Sheet metal cabinet differentiated thickness design method based on impact load distribution

Abstract

The present application belongs to the technical field of mechanical design, and relates to a sheet metal cabinet differentiated thickness design method based on impact load distribution. Impact load is applied to the three-dimensional model of the cabinet for finite element simulation, the first area where the stress exceeds the yield strength of the material is identified, and the stress optimization area where stress is concentrated due to geometric shape mutation is marked. Based on the geometric fairing algorithm, the geometric contour line of the stress optimization area is smoothed to obtain a first round of optimized model. The parts to be compensated are identified again by simulation, and the second round of optimized model is obtained by iteratively increasing the thickness of the plate locally. The impact simulation verification is performed on the second round of optimized model, and if the index is not met, the smoothing parameter is adjusted for reiteration. The present application realizes accurate optimization by associating simulation data with geometric characteristics, the geometric fairing with continuous curvature guides impact energy from the source, and the gradient thickening realizes smooth transition of stiffness, thereby improving material utilization efficiency and reducing overall weight on the premise of ensuring impact resistance reliability.

Description

Technical Field

[0001] This invention relates to the field of mechanical design technology, and in particular to a method for designing differentiated thicknesses of sheet metal cabinets based on impact load distribution. Background Technology

[0002] In high-risk application scenarios such as outdoor communication base stations, power monitoring, and rail transit, impact-resistant sheet metal cabinets play an important role in protecting the internal precision electronic equipment. These cabinets must be able to withstand impact loads such as accidental collisions and falling heavy objects to ensure the safe and stable operation of critical infrastructure.

[0003] For the impact resistance design of server rack structures, finite element simulation is commonly used in engineering practice for auxiliary analysis. By applying impact loads to the three-dimensional model of the rack, stress distribution and deformation data can be obtained, thereby identifying weak points in the structure. Based on the simulation results, the thickness of the sheet metal is locally increased in areas of high stress to improve impact resistance. This simulation-driven local reinforcement method avoids, to some extent, the weight and cost problems caused by the early one-size-fits-all approach of thickening the entire rack, allowing more material to be concentrated in areas of high stress.

[0004] However, simulation-based local reinforcement methods have certain limitations in practical applications; simulation analysis can only present the stress distribution results, and it is difficult to provide clear criteria for judging the complex causes of stress concentration. Since the cabinet structure contains a large number of bending, welding and assembly features, its stress field is often the result of the combined effects of geometry, material distribution and load conditions. How to accurately extract the area that needs to be optimized from the complex simulation results depends on the engineer's personal experience and lacks a systematic standard. Summary of the Invention

[0005] Therefore, the purpose of this invention is to overcome the shortcomings of existing technologies, such as the lack of systematic stress analysis methods, the reliance on experience in optimization processes, and low efficiency. It provides a method for designing differentiated thicknesses of sheet metal cabinets based on impact load distribution. This method achieves precise optimization by correlating simulation data with geometric features, guides impact energy at its source through geometrically compliant force flow with continuous curvature, and achieves a smooth transition in structure and stiffness through gradient thickening. This effectively improves material utilization efficiency and reduces overall weight while ensuring impact resistance reliability.

[0006] To address the aforementioned technical problems, this invention provides a method for designing differentiated thicknesses of sheet metal cabinets based on impact load distribution, comprising: S1. Apply an impact load to the 3D model of the cabinet and obtain the stress distribution through finite element simulation; based on the stress distribution, identify the first region where the stress exceeds the material yield strength, and mark the stress optimization zone in the first region where stress concentration is caused by abrupt changes in geometry; S2. The geometric contour lines of the stress optimization zone are smoothed based on the geometric smoothing algorithm to obtain the first round of optimization model; S3. Perform finite element simulation on the first round of optimization model, and identify the parts to be compensated where the stress exceeds the yield strength of the material based on the simulation results; S4. For the part to be compensated, the thickness of the plate is locally increased to obtain a thickness compensation model; and finite element simulation is performed on the thickness compensation model. According to the simulation results, if there are parts where the stress exceeds the yield strength of the material, the local thickness is increased iteratively until the stress in all parts is not greater than the yield strength of the material, and a second round of optimization model is obtained. S5. Perform finite element impact simulation on the second round of optimization model to obtain stress and deformation data, and determine whether the strength and deformation indicators are met. If they are met, design the cabinet based on the second round of optimization model; otherwise, adjust the smoothing parameters of the geometric smoothing algorithm and repeat steps S2 to S5.

[0007] Preferably, adjusting the smoothing parameters of the geometric smoothing algorithm includes: obtaining the nodes in the second round of optimization model where the stress exceeds the material yield strength in the impact simulation; determining whether the geometric position of the nodes exceeds the limit and whether it belongs to the transition fillet or continuous surface after smoothing in step S2; if so, obtaining the actual radius of curvature at the nodes exceeds the limit and determining whether the actual radius of curvature is less than the reference radius of curvature; if so, updating the smoothing parameters of the stress optimization zone to the reference radius of curvature; otherwise, updating the smoothing parameters to the product of the actual radius of curvature and the scaling factor; the scaling factor is 1.5~1.8.

[0008] Preferably, adjusting the smoothing parameters of the geometric smoothing algorithm further includes: if the over-limit node is not on the transition fillet or continuous surface, increasing the local thickness of the plate region where the over-limit node is located in the second round of optimization model.

[0009] Preferably, marking the stress optimization zone in the first region where stress concentration occurs due to abrupt changes in geometry includes: obtaining a set of stress-over-limit nodes in the first region where the stress exceeds the material's yield strength; obtaining the coordinates of all geometric contour lines in the 3D model of the cabinet and generating a set of geometric contour lines; calculating the shortest spatial distance from each node in the set of stress-over-limit nodes to each geometric contour line in the set of geometric contour lines, and marking nodes whose shortest spatial distance is less than the tolerance as nodes to be optimized; clustering the nodes to be optimized, merging spatially adjacent nodes to be optimized into the same continuous geometric region, and marking the geometric contour line segments covered by the continuous geometric region as stress optimization zones.

[0010] Preferably, the geometric contour line of the stress optimization zone is smoothed based on a geometric smoothing algorithm, including: obtaining a discrete point coordinate sequence of the contour line corresponding to the stress optimization zone; calculating the curvature between adjacent discrete points based on the discrete point coordinate sequence to generate a curvature sequence; identifying adjacent discrete point pairs in the curvature sequence where the curvature value changes abruptly, and marking the intervals corresponding to the adjacent discrete point pairs as curvature abrupt change intervals; extending to both sides of the contour line with the curvature abrupt change interval as the center to determine the interval to be smoothed; reconstructing the original contour line in the interval to be smoothed based on a NURBS curve fitting algorithm to generate a new contour line segment with continuous curvature in the curvature abrupt change interval, and replacing the original contour line in the interval to be smoothed with the new contour line segment to obtain an updated contour line.

