A structural collision topology optimization method based on finite element analysis for engineering applications

By combining geometric nonlinearity and material nonlinearity finite element analysis, the computational stability and accuracy problems of existing topology optimization under complex collision conditions are solved, and the synergistic optimization of lightweight and safety of crash-resistant structures is achieved.

CN123116673APending Publication Date: 2026-11-17ORDOS INST OF APPLIED TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611207456.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-11
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing topology optimization methods struggle to accurately describe structural responses under complex collision conditions, especially under conditions of large deformation and plastic yielding. Furthermore, they suffer from high computational costs, poor stability, and a lack of stable solution strategies applicable to engineering collision problems.

Method used

A finite element analysis method coupled with geometric nonlinearity and material nonlinearity is adopted, combining the Total Lagrange finite strain and J2 elastoplastic model. The nonlinear equilibrium equation is solved by Newton-Raphson iteration, a consistent adjoint sensitivity analysis is established, and topology optimization is achieved by using the moving asymptote algorithm and density filtering method.

Benefits of technology

It improves the computational stability and accuracy of topology optimization under complex working conditions, and can automatically identify the optimal induction groove layout of energy-absorbing structures, achieving synergistic optimization of lightweight and impact resistance performance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123116673A_ABST
    Figure CN123116673A_ABST
Patent Text Reader

Abstract

This invention belongs to the field of structural optimization design technology, specifically relating to a structural collision topology optimization method based on finite element analysis for engineering applications. The method first establishes a finite element model of the collision structure and divides the design region; it then uses a density-variable-based topology optimization model to describe material distribution; it establishes a large deformation geometric nonlinear analysis model considering finite strain, and combines this with a J2 elastoplastic constitutive model to establish a material nonlinear analysis model; it establishes an adjoint variable sensitivity analysis model to obtain the sensitivity information of the objective function with respect to the design variables; it updates the design variables using a moving asymptote algorithm; it iterates until the optimization convergence condition is met; and finally, it outputs the optimal topology structure that meets the requirements of lightweighting and impact resistance. This invention can comprehensively consider complex factors such as geometric nonlinearity, material nonlinearity, and collision contact, improving the accuracy and convergence stability of collision topology optimization calculations while reducing optimization computational costs.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of structural optimization design technology, specifically relating to a structural collision topology optimization method based on finite element analysis for engineering applications. Background Technology

[0002] With the increasing demands for lightweight structures and collision safety performance in the automotive, rail transportation, aerospace, and protective equipment industries, utilizing computer-aided optimization methods to achieve coordinated design of structural crashworthiness and lightweighting has become an important research direction in engineering. Structural collisions are typically accompanied by transient, large displacement, large rotation, local buckling, material plastic deformation, and complex contact phenomena. Traditional experience-based structural design methods struggle to achieve the optimal balance between lightweighting and crashworthiness. Therefore, there is an urgent need to establish structural optimization design methods that can directly address engineering collision conditions.

[0003] Topology optimization, a structural optimization technique capable of automatically searching for the optimal material distribution within a given design space, has been widely applied in mechanical, civil, aerospace, and automotive engineering fields. Most existing topology optimization methods are based on the linear elasticity assumption, which assumes small structural deformation and linear material behavior, and obtains the structural response through linear finite element analysis. However, in actual collision scenarios, structures often undergo significant large deformations and plastic yielding. Linear topology optimization struggles to accurately reflect the structural mechanical behavior during real collisions, leading to substantial deviations between the obtained optimization results and actual engineering conditions.

[0004] To improve the applicability of topology optimization under complex conditions, researchers have proposed a geometrically nonlinear topology optimization method, which improves the accuracy of structural analysis by considering large deformation effects. Simultaneously, a material nonlinear topology optimization method has also been proposed, incorporating an elastoplastic constitutive model into the optimization process. However, most existing methods only consider a single factor from geometric or material nonlinearity, lacking a comprehensive optimization framework that simultaneously considers the coupling effects of geometric and material nonlinearities. When a structure undergoes both large deformation and plastic yielding during a collision, a single nonlinear model struggles to accurately describe the actual structural response.

