A simulation modeling method for honeycomb wall barrier cells

Abstract

The present invention relates to the technical field of vehicles and discloses a simulation modeling method for honeycomb wall cell units. The method includes: establishing a discrete Y-shaped cross-section cell unit model based on the geometric symmetry of the honeycomb structure, and eliminating the interference of the air residue effect on the deformation law in traditional modeling by decomposing the honeycomb structure into discrete cell units; fitting the stress-strain curve of the honeycomb material based on the Y-shaped cross-section cell unit model, and realizing the coupled characterization of strain hardening, rate sensitivity and temperature softening effects through the inverse calibration of material constitutive parameters. The method reconstructs the honeycomb topology structure through discretized cells, suppresses the propagation of non-physical deformation, accurately restores the mechanical response characteristics during the dynamic compression process of the honeycomb material, solves the problem of distortion of the deformation mode caused by structural simplification in the traditional simulation model, and provides a high-precision basic model support for the digital modeling of automotive collision honeycomb barriers.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicles, and particularly to a method for simulating and modeling honeycomb wall cell elements. Background Art

[0002] In the field of automotive crash safety testing, honeycomb aluminum materials are widely used in the construction of wall models to simulate the collision energy absorption process due to their high compatibility with the mechanical properties of the vehicle front-end structure. However, traditional modeling methods have significant bottlenecks: although the equivalent material model simplifies calculations, it is difficult to capture the nonlinear deformation details of the honeycomb porous structure; the solid element model distorts the local buckling behavior due to excessive homogenization; while the shell element model, although able to restore geometric features, is limited by boundary effects such as adhesive failure and air entrapment, making it difficult to accurately reproduce the progressive collapse law during honeycomb crushing. Especially when the honeycomb is subjected to dynamic impact, existing models are prone to non-physical bottom deformation, which is significantly different from the regular folding mode dominated by the top of the physical object, severely restricting the credibility of simulation results.

[0003] On the other hand, the calibration accuracy of material constitutive parameters directly affects the simulation effectiveness. Traditional methods rely on local test data or empirical formulas, lacking the global mapping ability for complex dynamic responses, resulting in insufficient matching between stress-strain curves and test results. Especially under the coupling effect of strain rate sensitivity and temperature, the parameter inversion process is prone to falling into local optimal solutions, making it difficult to establish a strong correlation between material properties and macroscopic mechanical behaviors. How to construct a honeycomb simulation model that takes into account geometric details and physical authenticity while ensuring computational efficiency and realizing intelligent optimization of constitutive parameters has become a key problem in improving the level of crash safety design. Summary of the Invention

[0004] To solve the above technical problems, the present invention provides a method for simulating and modeling honeycomb wall cell elements. By establishing a discrete Y-shaped cross-section cell element model, the problem of air entrapment is eliminated, and the accuracy of the stress-strain curve is improved.

[0005] The present invention provides a method for simulating and modeling honeycomb wall cell elements, including:

[0006] Establishing a discrete Y-shaped cross-section cell element model based on the geometric symmetry of the honeycomb structure;

[0007] Fitting the stress-strain curve of the honeycomb material through the discrete Y-shaped cross-section cell element model;

[0008] The establishment of the discrete Y-shaped cross-section cell element model based on the geometric symmetry of the honeycomb structure includes:

[0009] Determining the minimum Y-shaped cell element based on the honeycomb structure;

[0010] Determine the discrete Y - type cross - section cell unit model based on the minimum Y - type cell unit;

[0011] Among them, the minimum Y - type cell unit is composed of two flange wall surfaces with an included angle of 120° bonded and connected;

[0012] The fitting of the stress - strain curve of the honeycomb material by the discrete Y - type cross - section cell unit model includes:

[0013] Determine the parameter space range of the material constitutive parameters of the minimum Y - type cell unit;

[0014] Generate multiple groups of parameter sampling sets within the parameter space range and conduct dynamic compression simulations to obtain corresponding simulated stress - strain data;

[0015] Train a parameter inversion model based on the mapping relationship between the simulated stress - strain data and the target test data;

[0016] Conduct reverse calibration on the test curve through the parameter inversion model, output the optimized constitutive parameters, and generate a fitted stress - strain curve.

[0017] Optionally, the minimum Y - type cell unit includes a first flange wall surface, a second flange wall surface, and a third flange wall surface. The first flange wall surface and the second flange wall surface are single - layer thickness shell units, and the third flange wall surface is a double - layer thickness shell unit.

[0018] Optionally, the third flange wall surface represents the bonded connection of two flange wall surfaces based on a single - layer thickness shell unit with double thickness.

[0019] Optionally, the third flange wall surface represents the bonded connection of two flange wall surfaces based on the contact of two single - layer thickness shell units.

[0020] Optionally, the third flange wall surface represents the bonded connection of two flange wall surfaces based on two single - layer thickness shell units and an adhesive layer located between the two single - layer thickness shell units.

