Multi-parameter collaborative optimization design method for lateral impact resistance of honeycomb sandwich panel battery pack

Abstract

The invention discloses a multi-parameter collaborative optimization design method for lateral impact resistance of a honeycomb sandwich panel battery pack, and relates to the technical field of structural safety and lightweight design of new energy automobile battery packs. An existing battery pack generally has the defects of insufficient light weight, limited energy absorption efficiency and lack of multi-parameter collaborative optimization with an upper box body, a lower box body and a stiffening beam in impact resistance research. Therefore, according to the honeycomb sandwich panel battery pack lateral impact resistance multi-parameter collaborative optimization design and parameter calculation method, a honeycomb sandwich panel finite element model is established, performance parameters are obtained, an optimal energy absorption combination is screened, then a battery pack box body finite element model is established, and multiple design variables are set; and constructing a function relationship between variables and responses by utilizing an agent model, and carrying out multi-objective optimization by combining an NSGA-II algorithm to finally obtain an optimal design parameter meeting a safety threshold and carrying out simulation verification. The method is suitable for battery pack anti-impact protection design work of the electric automobile under the lateral collision working condition.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of safety and lightweight design of new energy automobile battery pack structures, and particularly relates to a multi-parameter collaborative optimization design for lateral impact resistance of a honeycomb sandwich plate battery pack. BACKGROUND

[0002] With the acceleration of global energy transformation and the deepening of the "double carbon" target, the number of new energy vehicles in China has increased explosively, and fire accidents have also increased year by year, many of which are caused by impact load, mechanical abuse and thermal runaway fire. Among them, the threat of side impact to the battery pack is particularly prominent due to the short buffer zone of the vehicle body and the limited energy absorption space. Therefore, it is crucial to protect the impact resistance of the battery pack for the safety of passengers and vehicles.

[0003] In recent years, scholars have studied the response of battery packs under impact. Ma Bin et al. established an anisotropic model of battery pack bracket, box and battery, and found that increasing the thickness of the battery pack box and bracket can effectively improve the impact resistance of the battery pack. Qu Dongren et al. further analyzed the influence of stress and strain characteristics inside the battery under sharp foreign object impact conditions, and the results showed that the box material type is the key factor. Haris et al. evaluated the vibration and impact performance of the battery pack battery group through finite element analysis method, and found that the battery may be damaged under side impact, and the stress level of the battery needs to be reduced by increasing the energy absorption structure. In order to balance the lightweight and impact resistance requirements, the honeycomb sandwich panel becomes a candidate material due to its high specific strength, high specific stiffness and multi-cell core plastic energy dissipation characteristics, and its performance can be quantified by mass specific energy absorption, initial peak force, average platform stress and other indicators. Such structures have attracted widespread attention in the field of battery pack protection. Mustaffa et al. analyzed the damage characteristics of the battery group under lateral impact using finite element method, and proposed a honeycomb reinforced structure, and compared the performance of three reinforced materials: high-strength steel, stainless steel and aluminum alloy. The results showed that when using high-strength steel, stainless steel and aluminum alloy materials, the penetration depth is reduced by 22.4%, 20.2% and 19.4% respectively, and the total energy absorption of the battery group is reduced by 18.7%. Arslan et al. designed a honeycomb and auxetic hybrid structure to improve the impact resistance of the side beam of the battery pack, and combined ANN and NSGA-II algorithms for multi-objective optimization, the results showed that the weight of the optimized design was reduced by 23.9%, and the performance was improved by 3%. LI et al. designed seven different bionic honeycomb structures attached to the outside of the battery pack box, and through 10m / s rigid column lateral impact simulation, the deformation of the battery pack box, the maximum stress and the energy absorption of the honeycomb structure were calculated. The results showed that the bionic grass stem honeycomb structure improved the safety performance of the battery pack the most significantly, and after further parameter optimization of the structure, the deformation of the box was reduced by 30%, and the maximum stress was reduced by 10%. Ashvin et al. used a hexagonal honeycomb structure to design the frame of the battery pack box, and analyzed the mechanical properties of the battery pack under gravity load, and found that the modal frequency increased with the increase of the height of the honeycomb core, compared with the original design, the deformation of the honeycomb box was reduced by 40.57%, and the stress was reduced by 25.50%. In summary, the research on the impact load of the battery pack is mostly from the material properties, and the sandwich panel box is generally limited to the design of the parameters of the honeycomb, and a multi-parameter collaborative optimization system with the upper and lower boxes and internal reinforcing beams has not been established, resulting in insufficient analysis depth.

[0004] In summary, the existing technology is mostly focused on increasing the wall thickness, changing the material or introducing a single means such as honeycomb structure to improve the impact resistance of the battery pack, but there are generally defects such as insufficient lightweight, limited energy absorption efficiency, and lack of multi-parameter collaborative optimization with the upper and lower boxes and reinforcing beams. SUMMARY

[0005] To solve the defects of the battery pack impact resistance performance in the prior art, such as insufficient lightweight, limited energy absorption efficiency, and lack of multi-parameter collaborative optimization with the upper and lower boxes and the reinforcing beam, the technical scheme provided by the present application is as follows: A honeycomb sandwich panel battery pack lateral impact resistance multi-parameter collaborative optimization design parameter calculation method, comprising: a step of establishing a honeycomb sandwich panel finite element model and obtaining performance evaluation parameters as basic data for structure screening; a step of obtaining an energy absorption combination of the honeycomb sandwich panel that meets the preset conditions based on the performance evaluation parameters, and taking the combination as an input condition for modeling the battery pack box structure; a step of establishing a battery pack box finite element model based on the energy absorption combination of the honeycomb sandwich panel that meets the preset conditions and presetting design variables, and outputting structure performance indicators; a step of constructing a surrogate model based on the structure performance indicators and establishing a functional relationship between the design variables and the responses; a step of using the surrogate model as input, using the NSGA-II algorithm for multi-objective optimization, and outputting design parameters; a step of substituting the design parameters into the battery pack finite element model for verification and outputting optimization results.