[0005] Furthermore, collision topology optimization typically requires extensive nonlinear finite element iterative calculations, resulting in high computational costs and poor convergence stability. Especially when using the density method for topology optimization, low-density regions are prone to severe element distortion, ill-conditioned stiffness matrices, and nonlinear solution failures, leading to premature termination of the optimization process. Existing techniques for addressing these issues largely rely on empirical parameter adjustments and lack stable solution strategies applicable to engineering collision problems.

[0006] On the other hand, in the design of crash-resistant structures such as automotive energy-absorbing boxes, anti-collision beams, and longitudinal beams, local weakening structures such as guide grooves and window holes can effectively reduce the initial peak impact force and improve the progressive folding mode. Currently, the design of the hole shape, position, and layout of such structures still mainly relies on engineering experience and parametric experiments, making it difficult to obtain a globally optimal material distribution scheme. Existing research mostly focuses on parameter optimization for simplified thin-walled tubes, lacking collision topology optimization methods that can be directly extended to actual engineering energy-absorbing structures.

[0007] Therefore, there is an urgent need to propose a structural collision topology optimization method that can simultaneously consider geometric nonlinearity, material nonlinearity, and collision contact effects, and has good computational stability and engineering applicability, so as to achieve efficient and lightweight design of impact-resistant structures. Summary of the Invention

[0008] To address the problems in existing collision topology optimization techniques, such as difficulty in simultaneously considering the coupling effects of geometric nonlinearity and material nonlinearity, poor stability of nonlinear solutions, complex sensitivity analysis, and insufficient engineering applicability, this invention proposes a structural collision topology optimization method for engineering applications.

[0009] The technical solution of this invention is: A structural collision topology optimization method based on finite element analysis for engineering applications includes the following steps: S1. Obtain the geometric model of the collision-bearing structure to be optimized. Before performing finite element analysis, divide the design region into a design region and a non-design region. Discretized A finite number of elements; a continuous density variable is defined for each element. This is used to characterize the state of existence of materials within a unit, and , The minimum density value set to prevent singularity of the stiffness matrix; Represents a solid unit. This represents an empty element, with fixed solid material in the non-design region; and defines the elastic modulus, Poisson's ratio, J2 yield stress, hardening modulus, contact relationship, collision boundary conditions (rigid wall collision or velocity loading), contact relationship, and constraint conditions. S2. Discretize the design region into finite element elements, using the element density vector. As a topology design variable, this invention applies the SIMP interpolation model to the uniform elastoplastic tangent operator, rather than interpolating only the elastic Young's modulus. This establishes a mapping relationship between the material's uniform elastoplastic tangent matrix and density (the design variable) using SIMP interpolation. , in, The uniform elastoplastic tangent matrix of the element. For solid materials, a consistent elastoplastic tangent matrix is ​​formed. The uniform elastoplastic tangent matrix of empty elements. As a penalty factor; S3. Establish a finite element analysis model coupling geometric nonlinearity and material nonlinearity: A Total Lagrange finite strain description is used, and the deformation gradient F and Green-Lagrange strain are calculated from the displacement field. ,in The tensor is unit; for material nonlinearity, an isotropic J2 elastoplastic constitutive model and a back-mapping algorithm are used to update the plastic stress, which is strengthened by isotropic reinforcement. Evolutionary yield surface, in which, To adapt to equivalent plastic strain Evolving yield stress, The initial yield stress, To enhance the modulus; Construct a consistent tangent stiffness matrix based on the material tangent stiffness matrix and geometric stiffness matrix. ; S4. Solve the nonlinear equilibrium equations using the Newton-Raphson incremental iteration method: , in, It is a nonlinear internal force vector. For external force vectors, It is the global displacement vector, and the uniform elastic-plastic tangent matrix of the structure is assembled at the same time. The Jacobian value is used to solve the nonlinear equilibrium equations during the Newton-Raphson iteration process; the global nodal displacement vector of the element in the collision history is obtained by solving this value. Cauchy stress field Green-Lagrange strain field G, equivalent plastic strain field The structural response is equal; when the Jacobian value of an element is negative, the element distortion angle exceeds the threshold, or the maximum displacement increment exceeds the threshold, the incremental step size is automatically reduced and the solution is resolved. S5. Establish an optimization objective function with the goal of minimizing compliance or maximizing energy absorption: , in, It is determined by a given volume fraction Defined optimization constraints, and These are the volume of the unit and the total volume of the design area, respectively. It is the lower bound of the design variable; Regarding the objective function with respect to design variables Find the total derivative and introduce the adjoint variable. The adjoint equation is established, and the sensitivity of the objective function to the density variable is derived by solving for the adjoint variables. The output of S4 is reused to reconstruct the uniform elastoplastic tangent matrix of the structure. The adjoint equation is: , in, This is the transpose of the structural tangent stiffness matrix. S6. Update design variables based on sensitivity using the Moving Asymptote Algorithm (MMA). And employ weights based on the filter radius. Density filtering is performed by weighted averaging of neighborhood cell densities to suppress the dependence of checkerboard patterns on grids. S7. Substitute the updated density variable back into S3 to re-subdivide the stiffness and repeat S4–S6 until the maximum relative change of the design variable in two adjacent iterations is less than the first threshold C1, the relative change rate of the objective function is less than the second threshold C2, or the maximum number of iterations C3 is reached. S8. Output the optimal unit density distribution that satisfies the volume fraction constraint and collision energy absorption index. Perform geometric reconstruction on the distribution to generate a solid model for energy absorption box, anti-collision beam, longitudinal beam, battery pack protection structure or rail transit crash-resistant structure.