[0021] Optionally, the material constitutive parameters satisfy the following fluid stress equation:

[0022] ;

[0023] Among them, is the yield stress, is the hardening modulus, is the hardening index, is the strain - rate sensitivity coefficient, is the temperature softening index, is the flow stress, is the plastic strain, is the strain rate, reference strain rate, is the melting temperature, is the current temperature.

[0024] Optionally, the parameter inversion model is a neural network model, whose input is the feature vector of stress-strain data and output is the material constitutive parameter vector.

[0025] Optionally,

[0026] the generation of the parameter sampling set includes dynamically adjusting the sampling density in the parameter space according to the prediction error feedback of the parameter inversion model, so that the distribution of sampling points matches the parameter sensitivity;

[0027] The inverse calibration includes an iterative optimization process. When the matching degree between the fitted stress-strain curve and the target test curve does not reach the preset condition, the parameter sampling set is re-expanded and the parameter inversion model is updated to generate a new fitted curve until the matching condition is met.

[0028] The present invention has the following technical effects:

[0029] The honeycomb wall barrier cell simulation modeling method provided by the present invention suppresses the non-physical deformation problems caused by the residual air effect and the weakening of the bonding interface strength in the traditional model from the structural root through the construction mechanism of the discretized Y-shaped cross-section cell unit. In the traditional solution, due to the lack of consideration of the internal air flow characteristics of the honeycomb cell and the dynamic failure behavior of the adhesive, abnormal buckling occurs in the bottom area during the simulation, showing a significant deviation from the regular crushing mode dominated by the top in the physical impact test. This method uses the discretized modeling of geometrically symmetric Y-shaped cell units to accurately characterize the load transfer path between adjacent honeycomb walls, eliminate the interference of residual air on the deformation law, and make the simulation model strictly follow the real physical characteristics of folding preferentially at the top and remaining stable at the bottom during the dynamic compression process, significantly improving the spatial consistency of the deformation mode.

[0030] Through the differential design of the hierarchical thickness of the flange wall surface and the diversified modeling strategy of adhesive connection, a balance is achieved between the restoration of interfacial mechanical properties and computational efficiency. The hybrid configuration of single-layer and double-layer shell elements controls the local stiffness, avoiding both the bottom virtual displacement caused by the hourglass effect in the traditional solid element model and ensuring that the adhesive interface does not experience unexpected strength attenuation during compression. Combining the global optimization ability of the parameter inversion model, the neural network is used to decouple the coupling relationship of multi-dimensional constitutive parameters, effectively solving the problem of insufficient adaptability of traditional calibration methods in scenarios with wide strain rates and temperature gradients, and making the morphological characteristics of the simulation curve highly match the experimental results in the elastic yield, plastic flow, and densification stages. In addition, the adaptive sampling mechanism in the parameter space significantly improves the calibration efficiency by dynamically focusing on the highly sensitive parameter region, providing a standardized process support for the rapid modeling of different batches of honeycomb materials. The multi-path implementation scheme of adhesive connection flexibly adapts to different simulation accuracy requirements through strategies such as contact algorithms, equivalent stiffness characterization, and material property definition, ensuring the authenticity of the interface separation behavior while reducing the computational resource consumption caused by iterative correction of adhesive failure in the traditional scheme, and providing a solution that takes into account both robustness and practicality for the performance prediction of honeycomb barriers under complex working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the specific embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0032] Figure 1 Schematic flow chart of a method for simulating and modeling honeycomb barrier cells provided by an embodiment of the present invention;

[0033] Figure 2 Schematic diagram of the structure of a minimum Y-shaped cell unit provided by an embodiment of the present invention;

[0034] Figure 3 Schematic diagram of a constraint direction provided by an embodiment of the present invention;

[0035] Figure 4 Schematic diagram of the structure of a minimum Y-shaped cell unit provided by an embodiment of the present invention;

[0036] Figure 5 Schematic diagram of the structure of another minimum Y-shaped cell unit provided by an embodiment of the present invention;

[0037] Figure 6 Schematic diagram of the structure of another minimum Y-shaped cell unit provided by an embodiment of the present invention;

[0038] Figure 7 A comparison diagram of deformation modes provided by an embodiment of the present invention;

[0039] Figure 8 A schematic diagram of the comparison of average compressive stress provided by an embodiment of the present invention;

[0040] Figure 9 A schematic diagram of the deformation mode of a physical honeycomb;

[0041] Figure 10 A schematic diagram of the results of traditional modeling and simulation;

[0042] Figure 11 A schematic diagram of the results of the modeling and simulation of the present invention.

[0043] Reference numerals

[0044] 11. First flange wall surface; 12. Second flange wall surface; 13. Third flange wall surface; 14. Adhesive layer. Detailed implementation manners

[0045] To make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope protected by the present invention.