[0011] Preferably, when determining the area to be smoothed, the length ΔL extending to both sides of the contour line is set as follows: ; L represents the length of the interval where curvature changes abruptly; R avg The value represents the average radius of curvature of the contour line on both sides of the curvature abrupt change interval; k represents the adjustment coefficient, which ranges from 0.5 to 2.

[0012] Preferably, for the part to be compensated, a thickness compensation model is obtained by locally increasing the thickness of the board material, including: obtaining the board material region of the part to be compensated in the first round of optimization model, extracting the current thickness of the board material region and the geometric boundary information of the part to be compensated; based on the geometric boundary information, generating multiple concentric annular transition zones layer by layer outward from the part to be compensated, the multiple concentric annular transition zones together forming a thickness transition zone; increasing the board material thickness of the part to be compensated by a set thickness increment to obtain the compensated thickness; within the thickness transition zone, the thickness of each annular transition zone decreases layer by layer from the inner annular transition zone to the outer annular transition zone; the thickness of the innermost annular transition zone is equal to the compensated thickness, and the gradient thickness value of the outermost annular transition zone is equal to the current thickness; replacing the corresponding region in the first round of optimization model with the locally thickened region formed by the part to be compensated and the multiple concentric annular transition zones around it, to obtain the thickness compensation model.

[0013] Preferably, determining the number of layers in the concentric annular transition zone includes: obtaining the area A of the part to be compensated and the minimum circumscribed circle radius R of the boundary; and determining the initial number of layers N0 based on the area A and the minimum circumscribed circle radius R. ; Calculate the difference between the current thickness and the compensated thickness to obtain the thickness change ΔD; Calculate the number of adjustment layers N1 based on the thickness change: N1 = floor(ΔD / D*); Determine the number of layers N of the concentric annular transition zone based on the initial number of layers and the number of adjustment layers: N = max(N0, N1); Celi represents rounding up; floor represents rounding down; max represents taking the larger value; D* represents the maximum allowable thickness change of a single layer.

[0014] Preferably, the progressively decreasing method includes linear decreasing or S-shaped decreasing.

[0015] Preferably, the method for determining the layer-by-layer decrease is as follows: determine whether the edge stress of the part to be compensated is greater than the average stress of its region; if so, determine an S-shaped decrease; otherwise, determine a linear decrease.

[0016] The technical solution of the present invention has the following advantages compared with the prior art: The sheet metal cabinet differential thickness design method based on impact load distribution described in this invention achieves precise optimization by correlating simulation data with geometric features, guides impact energy at the source by transmitting force flow through geometric light with continuous curvature, and achieves a smooth transition in structure and stiffness through gradient thickening, thereby effectively improving material utilization efficiency and reducing overall weight while ensuring impact resistance reliability. Attached Figure Description

[0017] To make the content of this invention easier to understand, the invention will be further described in detail below with reference to specific embodiments and accompanying drawings, wherein... Figure 1 This is a flowchart of a sheet metal cabinet differential thickness design method based on impact load distribution in a preferred embodiment of the present invention; Figure 2 This is a flowchart of adjusting the smoothing parameters of the geometric smoothing algorithm in a preferred embodiment of the present invention; Figure 3 This is a flowchart illustrating the marking of the stress optimization zone in a preferred embodiment of the present invention; Figure 4 This is a flowchart of the smoothing process of the geometric contour line of the stress optimization zone in a preferred embodiment of the present invention; Figure 5 This is a flowchart of a thickness compensation model obtained by local thickening in a preferred embodiment of the present invention; Figure 6 This is a flowchart illustrating the determination of the number of concentric annular transition zone layers in a preferred embodiment of the present invention; Figure 7 This is a flowchart illustrating the process of determining the progressively decreasing method in a preferred embodiment of the present invention. Detailed Implementation

[0018] The present invention will be further described below with reference to the accompanying drawings and specific embodiments, so that those skilled in the art can better understand and implement the present invention. However, the embodiments described are not intended to limit the present invention.

[0019] The purpose of this invention is to provide a method for designing differentiated thicknesses of sheet metal cabinets based on impact load distribution, overcoming the shortcomings of existing technologies such as lack of systematic stress analysis methods, reliance on experience in the optimization process, and low efficiency.

[0020] The following combination Figure 1 The technical solutions of the present invention will be described in detail with specific embodiments. It should be noted that the embodiments are only used to explain the present invention and not to limit the present invention.

[0021] Step S1, Identify stress concentration areas: First, obtain the 3D model of the sheet metal cabinet to be designed. This model is stored in a computer-aided design data format and contains complete geometric information of the cabinet, such as panels, frames, doors, ventilation holes, and edge and corner features. Import the 3D model into finite element analysis preprocessing software, and mesh its surface to generate multiple analysis elements. The mesh density needs to be set reasonably according to the cabinet size and calculation accuracy requirements. Typically, the mesh is densified in areas of geometric abrupt changes to ensure calculation accuracy.

[0022] Subsequently, an impact load spectrum is applied to the model. The impact load spectrum is determined based on the actual application scenario of the cabinet, and may include impact waveforms of different directions and energy levels to simulate conditions such as accidental collisions or heavy object drops. The stress distribution of each analysis element under the impact load is obtained through calculation using a finite element solver.

[0023] Based on the stress distribution and using the material's yield strength as a benchmark, all regions where the stress value exceeds this yield strength are identified as the first region. However, these first regions may contain stress concentrations caused by abrupt changes in geometry, or stress exceedances caused by excessive local loads or insufficient foundation thickness. To accurately screen out the former, this step further performs the following operations: All nodes in the first region where the stress exceeds the material's yield strength are identified, generating a set of stress-over-limit nodes. Simultaneously, the geometric coordinates of all edge lines and corner lines are extracted from the 3D model of the cabinet, generating a set of geometric contour lines. The shortest spatial distance from each node in the stress-over-limit node set to each geometric contour line in the geometric contour line set is calculated. Nodes with a shortest spatial distance less than the tolerance are marked as nodes to be optimized. The tolerance can be set based on the average size of the finite element analysis cells, for example, 1.5 times the average size of the analysis cells, to ensure that nodes located near geometric abrupt changes but slightly offset can also be effectively captured.

[0024] Cluster analysis is performed on the nodes to be optimized, merging spatially adjacent nodes into the same continuous geometric region. Density clustering can be used, grouping nodes whose distance to each other is less than a threshold into the same cluster. The geometric contour segment covered by each cluster is marked as a stress optimization region. Through these steps, one or more stress optimization regions are ultimately obtained, which accurately correspond to the locations of stress concentration caused by abrupt geometric changes.

[0025] This step achieves accurate identification of specific types of regions where stress concentration is caused by abrupt changes in geometric shape by spatially matching stress-over-limit nodes with geometric contour lines and clustering and merging nodes to be optimized. This allows subsequent geometric smoothing processing to be applied precisely to the parts that truly need geometric modification, avoiding ineffective geometric modifications in flat or insufficient strength areas, thereby improving the targeting and efficiency of the entire optimization process.