[0010] Specifically, in S3, the yield function described by the J2 elastoplastic model for material nonlinearity is: , in, The von Mises equivalent stress.

[0011] Specifically, in S5, the specific expression for sensitivity is: , in, This is the global nodal displacement vector of the element; The space occupied by the unit in the reference configuration; Represents the nonlinear strain-displacement matrix of the element; The element-consistent tangent modulus for numerical reconstruction; the specific calculation method for sensitivity is as follows: First, obtain the strain and stress increments at two adjacent time steps. and ; Then, select m (m > 6) key stress-strain increments to construct a strain-stress increment matrix (6 × m). and ; , , Finally, using Approximately obtained Then, substituting this into the final expression yields... .

[0012] Specifically, in S6, the density expression after node filtering in density filtering is: , in, , and The first The filtered density, volume, and design variables of each unit.

[0013] Specifically, the iteration termination condition in S7 is expressed as follows: , , , in, , and The set convergence threshold; Indicates the number of optimization iterations. Indicates the maximum allowed number of iterations; and This represents the optimized and updated design variable vector and objective function; and This represents the vector of variables and the objective function before optimization and update.

[0014] A computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the method of the present invention.

[0015] A collision structure optimization system includes a memory and a processor. The processor calls a program in the memory to execute the method of the present invention, and is externally connected to a finite element solution module and a CAD reconstruction module to ultimately generate a CAD solid model.

[0016] The beneficial effects of this invention are: 1) This invention is the first to unify geometric nonlinearity, material nonlinearity and collision topology optimization into the same analysis framework, which can more accurately describe the structural response in the real collision process.

[0017] 2) This invention uses the Total Lagrange finite strain theory coupled with the J2 elastic-plastic model for analysis, which can simultaneously consider the influence of large deformation and plastic yield on the topology optimization results.

[0018] 3) This invention establishes a consistent adjoint sensitivity analysis method and provides an engineering approximate solution method, which avoids directly solving for displacement sensitivity and greatly improves the computational efficiency of large-scale nonlinear topology optimization.

[0019] 4) This invention proposes an adaptive convergence control strategy suitable for collision problems, which can effectively suppress solution failures caused by element distortion and improve the stability of the optimization process.

[0020] 5) This invention adopts a collaborative iterative implementation method using LS-DYNA and MATLAB, which has good engineering feasibility and can be directly applied to industrial collision analysis platforms.

[0021] 6) This invention can automatically identify the optimal induction groove layout in the energy-absorbing structure, significantly reducing the structural mass while ensuring collision safety, and achieving synergistic optimization of lightweighting and impact resistance.