[0006] Further, in a preferred embodiment, the performance evaluation parameters include initial peak force, total energy absorption, mass-specific energy absorption, and average platform stress.

[0007] Further, in a preferred embodiment, the energy absorption combination of the honeycomb sandwich panel that meets the preset conditions is a combination of a hexagonal honeycomb structure and a carbon fiber reinforced composite panel.

[0008] Further, in a preferred embodiment, the objective function of the multi-objective optimization includes the battery pack mass and the intrusion amount, and the constraint condition is that the extrusion force reaches 100 kN and the intrusion amount is less than 20 mm.

[0009] A honeycomb sandwich panel battery pack lateral impact resistance multi-parameter collaborative optimization design parameter calculation device is also provided, comprising: a module for establishing a honeycomb sandwich panel finite element model and obtaining performance evaluation parameters as basic data for structure screening; a module for obtaining an energy absorption combination of the honeycomb sandwich panel that meets the preset conditions based on the performance evaluation parameters, and taking the combination as an input condition for modeling the battery pack box structure; a module for establishing a battery pack box finite element model based on the energy absorption combination of the honeycomb sandwich panel that meets the preset conditions and presetting design variables, and outputting structure performance indicators; Based on the structural performance index, a proxy model is constructed to establish the functional relationship between the design variables and the response. Using the proxy model as input, the NSGA-II algorithm is used for multi-objective optimization and the design parameters are output. The design parameters are substituted into the battery pack finite element model for verification and the optimization results are output.

[0010] A honeycomb sandwich plate battery pack lateral impact resistance multi-parameter collaborative optimization design method is also provided, comprising: According to the optimization results output by the parameter calculation method, the steps of designing the honeycomb sandwich plate battery pack are performed.

[0011] A honeycomb sandwich plate battery pack lateral impact resistance multi-parameter collaborative optimization design device is also provided, comprising: According to the optimization results output by the parameter calculation method, the module for designing the honeycomb sandwich plate battery pack is performed.

[0012] A computer storage medium is also provided for storing a computer program, which causes the computer to execute the method when the computer program is read by the computer.

[0013] A computer is also provided, comprising a processor and a storage medium, which executes the method when the processor reads the computer program stored in the storage medium.

[0014] A computer program product is also provided as a computer program, which implements the method when executed.

[0015] Compared with the prior art, the technical solution provided by the present application has the following advantages: The present application adopts a combination of hexagonal honeycomb structure and CFRP panel to construct the battery pack box, which can significantly reduce the weight while maintaining the structural strength compared to the traditional aluminum alloy frame. The hexagonal honeycomb structure can more stably disperse the impact load during the crushing process due to its symmetry and uniform stress characteristics. Combined with the lightweight and high-strength CFRP panel, the battery pack achieves better energy absorption performance and lightweight effect than the traditional thick-walled aluminum alloy reinforcement method, thereby effectively improving the endurance of electric vehicles.

[0016] The present application optimizes the thickness, cross-sectional height and side length of the honeycomb core to improve the specific energy absorption level. Research shows that the traditional method of increasing the thickness of the metal box to improve impact resistance often results in a significant increase in weight. The present application improves the energy absorption efficiency by adjusting the structural parameters to achieve higher energy absorption capacity under limited mass, overcoming the contradiction between weight and energy absorption in existing research.

[0017] This scheme introduces multi-parameter collaborative optimization in the design of battery pack upper and lower box and stiffening beam, avoiding the limitations brought by single thickening or simply adding honeycomb structure. Through genetic algorithm for global optimization of multiple variables, the optimal combination of box thickness and honeycomb geometric parameters is determined, which reduces the extrusion force by 22.52% and meets the bearing requirements. Compared with existing research relying on single factor adjustment, a more balanced result between performance and safety is achieved.

[0018] This scheme introduces RBF approximation model to replace the traditional finite element successive iteration optimization, greatly improving the calculation efficiency. In existing research, the optimization process can only be tried within a limited parameter range due to the long simulation time. This scheme uses the approximation model to accelerate the optimization while ensuring accuracy, enabling parameter optimization in multi-dimensional space. The error in simulation verification is controlled within 5%, ensuring the feasibility and reliability of the optimization.

[0019] Under lateral impact conditions, the results before and after optimization show that the battery pack mass is reduced by 47.21%, and the energy absorption is increased by 34.43%, which exceeds the level of existing research that can only achieve 20% to 30% weight reduction or 10% to 20% energy absorption improvement. Although the intrusion amount increases, it is still below the safety threshold of 20mm, proving that lightweight and energy absorption performance are optimized while ensuring safety, breaking the limitations of existing single structure design that cannot balance safety and lightweight.

[0020] Suitable for battery pack impact protection design under lateral collision conditions of electric vehicles. BRIEF DESCRIPTION OF DRAWINGS

[0021] Figure 1 For sandwich plate quasi-static compression finite element model.

[0022] Figure 2 For honeycomb sandwich plate energy curve.

[0023] Figure 3 For evaluation index and factor relationship curve.

[0024] Figure 4 For battery pack finite element model.

[0025] Figure 5 For approximation model evaluation coefficient.

[0026] Figure 6 For actual M and predicted M inspection chart.

[0027] Figure 7 For actual SEA and predicted SEA inspection chart.

[0028] Figure 8 For actual F and predicted F inspection chart.

[0029] Figure 9 Actual D and predicted D test chart.

[0030] Figure 10 Correlation coefficient of variable and response.

[0031] Figure 11 X9 and X 10 Effect on quality.

[0032] Figure 12 Effect of X3 and X2 on quality specific energy.

[0033] Figure 13 X 10 Effect of X

[0034] Figure 14 Effect of X1 and X2 on intrusion.

