, leading to overestimation of risk. This embodiment obtains a normalized strain gradient by dividing the gradient values ​​on both sides by the elongation of their respective parent materials, allowing the gradients on both sides to be compared under the same plasticity criterion, thus eliminating the interference of the intrinsic plasticity difference of the parent materials on the judgment of the gradient danger level.

[0046] The specific operation steps of this implementation method are as follows.

[0047] The first step is to obtain the elongation of the base metal on both sides of the weld. Elongation is the engineering strain value of the base metal at fracture in a uniaxial tensile test, usually expressed as a percentage. This data can be obtained directly from the material certificate or incoming inspection report of the base metal. If the base metals on both sides of the weld are different grades of steel plates, find their respective elongation values. For example, if one side of the weld is 22MnB5 steel plate, the elongation is approximately 6%; the other side is DP980 steel plate, the elongation is approximately 12%. If the base metals on both sides are of the same grade but different thicknesses, the elongation values ​​are taken as the same. In this case, the normalization operation is equivalent to directly comparing the absolute values ​​of the gradients, and the method is still applicable. If the base metal elongation data is not provided upon arrival at the factory, tensile specimens of each base metal can be cut from the blank and measured according to the standards for tensile testing of metallic materials.

[0048] The second step involves dividing the calculated absolute value of the strain gradient on one side of the weld by the elongation of the base material on that side to obtain the normalized strain gradient. Specifically, the absolute value of the gradient is divided by the percentage elongation (without the percentage sign). For example, if the absolute value of the gradient on one side is 0.08 mm and the elongation of the base material on that side is 8%, then the normalized strain gradient is 0.08 divided by 8, which equals 0.01. The physical meaning of this normalized value is the strain gradient borne per unit elongation of the base material on that side, i.e., the ratio of the degree of deformation concentration to the base material's own plastic reserve. A larger normalized value indicates a higher degree of consumption of the base material's plastic reserve by deformation concentration, and a greater risk of fracture.

[0049] The third step is to perform the exact same operation on the other side of the weld to obtain the normalized strain gradient on that side. The calculation steps are the same as in the second step, taking the absolute value of the gradient on that side and dividing it by the elongation of the base material.

[0050] The fourth step is to compare the magnitudes of the normalized strain gradients on both sides. The original strain gradient corresponding to the larger normalized strain gradient—that is, the original gradient value used in the second step for that side—is taken as the strain gradient at that weld location. The reason for choosing the larger one is that the side with the larger normalized gradient is the side where the plastic reserve is more severely depleted by deformation, indicating a higher risk of fracture. Taking the original gradient value rather than the normalized value as the output is to maintain the physical dimensions of the strain gradient in the output, ensuring full compatibility with the calculation interfaces mentioned earlier or later, and without altering the calculation logic of the downstream equivalent fracture risk index.

[0051] The technical effect and principle of this embodiment lies in introducing the intrinsic plasticity parameter of the base material elongation into the comparison of strain gradients, so that the comparison of the danger levels of the gradients on both sides is based on the same plasticity benchmark. This normalization operation is equivalent to uniformly converting the deformation bearing capacity of the base materials on both sides to the same scale, and then using the gradient value to measure the degree of consumption of each bearing capacity, with the side with more severe consumption being taken as the final output.

[0052] In a preferred embodiment, prior to calculating the equivalent rupture risk index, the following step is further included: The strain gradient is compared with a preset gradient activation threshold; If the strain gradient is less than the gradient activation threshold, it is determined that the weld area has not yet entered a high deformation gradient state, which is used to indicate that the measured hardness value is included in the calculation with the first participation weight when calculating the equivalent fracture risk index. If the strain gradient is greater than or equal to the gradient activation threshold, it is determined that the weld area has entered a high deformation gradient state, which is used to indicate that the measured hardness value is included in the calculation with a second participation weight when calculating the equivalent fracture risk index. The second participation weight is greater than the first participation weight.

[0053] Specifically, when calculating the equivalent fracture risk index, four factors are involved: strain gradient, measured hardness value, elongation of the thicker base material, and the thickness ratio of the base materials on both sides of the weld. This calculation method, which involves all factors throughout, implicitly assumes that the contribution of the hardness softening factor to the fracture risk is constant under any deformation state. However, in the actual physical process of stamping integral door ring welded plates, the danger of weld hardness softening is not constant but closely related to the degree of deformation concentration. Under conditions of small stamping deformation and low strain gradient, the sheet deformation is relatively uniform. Even if the hardness in the weld area is low, the material can adapt to the deformation through uniform plastic flow. At this time, the fracture contribution of hardness softening is not significant, and the hardness factor should not occupy a large proportion in the risk assessment. When the stamping deformation increases and the strain gradient rises to a certain level, the deformation is highly concentrated in a narrow area near the weld. At this time, the heat-affected zone of the weld with low hardness becomes the breakthrough point of deformation concentration. The coupling effect of hardness softening and strain concentration significantly amplifies the fracture risk, and the participation of the hardness factor in the risk assessment should be increased. If these phase differences are ignored and the hardness factor is always included in the calculation with the same weight, the risk contribution of hardness softening will be overestimated in the low gradient stage, leading to false alarms; while in the high gradient stage, the coupling and amplification effect of hardness softening and strain concentration will be underestimated, leading to missed risks. The solution in this embodiment divides the state of strain gradient into low gradient and high gradient by preset a gradient activation threshold, and assigns different participation weights to the hardness factor in different levels, so that the risk assessment contribution of the hardness factor matches the actual degree of deformation concentration.