[0046] Aiming at the defects of the traditional honeycomb wall barrier modeling method in terms of distorted deformation mode and difficulty in balancing calculation efficiency and accuracy, the core of the present invention lies in proposing a collaborative optimization method based on discretized cell reconstruction and parameter inverse calibration. By decomposing the honeycomb structure into geometrically symmetric Y-shaped cross-section cell units, the non-physical deformation caused by the air residue effect in the continuum model is eliminated; at the same time, an intelligent mapping relationship between the material constitutive parameters and the global mechanical response is established, and the dynamic compression simulation data is used to drive the parameter inversion model, breaking through the adaptability bottleneck of the traditional calibration method in the wide strain rate scenario. The following combines the drawings and embodiments to detail the technical implementation path of the present invention, specifically including: three core steps of discretized cell topology construction, multi-mode modeling of the bonding interface, and inverse optimization of constitutive parameters, as well as the test verification results.

[0047] Figure 1 A schematic diagram of the process of a honeycomb wall barrier cell simulation modeling method provided by an embodiment of the present invention, and the method includes:

[0048] S1. Establish a discrete Y-shaped cross-section cell unit model based on the geometric symmetry of the honeycomb structure;

[0049] S2. Fit the stress-strain curve of the honeycomb material through a discrete Y-shaped cross-section cell unit model.

[0050] The core of the honeycomb wall barrier cell simulation modeling method lies in reconstructing the geometric and mechanical properties of the honeycomb structure through a discretization modeling strategy. Traditional modeling methods, due to the use of continuum elements or homogenization treatment, are difficult to accurately characterize the porous topological features of honeycomb aluminum, resulting in the deviation of the deformation path from the physical test law in compression simulation. To solve this problem, this method is based on the geometric symmetry of the honeycomb structure, and decomposes the overall honeycomb block into discrete Y-shaped cross-section cell units. The spatial arrangement law of each cell unit is determined through geometric symmetry analysis, and the connection relationship between adjacent units is defined through a local coordinate system, eliminating the non-physical deformation caused by air residue or bonding interface simplification in traditional models. Discretization modeling transforms the complex honeycomb structure into an extensible set of standardized units through topological decomposition, retains the true geometric shape of the honeycomb wall surface, and at the same time constructs a mechanical response network between units, making the load transfer path during compression closer to the actual material behavior.

[0051] In specific implementation, this embodiment can construct a discretization model based on a finite element preprocessing tool and perform dynamic compression simulation using an explicit dynamics solver. Figure 2 Schematic diagram of a minimum Y-shaped cell unit structure provided by an embodiment of the present invention Figure 3 Schematic diagram of a constraint direction provided by an embodiment of the present invention. For the geometric symmetry characteristics of the honeycomb structure, select Figure 2 The minimum Y-shaped cell unit (UC unit) shown is used as the basic modeling unit. After allocating a local coordinate system to the surface of each UC unit, normal displacement constraints x and tangential coupling constraints yz are applied to the three surface nodes according to the Figure 3 Shown constraint direction to eliminate the influence of the unit edge curvature on the deformation mode.

[0052] When fitting the stress-strain curve of the honeycomb material based on the discretization model, it is necessary to establish the mapping relationship between the material constitutive parameters and the global mechanical response. Traditional parameter calibration relies on empirical formulas or local test data, and it is difficult to capture the coupling relationship of strain hardening, rate sensitivity, and temperature softening effects at the same time. This method optimizes the material constitutive parameters through an inverse calibration process, combines dynamic compression simulation to generate stress-strain data of multiple parameter sampling sets, and constructs a parameter-response database covering a wide range of working conditions. The synergistic effect of discretization modeling and parameter inversion enables the simulation model to reproduce the deformation law of top-priority buckling during dynamic compression and predict the stress plateau characteristics at different strain rates, significantly improving the consistency between the simulation results and physical tests.

[0053] In some embodiments, establishing a discrete Y-shaped cross-section cell unit model based on the geometric symmetry of the honeycomb structure includes:

[0054] Determine the minimum Y - type cell unit based on the honeycomb structure;

[0055] Determine the discrete Y - type cross - section cell unit model based on the minimum Y - type cell unit;

[0056] Among them, the minimum Y - type cell unit is composed of two flange wall surfaces with an included angle of 120° bonded and connected.

[0057] The determination of the minimum Y - type cell unit is the basis of discretized modeling. The periodic symmetry characteristics of the honeycomb structure determine the spatial arrangement law of its micro - units. Through the microscopic geometric analysis of honeycomb aluminum, it is found that it is composed of multiple periodically arranged hexagonal cells, and each hexagonal cell can be decomposed into six Y - type cross - section units. The Y - type cross - section unit is composed of two flange wall surface groups, and the included angle between the flange wall surfaces is determined by the honeycomb symmetry. Taking a vertex of the hexagonal cell as the symmetry center, the adjacent flange wall surface groups extend and intersect along a specific angle to form the basic configuration of the Y - type cross - section. The connection method of the flange wall surface groups directly affects the overall stiffness and deformation mode of the honeycomb structure. In the modeling process, the bonded connection between the flange wall surfaces is realized through geometric constraints and contact algorithms to ensure that the load transfer between adjacent units is consistent with the interface behavior of real materials.