[0026] Step S2, Geometric Smoothing: For the geometric contour lines corresponding to the stress optimization zone marked in step S1, each geometric contour line is represented by a series of discrete point coordinate sequences in the CAD system. The discrete point coordinate sequences are extracted as the basic data for subsequent processing.

[0027] Based on the discrete point coordinate sequence, the curvature values ​​between adjacent discrete points are calculated to generate a curvature sequence. Pairs of adjacent discrete points in the curvature sequence that exhibit a step change in curvature value are identified, and the intervals corresponding to these pairs are marked as curvature abrupt change intervals. A step change refers to a curvature difference between two adjacent points exceeding a threshold value. This threshold can be dynamically determined based on the overall curvature distribution characteristics of the contour line, for example, by taking twice the standard deviation of the curvature sequence.

[0028] Centered on the curvature abrupt change interval, a predetermined length is extended to both sides of the contour line to determine the smoothing interval. The extension length must balance the smoothing effect and geometric conformity requirements, and is typically 1 to 2 times the length of the curvature abrupt change interval. Subsequently, the NURBS curve fitting algorithm is used to reconstruct the original contour line within the smoothing interval. An objective function is constructed, which includes two terms: a position deviation squared term and a curvature smoothing term. The position deviation squared term characterizes the closeness of the fitted curve to the original discrete point coordinate sequence, ensuring that the reconstructed curve does not deviate excessively from the original design; the curvature smoothing term characterizes the drastic curvature change of the fitted curve within the curvature abrupt change interval, forcing curvature continuity within the abrupt change interval. By solving for the control point coordinates and corresponding weighting factors that minimize the objective function value, a new contour line segment with continuously changing curvature values ​​within the curvature abrupt change interval is generated. The new contour line segment replaces the original contour line within the smoothing interval to obtain the updated geometric contour line. Traverse all geometric contour lines corresponding to the stress optimization zone and perform the above processing until all curvature abrupt change intervals have been processed, generating a geometrically smooth model with continuous curvature transition characteristics, which serves as the first round of optimization model.

[0029] The purpose of this geometric smoothing step is to divert impact energy from the root of force transmission by utilizing the curvature continuity of the geometric contour lines, thereby reducing stress peaks caused by geometric abrupt changes. Furthermore, by limiting the smoothing range to local areas near the curvature abrupt change interval, it maximizes the preservation of well-designed straight lines or large-curvature arcs on the contour lines, achieving a balance between eliminating defects and preserving the design.

[0030] Step S3, Identify the area to be compensated: Apply the same impact load spectrum as in step S1 to the first-round optimized model obtained in step S2, and perform finite element simulation again to obtain an updated stress distribution cloud map. Based on the stress distribution cloud map, using the material yield strength as a benchmark, identify all nodes whose stress values ​​exceed the yield strength, and mark the regions where these nodes are located as areas to be compensated. The areas to be compensated identified at this time are regions where stress still exceeds the standard after geometric smoothing, and their cause is local strength deficiency rather than geometric defects.

[0031] Although step S2 reduces the stress peak caused by geometric abrupt changes, the scope of geometric optimization is mainly limited to stress concentration problems caused by shape defects. For parts that cannot withstand impact loads due to insufficient thickness of the base plate, even if the geometry is smoother, their load-bearing capacity will not change fundamentally. Therefore, after completing geometric smoothing, the parts to be compensated identified by simulation again in this step are precisely the areas that geometric smoothing cannot cover and must be solved by increasing material strength.

[0032] Step S4, compensate for local thickness: For the board area in the first-round optimization model where the area to be compensated identified in step S3 is located, extract the current thickness of the board area; increase the board thickness of the area to be compensated by a set thickness increment to obtain the compensated thickness. The set thickness increment can be set according to the standard specifications of the board, for example, 0.5mm or 1.0mm. Replace the original thickness of the area to be compensated with the compensated thickness to obtain the thickness compensation model.

[0033] Finite element analysis (FEM) is performed on the thickness compensation model to obtain updated stress distribution data. Based on the simulation results, if there are any locations where the stress value still exceeds the material's yield strength, these new locations are identified as areas to be compensated, and the process of locally increasing the plate thickness is repeated. This iterative process continues until the stress value in all locations of the current model is no greater than the material's yield strength, resulting in the second round of optimized model.

[0034] This step uses an iterative thickening method, with each material input verified by simulation to ensure that the strength standard is met with the minimum amount of material used. This avoids the problem of rework due to insufficient thickening or material waste due to excessive thickening, achieving the optimal configuration of material usage and effectively controlling the overall weight while meeting the impact resistance requirements.

[0035] Step S5, Comprehensive effect verification and iteration: Apply the same impact load spectrum as in step S1 to the second-round optimized model obtained in step S4, perform finite element impact simulation, and obtain stress distribution cloud map and deformation data of each node. Extract the maximum stress value of all nodes based on the stress distribution cloud map, and extract the maximum deformation value of all nodes based on the deformation data; compare the maximum stress value with the allowable stress value, and compare the maximum deformation value with the allowable deformation value.

[0036] If the maximum stress value is less than or equal to the allowable stress value, and the maximum deformation value is less than or equal to the allowable deformation value, then the second round of optimization model is determined to meet the strength and deformation indices. The second round of optimization model is then used as the final cabinet structure design model to guide the production and manufacturing of the cabinet.

[0037] If the maximum stress exceeds the allowable stress, or the maximum deformation exceeds the allowable deformation, the current model is deemed not to meet the design requirements. In this case, the smoothing parameters of the geometric smoothing algorithm are adjusted, for example, by increasing the weight coefficient of the curvature smoothing term in the objective function, or by increasing the extension length of the interval to be smoothed. After adjusting the smoothing parameters, steps S2 to S5 are re-executed until an optimized model that meets the design specifications is obtained.

[0038] In summary, the sheet metal cabinet differential thickness design method based on impact load distribution in this invention achieves precise optimization by correlating simulation data with geometric features, guides impact energy at its source by transmitting force flow through geometric light with continuous curvature, and achieves a smooth transition in structure and stiffness through gradient thickening. Thus, while ensuring impact resistance reliability, it effectively improves material utilization efficiency and reduces overall weight.

[0039] In the above embodiments, when the simulation verification fails and the smoothing parameters of the geometric smoothing algorithm need to be adjusted, how to determine the specific direction and magnitude of the parameter adjustment, and how to distinguish the stress exceeding the standard due to different reasons and take differentiated coping strategies.