[0054] The specific operation steps of this implementation method are as follows.

[0055] The first step is to compare the calculated strain gradient with a preset gradient activation threshold. The gradient activation threshold is the boundary value that distinguishes between low deformation gradient states and high deformation gradient states. This comparison operation is performed before calculating the equivalent fracture risk index, serving as a judgment step in determining the weighting of the stiffness factor.

[0056] The second step involves adopting different weighting strategies based on the comparison results. If the strain gradient is less than the gradient activation threshold, the weld region is determined not to have entered a high deformation gradient state. In this state, deformation is relatively uniform, and the contribution of weld hardness softening to fracture is insignificant. At this point, the measured hardness value is used as the first participating weight in the calculation of the equivalent fracture risk index. The first participating weight is a preset benchmark value, whose function is to keep the contribution of the hardness factor to the risk index at a low baseline level in the low gradient state. Risk assessment is mainly driven by three factors: strain gradient, base metal elongation, and thickness ratio.

[0057] If the strain gradient is greater than or equal to the gradient activation threshold, the weld region is determined to have entered a high deformation gradient state. In this state, deformation is highly concentrated near the weld, and the hardness softening zone becomes a weak point in the deformation concentration, with a significant amplification effect from the coupling of hardness softening and strain concentration. Therefore, when calculating the equivalent fracture risk index, the measured hardness value is included in the calculation with a second participation weight. This second participation weight is greater than the first participation weight, and its function is to amplify the contribution of the hardness factor to the risk index under high gradient conditions, reflecting the true physical effect of the coupling amplification of fracture risk by hardness softening and strain concentration.

[0058] The third step involves selecting the corresponding weights based on the judgment results, then combining the selected weights with the measured hardness values ​​to proceed to the calculation of the equivalent fracture risk index. The specific combination method of the weights in the calculation depends on the fusion formula of the equivalent fracture risk index. For example, in the product form of the fusion formula, the weights serve as the exponential coefficient of the hardness softening term; in the weighted summation form of the fusion formula, the weights serve as the multiplier of the hardness term.

[0059] The technical effect and principle of this implementation method lies in upgrading the hardness softening factor from static participation to conditionally triggered participation, enabling the fracture risk assessment model to more accurately match the actual physical process of welded plate stamping. By introducing a gradient activation threshold, the hardness softening factor is suppressed in the low gradient stage, avoiding false alarms for normal production conditions; it is activated and amplified in the high gradient stage, enhancing the detection sensitivity for high-risk conditions.

[0060] In a preferred embodiment, the gradient activation threshold is determined in the following manner: From the stamped part, select a region on the thinner side of the base material on both sides of the weld, and whose distance from the weld is greater than a preset distance threshold, as the reference base material region; Within the reference base material region, principal strain values ​​of multiple grids are extracted along the direction perpendicular to the weld. The average value of the principal strain values ​​of the multiple grids within the reference base material region is calculated and used as the base material reference strain value. The reference strain value of the parent material is used as the gradient activation threshold.

[0061] Specifically, the gradient activation threshold is a preset judgment limit value used to distinguish between low deformation gradient states and high deformation gradient states. The setting of this threshold directly determines when the hardness factor is activated and amplified. If the threshold is set too high, the hardness factor will not be activated when it should be, leading to missed high-risk warnings; if the threshold is set too low, the hardness factor will be activated when deformation is still uniform, leading to frequent false alarms. Traditionally, process engineers set a fixed value based on experience, but this approach has two problems: first, different parts and dies have different deformation characteristics, making a single general empirical value difficult to apply to all working conditions; second, empirical settings lack objective basis and cannot adaptively adjust with changes in production conditions. This solution does not input the threshold externally, but instead extracts a reference value from the mesh strain data already measured on the stamped part itself as the threshold, achieving self-calibration of the threshold and enabling it to automatically adapt to the actual conditions of the part and die.

[0062] The specific operation steps of this implementation method are as follows.