[0058] The expansion of the minimum Y - type cell unit is completed through replication and spatial transformation. Based on the periodic characteristics of the honeycomb, the single Y - type unit is arranged in an array along the spatial extension direction of the honeycomb block, and the connection angle and spacing between the units are adjusted to match the geometric topology of the physical honeycomb. The connection relationship between the units of the discretized model is defined through shared nodes or contact pairs, avoiding the stiffness distortion caused by unit continuity in traditional integral modeling. The discretization strategy suppresses the bottom virtual displacement phenomenon in the compression simulation through local unit stiffness regulation, making the deformation strictly limited to the top area, which is consistent with the crushing law observed in the experiment.

[0059] In some embodiments, fitting the stress - strain curve of the honeycomb material through the discrete Y - type cross - section cell unit model includes:

[0060] Determine the parameter space range of the material constitutive parameters of the minimum Y - type cell unit;

[0061] Generate multiple groups of parameter sampling sets within the parameter space range and conduct dynamic compression simulations to obtain the corresponding simulated stress - strain data;

[0062] Train the parameter inversion model based on the mapping relationship between the simulated stress - strain data and the target experimental data;

[0063] Inverse - calibrate the experimental curve through the parameter inversion model, output the optimized constitutive parameters and generate the fitted stress - strain curve.

[0064] The inverse calibration of the constitutive parameters of the honeycomb material needs to be achieved through the collaborative optimization of parameter space sampling and dynamic compression simulation. The determination of the parameter space range needs to comprehensively consider the manufacturing process and mechanical properties of aluminum honeycomb, covering the physically reasonable ranges of key parameters such as yield stress, hardening modulus, and strain rate sensitivity coefficient. The generation of the parameter sampling set adopts Latin hypercube design to ensure that the sampling points are evenly distributed in the multi-dimensional space and avoid the omission of local parameter combinations. After each set of parameters is input into the finite element model, stress-strain data are obtained through dynamic compression simulation. The simulation conditions need to be consistent with the loading rate and boundary constraints of the physical test to ensure data comparability. The simulation data set extracts feature vectors through preprocessing, including key indicators such as elastic modulus, yield platform length, and densification strain, to construct a mapping relationship library of parameters-responses.

[0065] The training of the parameter inversion model is based on a deep neural network architecture. The input layer corresponds to the feature vector of the stress-strain curve, and the output layer is the constitutive parameter vector. Among them, the feature vectors of the input layer can include the slope of the elastic section, the average stress of the plastic platform, and the densification strain rate, which are extracted through segmented sampling. The network optimizes the weights through the backpropagation algorithm, and the loss function takes into account both the absolute error of the parameters and the similarity of the curve shape. The trained model can analyze the optimal combination of constitutive parameters from the test curve, breaking through the dependence on empirical formulas in traditional calibration methods. During the inverse calibration process, if the matching degree between the fitted curve and the test data is insufficient, it is necessary to expand the parameter sampling set and retrain the model, and gradually reduce the uncertainty of the parameter space through iterative optimization. This process realizes parameter update and simulation task scheduling through an automated script, significantly reducing the cost of manual intervention and providing a standardized process for the rapid modeling of different batches of honeycomb materials.

[0066] Continue to refer to Figure 2 In some embodiments, the minimum Y-shaped cell unit includes a first flange wall surface 11, a second flange wall surface 12, and a third flange wall surface 13. The first flange wall surface 11 and the second flange wall surface 12 are single-layer thickness shell units, and the third flange wall surface 13 is a double-layer thickness shell unit.

[0067] The thickness design of the flange wall directly affects the mechanical properties and computational efficiency of the model. In the traditional solid element model, the homogeneous thickness cannot characterize the stiffness difference of the bonded connection, resulting in weakened interface strength and hourglass effect. The first flange wall 11 and the second flange wall 12 adopt single-layer thickness shell elements to simulate the real thickness of the honeycomb aluminum foil; the third flange wall 13 adopts double-layer thickness shell elements to equivalently characterize the bonded connection of adjacent aluminum foils. The double-layer thickness shell element is realized by geometric superposition, and its equivalent stiffness is consistent with the real bonding interface, avoiding the iterative convergence problem of the contact algorithm. Buckling occurs preferentially in the single-layer shell element region, forming an initial folding band at the top; the buckling propagation is delayed in the double-layer shell element region due to the increased stiffness, showing a stepped progressive deformation. The mixed thickness design optimizes the overall mechanical response through local stiffness regulation, suppresses abnormal deformation at the bottom, and at the same time reduces the number of interface contact pairs and the computational complexity.

[0068] The simulation results of the mixed thickness model show that the deformation process strictly follows the physical law of preferential crushing at the top, and the bottom region remains stable due to the enhanced stiffness. This design significantly improves the computational efficiency while maintaining high accuracy, providing a feasible solution for large-scale honeycomb wall barrier modeling in engineering scenarios.

[0069] Figure 4 The figure shows a schematic diagram of a minimum Y-shaped cell unit structure provided by an embodiment of the present invention. In some embodiments, the third flange wall 13 characterizes the bonded connection of two flange walls based on a single-layer thickness shell element with double thickness. The single-layer thickness of the shell element is denoted as h, so Figure 4 the thickness of the single-layer thickness shell element with double thickness is 2h in the figure.