[0040] Specifically, the stress-over-limit nodes that still exist in the second round of optimization model may have multiple causes. These nodes may be located on transition fillets or continuous surfaces that have already been processed in step S2, indicating that the current smoothing parameters are still insufficient to completely eliminate stress concentration. They may also be located in flat plate areas or other non-geometrically abrupt locations that have not undergone geometric smoothing, indicating that the plate thickness in these areas is insufficient. If the same parameter adjustment method is used for these two different causes of stress over-limit, increasing the thickness in terms of geometry will be ineffective, and adjusting the radius of curvature in terms of thickness will also be ineffective.

[0041] To address the problem of inability to distinguish the causes of stress exceedance during iterative optimization, leading to blind adjustments of geometric and thickness parameters and reduced iterative efficiency, a preferred embodiment of this invention provides a differentiated thickness design method, which is described below in conjunction with... Figure 2 The technical solution of the present invention will be described in detail with reference to specific embodiments.

[0042] First, obtain all nodes in the second-round optimized model where the stress exceeds the material's yield strength during impact simulation, and record these nodes as out-of-limit nodes; for each out-of-limit node, perform the following judgment and processing: Case 1: The out-of-limit node is located on the transition fillet or continuous surface after the smoothing process in step S2: Determine whether the geometric position of the out-of-limit node belongs to the transition fillet or continuous surface after the smoothing process in step S2. This determination can be made by comparing the coordinates of the out-of-limit node with the spatial range of the new contour line segment generated in step S2. If the out-of-limit node is located on the transition fillet or continuous surface, it indicates that the geometric smoothing process in that area is insufficient, and the current smoothing parameters have not completely eliminated stress concentration.

[0043] At this point, the actual radius of curvature at the over-limit node is obtained. The actual radius of curvature refers to the radius of curvature at the location of the node in the current model, which can be obtained by calculating the geometric parameters of the contour line near the node. The actual radius of curvature is then compared with the reference radius of curvature. The reference radius of curvature is a preset reference value that can be determined based on the characteristics of the cabinet material and the impact load level. For example, it can be the minimum allowable radius of curvature corresponding to the bending fatigue limit of the material.

[0044] Determine whether the actual radius of curvature is smaller than the reference radius of curvature: If the actual radius of curvature is smaller than the reference radius of curvature, it indicates that the current radius of curvature is too small, the corners are too sharp, and stress concentration occurs. In this case, the smoothing parameter of the stress optimization zone is updated to the reference radius of curvature. This means that in step S2 of the next iteration, the reference radius of curvature will be directly used as the target value to smooth the region, raising the radius of curvature to at least the safe lower limit.

[0045] If the actual radius of curvature is greater than or equal to the reference radius of curvature, it indicates that the current radius of curvature has reached above the safety lower limit, but stress exceedance still exists, suggesting that the radius of curvature needs to be further increased to achieve smoother force flow transmission. In this case, the smoothing parameter is updated to the product of the actual radius of curvature and a scaling factor. The scaling factor is between 1.5 and 1.8, for example, 1.6. This means that in the next iteration, the current radius of curvature is multiplied by a factor of 1.5 to 1.8 to generate a larger target radius of curvature, providing a more fully smooth transition.

[0046] Case 2: The over-limit node is not located on the transition fillet or continuous surface: If the assessment results indicate that the excessive stress at the node is not located on a transition fillet or continuous curved surface, it means that the excessive stress at the node is not caused by insufficient geometric smoothness, but rather by insufficient plate thickness in that area to withstand the impact load. In this case, the local thickness of the plate area where the excessive stress node is located should be increased.

[0047] Specifically, the current thickness of the plate region where the out-of-limit node is located in the second round of optimization model is obtained. This current thickness is then increased by a set increment step to generate an updated local thickness value. The increment step can be set according to the plate specifications, for example, 0.5 mm or 1.0 mm. The updated local thickness is used as the initial reference value for step S4 in the next iteration. That is, when step S4 is re-executed, the thickness compensation for this region will start from this updated thickness value.

[0048] After completing the classification and processing of all the above-mentioned out-of-limit nodes, the updated smoothing parameters (including the updated curvature radius determined for each stress optimization zone) and the updated local thickness value are used as inputs, and the process returns to step S2 to start a new round of optimization iteration.

[0049] Determining whether an out-of-limit node belongs to a transition fillet or continuous surface after the smoothing process in step S2 essentially involves attributing the cause of the stress exceeding the limit: for nodes located on smooth surfaces, stress exceeding the limit indicates insufficient geometric smoothness; for nodes not located on smooth surfaces, stress exceeding the limit indicates insufficient thickness. This attribution of causes allows subsequent adjustment measures to accurately address the root cause of the problem.

[0050] When the actual radius of curvature is smaller than the reference radius of curvature, the reference radius of curvature is directly used as the update value. The reference radius of curvature is a safety lower limit determined based on material properties and load conditions; a value smaller than this indicates an unacceptable geometric defect that must be immediately corrected to a safe level. When the actual radius of curvature is larger than the reference radius of curvature, it is further increased by multiplying by a scaling factor. The scaling factor is between 1.5 and 1.8, an empirical range derived from numerous simulation experiments: a value less than 1.5 results in too small an adjustment range, potentially requiring multiple iterations to converge; a value greater than 1.8 results in too large an adjustment range, which may lead to an excessive increase in the radius of curvature, altering the original design features and increasing unnecessary manufacturing complexity.

[0051] Through the above technical solution, the present invention achieves accurate attribution and classification of the causes of stress exceeding the standard in the iterative optimization process, so that the adjustment of geometric smoothing parameters and the compensation of local thickness each act on the real problem area, avoiding invalid iterations caused by blind trial and error, and effectively improving the convergence speed of the optimization process.

[0052] In the above embodiment, step S1 identifies the stress optimization zone. In the actual simulation process, how can we accurately and objectively identify the areas where stress concentration is caused by abrupt changes in geometry from the finite element simulation data, rather than simply marking all areas with excessive stress as optimization objects indiscriminately?

[0053] Specifically, the stress distribution output by finite element simulation contains stress data from tens of thousands of nodes. Among these, nodes with stress exceeding the material's yield strength may be located throughout the cabinet. The causes of these stress-over-limit nodes are complex: some are located at geometrically abrupt changes such as corners and edges, where the stress exceedance is mainly caused by the geometry; others are located in flat plate areas, where the stress exceedance is mainly caused by excessive local loads or insufficient foundation thickness. If these two types of nodes are not distinguished, subsequent geometric smoothing processing will not be able to accurately target the areas that truly require geometric modification, resulting in ineffective geometric modifications in flat plate areas or omissions in areas of geometric abrupt changes, leaving the problem unresolved.

[0054] To address the challenge of objectively and accurately identifying stress concentration regions caused by geometric abrupt changes from massive simulation data, which leads to a lack of specificity in subsequent optimization measures and a tendency to omissions or misjudgments, a preferred embodiment of the present invention provides a differentiated thickness design method, which is described below in conjunction with... Figure 3 The technical solution of the present invention will be described in detail with reference to specific embodiments.