[0063] The first step is to select a reference base material area from the stamped part. The selection of the reference base material area follows two principles: First, select the thinner side of the base material on both sides of the weld, because the thinner side deforms more and has a higher strain level during stamping. Using the normal deformation level of this side as a reference is more conservative and safer than using the thicker side. Second, select an area whose distance from the weld is greater than a preset distance threshold. The preset distance threshold ensures that this area is far from the weld heat-affected zone and deformation concentration zone. Its strain state represents the uniform deformation level of the base material under normal stamping conditions, rather than local strain concentration caused by the presence of the weld. The preset distance threshold is determined based on the typical width of the weld heat-affected zone, usually five to ten times the weld width. For example, for a weld with a width of 1.5 mm, the distance threshold can be 10 mm. Within the area that meets the above two conditions, delineate a continuous area containing multiple grids as the reference base material area.

[0064] The second step involves extracting principal strain values ​​from multiple meshes within the reference base material region, perpendicular to the weld direction. The extraction direction is consistent with the extraction direction used in calculating the strain gradient, both perpendicular to the weld direction. These extracted principal strain values ​​are then grouped into a sequence, and the arithmetic mean of this sequence is calculated. Specifically, the sum of all principal strain values ​​in the sequence is divided by the number of meshes; the result is the base material reference strain value. The physical meaning of the base material reference strain value is the level of uniform deformation occurring in the normal base material region far from the weld influence under current stamping process conditions, reflecting the reference strain of the base material under normal service conditions.

[0065] The third step is to directly use the base metal reference strain value as the gradient activation threshold. No multiplication by any coefficient or adjustment is required. The physical logic behind this approach is that when the strain gradient near the weld has not yet exceeded the uniform deformation level of the normal base metal, the deformation concentration in the weld area is still within the normal deformation range of the base metal, and the risk of hardness softening has not been activated. When the strain gradient near the weld exceeds the uniform deformation level of the normal base metal, the deformation concentration in the weld area is significantly different from the uniform deformation of the base metal, and the risk of hardness softening is activated. The base metal reference strain value naturally distinguishes the boundary between normal deformation and abnormal concentration.

[0066] The technical effect and principle of this implementation method lies in transforming the determination of the gradient activation threshold from external empirical input to self-calibration based on the measured data of the part itself, enabling the threshold to automatically adapt to the part structure, mold condition, and stamping process conditions. The selection of the reference base material area follows the principle of using the thinner side of the base material away from the weld, ensuring that the reference data represents the normal deformation level of the base material rather than the abnormal deformation of the weld-affected area. The threshold is dynamically updated with each part, adapting to the drift of the reference strain level caused by factors such as mold temperature changes and batch differences in sheet metal during production, ensuring that the activation judgment of the hardness factor always remains consistent with the current production state.

[0067] In a preferred embodiment, the first participation weight is a preset benchmark value; the second participation weight is determined by calculating the ratio of the strain gradient to the benchmark strain value of the base material as the weld deformation concentration multiple; the second participation weight is equal to the product of the first participation weight and the weld deformation concentration multiple.

[0068] Specifically, the gradient activation threshold is determined to be the base material reference strain value. When the strain gradient is less than this threshold, the hardness factor participates in the calculation of the equivalent fracture risk index with the first participation weight; when the strain gradient is greater than or equal to this threshold, the hardness factor participates in the calculation with the second participation weight, and the second participation weight is greater than the first participation weight. If the first participation weight is set arbitrarily, the reference value of the entire graded control system lacks an objective basis; if the amplification factor of the second participation weight remains fixed, it is impossible to distinguish the risk level differences under different deformation concentration degrees. This scheme sets the first participation weight to a preset reference value and dynamically links the second participation weight to the weld deformation concentration factor, making the weight amplification degree proportional to the severity of deformation concentration, thus achieving refined control of the hardness factor participation degree.

[0069] The specific operation steps of this implementation method are as follows.

[0070] The first step is to set the first participation weight to a preset baseline value. The first participation weight is the participation coefficient of the hardness factor under low deformation gradient conditions. Its function is to maintain a low baseline contribution level of the hardness factor during the uniform deformation stage. The baseline value is set as follows: During the mold trial stage, select a batch of qualified parts confirmed to have no cracking risk. Obtain the values ​​of each factor according to the previous steps. Adjust the participation coefficient of the hardness factor under low gradient conditions so that the calculated equivalent cracking risk index is exactly below the critical risk threshold. This participation coefficient can then be used as the baseline value. A typical value for the baseline value is a constant between 0.5 and 1.0. Once the baseline value is determined during the mold trial stage, it remains unchanged during subsequent mass production.