[0070] The bonded connection of the third flange wall 13 is realized by a double-layer thickness shell element. This design equivalently represents the bonding interface of adjacent honeycomb aluminum foils as a single shell element structure. The traditional contact algorithm requires defining complex contact pairs and failure criteria, which easily leads to difficulties in computational convergence. By the principle of geometric superposition, this method combines the bonding regions of two single-layer shell elements into a double-layer thickness shell element, and its equivalent bending stiffness is consistent with the bending characteristics of the real bonding interface. The thickness parameter of the double-layer shell element is determined according to the bonding process of the honeycomb aluminum foil to ensure that the stiffness distribution in the interface region matches the mechanical behavior of the physical material. In the dynamic compression simulation, the double-layer shell element region forms a local stiffness strengthening band due to the increased thickness, suppressing the diffusion of unexpected deformation caused by premature failure of the bonding interface. This equivalent modeling method significantly improves the simulation efficiency by reducing the number of contact pairs and the iterative calculation requirements, and at the same time avoids the model reconstruction cost caused by the correction of bonding failure in the traditional contact algorithm.

[0071] The geometric continuity of the double-layer thickness shell element ensures the integrity of the load transfer path. During the compression process, the double-layer shell element of the third flange wall 13 serves as a mechanical transition region, evenly dispersing the stress wave generated by the top buckling to adjacent elements, and avoiding mesh distortion caused by local stress concentration. The simulation results show that this design can accurately reproduce the progressive failure behavior of the honeycomb material at the bonding interface, making the deformation mode of the overall model highly consistent with the laminated crushing characteristics observed in the physical test.

[0072] Figure 5 It is a schematic diagram of another minimum Y-shaped cell unit structure provided by an embodiment of the present invention. In some embodiments, the third flange wall 13 represents the bonding connection of the two flange walls based on the contact of two single-layer thickness shell elements.

[0073] The bonding connection of the third flange wall 13 is realized through the contact algorithm of two single-layer shell elements, and this scheme retains the true geometric shape of the honeycomb aluminum foil. The two single-layer shell elements are arranged in parallel along the bonding interface, and the contact surface defines the bonding behavior through master-slave nodes. The master node bears the normal contact force and tangential friction stress, and the slave node is synchronized with the movement of the master node through the constraint equation to ensure the continuity of the interface load transfer. A bonding strength threshold is set in the contact algorithm, and when the interface stress exceeds the threshold, failure separation is triggered to simulate the peeling behavior of the real bonding layer. This design avoids the problem of interface strength distortion caused by stiffness averaging in the traditional equivalent thickness model through refined contact definition.

[0074] The parameter settings of the contact algorithm need to match the bonding process parameters of the honeycomb aluminum foil. The normal contact stiffness is calibrated according to the compression modulus of the bonding layer, and the tangential friction coefficient is determined through the interface shear test data. In the dynamic compression simulation, the failure threshold of the contact pair controls the separation timing of the bonding interface, making the deformation process show the characteristics of staged buckling expansion. While retaining the mechanical details of the interface, this scheme reduces the global calculation amount through the localization definition of the contact pair, and is applicable to the simulation of medium and small honeycomb models with strict accuracy requirements for interface failure behavior.

[0075] Figure 6 It is a schematic diagram of another minimum Y-shaped cell unit structure provided by an embodiment of the present invention. In some embodiments, the third flange wall 13 represents the bonding connection of the two flange walls based on two single-layer thickness shell elements and a bonding layer 14 located between the two single-layer thickness shell elements.

[0076] The bonded connection of the third flange wall surface 13 is realized through two single-layer shell elements and an intermediate bonding layer 14. This solution uses independent unit layers to accurately characterize the interfacial mechanical properties. The bonding layer 14 is composed of thin solid elements, and its thickness is consistent with the actual adhesive coating process. The bonding layer elements are connected to the upper and lower single-layer shell elements through shared nodes to form a "sandwich" structure. For example, a solid element with a thickness of 0.05 mm can be inserted between the single-layer shell elements, and the elastic-plastic constitutive model of the adhesive is used to define the interfacial mechanical behavior, and the separation process is characterized by the normal / tangential stress function. The material properties of the bonding layer are defined according to the tensile-shear coupling response of the adhesive, and the normal strength and tangential peel threshold are calibrated through interface test data. Through the introduction of independent bonding layer elements, this design can accurately simulate the progressive failure process of the bonding interface under complex loads, including normal peel, shear slip, and mixed-mode fracture.

[0077] The mechanical response of the bonding layer elements directly affects the deformation mode of the overall model. In the dynamic compression simulation, the plastic strain accumulation of the bonding layer 14 triggers interface failure, resulting in the gradual separation of the upper and lower shell elements. This failure mechanism is consistent with the evolution process of the bonding layer 14 from elastic deformation to fracture in physical tests, enabling the simulation model to capture the stress relaxation characteristics of honeycomb materials in the later stage of crushing. Through the refined modeling of the bonding layer elements, this solution provides a high-fidelity simulation tool for studying the influence of interface failure on the overall mechanical behavior of honeycombs, and is suitable for accurate prediction of honeycomb performance under extreme working conditions in fields such as aerospace.