[0055] First, all nodes in the first region where the stress exceeds the material's yield strength are identified. These nodes are recorded as stress-over-limit nodes, and a set of stress-over-limit nodes is generated. This set contains all nodes in the first region where the stress exceeds the limit, regardless of whether they are located at geometrically abrupt changes or in flat plate regions.

[0056] Secondly, the coordinate data of all geometric contour lines in the 3D model of the server rack are obtained to generate a set of geometric contour lines. These geometric contour lines include all the edge lines and corner lines of the server rack and are the main carriers of geometric abrupt changes. These contour lines are typically stored as discrete point sequences in the CAD model and can be directly extracted through the secondary development interface of the CAD software.

[0057] Subsequently, the shortest spatial distance from each node in the set of stress-over-limit nodes to each geometric contour line in the set of geometric contour lines is calculated. This calculation is implemented using a point-to-line distance algorithm in spatial geometry. For each geometric contour line, the distance from the node to each point on that contour line is calculated and the minimum value is taken. After calculation, each stress-over-limit node will obtain a set of distance values, corresponding to its distance to each geometric contour line.

[0058] Nodes whose shortest spatial distance is less than the tolerance are marked as nodes to be optimized. The tolerance value needs to be set reasonably to balance the accuracy and comprehensiveness of identification. Too small a tolerance may cause nodes that should be affected by geometrical abrupt changes to be missed; too large a tolerance may cause nodes in plate areas far from geometrical abrupt changes to be mistakenly selected. In this embodiment, the tolerance can be determined based on the average size of the analysis elements in the finite element analysis, for example, 1.5 to 2 times the average size of the analysis elements. The principle is that the stress concentration effect is usually still significant within a range of about one mesh element away from the geometrical abrupt change. Setting the tolerance at a multiple of the analysis element size can effectively capture the affected nodes within this range.

[0059] After the above screening, a batch of nodes to be optimized was obtained. These nodes all meet two conditions: first, the stress exceeds the material's yield strength; second, they are located within the tolerance range near the geometric contour line. However, these nodes are still discrete and have not yet formed a continuous geometric region. Therefore, clustering is required for these nodes to be optimized.

[0060] The goal of clustering is to merge spatially adjacent nodes to be optimized into the same continuous geometric region. This implementation uses a density-based clustering algorithm: starting from any node to be optimized, it searches for all neighboring nodes whose spatial distance is less than the clustering threshold, and assigns the found nodes to the same node cluster; then, using the newly assigned node as a new starting point, it continues to search for neighboring nodes until no new nodes can be found, forming a complete node cluster. The clustering threshold can be set according to the mesh density, for example, twice the average size of the analysis cells, to ensure that nodes belonging to the same continuous stress region can be correctly merged.

[0061] Repeat the above process until all nodes to be optimized are assigned to their respective node clusters, generating one or more node clusters. Each node cluster represents a spatially contiguous region of stress concentration.

[0062] Finally, the geometric contour segments covered by each node cluster are marked as a stress optimization zone. Specifically, the coordinate range of all nodes in the node cluster is obtained, and the start and end positions of this range on the geometric contour line are determined. The contour segment between the start and end positions is extracted as the stress optimization zone corresponding to that node cluster. In this way, each node cluster corresponds to a continuous geometric contour segment, and these segments are the objects that need to be geometrically smoothed in the subsequent step S2.

[0063] Through the above steps, one or more stress optimization zones are finally obtained. These zones accurately correspond to the locations where stress concentration occurs due to abrupt changes in geometry and are presented in the form of continuous geometric contour segments, which can be directly used for subsequent geometric smoothing processing.

[0064] The present invention transforms the abstract phenomenon of excessive stress into a quantifiable problem of geometric feature screening, enabling objective and accurate identification of a specific type of region where stress concentration is caused by abrupt changes in geometric shape. This allows subsequent geometric smoothing processing to be precisely applied to the parts that truly require geometric modification.

[0065] In the above embodiments, the actual simulation process faces the problem of how to make geometric smoothing effectively eliminate stress concentration while preserving the original design features to the greatest extent and avoiding excessive modification that leads to geometric distortion.

[0066] Specifically, simple rounding only ensures geometric continuity and cannot achieve a smooth transition of curvature. In the process of impact force transmission, there are still turning points, and the effect of alleviating stress concentration is limited. In addition, uniformly rounding the entire contour line will change the originally well-designed straight segments, which may affect the installation space of internal equipment or the fit between adjacent components.

[0067] To address the problems of difficulty in accurately locating defects, inability to achieve curvature continuity, and susceptibility to excessive design modifications during geometric smoothing, the preferred embodiment of this invention provides a differentiated thickness design method, which is described below in conjunction with... Figure 4 The technical solution of the present invention will be described in detail with reference to specific embodiments.

[0068] First, obtain the discrete point coordinate sequence of the contour line corresponding to the stress optimization zone. The stress optimization zone is a continuous geometric contour line segment obtained in step S1. In CAD systems, these contour lines are usually stored in the form of parametric equations or discrete point sequences. For numerical calculation, they need to be discretized into a series of coordinate points. This can be done by directly exporting discrete points from the CAD model, or by generating a discrete point sequence through equidistant sampling of the parametric equations. The density of the discrete points needs to accurately reflect the geometric characteristics of the contour line; typically, 1-2 points per millimeter are used, or sampling can be intensified in areas of drastic curvature changes.

[0069] Secondly, based on the discrete point coordinate sequence, the curvature between adjacent discrete points is calculated to generate a curvature sequence. Curvature can be calculated using the three-point or five-point method of numerical differentiation. For three adjacent points, an approximate curvature value at the midpoint can be calculated. By traversing all discrete points, the curvature value corresponding to each point is obtained, and these values ​​are arranged in order to generate the curvature sequence. The curvature sequence reflects the degree of curvature at various locations on the contour line and is the basis for identifying geometric abrupt changes.

[0070] Subsequently, adjacent discrete point pairs in the curvature sequence where the curvature value undergoes a step change are identified, and the intervals corresponding to these adjacent discrete point pairs are marked as curvature abrupt change intervals. A step change refers to the curvature difference between two adjacent points exceeding a preset abrupt change threshold. Setting the abrupt change threshold is crucial; this implementation uses a dynamic threshold method: calculating the mean and standard deviation of the entire curvature sequence, and setting the abrupt change threshold to the mean plus two to three times the standard deviation. Adjacent point pairs whose curvature values ​​exceed this threshold indicate a drastic curvature change, i.e., a geometric abrupt change. The intervals corresponding to these adjacent point pairs on the contour line (i.e., the line segment from the first point to the second point) are marked as curvature abrupt change intervals. Multiple curvature abrupt change intervals may exist within a stress optimization zone, and each needs to be marked separately.