[0035] Figure 15 Flowchart based on NSGA-II algorithm optimization.

[0036] Figure 16 Optimized solution set.

[0037] Figure 17 Comparison of intrusion and extrusion pressure before and after optimization.

[0038] Figure 18 Comparison of quality specific energy before and after optimization. DETAILED DESCRIPTION

[0039] In order to make the advantages and benefits of the technical solutions provided by the present application more clear, the technical solutions provided by the present application will be further described in detail in combination with the drawings. Specifically: Embodiment one, the embodiment provides a kind of honeycomb sandwich plate battery pack lateral impact resistance multi-parameter collaborative optimization design parameter calculation method, comprising: The step of establishing a honeycomb sandwich plate finite element model and obtaining performance evaluation parameters as basic data for structure screening; Based on performance evaluation parameters, obtain the energy absorption combination of honeycomb sandwich plate that meets the preset condition, and the combination is used as the input condition of battery pack box structure modeling step; On the basis of the energy absorption combination of honeycomb sandwich plate that meets the preset condition, the step of establishing a battery pack box finite element model and presetting design variables, and outputting structure performance index; Based on structure performance index, the step of constructing proxy model and establishing the functional relationship between design variables and response; The step of using proxy model as input, using NSGA-II algorithm for multi-objective optimization and outputting design parameters; the step of substituting the design parameters into the battery pack finite element model for verification and outputting the optimization results.

[0040] The performance evaluation parameters include initial peak force, total energy absorption, mass-specific energy absorption, and average plateau stress.

[0041] The energy absorption combination of the honeycomb sandwich panel that meets the preset condition is a combination of a hexagonal honeycomb structure and a carbon fiber reinforced composite panel.

[0042] The objective function of the multi-objective optimization includes battery pack mass and intrusion amount, and the constraint condition is that the extrusion force reaches 100 kN and the intrusion amount is less than 20 mm.

[0043] A honeycomb sandwich panel battery pack lateral impact resistance multi-parameter collaborative optimization design method is also provided, comprising: According to the optimization results output by the parameter calculation method, the step of designing the honeycomb sandwich panel battery pack is performed.

[0044] Embodiment two, the technical solution provided by embodiment one is further described in detail, specifically: The technical solution provided by the embodiment is generally divided into six stages of honeycomb sandwich panel modeling and performance evaluation, orthogonal test screening, box finite element modeling, multi-parameter proxy modeling, optimization design and verification, and the output and input between stages are closely connected to form a complete optimization design process.

[0045] First, a honeycomb sandwich panel finite element model is established and performance evaluation is performed.

[0046] The honeycomb sandwich panel is composed of an upper panel, a honeycomb core layer, and a lower panel, and the core layer can be selected from three typical structures of hexagonal, rectangular, and corrugated. In the finite element software LS-DYNA, shell elements are used for modeling, the thickness of the honeycomb core layer is 0.05 mm, the height is 10 mm, and the panel material is aluminum alloy. The upper and lower rigid walls are used to simulate the compression boundary conditions, and the honeycomb and the panel are connected through node binding. By applying quasi-static compression to the honeycomb sandwich panel, performance evaluation parameters such as initial peak force, total energy absorption, mass-specific energy absorption, and average plateau stress are obtained. These parameters are used as input data for the next step of orthogonal test analysis.

[0047] Second, the optimal energy absorption combination of the honeycomb sandwich panel is screened based on the orthogonal test design.

[0048] By designing orthogonal test table, setting honeycomb shape, panel material and thickness three factors, each factor is set three levels. With total energy absorption, specific energy absorption, peak force and average crushing force as evaluation index, the influence degree of different factors on performance is compared by range analysis method. The test results show that the hexagonal honeycomb structure can more evenly disperse the load in the compression process, and shows a stable deformation mode; carbon fiber reinforced composite material (CFRP) panel can further improve the specific energy absorption performance due to its high specific strength and low density. The final selected optimal combination is hexagonal honeycomb structure and CFRP panel, which provides parameter basis for subsequent battery pack box modeling.

[0049] Thirdly, on the basis of the optimal energy absorption combination, the finite element model of battery pack box is established.

[0050] The battery pack box is composed of upper and lower boxes, sandwich panels and internal stiffening beams, and non-bearing parts such as lifting lugs are not involved in the calculation. Taking hexagonal honeycomb + CFRP panel as the basic sandwich panel component, setting honeycomb thickness X1, honeycomb height X2, honeycomb edge length X3, CFRP different layer thickness X4~X7, upper box thickness X8, lower box thickness X9 and stiffening beam thickness X10 as design variables. In finite element simulation, the mass M, mass specific energy absorption SEA, extrusion force F and intrusion amount D of battery pack are output by lateral extrusion of the box structure by rigid column. These indicators not only reflect the lightweight effect of the structure, but also reflect the impact resistance, which provides basic data for the construction of proxy model.

[0051] Fourthly, the proxy model is constructed to replace the complex finite element model.

[0052] In order to reduce the calculation amount caused by multiple iterations, Latin hypercube sampling method is used to select 72 sample points in the value space of design variables, and the response data is obtained by finite element simulation. Using these data to construct three common proxy models, namely radial basis function model (RBF), Kriging model and response surface model (RSM). By comparing the prediction accuracy, it is found that the fitting effect of RBF model on specific energy absorption, extrusion force and intrusion amount is the best, so RBF model is selected as the substitute model for optimization calculation. This proxy model can accurately reflect the functional relationship between design variables and performance indicators, and provide an efficient tool for subsequent optimization.

[0053] Fifthly, NSGA-II algorithm is used for multi-objective optimization.