[0078] Figure 7 This is a comparison diagram of deformation modes provided by an embodiment of the present invention. As Figure 7 shown, the simulated deformation modes of the three models (double-thickness shell elements, contact algorithm, bonding layer solid elements) are all consistent with the results of actual physical tests, verifying the accuracy of the present invention in terms of top-priority buckling characteristics. Model 1 is Figure 4 the corresponding embodiment, Model 2 is Figure 5 the corresponding embodiment, and Model 3 is Figure 6 the corresponding embodiment. Among them, the double-thickness shell element model (corresponding to Figure 4 the embodiment) has the most regular deformation propagation path, indicating that its bonding interface stiffness strengthening strategy effectively suppresses asymmetric folding.

[0079] Figure 8 This is a schematic diagram of the comparison of average compression stress provided by an embodiment of the present invention. Figure 8 The average collision intensities of the three models are further compared with the test results. Model 1 is Figure 4 the corresponding embodiment, Model 2 is Figure 5 the corresponding embodiment, and Model 3 is Figure 6 the corresponding embodiment. The double-thickness shell element model (corresponding to Figure 4The error between the collision intensity curve in the embodiment of () and the test value in the plastic plateau stage is less than 5%, which is significantly better than the traditional modeling method (error > 15%). It proves that through the discretized cell reconstruction and parameter inversion optimization, the present invention can accurately characterize the dynamic energy absorption characteristics of honeycomb materials.

[0080] The calculation times of three models (i.e., Figure 4 , 5 , 6) are compared in Table 1. Model 1 only takes 10 minutes. Among them, Model 1 is the model corresponding to Figure 4 , Model 2 is the model corresponding to Figure 5 , and Model 3 is the model corresponding to Figure 6 .

[0081] Table 1 - Comparison of processing times of three models

[0082]

[0083] In some embodiments, the material constitutive parameters satisfy the following fluid stress equation:

[0084] ;

[0085] Wherein, is the yield stress, is the hardening modulus, is the hardening index, is the strain rate sensitivity coefficient, is the temperature softening index, is the flow stress (elastic + plastic components), is the plastic strain (true strain), is the strain rate, is the reference strain rate, is the melting temperature in Kelvin degrees. The temperature calculation is assumed to be carried out under adiabatic conditions, is the current temperature, . is the specific heat per unit volume, is the initial temperature (unit: Kelvin), is the heat per unit volume under adiabatic conditions. Wherein, , , , , have value ranges of 100 - 500, 40 - 800, 0.01 - 0.6, 0.01 - 1, and 1 - 2 respectively.

[0086] The material constitutive equation needs to simultaneously characterize the strain hardening, rate sensitivity, and temperature softening effects of aluminum honeycomb. The mathematical form of the equation needs to satisfy the nonlinear characteristics of the stress-strain relationship in dynamic compression simulation, and its core terms include the strain hardening term, rate strengthening term, and temperature softening term. The strain hardening term describes the stress increase caused by the accumulation of plastic strain through a power function, and the hardening index controls the rising slope of the curve. The rate strengthening term uses a logarithmic function to relate the strain rate and the stress increment, reflecting the improvement of the material's resistance to deformation at high loading rates. The temperature softening term simulates the attenuation of the material's strength caused by the increase in temperature through an exponential function, and the softening index determines the attenuation rate. The physical meanings of the parameters in the equation need to correspond to the micro-mechanical behavior of aluminum honeycomb. For example, the yield stress corresponds to the initial plastic deformation threshold, and the hardening modulus reflects the resistance to dislocation movement.

[0087] The calibration of the constitutive parameters needs to combine the experimental data of multiple working conditions. The strain hardening and rate sensitivity characteristics are obtained through constant strain rate compression tests, and the temperature softening effect is calibrated through variable temperature tests. In the simulation model, the parameter combination of the equation needs to ensure the consistency of the stress-strain curve morphology under different loading conditions. For example, the stress plateau at high strain rates needs to match the steep rising characteristics of the experimental curve through the adjustment of the rate sensitivity coefficient, and the stress drop caused by the increase in temperature needs to be reproduced through the optimization of the softening index. The multi-term coupling mechanism of the equation enables it to adapt to the simulation requirements of a wide range of working conditions, providing a reliable theoretical framework for the joint analysis of transient impact and thermo-mechanical coupling effects in vehicle collision scenarios.

[0088] In some embodiments, the parameter inversion model is a neural network model, whose input is the feature vector of stress-strain data, and the output is the material constitutive parameter vector.

[0089] The architecture design of the neural network model needs to balance the accuracy and computational efficiency of parameter inversion. The input layer receives the feature vector of the preprocessed stress-strain curve, including key morphological indexes such as the slope of the elastic segment, the length of the plastic plateau, and the densification strain rate. The hidden layer adopts a fully connected structure to realize the high-dimensional mapping of features through a non-linear activation function. The output layer corresponds to the constitutive parameter vector, and the node values are directly related to physical quantities such as the yield stress and the hardening modulus. The training data of the network comes from the dynamic compression simulation results of the parameter sampling set, and the data set needs to cover the mechanical response range of honeycomb materials under typical working conditions. During the training process, the design of the loss function needs to balance the absolute error of the parameters and the similarity of the curve morphology. For example, the dynamic time warping algorithm is introduced to quantify the morphological differences between curves, avoiding the local optimal trap caused by a single mean square error.