[0022] 7) This invention is applicable to the design of automotive energy-absorbing boxes, longitudinal beams, anti-collision beams, battery pack protective structures, rail transit crash-resistant structures, aerospace energy-absorbing structures, and other engineering structures subjected to impact loads, and has broad engineering application prospects. Attached Figure Description

[0023] Figure 1 This is a flowchart for implementing collision topology optimization in the project.

[0024] Figure 2 Diagram showing the dimensional parameters and operating conditions for the collision topology optimization model.

[0025] Figure 3 The iterative process of the objective function for collision topology optimization and the diagram of the topology evolution process.

[0026] Figure 4 This is a comparison chart of the performance of the structure before and after topology optimization. Detailed Implementation

[0027] The present invention will now be described in detail with reference to the accompanying drawings.

[0028] The purpose of this invention is to establish a unified geometry-material nonlinear coupling topology optimization framework, which, while ensuring computational stability, achieves synergistic optimization between lightweighting and crashworthiness of collision structures, and can be directly applied to the design of automotive energy-absorbing boxes, longitudinal beams, anti-collision beams, battery pack protective structures, and other engineering crashworthiness structures.

[0029] like Figure 1 As shown, the present invention mainly includes the following steps: Step S1: Establish the finite element model of the collision structure A finite element model is established based on the geometry of the structure to be optimized, and the design region and non-design region are determined. The design region is used for material distribution optimization, while the original structure remains unchanged in the non-design region.

[0030] In the finite element model, material parameters, contact relationships, boundary conditions, and collision load conditions are defined. Material parameters include elastic modulus, Poisson's ratio, yield stress, hardening modulus, and density; collision loads can be applied through rigid wall collisions, displacement loading, or velocity loading. Figure 2 In the implementation scheme shown, the rigid wall has a mass of 0.3 tons and impacts the rectangular cross-section square tube at a speed of 1 m / s. Although the bottom of the square tube has 6 degrees of freedom of the unit node, they are all constrained. The dimensions of the square tube are shown in the figure: 80 mm in length, 60 mm in width, and 60 mm in height.

[0031] Step S2: Establish a topology optimization model based on density variables The design region is discretized into multiple finite element elements, and a continuous density variable is defined for each element. The density variable takes values ​​in the following range: in, The minimum density value set to prevent the stiffness matrix from becoming singular.

[0032] Instead of interpolating only the elastic Young's modulus, the SIMP interpolation model is applied to the uniform elastoplastic tangent operator, thus establishing a mapping relationship between the material's uniform elastoplastic tangent matrix and density (design variable) using SIMP interpolation. , in, The uniform elastoplastic tangent matrix of the element. For solid materials, a consistent elastoplastic tangent matrix is ​​formed. The uniform elastoplastic tangent matrix of empty elements. As a penalty factor; Step S3: Establish a coupled analysis model of geometric nonlinearity and material nonlinearity. A geometric nonlinear analysis model was established using the Total Lagrange finite strain description method, with all kinematic quantities defined relative to the initial reference configuration.

[0033] The deformation gradient F is calculated based on the displacement field, and the Green-Lagrange strain tensor is further calculated. : in, It is a unit tensor.

[0034] The material nonlinearity is described using the J2 elastoplastic model, and its yield function is: in, For von Mises equivalent stress, To adapt to equivalent plastic strain Evolving yield stress.

[0035] Adopting the principle of isotropic reinforcement: in, The initial yield stress, To enhance the modulus.

[0036] Plastic stress updates are performed using a return mapping algorithm, and a consistent tangent stiffness matrix is ​​constructed.

[0037] Step S4: Solve the nonlinear finite element equations Establish the structural nonlinear equilibrium equations: in, For internal force vectors, This is the external force vector.

[0038] The displacement vector u is solved using the Newton-Raphson incremental iterative method, and the uniform tangent stiffness matrix is ​​used as the iterative Jacobian matrix.

[0039] When convergence difficulties occur during the iteration process, the stability of the finite element model is judged based on the Jacobian value of the element, the degree of element distortion, and the maximum displacement change. If the stability condition is met, the subsequent optimization iteration continues; otherwise, the incremental step size is automatically reduced and the calculation is repeated.