[0054] The RBF agent model is taken as an input of optimization calculation, and the battery pack mass and intrusion amount are taken as objective functions, and the constraint condition is that the extrusion pressure is not less than 100 kN and the intrusion amount is less than a safety threshold of 20 mm. The optimization process adopts the setting of a population size of N, a crossover probability of 0.9 and a mutation probability of 0.1, and obtains a Pareto optimal solution set through non-dominated sorting and congestion degree determination. In the iteration process, the optimal solution gradually approaches the real Pareto frontier. Finally, a group of parameters are obtained, that is, the honeycomb thickness is 1.13 mm, the honeycomb height is 10 mm, the honeycomb side length is 8.24 mm, the CFRP layer thickness is 0.2 / 0.2 / 0.5 / 0.2 mm, the upper box thickness is 1 mm, the lower box thickness is 1.13 mm, and the reinforcing beam thickness is 2.14 mm.

[0055] In the sixth step, the optimal design parameters are substituted into the finite element model for verification.

[0056] The above optimal parameters are input into the battery pack finite element model to obtain the simulation results after optimization. The results show that the battery pack mass decreases from 44.86 kg before optimization to 23.68 kg, a decrease of 47.21%; the mass specific energy increases from 307.61 J / kg to 413.52 J / kg, an increase of 34.43%; the extrusion pressure decreases from 136.07 kN to 111.06 kN, a decrease of 22.52%; and the intrusion amount increases from 11.54 mm to 14.71 mm, but is still lower than the safety threshold of 20 mm. It shows that the optimization design significantly improves the lightweight effect while ensuring the safety of the structure.

[0057] The honeycomb sandwich plate battery pack lateral impact multi-parameter collaborative optimization design method of the application can effectively overcome the shortcomings of the prior art which relies on thickening or single honeycomb structure improvement, and realize the dual improvement of lightweight and safety performance of the battery pack under lateral impact working conditions.

[0058] wherein, Figure 1 It is a schematic diagram of a honeycomb sandwich plate quasi-static compression finite element model. The diagram shows that the honeycomb sandwich plate is composed of an upper panel, a honeycomb core layer and a lower panel, and rigid walls are arranged on the upper and lower sides to simulate the external compression working condition. By applying a displacement load on the upper rigid wall, the deformation mode and energy absorption characteristics of the honeycomb core layer during compression can be obtained, and the diagram directly reflects the boundary conditions and stress mode of the sample.

[0059] Figure 2 It is an energy curve diagram of a honeycomb sandwich plate. The diagram shows the energy distribution of the hexagonal, rectangular and corrugated honeycomb sandwich plates during the quasi-static compression process. The curve is smooth and the hourglass energy accounts for less than 10% of the total internal energy, indicating that the calculation accuracy of the finite element model meets the requirements. By comparing the energy absorption curves of the three types of structures, the energy absorption capacity of different honeycomb geometries can be directly evaluated.

[0060] Figure 3 The figure is a curve of evaluation index and factor. This figure visualizes the orthogonal test results, showing the influence trend of the three factors of honeycomb shape, panel material and thickness on the total energy absorption, peak force, specific energy absorption and average crushing force. Through the trend of the curve, the importance of each factor on the performance index and the optimal combination can be determined intuitively.

[0061] Figure 4 The figure is a schematic diagram of the finite element model of the battery pack. This figure shows the overall structure of the battery pack box, including the upper box, the lower box, the honeycomb sandwich panel and the internal strengthening beam. This model is used to simulate the lateral rigid column extrusion condition, and outputs key response parameters such as mass, specific energy absorption, extrusion force and intrusion amount, which is the basis for subsequent multi-parameter optimization.

[0062] Figure 5 The figure is an evaluation coefficient diagram of the approximation model. This figure shows the evaluation index R 2 and error comparison of RBF, Kriging and RSM three types of surrogate models in fitting the finite element results of the battery pack. The results show that the fitting accuracy of RBF model on mass specific energy absorption, extrusion force and intrusion amount is optimal, which is suitable as the surrogate model for subsequent optimization.

[0063] Figures 6-9 The figure is a comparison of actual value and predicted value. The four response indicators of mass, mass specific energy absorption, extrusion force and intrusion amount are compared respectively, showing the degree of agreement between the predicted value of the surrogate model and the actual value of the finite element simulation. The data points in the figure are close to the diagonal line, indicating that the RBF model can accurately represent the relationship between variables and responses.

[0064] Figure 10 The figure is a correlation coefficient diagram of variables and responses. This figure shows the correlation between design variables and performance responses in the form of matrix correlation coefficient. The positive and negative coefficients in the figure reveal the influence direction and strength of different variables on mass, specific energy absorption, extrusion force and intrusion amount, providing a reference for further determining the optimization focus.

[0065] Figures 11-14 The figure is a surface plot of response with main variable change. This figure contains four subplots, which respectively show the relationship between mass and lower box and strengthening beam thickness, the relationship between specific energy absorption and honeycomb geometric size, the relationship between extrusion force and honeycomb height and strengthening beam thickness, and the relationship between intrusion amount and honeycomb thickness and height. The surface plot intuitively reflects the interaction of design variables and its influence law on performance indicators.

[0066] Figure 15Figure 1 is an optimization flowchart based on the NSGA-II algorithm. The figure shows the main steps of the optimization, including initial population generation, non-dominated sorting, crossover mutation operation, and Pareto solution set selection process. Through this flowchart, efficient solution of the multi-objective optimization of the battery pack box can be achieved.

[0067] Figure 16 Figure 2 is an optimization solution set distribution diagram. The figure shows the process of the population gradually converging to the Pareto frontier during the iteration process. The distribution of the point set indicates that the selected optimization algorithm can effectively balance the battery pack mass and intrusion amount objectives and output multiple feasible solutions for selection.

[0068] Figure 17 Figure 3 is a comparison chart of the battery pack intrusion amount and extrusion force curves before and after optimization. The figure shows that the extrusion force peak of the optimized battery pack is reduced by 22.52% compared to before optimization, and while the intrusion amount increases when reaching the 100 kN load requirement, it is still less than the 20 mm safety threshold, verifying the rationality of the optimized design.