[0090] The inversion ability of the neural network depends on the effective representation of feature vectors. The feature extraction of the stress-strain curve is realized by segment sampling with a sliding window and principal component analysis, which eliminates noise interference and retains the key morphological information of the curve. The trained network can parse out the implicit combination of material parameters from the fluctuation characteristics of the test curve. For example, the hardening index can be inferred from the steepness of the plastic plateau, or the temperature softening effect can be judged from the hysteresis of the densification strain. This model breaks through the dependence on prior formulas of traditional inversion methods and is especially good at dealing with non-linear problems with multi-parameter coupling, such as the stress relaxation characteristics under the combined action of strain rate and temperature. Simulation verification shows that the constitutive parameters inverted by the neural network can make the fitting curve highly consistent with the test data within a wide range of working conditions, providing reliable support for the modeling of honeycomb materials under complex loading conditions.

[0091] In specific implementation, the initial constitutive parameters can be obtained through the following two methods:

[0092] 1) Based on the fitting function of finite element software and the JOHNSON-COOK model, input the yield strength, ultimate strength and necking strain of the material, and automatically generate the parameters;

[0093] 2) Use the parameter optimization tool to iteratively solve the optimal parameter combination by the least square method through experimental stress-strain data. The above parameters are used as the initial sampling set for neural network training.

[0094] In some embodiments,

[0095] The generation of the parameter sampling set includes dynamically adjusting the sampling density in the parameter space according to the prediction error feedback of the parameter inversion model, so that the distribution of sampling points matches the parameter sensitivity;

[0096] The specific adjustment rules are as follows:

[0097] If the prediction error of the parameter inversion model exceeds the preset threshold (such as 10%), reduce the original parameter sampling interval by 50% and resample;

[0098] If the model accuracy meets the requirements but the calculation time is too long, expand the sampling interval to 2 times the original interval.

[0099] Example:

[0100] Taking the calibration of the yield stress parameter A as an example, the initial sampling range is 100 - 500 MPa, and the interval is 100 MPa (taking values 100, 200, 300, 400, 500).

[0101] If the fitting error exceeds 10% after the first inversion, reduce the interval to 50 MPa and resample (taking values 100, 150, 200,..., 500);

[0102] If the error is less than 5% but the single simulation takes more than 24 hours, the interval is expanded to 200 MPa (taking values of 100, 300, 500).

[0103] Inverse calibration includes an iterative optimization process. When the matching degree between the fitted stress-strain curve and the target test curve does not reach the preset condition, the parameter sampling set is re-expanded and the parameter inversion model is updated to generate a new fitted curve until the matching condition is met.

[0104] The dynamic adjustment mechanism of the parameter sampling set realizes iterative optimization through feedback control. The range of the initial parameter space is determined according to the material process and historical test data, and the Latin hypercube design is used to generate uniformly distributed sampling points. The matching degree between the simulation curve corresponding to each group of parameters and the target test curve is quantified by similarity scoring, and the scoring result is fed back to the sampling controller. The controller identifies the parameter sensitive area according to the scoring distribution, increases the sampling density in the high-sensitive area in subsequent iterations, and reduces the redundant sampling in the low-sensitive area at the same time. For example, if a small change in a certain parameter subspace causes a significant fluctuation in the matching degree, it is determined that the area is highly sensitive and the number of sampling points needs to be increased by adaptive mesh refinement.

[0105] The iterative optimization process gradually converges to the optimal parameter solution through closed-loop feedback. After each iteration, the newly added sampling data is used to update the parameter inversion model to improve its prediction accuracy in the high-sensitive area. If the matching degree of the fitted curve output by the current inversion model does not reach the threshold, the parameter space expansion mechanism is triggered, such as expanding the parameter boundary along the sensitive direction or introducing new coupling dimensions. This process continues until the matching degree meets the preset condition or the number of iterations is exhausted. The dynamic adjustment strategy significantly improves the calibration efficiency by balancing "exploration" (sampling in unknown areas) and "exploitation" (refinement in high-sensitive areas). Engineering applications show that this mechanism can greatly reduce the total number of simulation times for parameter calibration, while ensuring the global optimality of the inverted parameters, providing core technical support for the rapid consistency evaluation of multi-batch honeycomb materials.

[0106] In specific implementation, the test can use commercial 3003 aluminum honeycomb samples with specifications of 250mm×250mm×150mm and a wall thickness of 19mm. A 72kg indenter is used to impact and test at a speed of 5.24m / s, and the compressive stress is calculated by σ = F / A, where F is the impact force and A is the out-of-plane cross-sectional area. Figure 9 It is a schematic diagram of the deformation mode of the physical honeycomb Figure 10 It is a schematic diagram of the traditional modeling and simulation results Figure 11 It is a schematic diagram of the modeling and simulation results of the present invention. As Figures 9 - 11 shown, the physical deformation mode is highly consistent with the simulation results of the present invention in terms of the top-priority buckling characteristics, while the traditional model shows non-physical deformation at the bottom.