[0071] The second step is to calculate the weld deformation concentration factor. The weld deformation concentration factor is defined as the ratio of the strain gradient to the base metal reference strain value. The calculation methods for both the strain gradient and the base metal reference strain value have been described previously. The physical meaning of this ratio is how many times the degree of deformation concentration near the weld is compared to the normal uniform deformation level of the base metal. A larger factor indicates a more concentrated deformation at the weld, and a higher risk of tearing in the hardening softening zone. For example, if the strain gradient is 0.12 per millimeter and the base metal reference strain value is 0.04, then the weld deformation concentration factor is 3, meaning that the degree of deformation concentration at the weld is three times the normal uniform deformation of the base metal.

[0072] The third step is to determine the second participation weight. The second participation weight is equal to the product of the first participation weight and the weld deformation concentration factor. For example, if the baseline value of the first participation weight is 0.8 and the weld deformation concentration factor is 3, then the second participation weight is 0.8 multiplied by 3, which equals 2.4. The physical logic of this calculation method is that when the deformation concentration factor is 1, that is, the degree of deformation concentration at the weld is exactly equal to the normal deformation level of the base material, the strain gradient is equal to the gradient activation threshold, and it is at the boundary between low and high gradients. The second participation weight is equal to the first participation weight, and the participation degree of the hardness factor smoothly connects with the low gradient state without abrupt changes. When the deformation concentration factor is greater than one, the deformation at the weld is more concentrated than the uniform deformation of the base material. The larger the factor, the more intense the concentration. The second participation weight increases linearly accordingly, and the participation degree of the hardness factor is proportionally amplified, so that the risk index can reflect the real physical effect of the intensified coupling between deformation concentration and hardness softening.

[0073] The fourth step involves using the determined first or second participation weight to weight the hardness factor when calculating the equivalent fracture risk index. In the product-form fusion formula, the weights are applied by multiplying the hardness softening ratio term by the selected weight and then multiplying it together. In the weighted summation form fusion formula, the normalized value of the hardness term is multiplied by the selected weight and then added to the other terms.

[0074] The technical effect and principle of this embodiment lies in linking the second participation weight to the weld deformation concentration factor, so that the amplification degree of the activated hardness factor is proportional to the severity of the deformation concentration. The more concentrated the deformation, the more dangerous the hardness softening, and the greater the weight of the hardness factor in risk assessment. This dynamic weight design avoids the problems caused by a fixed amplification factor, namely, excessive weight leading to false alarms under moderate concentration levels, and insufficient weight leading to missed alarms under extreme concentration levels.

[0075] In a preferred embodiment, the critical risk threshold takes different values ​​on the straight weld segment and the curved weld segment, wherein the critical risk threshold on the curved weld segment is lower than the critical risk threshold on the straight weld segment.

[0076] Specifically, after the equivalent fracture risk index is calculated, it needs to be compared with a preset critical risk threshold. If it exceeds the threshold, a fracture risk is determined to exist. If the same critical risk threshold is used for the entire weld, the differences in fracture risk tolerance due to different weld geometries are ignored. The stress state of a straight weld segment during the stamping process is mainly uniaxial tension, and the stress state is relatively simple, resulting in a relatively high fracture limit of the weld material under uniaxial tension. Curved weld segments, on the other hand, are subjected to multiple load components including tension, bending, and shear during the stamping process, resulting in a complex stress state. The fracture limit of the weld under complex stress states is significantly lower than that under uniaxial tension. If the same critical risk threshold is used for curved weld segments as for straight weld segments, when the equivalent fracture risk index of a curved weld segment does not reach the threshold of the straight segment but is actually in a dangerous state, the system will misjudge it as safe, leading to missed risk reports. This solution utilizes the completed weld segmentation results to set different critical risk thresholds for straight and curved weld segments, with the threshold for curved weld segments being lower than that for straight weld segments. This ensures that the judgment criteria for risk assessment match the actual fracture tolerance of the weld area.

[0077] The specific operation steps of this implementation method are as follows.

[0078] The first step is to determine the critical risk threshold for the straight weld segment. The threshold determination method for the straight weld segment is as follows: During the mold trial stage, collect equivalent fracture risk index data from multiple production batches along the straight weld segment. Screen out the index values ​​of qualified parts that have not fractured, and calculate the statistical distribution characteristics of these qualified index values. Take the average of the qualified indices plus three times the standard deviation as the critical risk threshold for the straight weld segment. The statistical meaning of this threshold is that, under normal qualified production conditions, the equivalent fracture risk index of the straight weld segment has a 99.7% probability of falling within this threshold; index values ​​exceeding this threshold can be considered abnormal.