[0069] Figure 18 Figure 4 is a comparison chart of the battery pack mass-specific energy absorption curves before and after optimization. The figure shows that the mass-specific energy absorption of the optimized honeycomb sandwich panel box is significantly higher than that of the pre-optimization design during the lateral impact process, indicating that the optimization result can enhance the energy absorption performance while achieving lightweight.

[0070] Embodiment Three, the above-mentioned technical solutions are further described in detail through specific embodiments, specifically: By adopting a honeycomb sandwich panel to replace the traditional battery pack aluminum alloy box frame, a multi-component parameterized model is established, including the honeycomb core layer, the panel, the upper and lower box, and the reinforcing beam. Based on LS-DYNA explicit dynamics simulation, the evolution law of the battery pack's mass, mass-specific energy absorption, extrusion force, and intrusion amount under lateral impact conditions is analyzed, and the safety of the battery pack is evaluated accordingly, and then the optimization focus is determined.

[0071] 1 Modeling method and evaluation of honeycomb sandwich panel structure 1.1 Establishment of sandwich panel structure model The honeycomb sandwich panel is composed of an upper panel, a honeycomb core layer, and a lower panel. In the ANSYS / LSDYNA software, the honeycomb core layer structure is established as a finite element model of a regular hexagon, a rectangle, and a corrugated shape, respectively, as shown in Figure 1The two rigid walls are arranged on the upper and lower sides of the honeycomb, respectively. The keyword "RIGIDWALL" is used to apply a downward speed of 0.5 m / s to the upper rigid wall to simulate the quasi-static compression process. The "SINGLE_SURFACE" is used for the contact between the honeycombs. The "TIED_NODE_TO_SURFACE" is used for the contact between the honeycomb and the upper and lower panels. The dynamic friction coefficient is 0.2, the static friction coefficient is 0.15, the material constitutive relation uses the ideal elastoplastic model MAT03_PLASTIC_KINEMATIC, and the shell element is used. In order to obtain the stable mechanical parameters of the honeycomb structure, the in-plane size of the honeycomb sample should be at least N×N = 9×9

[15] . The length of the sandwich panel sample is 100 mm×100 mm, the number of honeycomb core layers is N×N = 13×15, the single honeycomb wall thickness is t = 0.05 mm, the height is 10 mm, and the materials of the honeycomb core layer and the upper and lower panels are AL5052. The Young's modulus E = 69 GPa, the density ρs = 2.73 g·cm-3, the Poisson's ratio v = 0.33, the static yield strength O = 123 MPa, and the ETAN = 50 MPa.

[0072] 1.2 Impact resistance and energy absorption evaluation index When the honeycomb sandwich panel is subjected to external impact, the impact force does work on the sandwich panel. In this process, the honeycomb panel and the honeycomb core layer deform to absorb energy and slow down the impact. In this analysis, since the impact speed is not high and the honeycomb core layer hole wall is smooth, the energy consumed and broken due to thermal effects and friction is not considered. The work done by the external impact on the sandwich panel is entirely converted into the internal energy generated by the honeycomb core layer and the upper and lower panels. The evaluation index of the energy absorption of the honeycomb sandwich panel structure mainly includes the initial peak force (F), the total energy absorption (EA), the specific energy absorption (SEA), and the average platform stress (σ).

[0073] The initial peak force (F) is the maximum resistance generated by the structure in the initial deformation stage during the axial crushing process of the honeycomb material. The total energy absorption reflects the total energy absorbed by the honeycomb during the extrusion process, which can be obtained from the force-displacement integral curve as follows: (1) In the formula, d is the crushing displacement. F ( x ) represents the size of the crushing force of the structure with the crushing displacement.

[0074] The specific energy absorption reflects the characteristics of the material during the energy absorption process under the action of impact. The calculation formula is as follows: (2) In the formula, M is the total mass of the honeycomb sandwich panel.

[0075] The average platform stress is an important indicator to measure the stable progressive buckling of honeycomb sandwich panel, and its formula is: (3) In the formula, F is the average compression force in the interval of 20% to 70% of the compression displacement, and A is the panel cross-sectional area 1.3 Finite element result evaluation In the finite element calculation, the high-efficiency and time-saving Gaussian element single-point integration calculation method is generally used, but this method may have an hourglass mode, i.e. zero strain and zero stress are generated in the calculation process, and after interpolation calculation, the internal energy is zero, which leads to energy non-conservation in the calculation process, affects the calculation accuracy and even calculation error, so it is necessary to judge the hourglass energy change in the calculation process, and the judgment standard is whether the hourglass energy generated by the hourglass mode can exceed 10% of the total internal energy

[16] . The output energy change curves of the three honeycomb sandwich panel structures in the calculation are shown in Figure 2 , the hourglass energy generated by the hexagonal, rectangular and corrugated shapes accounts for 6.28%, 5.52% and 2.40% of the total internal energy respectively, and the internal energy curves change smoothly without mutation, and the model meets certain accuracy requirements.

[0076] 2 Orthogonal test 2.1 Orthogonal test design In the exploration of the energy absorption of honeycomb sandwich panel structure, different honeycomb shapes affect the deformation mode and failure mechanism under impact, different panel materials affect the peak force at the initial contact and the lightweight degree of the structure, and different thicknesses affect the overall stiffness and energy absorption rate of the structure, so three factors are set in this paper, i.e. honeycomb shape, panel material type and thickness, each factor has 3 levels, and the test scheme is L9(3 4 ), and the specific factors and levels are shown in Table 1, in which the steel Q235 has a Young's modulus E=210 GPa, a density p s =7.85 g.cm -3 , and a Poisson's ratio v=0.3, and the mechanical property parameters of the carbon fiber reinforced composite material (CFRP) used are shown in Table 2.