[0071] By identifying curvature abrupt change intervals, geometric defects can be accurately located; this allows smoothing processing to focus on the local areas that truly need modification, rather than blindly modifying the entire outline, thus ensuring a smoothing effect while preserving the original design features to the greatest extent possible.

[0072] Next, using the curvature abrupt change interval as the center, extend to both sides of the contour line to determine the area to be smoothed. The purpose of this extension is to provide sufficient transition space for smoothing processing, allowing the reconstructed curve to smoothly blend into the original contour line. The determination of the extension length must balance the smoothing effect and geometric conformation requirements. In this embodiment, the extension length is taken as 1 to 2 times the length of the curvature abrupt change interval, and the specific value can be dynamically adjusted according to the severity of the curvature abrupt change: the more severe the curvature change, the greater the extension length. After extension, the entire contour line segment from the left extension endpoint to the right extension endpoint is the area to be smoothed.

[0073] By extending to both sides to determine the area to be smoothed, a smooth connection between the reconstructed curve and the original contour line is achieved, avoiding new curvature discontinuities caused by improper selection of connection points, and ensuring the geometric quality of the entire contour line after smoothing.

[0074] Then, based on the NURBS curve fitting algorithm, the original contour lines within the region to be smoothed are reconstructed. NURBS (Non-Uniform Rational B-Spline) curves are a widely used curve-surface representation in computer-aided design, possessing good local support and flexibility. The reconstruction process includes the following sub-steps: The objective function for constructing the NURBS curve comprises two terms: a squared position deviation term and a curvature smoothing term. The squared position deviation term characterizes the closeness of the fitted curve to the original discrete point coordinate sequence within the curvature abrupt change interval, ensuring that the reconstructed curve does not deviate excessively from the original design. The curvature smoothing term characterizes the drastic curvature change of the fitted curve within the curvature abrupt change interval, forcing a smooth transition of curvature within this interval. By adjusting the weighting coefficients of these two terms, the relationship between positional accuracy and curvature smoothness can be balanced.

[0075] The goal is to find the coordinates of the control points and their corresponding weights on the NURBS curve that minimizes the objective function. This is an optimization problem, which can be solved using the least squares method combined with gradient descent. During the solution process, the tangent vectors at the two endpoints of the section to be smoothed are used as boundary constraints to ensure that the reconstructed curve smoothly connects to the original contour line at the endpoints. Through iterative optimization, a set of optimal control point coordinates and weights is obtained.

[0076] Based on the obtained control points and weight factors, a NURBS curve is generated. This curve achieves continuous change in curvature value within the curvature abrupt change interval; that is, the curvature value transitions smoothly from the start to the end of the abrupt change interval without step jumps. This NURBS curve is used as the new contour line segment.

[0077] By introducing a dual-objective NURBS reconstruction that includes positional deviation and curvature smoothing, we can achieve synergistic optimization of geometric conformity and mechanical properties. This results in an optimized contour line that has both good geometric accuracy and excellent mechanical properties, avoiding one-sided optimization that sacrifices one aspect for another.

[0078] Finally, the original contour lines within the smoothing interval are replaced with new contour lines to obtain updated contour lines. Specifically, the portion of the original contour lines corresponding to the smoothing interval is deleted and replaced with newly generated NURBS curve segments, ensuring geometric continuity at the connection points at both ends.

[0079] Traverse all geometric contour lines corresponding to the stress optimization zone, and perform the above steps for each contour line until all curvature abrupt change intervals have been processed, generating a geometrically smooth model with continuous curvature transition characteristics, which serves as the first round of optimization model.

[0080] Through the above-mentioned preferred solutions, the present invention achieves precise local smoothing of geometric contour lines, accurately locates the position of curvature change and reconstructs only the local area, and retains the original design features to the maximum extent while ensuring curvature continuity, thereby eliminating stress concentration from the geometric root.

[0081] Furthermore, when determining the area to be smoothed, the length ΔL extending to both sides of the contour line is set as follows: ; L represents the length of the interval where curvature changes abruptly; R avg The value represents the average radius of curvature of the contour line on both sides of the curvature abrupt change interval; k represents the adjustment coefficient, which ranges from 0.5 to 2.

[0082] Specifically, a curvature abrupt change interval is the contour segment corresponding to a pair of adjacent discrete points in a curvature sequence where the curvature value undergoes a step change. The length of the curve between the start and end points of this interval is calculated as the interval length L.

[0083] In this embodiment, the average radius of curvature R avg The calculation interval is defined as the range on both sides of the curvature abrupt change interval, which is roughly the same length as the abrupt change interval. In practice, this can be automated through programming: using the starting point of the curvature abrupt change interval as a reference, search to the left for contour segments with a cumulative arc length of L, calculate the radius of curvature at each point on this segment, and take the average value; similarly, obtain the average radius of curvature on the right side, and then combine the average values ​​from both sides as R. avg .

[0084] The adjustment coefficient k is determined based on the different materials of the plate. The principle for its value is that brittle materials are sensitive to stress concentration. Once the stress exceeds the yield strength, brittle fracture is likely to occur. A longer transition range is needed to ensure smooth force flow, so a larger k value is taken. Tough materials have better plastic deformation ability and can redistribute stress through local yielding. They have a higher tolerance for stress concentration, so a smaller k value can be taken.

[0085] For highly brittle materials (such as cast aluminum, cast iron, and high-carbon steel), a value of 1.8 is preferred; for moderately brittle materials (such as ordinary cast steel and quenched steel), a value of 1.6 is preferred; for ordinary structural steel (such as Q235, Q345, and 16Mn), a value of 1.2 is preferred; for highly tough materials (such as aluminum alloys and austenitic stainless steel), a value of 1.0 is preferred; and for ultra-tough materials (such as copper alloys, pure aluminum, and low-carbon steel), a value of 0.7 is preferred.

[0086] Next, starting from the beginning of the curvature abrupt change interval, extend a length ΔL away from the interval to determine the left extension endpoint; starting from the end of the curvature abrupt change interval, extend a length ΔL away from the interval to determine the right extension endpoint.

[0087] Finally, the entire contour segment between the left and right extension endpoints is defined as the smoothing region. This region includes the curvature abrupt change region itself and the transition region extending ΔL to both sides, providing ample smoothing space for subsequent NURBS curve reconstruction.

[0088] When implementing step S4 to locally increase the thickness of the plate at the area to be compensated, a problem arises: how to avoid creating new stress concentration points due to the increased thickness while increasing the local thickness.

[0089] Specifically, if a uniformly thick patch is directly added to the area to be compensated, there is a significant thickness abrupt change between the thickened and original areas. Under impact loads, this abrupt thickness change becomes a new stress concentration point, potentially leading to edge cracking or delamination failure. To address this issue, a preferred embodiment of the present invention provides a differentiated thickness design method, which is described below in conjunction with... Figures 5-7The technical solution of the present invention will be described in detail with reference to specific embodiments.