[0079] The second step is to determine the critical risk threshold for curved weld segments. Due to the complex stress state, curved weld segments have a lower fracture tolerance than straight weld segments. The threshold for curved weld segments is also based on measured data from the trial molding phase. Equivalent fracture risk index data from multiple production batches on the curved weld segment are collected. The index values ​​of qualified parts that did not fracture are selected, and their average value plus three times the standard deviation is calculated as the critical risk threshold for the curved weld segment. Since the distribution range of the qualified index for curved weld segments is naturally lower than that for straight weld segments, the calculated threshold for curved weld segments will inevitably be lower than that for straight weld segments. Instead of multiplying the threshold for straight segments by a discount factor to determine the threshold for curved segments, it is directly derived from the qualified data of the curved segments themselves, avoiding the problem of subjectively setting the discount factor.

[0080] The third step, during mass production, involves assessing the fracture risk at a specific weld location. First, it's determined whether the location belongs to a straight weld segment or a curved weld segment. This assessment directly relates to the completed weld segmentation. If the location is a straight weld segment, the equivalent fracture risk index is compared to the critical risk threshold for straight weld segments; if the location is a curved weld segment, the equivalent fracture risk index is compared to the critical risk threshold for curved weld segments. The comparison results are processed in the same way as described above; if the index exceeds the corresponding threshold, a fracture risk is identified.

[0081] The technical effect and principle of this implementation method lies in utilizing an established weld geometry segmentation system to refine the critical risk threshold from a single value into segmented values ​​that match the weld geometry. Curved weld segments, due to their complex stress states and low fracture tolerance, are subject to stricter judgment criteria; straight weld segments, due to their simple stress states and high fracture tolerance, are subject to relatively lenient judgment criteria. This differentiated threshold setting method ensures that the correspondence between the equivalent fracture risk index and the actual occurrence of fracture remains consistent across different weld segments, avoiding underreporting in curved segments and false alarms in straight segments caused by a uniform threshold. Both thresholds are calibrated based on actual production qualification data for their respective segments, requiring no external experience input, and are consistently used in production after calibration.

[0082] In a preferred embodiment, obtaining the measured hardness value of the weld area includes the following steps: On the stamped part, the initial hardness values ​​of multiple measuring points are obtained sequentially along the weld seam. The multiple measuring points are distributed on the weld seam centerline and at positions offset by a preset distance on both sides of the weld seam centerline. The minimum value among the initial hardness values ​​from multiple measuring points is selected as the measured hardness value of the weld area.

[0083] Specifically, in actual production, the width of the laser-welded seam in an integrated door ring is typically only one to two millimeters. The indenter diameter of a portable hardness tester is on the same order of magnitude as the weld width. When applying hardness tester pressure to the weld, it is difficult for operators to visually align the indenter precisely with the weld centerline. The actual landing point of the indenter may partially press on the weld, partially on the heat-affected zone of the base material, or even completely off-center onto the base material. Since the hardness of the base material is usually higher than that of the heat-affected zone of the weld, indenter deviation will lead to an overestimation of the measured hardness value, meaning the measured hardness value is higher than the true hardness of the weld. When the degree of hardness softening is underestimated, the calculated equivalent fracture risk index is also lower, potentially leading to missed detections of fracture risk and resulting in batch scrap. Improving the operator's alignment accuracy is impractical for narrow welds; therefore, a measurement strategy is needed that does not rely on precise positioning but can reliably approximate the true minimum hardness of the weld.

[0084] The specific operation steps of this implementation method are as follows.

[0085] The first step involves sequentially acquiring initial hardness values ​​at multiple measuring points along the weld seam on the stamped part. The measuring points are arranged as follows: at least two points are placed on the weld centerline; at least one point is placed at a predetermined distance offset to one side of the weld centerline; and at least one point is also placed at the same predetermined distance offset to the other side of the weld centerline. The predetermined distance is determined based on the width of the heat-affected zone (HAZ) of the laser welding process. The HAZ width is typically 0.5 to 1.5 mm, and the predetermined distance is half the width of the HAZ, ensuring that the offset measuring points cover the HAZ area. For example, if the weld width is 1.5 mm and the HAZ width is approximately 1 mm, and the predetermined distance is 0.5 mm, then the three measuring points would be located at the weld centerline, 0.5 mm to the left, and 0.5 mm to the right. The total number of measuring points should be no less than three; the narrower the weld and the greater the hardness gradient, the more measuring points should be used. Each measuring point is read after applying a test force using a portable Vickers or Rockwell hardness tester.

[0086] The second step involves selecting the minimum value from the initial hardness values ​​at multiple measuring points as the measured hardness value for the weld area. The physical logic behind selecting the minimum value is based on the following facts: the hardness distribution of laser-welded seams exhibits a gradient characteristic. The weld center is the hardest due to the rapid cooling effect of the laser, the heat-affected zone (HAZ) is the softest due to tempering, and the base metal hardness lies between the two. When the indenter is precisely aligned with the weld centerline, the measured hardness is the higher hardness at the weld center; when the indenter deviates from the HAZ, the measured hardness is the lower hardness of the HAZ; and when the indenter is completely off-center from the base metal, the measured hardness is the medium hardness of the base metal. Fracture always begins at the location of the HAZ, where the microstructure is softest and the strength is lowest. Therefore, selecting the minimum value from multiple measuring points is equivalent to automatically filtering the hardness data closest to the softened location of the HAZ. This minimum value represents the true lowest hardness on the weld cross-section and is the hardness value that best reflects the risk of fracture.