[0077]

[0078] For the performance index evaluation of the sandwich panel, the total absorbed energy in the compression process is used to evaluate the total energy absorption, the peak force is used to reflect the maximum load strength in the initial compression stage, the mass-specific energy absorption is used to reflect the lightweight degree, and the average compression force is used to characterize the load stability in the whole deformation process to prevent the structure from collapsing and failing instantaneously. The specific evaluation results are shown in Table 3.

[0079]

[0080] 2.2 Range analysis According to the test results in Table 3, the total energy absorption in the compression of the sandwich panel is analyzed by using the range analysis method, and the range table obtained is shown in Table 4. Among them i The average value of the results corresponding to the horizontal is k i (1, 2, 3), and the range R is determined by the difference between the maximum and minimum average values in the same factor. The greater the range, the greater the influence of the factor on the evaluation index, and the smaller the range, the lighter the influence of the factor on the evaluation index.

[0081]

[0082] In order to more intuitively analyze the significance, the changes of each evaluation index with the levels of the factors are as shown in Figure 3 .

[0083] The primary and secondary relationships of the three factors on the four indexes are as follows: in the analysis of total energy absorption, the three factors affecting from primary to secondary are C, A and B according to the range R. For ki, the maximum total energy absorption is C3, the thicker the honeycomb sandwich thickness, the larger the deformation area, and the more the energy absorption. The maximum total energy absorption of factor A is A1 hexagonal structure. The hexagonal structure has high symmetry and uniform distribution of wall surface, and can more evenly disperse the load and deform more stably in the crushing process, thereby absorbing energy. The maximum total energy absorption of factor B is B2, and the steel has high yield strength and tensile strength, which can withstand greater bending stress and can more effectively transmit to the honeycomb core to cause more sufficient plastic deformation. Therefore, the best total energy absorption combination is C3A1B2, and by analogy, the factors from primary to secondary for peak force are C, A and B, and the corresponding best levels are C3A3B3. The factors from primary to secondary for specific energy absorption are B, A and C, and the corresponding best levels are B1A1C2. The factors from primary to secondary for average crushing force are C, B and A, and the corresponding best levels are C3B3A2. For the index evaluation of the honeycomb sandwich panel structure, specific energy absorption > peak force > total energy absorption > crushing force, and thus the hexagonal core layer and carbon fiber panel are the best combination.

[0084] 3 Impact model design of battery pack box of honeycomb sandwich panel structure 3.1 Parameter definition In the actual operation of electric vehicles, the battery pack box as the key load-bearing structure of the power battery system will inevitably be subjected to mechanical loads from multiple directions. Especially in complex road conditions or sudden collision events, the box structure may bear stress exceeding the design threshold, resulting in material yield, structural deformation, or even local buckling failure.

[0085] The side impact condition mainly applies force load or speed to the rigid column to extrude the battery pack, and judges whether the intrusion of the battery pack box is less than the safety threshold of the box and the battery module. The safety distance is 20 mm. After determining the factor level, the battery pack is modeled, and its finite element model is as follows Figure 4 In the calculation, the lug is mainly used to fix the battery pack to the vehicle body, so it is not considered as a force-bearing part. According to the analysis in the foregoing, the panel material is selected as CFRP, and the sandwich structure is hexagonal. The thickness of the hexagon, the cross-sectional height X2, the cross-sectional length X3, the 0° carbon fiber layer thickness X4, the 45° layer thickness X5, the -45° layer thickness X6, the 90° layer thickness X7, the upper box thickness X8, the lower box thickness X9, and the stiffener thickness X10 are determined as design variables. M represents the total mass of the upper box, the lower box, the stiffener, and the sandwich plate box (battery pack box), SEA is the specific energy absorption during loading, F is the extrusion force, and D is the intrusion.

[0086] 3.2 Construction of surrogate model The number of finite element analysis units of a complex model is large, and the running time is long. In the optimization design, each adjustment needs to be recalculated, and the time input increases. The approximate model method can use a substitute model to approximate the real situation, simplify the model optimization process, speed up the optimization iteration, and reduce the calculation time cost.

[0087] 3.2.1 Sampling point selection In the finite element analysis process, in order to quantify the relationship between the input variables and the output response, DOE experimental design is needed to obtain key information with fewer test times. In this study, the Latin hypercube design method is used, which ensures good space filling of sample points in multi-dimensional parameter space, and can cover all variables, thereby reducing errors. The total number of tests is determined to be 72 times, and each is simulated and solved. The specific test points and solution results are shown in Table 5.

[0088]

[0089] 3.2.2 Model fitting and accuracy evaluation Currently, the commonly used surrogate model methods are Krigring, RBF, and RSM. The relationship between the design variables and the fitted response is obtained by using these three models. Based on the 72 sample points obtained by Latin cube sampling, 20 sample points are extracted, and the fitting accuracy of the three surrogate models is compared. The evaluation index is basically as follows: (4) (5) In the formula:y k for the simulation actual value; for the simulation predicted value; for the average of the simulation actual value; for the evaluation index R 2 and e RMS , the values of which respectively tend to 1 and 0, indicating that the fitting accuracy of the corresponding surrogate model is better. Among them, it is required that R 2 > 0.9 to preliminarily adopt the model. Based on 20 sample points, the R 2 and e RMS as shown in Table 6.

[0090]

[0091] In the model selection process, the R 2 is the main selection index, as shown in Table 7. Figure 5 For the quality, the fitting accuracy of RSM is greater than that of RBF and Krigring, and for the quality ratio of absorbed energy, extrusion force and intrusion amount, the fitting accuracy of RBF is greater than that of RSM and Krigring. The difference in fitting accuracy of RBF and RSM for the quality is 0.007, and both are greater than 0.998, while the fitting of RBF for the quality ratio of absorbed energy, extrusion force and intrusion amount is the best.