[0090] First, the plate region where the part to be compensated is located in the first round of optimization model is obtained, and the current thickness of the plate region and the geometric boundary information of the part to be compensated are extracted. The current thickness refers to the original thickness of the region before thickening, and the geometric boundary information includes the shape, size, and outline of the part to be compensated, which is used to determine the thickening range later.

[0091] Based on the geometric boundary information, multiple concentric annular transition zones are generated layer by layer outward from the area to be compensated. These multiple concentric annular transition zones together constitute the thickness transition region. The shape of the concentric annular transition zones matches the geometric contour of the area to be compensated. For example, for a circular area to be compensated, the transition zone is a concentric ring; for irregular shapes, the transition zone can be generated using an equidistant offset method. Each transition zone has an independent thickness, collectively forming a gradient region where the thickness gradually changes from the center to the edge.

[0092] Increase the thickness of the plate in the area to be compensated by a set thickness increment to obtain the compensated thickness. The set thickness increment can be determined according to the standard specifications of the plate, such as 0.5mm, 1.0mm, or 1.5mm. The specific value needs to take into account the material characteristics and the degree of stress exceeding the standard. The more severe the stress exceeding the standard, the larger the increment can be.

[0093] Within the thickness transition zone, the thickness of each annular transition zone decreases progressively from the innermost to the outermost annular transition zone. The thickness of the innermost annular transition zone is equal to the compensated thickness, and the thickness of the outermost annular transition zone is equal to the current thickness. The thickness of each intermediate transition zone is determined according to a preset decreasing pattern to ensure a smooth thickness transition from the center to the edge without abrupt changes.

[0094] The corresponding region in the first-round optimization model is replaced by a locally thickened region consisting of the area to be compensated and multiple concentric annular transition zones around it, resulting in a thickness compensation model. This model includes the compensated thickness of the central part and the thickness distribution of the surrounding gradient transition.

[0095] When the thickness changes abruptly, the stiffness of the structure also changes abruptly. Under impact loads, the point of stiffness abrupt change becomes an obstacle to force flow transmission, leading to stress concentration. By constructing multiple annular transition zones with varying thickness layer by layer, a gradient transition of stiffness is achieved, allowing force flow to be smoothly transmitted from the thickened region to the original region. The annular structure ensures uniformity in all directions of the transition, avoiding directional weak points.

[0096] When generating concentric annular transition zones, it is necessary to determine the specific number of layers in the transition zone: First, obtain the area A of the part to be compensated and the minimum circumscribed circle radius R of its boundary. Area A reflects the size of the part to be compensated, and the minimum circumscribed circle radius R reflects its overall span.

[0097] Secondly, based on the area A and the minimum circumscribed circle radius R, determine the initial number of layers N0. The calculation formula is: ; Here, ceil represents rounding up to ensure that the number of layers is an integer and at least 1. The physical meaning of this formula is: the larger the feature size of the area to be compensated, the more transition layers are needed to smoothly transition to the surrounding area; the larger the radius of the circumscribed circle, the narrower and longer the area, and the required number of layers should be adjusted accordingly.

[0098] Then, the difference between the current thickness and the compensated thickness is calculated to obtain the thickness change ΔD = compensated thickness - current thickness. ΔD reflects the total amount of thickness that needs to be increased.

[0099] Next, the number of adjustment layers N1 is calculated based on the thickness variation. The calculation formula is: N1 = floor(ΔD / D*). D* is the preset maximum allowable thickness variation per layer, that is, the maximum allowable thickness variation value of each transition zone, which is usually 0.5mm or 1.0mm; floor indicates rounding down. The physical meaning of this formula is that the larger the total thickness variation, the more layers are needed to distribute the gradient, so as to avoid stress concentration caused by excessive variation in a single layer.

[0100] Finally, the number of layers N in the concentric annular transition zone is determined based on the initial number of layers N0 and the adjusted number of layers N1: N = max( N0, N1 ), which means taking the larger of the two values ​​as the final number of layers. This ensures that there are enough layers to cover the geometry of the part to be compensated (as guaranteed by N0) and that the thickness gradient is within a reasonable range (as guaranteed by N1), thus achieving a comprehensive consideration of geometric and thickness factors.

[0101] Limitations on thickness reduction method: When setting the gradient thickness, there are two modes for decreasing the thickness layer by layer: linear decrease or S-shaped decrease.

[0102] In the linear decreasing pattern, the thickness of each layer decreases according to an arithmetic progression, meaning each layer is reduced by the same amount of thickness. This decreasing method is simple and easy to implement, and is suitable for situations where the stress distribution is relatively uniform.

[0103] In the S-shaped decreasing pattern, the thickness change is characterized by gradual changes at both ends and rapid changes in the middle. That is, the thickness changes slowly near the center and edges, and changes more rapidly in the middle region. This decreasing pattern can achieve a smooth transition of stress within a shorter transition zone, making it suitable for situations with severe stress concentration and the need for rapid attenuation.

[0104] Rules for choosing the decreasing method: First, the stress distribution characteristics of the area to be compensated are obtained, including the peak stress at the center and the edge stress at the edges. Simultaneously, the average stress value of the region containing the area to be compensated is obtained.

[0105] Next, it is determined whether the edge stress of the area to be compensated is greater than the average stress of its region. If so, it indicates that the high stress is not only concentrated in the central area but also extends to the edge region, with a wide stress distribution and slow attenuation. In this case, an S-shaped decreasing pattern that can provide a longer gradual decrease area is required. If the edge stress is less than or equal to the average stress, it indicates that the stress is mainly concentrated in the central area, and the stress in the edge region has significantly attenuated. A linear decreasing pattern can meet the requirements. This judgment rule ensures that the choice of decreasing method matches the stress distribution characteristics.

[0106] Comparing edge stress with mean stress is essentially a quantitative assessment of the stress distribution pattern. When edge stress is greater than mean stress, it indicates that the stress distribution curve exhibits a broad peak shape, with a wide range of high stress areas, requiring a longer transition region for smooth attenuation. When edge stress is less than mean stress, it indicates that the stress distribution curve exhibits a sharp peak shape, with high stress concentrated in the center, and the stress in the edge area has significantly decreased; linear decrease is sufficient to meet the requirements. The characteristic of S-shaped decrease is that it is gentle at both ends and rapid in the middle, which perfectly matches the stress attenuation requirement of a broad peak shape; the characteristic of linear decrease is that it changes uniformly throughout the entire transition region, matching the rapid attenuation requirement of a sharp peak shape.

[0107] The above-described solution of the present invention achieves a smooth transition of thickness by constructing a thickness transition zone centered on the part to be compensated and setting a gradient change law that gradually decreases from the compensated thickness value to the original thickness value. This solution not only solves the original stress concentration problem but also proactively avoids new stress concentrations that may be caused by abrupt changes in thickness.