[0087] The third step is to use the selected minimum value as the measured hardness value of the weld area for the calculation of the equivalent fracture risk index.

[0088] The technical effect and principle of this implementation method lie in transforming the measurement of narrow weld hardness from a single-point measurement relying on precise operator positioning to a multi-point coverage plus minimum value screening strategy that tolerates positioning deviations. Operators do not need to precisely align on the millimeter-narrow weld; they only need to mark multiple points on the weld and its adjacent area. The system automatically identifies the softest location by taking the minimum value. The physical basis of this method is the fundamental principle of fracture mechanics that fracture begins at the weakest point. Assessing risk using the hardness of the softest point aligns with the safety principle, which dictates that it is better to overestimate risk than to underestimate it.

[0089] In a preferred embodiment, before stamping the laser-welded blank with the grid pattern, the method further includes the following steps: From the laser-welded plate blank that has completed the mesh pattern production but has not yet been stamped, select multiple meshes located on both sides of the weld and adjacent to the weld as initial reference meshes. Measure the initial deformation of the initial reference mesh; If the initial deformation exceeds the preset initial deformation threshold, the deformation offset of the initial reference mesh is recorded. Measuring the deformation of each grid on both sides of the weld seam on the stamped part includes the following steps: Subtract the deformation offset corresponding to the grid from the deformation measured after stamping to obtain the net deformation of the grid caused by stamping.

[0090] Specifically, after welding and mesh fabrication are completed, the integrated door ring laser-welded plates are typically not immediately put into the stamping process, but are stored and transported in a stacked manner. Due to the different thicknesses of the base materials on both sides of the weld, there is a gap between the plates at the weld position during stacking. The contact pressure between the thinner base material and the upper plate is small, while the contact pressure between the thicker base material and the upper plate is large. Under the weight of the stack and the vibration during transportation, a small amount of elastic pre-deformation may occur near the weld. This pre-deformation is almost invisible to the naked eye, but it causes a small initial deformation of the mesh pattern pre-fabricated on the blank before stamping near the weld. If this initial deformation is ignored and the mesh deformation measured after stamping is directly attributed entirely to the stamping process, the actual deformation near the weld will be systematically overestimated, leading to an overestimation of the equivalent fracture risk index and triggering false alarms.

[0091] This embodiment does not attempt to eliminate pre-deformation at the physical level, as that would require changing the stacking method or adding interlayer pads, which is costly and affects production efficiency. Instead, it employs a data-level compensation strategy. Before stamping, the initial deformation of the initial reference grid is measured and recorded. After stamping, this initial deformation is subtracted from the total deformation to obtain the net deformation, thus eliminating the influence of pre-deformation at the data level. This approach is similar to performing a tare and zeroing operation before using measuring instruments, except that the tare operation is performed post-hoc at the data level.

[0092] The specific operation steps of this implementation method are as follows.

[0093] The first step is to select initial reference grids from the laser-welded blank, where the grid pattern has been completed but the stamping process has not yet begun. The initial reference grids are selected from the grids on both sides of the weld and immediately adjacent to it. "Immediately adjacent to the weld" means that the perpendicular distance between the grid centerline and the weld centerline is within a preset range, determined based on the grid side length and the width of the weld heat-affected zone. The purpose of selecting multiple initial reference grids is to reflect the overall pre-deformation situation near the weld through multi-point measurements, rather than relying solely on single-point measurements. The number of initial reference grids should be no less than three on each side of the weld, i.e., no less than six on both sides of the weld. The selected grids should cover the grid area used in subsequent post-stamping strain gradient calculations on the blank to ensure a one-to-one correspondence for compensation corrections.

[0094] The second step is to measure the initial deformation of the initial reference grid. The equipment used for this measurement is the same as that used to measure the grid deformation after stamping, namely, a grid strain measurement system. The blank with the grid pattern is placed on the measurement platform, and an image of the initial reference grid area is captured using a high-resolution industrial camera. The grid strain measurement system calculates the initial deformation of each initial reference grid by identifying the current position of the grid nodes and comparing it with the standard positions during grid fabrication. The specific calculation method for the initial deformation is the same as that for the grid deformation after stamping, i.e., solving for the deformation gradient tensor through the displacement of the four corner points, and then extracting the principal strain value as the deformation of that grid. The initial deformation is usually a very small value, at the micro-strain level. If the blank does not undergo measurable pre-deformation during stacking, the initial deformation is close to zero.