[0040] Step S5: Establish the objective function and constraints An optimization objective function is established based on the structural design requirements. The objective function may be the minimum flexibility, the maximum energy absorption, or a combination thereof.

[0041] Simultaneously establish volume fraction constraints: in, For unit volume, For the total volume of the design area, Given a volume fraction. In, for example Figure 2 In the implementation scheme shown, volume fraction Take 0.8, Step S6: Establish the accompanying sensitivity analysis model Taking the total derivative of the objective function with respect to the design variables yields the final sensitivity expression for the objective function with respect to the design variables: in, This is the global nodal displacement vector of the element; The space occupied by the unit in the reference configuration; Represents the nonlinear strain-displacement matrix of the element; The element-consistent tangent modulus is used for numerical reconstruction; the specific calculation method is as follows: First, select the strain and stress increments of two adjacent time steps output by commercial software (such as LS-DYNA). and ; In such Figure 3 In the embodiment shown, m=6 key stress-strain increments are selected to form a strain-stress increment matrix (6×6). and ; Finally, using Approximately obtained Then, substituting this into the final expression yields... .

[0042] Step S7: Update design variables The Moving Asymptote Algorithm (MMA) is used to update the design variables based on the sensitivity information.

[0043] Simultaneously, a density filtering method is used to perform weighted averaging on the design variables: , in, , and The first The filtered density, volume, and design variables of each unit.

[0044] Step S8: Iteratively optimize and output the results Repeat steps S4 to S7 until any of the following conditions are met: (1) The maximum change in design variables between two consecutive iterations is less than a preset threshold; (2) The rate of change of the objective function in two consecutive iterations is less than a preset threshold; (3) Optimize the number of iterations to reach the maximum allowed value.

[0045] After 14 iterations, the final output is the optimal topology that meets the requirements for lightweight design and collision performance, as shown below. Figure 2As shown, and based on the optimization results, engineering geometry reconstruction is performed to obtain the following... Figure 3 The final combined structure and physical model are shown. Finally, compression tests were conducted on the original physical model and the optimized physical model, and the corresponding force-displacement curves were output as follows: Figure 3 As shown in the figure. The results indicate that, thanks to the topology-optimized induced hole structure, the optimized square tube achieves uniform multi-layer progressive crushing, effectively suppressing the sudden drop in bearing capacity and eliminating drastic load fluctuations. The load curve is stable, and the specific energy absorption increases linearly. Compared with the original square tube, it has a more stable deformation mode and more reliable impact resistance and energy absorption characteristics. At the same time, due to the removal of some mass by the induced hole, a weight reduction of nearly 5% is achieved.