[0108] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A method for designing differentiated thickness of sheet metal cabinets based on impact load distribution, characterized in that... ,include: S1. Apply an impact load to the 3D model of the cabinet and obtain the stress distribution through finite element simulation; based on the stress distribution, identify the first region where the stress exceeds the material yield strength, and mark the stress optimization zone in the first region where stress concentration is caused by abrupt changes in geometry; S2. The geometric contour lines of the stress optimization zone are smoothed based on the geometric smoothing algorithm to obtain the first round of optimization model; S3. Perform finite element simulation on the first round of optimization model, and identify the parts to be compensated where the stress exceeds the yield strength of the material based on the simulation results; S4. For the part to be compensated, the thickness of the plate is locally increased to obtain a thickness compensation model; and finite element simulation is performed on the thickness compensation model. According to the simulation results, if there are parts where the stress exceeds the yield strength of the material, the local thickness is increased iteratively until the stress in all parts is not greater than the yield strength of the material, and a second round of optimization model is obtained. S5. Perform finite element impact simulation on the second round of optimization model to obtain stress and deformation data, and determine whether the strength and deformation indicators are met. If they are met, design the cabinet based on the second round of optimization model; otherwise, adjust the smoothing parameters of the geometric smoothing algorithm and repeat steps S2 to S5.

2. The sheet metal cabinet differential thickness design method based on impact load distribution according to claim 1, characterized in that, Adjust the smoothing parameters of the geometry smoothing algorithm, including: Obtain the nodes in the second round of optimization model where the stress exceeds the material yield strength in the impact simulation, and determine whether the geometric position of the nodes exceeds the limit and belongs to the transition fillet or continuous surface after the smoothing process in step S2. If so, obtain the actual radius of curvature at the out-of-limit node, and determine whether the actual radius of curvature is less than the reference radius of curvature: If so, the smoothing parameter of the stress optimization zone is updated to the reference radius of curvature; otherwise, the smoothing parameter is updated to the product of the actual radius of curvature and the scaling factor; the scaling factor is 1.5~1.

8.

3. The method for differentiated thickness design of sheet metal cabinets based on impact load distribution according to claim 2, characterized in that, Adjusting the smoothing parameters of the geometric smoothing algorithm also includes: if the over-limit node is not on the transition fillet or continuous surface, increasing the local thickness of the plate region where the over-limit node is located in the second round of optimization model.

4. The method for differentiated thickness design of sheet metal cabinets based on impact load distribution according to claim 1, characterized in that, The stress optimization zone in the first region, where stress concentration occurs due to abrupt changes in geometry, is marked, including: Obtain the set of stress-over-limit nodes in the first region where the stress exceeds the material's yield strength; Obtain the coordinates of all geometric contour lines in the 3D model of the cabinet, and generate a set of geometric contour lines; Calculate the shortest spatial distance from each node in the set of stress-exceeding nodes to each geometric contour line in the set of geometric contour lines, and mark the nodes whose shortest spatial distance is less than the tolerance as nodes to be optimized; Cluster the nodes to be optimized, merge spatially adjacent nodes to be optimized into the same continuous geometric region, and mark the geometric contour line segments covered by the continuous geometric region as stress optimization areas.

5. The method for differentiated thickness design of sheet metal cabinets based on impact load distribution according to claim 1, characterized in that, The geometric contour lines of the stress optimization zone are smoothed based on a geometric smoothing algorithm, including: Obtain the discrete point coordinate sequence of the contour line corresponding to the stress optimization zone; Based on the discrete point coordinate sequence, the curvature between adjacent discrete points is calculated to generate a curvature sequence; adjacent discrete point pairs in the curvature sequence where the curvature value undergoes a step change are identified, and the intervals corresponding to the adjacent discrete point pairs are marked as curvature abrupt change intervals; Using the curvature abrupt change interval as the center, extend to both sides of the contour line to determine the interval to be smoothed; Based on the NURBS curve fitting algorithm, the original contour line in the interval to be smoothed is reconstructed to generate a new contour line segment with continuous curvature in the interval with abrupt curvature change. The original contour line in the interval to be smoothed is replaced with the new contour line segment to obtain the updated contour line.

6. The method for differentiated thickness design of sheet metal cabinets based on impact load distribution according to claim 5, characterized in that, When determining the area to be smoothed, the length ΔL extending to both sides of the contour line is set as follows: ； L represents the length of the interval where curvature changes abruptly; R avg The value represents the average radius of curvature of the contour line on both sides of the curvature abrupt change interval; k represents the adjustment coefficient, which ranges from 0.5 to 2.

7. The method for differentiated thickness design of sheet metal cabinets based on impact load distribution according to claim 1, characterized in that, For the area to be compensated, a thickness compensation model is obtained by locally increasing the thickness of the plate, including: Obtain the plate region of the part to be compensated in the first round of optimization model, and extract the current thickness of the plate region and the geometric boundary information of the part to be compensated; Based on the geometric boundary information, multiple concentric annular transition zones are generated outward layer by layer from the center of the part to be compensated, and the multiple concentric annular transition zones together constitute the thickness transition zone. The thickness of the plate in the part to be compensated is increased by a set thickness increment to obtain the compensated thickness; Within the thickness transition region, the thickness of each annular transition zone decreases layer by layer from the inner annular transition zone to the outer annular transition zone; the thickness of the innermost annular transition zone is equal to the compensated thickness, and the gradient thickness value of the outermost annular transition zone is equal to the current thickness. The corresponding region in the first round of optimization model is replaced by the locally thickened region formed by the part to be compensated and the multiple concentric annular transition zones around it, to obtain a thickness compensation model.

8. The method for differentiated thickness design of sheet metal cabinets based on impact load distribution according to claim 7, characterized in that, Determining the number of layers in the concentric annular transition zone includes: Obtain the area A of the part to be compensated and the minimum circumcircle radius R of its boundary; Based on the area A and the minimum circumscribed circle radius R, determine the initial number of layers N0: ; Calculate the difference between the current thickness and the compensated thickness to obtain the thickness change ΔD; Calculate the number of adjustment layers N1 based on the thickness change: N1 = floor(ΔD / D*); The number of layers N in the concentric annular transition zone is determined based on the initial number of layers and the adjusted number of layers: N = max(N0, N1). Celi indicates rounding up; floor indicates rounding down; max indicates taking the larger value; D* indicates the maximum allowable thickness variation of a single layer.

9. The method for differentiated thickness design of sheet metal cabinets based on impact load distribution according to claim 7, characterized in that, The method of decreasing layer by layer includes linear decrease or S-shaped decrease.

10. The method for differentiated thickness design of sheet metal cabinets based on impact load distribution according to claim 9, characterized in that, The method for determining the layer-by-layer reduction is as follows: determine whether the edge stress of the part to be compensated is greater than the average stress of its region; if so, determine the S-shaped reduction; otherwise, determine the linear reduction.

Images