[0095] The third step involves comparing the initial deformation of each initial reference grid with a preset initial deformation threshold. The initial deformation threshold is the boundary value used to determine whether the pre-deformation has reached the point where compensation correction needs to be activated. This threshold is set based on the minimum resolvable deformation of the grid strain measurement system. When the initial deformation is less than the noise level of the measurement system, the measurement result is unreliable and no compensation is needed; when the initial deformation is greater than or equal to the noise level of the measurement system, the measurement result is reliable and compensation is required. A typical initial deformation threshold is two to three times the measurement accuracy of the grid strain measurement system.

[0096] The fourth step involves applying different processing methods based on the comparison results. For initial reference meshes where the initial deformation exceeds the initial deformation threshold, the deformation offset of that mesh is recorded. The deformation offset is the initial deformation measured for that mesh, including two parameters: the principal strain value and the principal strain direction. These parameters are recorded in a data table, corresponding one-to-one with the mesh position identifiers, serving as the pre-deformation compensation data file for the blank. For meshes where the initial deformation does not exceed the initial deformation threshold, the deformation offset is recorded as a default zero value or a null value, and is not included in subsequent compensation.

[0097] The fifth step, after stamping, involves measuring the deformation of each grid on both sides of the weld seam on the stamped part. In this measurement step, when measuring the total deformation of a grid after stamping, the pre-deformation compensation data file of the blank is first queried to determine if the grid had a recorded deformation offset before stamping. If recorded, the total deformation of the grid measured after stamping is subtracted from the corresponding deformation offset. This subtraction operation must be performed at the strain tensor level, not simply by numerical subtraction. Specifically, the deformation gradient tensor of the grid measured after stamping is multiplied by the inverse tensor of the pre-deformation gradient tensor calculated from the deformation offset to obtain the net deformation gradient tensor of the grid caused by stamping. The net principal strain value is then calculated from the net deformation gradient tensor. If the grid is not recorded or the record is zero in the pre-deformation compensation data file, the deformation measured after stamping is directly used as the net deformation. The net deformation replaces the original measured deformation and participates in the subsequent construction of the principal strain value sequence and strain gradient calculation.

[0098] The technical effect and principle of this implementation method lies in solving the problem of interference with measurement accuracy caused by the pre-deformation of stacked unequal-thickness welded plates. This is achieved by adding an initial deformation measurement of the mesh before stamping and performing data-level deduction compensation during post-stamping measurement. This method does not require changes to existing stacking and transportation methods, nor does it add additional production processes or hardware investment. It only adds a mesh image acquisition operation after mesh fabrication and before stamping begins. This operation uses the same equipment and has the same workflow as the post-stamping mesh measurement. The pre-deformation compensation data file follows the blank and is automatically retrieved during post-stamping measurement, achieving precise correction with one-plate-one-compensation correction.

[0099] The above description is merely a preferred embodiment of this application and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in this application is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the inventive concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions disclosed in this application.

Claims

1. A multi-factor fusion prediction method for the stamping fracture risk of an integrated door ring laser-welded plate, characterized in that, Includes the following steps: A grid pattern is made on both sides of the weld seam of the laser-welded plate blank. The grid pattern covers the weld seam and the base material area on both sides of the weld seam. The laser-welded plate blank is formed by connecting at least two steel plates of different thicknesses or materials through laser welding. The laser-welded blank with the grid pattern is stamped to obtain the stamped part; The deformation of each grid on both sides of the weld on the stamped part is measured, the principal strain value of each grid arranged perpendicular to the weld direction is obtained, and the strain gradient on both sides of the weld is calculated. The strain gradient is the rate of change of the principal strain value of multiple grids extending from the weld position to the base material direction on one side. Obtain the measured hardness value of the weld area and the material elongation of the thicker base material on both sides of the weld. Based on the strain gradient, the measured hardness value, the material elongation of the thicker base material, and the thickness ratio of the base materials on both sides of the weld, the equivalent fracture risk index is calculated. The equivalent fracture risk index is compared with a preset critical risk threshold. If the equivalent fracture risk index exceeds the critical risk threshold, it is determined that there is a risk of stamping fracture.

2. The multi-factor fusion prediction method for stamping fracture risk of integrated door ring laser-welded plate according to claim 1, characterized in that, The process of creating a grid pattern on both sides of the weld seam of a laser-welded plate blank includes the following steps: Based on the weld seam orientation of the laser-welded plate blank, the entire weld seam is divided into straight weld seam segments and curved weld seam segments; A grid pattern is made on the straight weld section and the curved weld section respectively. The grid on the straight weld section is a square grid, and the grid on the curved weld section is a radial grid. The grid lines of the radial grid are arranged along the weld tangent direction and the weld normal direction respectively, so as to accommodate the deformation difference of the curved weld section along the weld tangent direction and the weld normal direction during the stamping process.