[0092] In summary, the RBF surrogate model is used for fitting, and the model is used to replace the actual battery pack box under lateral impact. The function relationship between the variables and the response is obtained, and the distribution of the response values of the 20 sample points under the model is as shown in Figures 6-9 , and RBF will be used for subsequent multi-objective optimization design.

[0093] 3.2.3 Parameter relationship analysis To fully study the influence relationship between design variables and response values, the correlation coefficient Figure 10 is drawn to clarify the mutual relationship between the design variables and the correlation between them and the response values. The positive value is the positive correlation, and the negative value is the negative correlation. It can be seen that the mutual relationship between the design variables is weak as a whole, while the correlation between the variables and the response is close. Specifically, the correlation between the mass and the thickness of the reinforcement beam, the lower box and the upper box is larger; the correlation between the quality ratio of absorbed energy and the height and side length of the honeycomb core layer is larger; the correlation between the extrusion force and the thickness of the reinforcement beam, the thickness of the lower box and the height of the honeycomb is larger; the correlation between the intrusion amount and the height of the honeycomb, the thickness of the honeycomb and the thickness of the reinforcement beam is larger.

[0094] As shown in Table 8. Figures 11-14To further determine the variation of the four responses with the important design variables, a 3D surface as shown in Fig. 4 is established. Figures 11-14 As seen from a, the mass is positively correlated with the thickness of the lower box body and the thickness of the reinforcing beam, that is, the thickness increases, the mass also increases, and the mass increase caused by the increase of one unit of the thickness of the reinforcing beam is greater than the mass increase caused by the same increase of the thickness of the lower box body. As seen from b, the mass-specific energy is negatively correlated with the height and the side length of the honeycomb, and the increase of the size will lead to the decrease of the mass-specific energy, and the influence trend is consistent. As seen from c, the extrusion force decreases with the increase of the height of the honeycomb, but significantly increases with the increase of the thickness of the reinforcing beam, and the influence of the thickness of the reinforcing beam on the extrusion force is more significant. As seen from d, the intrusion amount decreases with the increase of the height of the honeycomb, and the rate of decrease presents a change from slow to fast; at the same time, the intrusion amount also decreases with the increase of the thickness of the honeycomb.

[0095] 4. Multi-parameter multi-objective optimization using genetic algorithm 4.1 Optimization model Through the analysis of the battery pack under the side impact working condition, it is known that the parameters of the upper and lower box bodies and the honeycomb are of great significance to further improve the side impact resistance of the battery pack box body while reducing the mass. When designing the same, it is necessary to ensure that the extrusion force reaches 100 kN, the intrusion amount at the same time is less than the distance between the box body and the battery, and the light weight is also ensured. The thicker the parameter thickness is, the smaller the intrusion amount will be, but the mass will increase, and the two are restricted, so it is necessary to perform multi-objective optimization design.

[0096] The multi-objective optimization mathematical model of the embodiment is as follows

[0097] (6)

[0098] In the formula, and are the intrusion amount objective function and the mass objective function of the battery pack box body respectively, D x ) and M x are the intrusion amount objective function and the mass objective function of the battery pack box body respectively, X L is the minimum range value X U is the maximum range value, SEA is the mass-specific energy, F is the extrusion force.

[0099] 4.2 Optimization process ​​The NSGA-Ⅱ algorithm is used to optimize the battery pack box. The NSGA-Ⅱ algorithm is the most representative algorithm based on the Pareto dominance rule. To avoid the non-dominated solution distribution too concentrated to affect the algorithm performance, the non-dominated solution density is considered, and the crowding strategy is adopted, which makes the algorithm have great advantages in maintaining the diversity of non-dominated solutions and preventing local optimal solutions. First, the initial population is randomly generated by the system, and after non-dominated sorting, selection, crossover and mutation, the initial population is generated. The population size is N. Then the parent population and the child population are combined to form a new mating pool with a population size of 2N. After fast non-dominated sorting in this population, the appropriate individuals are selected into the next generation population according to the crowding degree, and the population size is N. At this time, the population evolution is completed, and the specific process is shown in Figure 15 .

[0100] 4.3 Optimization results and verification The NSGA-Ⅱ non-dominated sorting genetic algorithm is used to optimize the objective function. The population size is 25, the crossover probability is 0.9, the crossover distribution index is 10, and the mutation distribution index is 20. The optimized solution set is shown in Figure 16 .

[0101] Figure 16 The feasible solution is constantly approaching the Pareto solution in the iteration process, and the specified optimal solution is obtained by calculation. The corresponding parameters are honeycomb thickness 1.13 mm, height 10 mm, side length 8.24 mm, 0° layer 0.2 mm, 45° layer 0.2, -45° layer 0.5 mm, 90° layer 0.2 mm, upper box thickness 1 mm, lower box thickness 1.13 mm, and reinforcement beam thickness 2.14 mm.

[0102] 4.3.1 Comparative analysis of optimization results The optimized results are shown in Table 7. Compared with the optimization before, the mass of the battery pack box after optimization is reduced by 47.21%, the specific energy absorption is increased by 34.43%, and the intrusion amount is increased by 27.47%. The parameters are substituted into the model for simulation verification. The maximum error between the optimization fitting results and the simulation results is 4.86%. Within the error range, it is proved that the optimal design variables are reliable.

[0103]

[0104] The intrusion amount and extrusion displacement curve before and after optimization are shown in Figure 17As shown, the extrusion peak force before optimization is 136.07 kN, and after optimization is 111.06 kN, which is reduced by 22.52%, meeting the design bearing requirement of 100 kN of the battery pack box, and by comparing the intrusion amount when reaching 100 kN, it is found that the intrusion amount before optimization is less than that after optimization, but still lower than the safety limit of 20 mm, because the aluminum alloy box frame before optimization is isotropic material mainly relying on material bending stiffness to resist deformation, while the honeycomb sandwich panel presents out-of-plane compression-bending coupling deformation under lateral impact, and the equivalent elastic modulus after optimization is lower than that before optimization, so the intrusion amount is larger while bearing the same extrusion force.