[0107] An embodiment of the present application may also be a computer-readable storage medium, which includes computer program instructions that cause a computer to execute the steps of the method for controlling the braking consistency between the tractor and the trailer in the tractor provided in any embodiment of the present application.

[0108] The computer program instructions can be written in any combination of one or more programming languages to write program code for performing the operations of the embodiments of the present application. The programming languages include object-oriented programming languages such as Java, C++, etc., and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computing device, partially on the user's device, executed as an independent software package, partially on the user's computing device and partially on a remote computing device, or entirely on a remote computing device or server.

[0109] The computer-readable storage medium can adopt any combination of one or more readable media. The readable medium can be a readable signal medium or a readable storage medium. The readable storage medium can, for example, include but is not limited to an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples (non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0110] It should be noted that the terms used in the present application are only for describing specific embodiments and do not limit the scope of the present application. As shown in the specification and claims of the present application, unless the context clearly indicates an exception, words such as "a", "an", "one", and / or "the" are not specifically singular and may also include plural. The term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method or device including a series of elements not only includes those elements but also includes other elements not explicitly listed, or further includes elements inherent to such process, method or device. Without further limitation, an element defined by the statement "including a..." does not exclude the existence of additional identical elements in the process, method or device including the element.

[0111] In this text, specific examples are used to elaborate on the principles and implementation modes of this application. The description of the above embodiments is only used to help understand the method and its core idea of this application. The above are only the preferred implementation modes of this application. It should be noted that due to the limited nature of written expression and the objectively infinite specific structures, for those of ordinary skill in the art in this technical field, without departing from the principles of this application, several improvements, refinements or changes can also be made, or the above technical features can be combined in an appropriate manner; these improvements, refinements, changes or combinations, or directly applying the concept and technical solution of the invention to other occasions without improvement, shall all be regarded as the protection scope of this application.

Claims

1. A simulation modeling method for honeycomb wall barrier cells, characterized in that Including: Establishing a discrete Y-shaped cross-section cell unit model based on the geometric symmetry of the honeycomb structure; Fitting the stress-strain curve of the honeycomb material through the discrete Y-shaped cross-section cell unit model; The establishing of the discrete Y-shaped cross-section cell unit model based on the geometric symmetry of the honeycomb structure includes: Determining the minimum Y-shaped cell unit based on the honeycomb structure; Determining the discrete Y-shaped cross-section cell unit model based on the minimum Y-shaped cell unit; Wherein, the minimum Y-shaped cell unit is formed by bonding and connecting two flange wall surfaces with an included angle of 120°; The fitting of the stress-strain curve of the honeycomb material through the discrete Y-shaped cross-section cell unit model includes: Determining the parameter space range of the material constitutive parameters of the minimum Y-shaped cell unit; Generating multiple groups of parameter sampling sets within the parameter space range and performing dynamic compression simulation to obtain corresponding simulated stress-strain data; Training a parameter inversion model based on the mapping relationship between the simulated stress-strain data and the target test data; Performing reverse calibration on the test curve through the parameter inversion model, outputting optimized constitutive parameters and generating a fitted stress-strain curve; The material constitutive parameters satisfy the following fluid stress equation: ； Among them, is the yield stress, is the hardening modulus, is the hardening index, is the strain rate sensitivity coefficient, is the temperature softening index, is the flow stress, is the plastic strain, is the strain rate, is the reference strain rate, is the melting temperature, is the current temperature; The parameter inversion model is a neural network model, whose input is the feature vector of the stress-strain data and the output is the material constitutive parameter vector; The generation of the parameter sampling set includes dynamically adjusting the sampling density within the parameter space according to the prediction error feedback of the parameter inversion model to make the distribution of the sampling points match the parameter sensitivity; The reverse calibration includes an iterative optimization process. When the matching degree between the fitted stress-strain curve and the target test curve does not reach the preset condition, the parameter sampling set is re-expanded and the parameter inversion model is updated to generate a new fitted curve until the matching condition is met.

2. The method according to claim 1, wherein The minimum Y-shaped cell unit includes a first flange wall surface, a second flange wall surface and a third flange wall surface. The first flange wall surface and the second flange wall surface are single-layer thickness shell units, and the third flange wall surface is a double-layer thickness shell unit.

3. The method according to claim 2, wherein The third flange wall surface characterizes the bonding connection of the two flange wall surfaces based on the single-layer thickness shell unit with double thickness.

4. The method according to claim 2, wherein The third flange wall surface characterizes the bonding connection of the two flange wall surfaces based on the contact of the two single-layer thickness shell units.

5. The method according to claim 2, characterized in that The third flange wall surface characterizes the bonding connection of the two flange wall surfaces based on the two single-layer thickness shell units and the bonding layer located between the two single-layer thickness shell units.

Images