3. The multi-factor fusion prediction method for stamping fracture risk of integrated door ring laser-welded plate according to claim 1, characterized in that, The calculation of the strain gradient on both sides of the weld includes the following steps: In a series of grids arranged perpendicular to the weld direction on one side of the weld, starting from the grid closest to the weld centerline, the difference between the principal strain values ​​of two adjacent grids is taken in sequence in the direction away from the weld to obtain a set of principal strain changes piecewise. Calculate the maximum value among the absolute values ​​of the segmental changes in the principal strain, and take it as the maximum inter-segment change on that side; Divide the maximum inter-segment variation by the distance between the center points of two adjacent grids to obtain the strain gradient on that side; Perform the above steps on both sides of the weld, and take the larger absolute value of the strain gradient on both sides as the strain gradient on both sides of the weld.

4. The multi-factor fusion prediction method for stamping fracture risk of integrated door ring laser-welded plate according to claim 3, characterized in that, Taking the larger absolute value of the strain gradients on both sides as the strain gradient on both sides of the weld includes the following steps: Obtain the elongation of the base materials on both sides of the weld; Divide the absolute value of the strain gradient on one side of the weld by the elongation of the base material on that side to obtain the normalized strain gradient on that side. Perform the same operation on the other side of the weld to obtain the normalized strain gradient on the other side. The original strain gradient corresponding to the larger of the normalized strain gradients on both sides is taken as the strain gradient on both sides of the weld.

5. The multi-factor fusion prediction method for stamping fracture risk of integrated door ring laser-welded plate according to claim 1, characterized in that, Before calculating the equivalent rupture risk index, the following steps are also included: The strain gradient is compared with a preset gradient activation threshold; If the strain gradient is less than the gradient activation threshold, it is determined that the weld area has not yet entered a high deformation gradient state, which is used to indicate that the measured hardness value is included in the calculation with the first participation weight when calculating the equivalent fracture risk index. If the strain gradient is greater than or equal to the gradient activation threshold, it is determined that the weld area has entered a high deformation gradient state, which is used to indicate that the measured hardness value is included in the calculation with a second participation weight when calculating the equivalent fracture risk index. The second participation weight is greater than the first participation weight.

6. The multi-factor fusion prediction method for stamping fracture risk of integrated door ring laser-welded plate according to claim 5, characterized in that, The gradient activation threshold is determined in the following way: From the stamped part, select a region on the thinner side of the base material on both sides of the weld, and whose distance from the weld is greater than a preset distance threshold, as the reference base material region; Within the reference base material region, principal strain values ​​of multiple grids are extracted along the direction perpendicular to the weld. The average value of the principal strain values ​​of the multiple grids within the reference base material region is calculated and used as the base material reference strain value. The reference strain value of the parent material is used as the gradient activation threshold.

7. The multi-factor fusion prediction method for stamping fracture risk of integrated door ring laser-welded plate according to claim 6, characterized in that, The first participation weight is a preset benchmark value; the second participation weight is determined by calculating the ratio of the strain gradient to the benchmark strain value of the base material as the weld deformation concentration multiple; the second participation weight is equal to the product of the first participation weight and the weld deformation concentration multiple.

8. The multi-factor fusion prediction method for stamping fracture risk of integrated door ring laser-welded plate according to claim 2, characterized in that, The critical risk threshold takes different values ​​on the straight weld segment and the curved weld segment, wherein the critical risk threshold on the curved weld segment is lower than the critical risk threshold on the straight weld segment.

9. The multi-factor fusion prediction method for stamping fracture risk of integrated door ring laser-welded plate according to claim 1, characterized in that, Obtaining the measured hardness value of the weld area includes the following steps: On the stamped part, the initial hardness values ​​of multiple measuring points are obtained sequentially along the weld seam. The multiple measuring points are distributed on the weld seam centerline and at positions offset by a preset distance on both sides of the weld seam centerline. The minimum value among the initial hardness values ​​from multiple measuring points is selected as the measured hardness value of the weld area.

10. The multi-factor fusion prediction method for stamping fracture risk of integrated door ring laser-welded plate according to claim 1, characterized in that, Before stamping the laser-welded plate blank with the grid pattern, the method further includes the following steps: From the laser-welded plate blank that has completed the mesh pattern production but has not yet been stamped, select multiple meshes located on both sides of the weld and adjacent to the weld as initial reference meshes. Measure the initial deformation of the initial reference mesh; If the initial deformation exceeds the preset initial deformation threshold, the deformation offset of the initial reference mesh is recorded. Measuring the deformation of each grid on both sides of the weld seam on the stamped part includes the following steps: Subtract the deformation offset corresponding to the grid from the deformation measured after stamping to obtain the net deformation of the grid caused by stamping.

Images