[0105] The mass specific energy absorption before and after optimization changes with time as shown in Figure 18 As shown, the sandwich panel box after optimization has partial buckling of the honeycomb when subjected to lateral impact, absorbs more energy, and the mass after optimization is smaller, so in the subsequent loading process, the mass specific energy absorption is more.

[0106] 5 Conclusion The present embodiment is based on a multi-parameter modeling method, taking the box of a certain battery pack as the research object, replacing the frame type aluminum alloy box with a honeycomb sandwich panel box, in order to further reduce the weight while improving its side impact resistance, taking the mass and intrusion amount of the battery pack box as the target, through simulation calculation under side impact working condition, and using NSGA-II algorithm for optimization analysis, the following conclusions are obtained: 1) The honeycomb sandwich panel box composed of hexagonal honeycomb structure and CFRP panel has the best energy absorption effect, has excellent mechanical properties when responding to side impact, and has lighter mass and higher specific energy absorption than the traditional aluminum alloy battery pack box frame.

[0107] 2) In order to enhance the side impact resistance of the battery pack, the thickness of the lower box and the stiffener can be increased to improve the stiffness, and the edge box can also be enhanced by increasing the thickness or adding stiffeners to enhance the bending capacity.

[0108] 3) After optimization by algorithm, the mass of the battery pack box is reduced by 47.21%, and the intrusion amount is increased by 27.47%, but still within the safety threshold. Compared with the actual calculation, the relative error interval is 1.22%~4.86%, which can be accepted. This design can achieve the purpose of lightweight under the premise of ensuring the lateral impact performance, which has positive significance for improving the endurance mileage of electric vehicles.

[0109] The technical solutions of the present application are described in further detail through several specific embodiments above, in order to highlight the advantages and benefits of the technical solutions provided by the present application. However, the above several specific embodiments are not used as a limitation to the present application, and any reasonable modifications and improvements, combinations and equivalent replacements, etc. of the present application within the scope of the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A method for calculating multi-parameter collaborative optimization design parameters for lateral impact resistance of a honeycomb sandwich panel battery pack, characterized in that, include: Establishing a finite element model of a honeycomb sandwich panel and obtaining performance evaluation parameters are steps that provide basic data for structural screening. Based on performance evaluation parameters, the energy absorption combination of the honeycomb sandwich panel that meets the preset conditions is obtained, and this combination is used as the input condition for modeling the battery pack box structure. Based on the energy-absorbing combination of the honeycomb sandwich panel that meets the preset conditions, the steps are as follows: establish a finite element model of the battery pack box and preset design variables, and output structural performance indicators. The steps to construct a proxy model and establish a functional relationship between design variables and response based on structural performance indicators; The steps involve using a surrogate model as input, employing the NSGA-II algorithm for multi-objective optimization, and outputting design parameters. The steps involve substituting the design parameters into the finite element model of the battery pack for verification and outputting the optimization results.

2. The method for calculating multi-parameter collaborative optimization design parameters for lateral impact resistance of a honeycomb sandwich panel battery pack according to claim 1, characterized in that, The performance evaluation parameters include initial peak force, total absorbed energy, energy absorbed by mass ratio, and average plateau stress.

3. The method for calculating multi-parameter collaborative optimization design parameters for lateral impact resistance of a honeycomb sandwich panel battery pack according to claim 1, characterized in that, The energy-absorbing combination of the honeycomb sandwich panel that meets the preset conditions is a combination of a hexagonal honeycomb structure and a carbon fiber reinforced composite material panel.

4. The method for calculating multi-parameter collaborative optimization design parameters for lateral impact resistance of a honeycomb sandwich panel battery pack according to claim 1, characterized in that, The objective function of the multi-objective optimization includes battery pack mass and intrusion amount, with constraints that the extrusion pressure reaches 100kN and the intrusion amount is less than 20mm.

5. A multi-parameter collaborative optimization design parameter calculation device for lateral impact resistance of a honeycomb sandwich panel battery pack, characterized in that, include: A finite element model of a honeycomb sandwich panel is established and performance evaluation parameters are obtained, serving as a module to provide basic data for structural screening. Based on performance evaluation parameters, an energy-absorbing combination of honeycomb sandwich panels that meets preset conditions is obtained, and this combination is used as a module for input conditions in the battery pack box structure modeling. Based on the energy-absorbing combination of the honeycomb sandwich panel that meets the preset conditions, a finite element model of the battery pack box is established and the design variables are preset, and a module that outputs structural performance indicators is generated. Based on structural performance indicators, a module is constructed to build a proxy model and establish the functional relationship between design variables and response; A module that uses a surrogate model as input, employs the NSGA-II algorithm for multi-objective optimization, and outputs design parameters; This module verifies the design parameters by substituting them into the finite element model of the battery pack and outputs the optimization results.

6. A multi-parameter collaborative optimization design method for lateral impact resistance of a honeycomb sandwich panel battery pack, characterized in that, include: The steps for designing a honeycomb sandwich panel battery pack are based on the optimization results output by the parameter calculation method described in claim 1.

7. A multi-parameter collaborative optimization design device for lateral impact resistance of a honeycomb sandwich panel battery pack, characterized in that, include: The module for designing a honeycomb sandwich panel battery pack is based on the optimization results output by the parameter calculation method described in claim 1.

8. A computer storage medium for storing computer programs, characterized in that, When the computer program is read by the computer, the computer executes the method of claim 1.

9. A computer, comprising a processor and a storage medium, characterized in that, When the processor reads the computer program stored in the storage medium, the computer executes the method of claim 1.

10. A computer program product, as a computer program, is characterized by: When the computer program is executed, it implements the method of claim 1.

Images