Claims

1. A structural collision topology optimization method based on finite element analysis for engineering applications, characterized in that, Includes the following steps: S1. Obtain the geometric model of the collision-bearing structure to be optimized. Before performing finite element analysis, divide the design region into a design region and a non-design region. Discretized A finite number of elements; a continuous density variable is defined for each element. This is used to characterize the state of existence of materials within a unit, and , The minimum density value set to prevent singularity of the stiffness matrix; Represents a solid unit. This represents an empty element, where the solid material is fixed in the non-design region; and defines the elastic modulus, Poisson's ratio, J2 yield stress, hardening modulus, contact relationship, collision boundary conditions in the form of rigid wall collision or velocity loading, contact relationship and constraint conditions; S2, using the unit density vector As a design variable for topology optimization, SIMP interpolation is used to establish the mapping relationship between the material-consistent elastoplastic tangent matrix and density: , in, The uniform elastoplastic tangent matrix of the element. For solid materials, a consistent elastoplastic tangent matrix is ​​formed. The uniform elastoplastic tangent matrix of empty elements. As a penalty factor; S3. Establish a finite element analysis model coupling geometric nonlinearity and material nonlinearity: A Total Lagrange finite strain description is used, and the deformation gradient F and Green-Lagrange strain are calculated from the displacement field. ,in The tensor is unit; for material nonlinearity, an isotropic J2 elastoplastic constitutive model and a back-mapping algorithm are used to update the plastic stress, which is strengthened by isotropic reinforcement. Evolutionary yield surface, in which, To adapt to equivalent plastic strain Evolving yield stress, The initial yield stress, To enhance the modulus; S4. Solve the nonlinear equilibrium equations using the Newton-Raphson incremental iteration method: , in, It is a nonlinear internal force vector. For external force vectors, It is the global displacement vector, and after assembling the uniform elastic-plastic tangent matrix of the structure, it is... Jacobian values ​​are used to solve the nonlinear equilibrium equations during the Newton-Raphson iteration process; the global nodal displacement vectors of the elements in the collision history are obtained by solving these equations. Cauchy stress field Green-Lagrange strain field G, equivalent plastic strain field When the Jacobian value of an element is negative, the element distortion angle exceeds the threshold, or the maximum displacement increment exceeds the threshold, the incremental step size is automatically reduced and the solution is resolved. S5. Establish an optimization objective function with the goal of minimizing compliance or maximizing energy absorption: , in, It is determined by a given volume fraction Defined optimization constraints, and These are the volume of the unit and the total volume of the design area, respectively. It is the lower bound of the design variable; Regarding the objective function with respect to design variables Find the total derivative and introduce the adjoint variable. The adjoint equation is established, and the sensitivity of the objective function to the density variable is derived by solving for the adjoint variables. The output of S4 is reused to reconstruct the uniform elastoplastic tangent matrix of the structure. The adjoint equation is: in, This is the transpose of the structural tangent stiffness matrix; S6. Update design variables based on sensitivity using the Moving Asymptote Algorithm (MMA). And employ weights based on the filter radius. Density filtering is performed by weighted averaging of neighborhood cell densities to suppress the dependence of checkerboard patterns on grids. S7. Substitute the updated density variable back into S3 to re-subdivide the stiffness and repeat S4–S6 until the maximum relative change of the design variable in two adjacent iterations is less than the first threshold C1, the relative change rate of the objective function is less than the second threshold C2, or the maximum number of iterations C3 is reached. S8. Output the optimal unit density distribution that satisfies the volume fraction constraint and collision energy absorption index. Perform geometric reconstruction on the distribution to generate a solid model for energy absorption box, anti-collision beam, longitudinal beam, battery pack protection structure or rail transit crash-resistant structure.

2. The structural collision topology optimization method based on finite element analysis for engineering applications as described in claim 1, characterized in that, In S3, the yield function described by the J2 elastoplastic model for material nonlinearity is: , in, For von Mises equivalent stress, Equivalent plastic strain is a historical variable that characterizes cumulative irreversible deformation.

3. The structural collision topology optimization method based on finite element analysis for engineering applications as described in claim 1, characterized in that, In S5, the specific expression for sensitivity is: , in, This is the global nodal displacement vector of the element; The space occupied by the unit in the reference configuration; Represents the nonlinear strain-displacement matrix of the element; The element-consistent tangent modulus for numerical reconstruction; the specific calculation method for sensitivity is as follows: Obtain the strain and stress increments at two adjacent time steps and ; Then, m key stress-strain increments are selected to form a strain-stress increment matrix. and : , , Finally, using Approximately obtained Then, substituting into the final expression, we obtain... .

4. The structural collision topology optimization method based on finite element analysis for engineering applications as described in claim 1, characterized in that, In S6, the density expression after node filtering in density filtering is: , in, , and The first The filtered density, volume, and design variables of each unit.

5. The structural collision topology optimization method based on finite element analysis for engineering applications according to claim 1, characterized in that, The iteration termination condition in S7 is specifically expressed as follows: , , , in, , and The set convergence threshold; Indicates the number of optimization iterations. Indicates the maximum allowed number of iterations; and This represents the optimized and updated design variable vector and objective function; and This represents the vector of variables and the objective function before optimization and update.

6. A computer-readable storage medium, characterized in that, The device contains a computer program that, when executed by a processor, implements the steps of the method according to any one of claims 1 to 5.

7. A collision structure optimization system, characterized in that, It includes a memory and a processor, wherein the processor calls a program in the memory to execute the method described in any one of claims 1 to 5, and is externally connected to a finite element solution module and a CAD reconstruction module to finally generate a CAD solid model.