A composite lightweight design and optimization method for automobile floor

Abstract

A composite lightweight design and optimization method for automobile floor, the present application determines the shape of the layer block, the thickness of the layer, the thickness of the layer block, the number of layers of the layer, the sequence of the layer by free size optimization, size optimization, layer sequence optimization under the premise of complying with the manufacturing process; In the layer sequence optimization, the continuous variable discretization rounding strategy is adopted to handle the layer number of the downward rounding layer block, to ensure the floor bearing performance, the continuous fiber layering strategy is adopted to reduce the layer loss phenomenon, and the interlayer strategy is adopted to reasonably distribute the layer sequence; Through multi-objective optimization, other parts of the body-in-white are optimized and designed, and the maximum advantage of the composite floor is played, wherein the entropy weight grey correlation degree analysis method can objectively find the optimal scheme in many schemes of multi-objective optimization, the present application realizes the automobile floor material-structure-process integrated collaborative optimization design and the like while ensuring the performance of the floor and realizing the lightweight design of the floor.

Description

Technical Field

[0001] This invention relates to the field of automotive lightweighting, and more specifically to a lightweight design and optimization method for composite materials used in automotive flooring. Background Technology

[0002] Safety, energy conservation, and environmental protection are the three major themes of modern automotive engineering technology. With the rapid growth of car ownership, a series of social problems such as excessive energy consumption and environmental pollution have emerged. Energy conservation is typically achieved through two approaches: developing engine technology to improve thermal efficiency and developing lightweight technologies. Given the increasing difficulty of improving engine technology, lightweight technology has become the most important way to solve the two major problems of energy conservation and environmental protection. Fuel consumption increases with the increase in vehicle weight; lightweighting can effectively alleviate energy consumption and environmental pollution problems. As one of the three major assemblies of a car, the body plays a crucial role in bearing loads, protecting occupant safety, creating a good aerodynamic environment, and providing cushioning, shock absorption, and vibration reduction. The body accounts for 30% to 40% of the total vehicle weight, and lightweighting the body plays a vital role in the overall vehicle lightweighting. Lightweighting lowers the vehicle's center of gravity, enhancing stability and comfort during driving; it reduces various resistances during driving, improving vehicle power; and it also reduces initial kinetic energy, shortens braking distance, and improves vehicle safety performance. Currently, commonly used lightweight optimization design methods mainly involve three aspects: the application of high-strength lightweight materials, the optimization of vehicle body structure, and the application of new process technologies.

[0003] High-strength, lightweight materials commonly used in lightweight design include high-strength steel, aluminum alloys, magnesium alloys, and carbon fiber composites. Among these, carbon fiber composites are structural materials made by combining carbon fibers with matrices such as resins, metals, and ceramics. Due to their low specific gravity, high rigidity, high strength, high modulus, corrosion resistance, fatigue resistance, good damping resistance, and designable properties, carbon fiber composites are widely used in transportation, medicine, chemical machinery, and sports. Because of their significant lightweighting effect, carbon fiber composites are increasingly being used in the automotive industry.

[0004] As a crucial load-bearing structural component of the vehicle's floor, the automotive floor plays a vital role in protecting the safety of the passenger compartment. Therefore, lightweight design of the floor has significant engineering value. Applying carbon fiber composite materials to automotive floor design can increase vehicle body rigidity and strength, improve overall vehicle crashworthiness, and absorb more energy in frontal and side collisions, thus protecting the passenger compartment. Furthermore, carbon fiber composite floor layers can protect the battery pack, preventing it from catching fire or exploding in a collision, thereby reducing secondary injuries to occupants.

[0005] Vehicle body structure optimization is categorized into topology optimization, size optimization, shape optimization, and morphology optimization based on the type of design variables. Traditional structural optimization methods primarily involve small, localized modifications to individual vehicle components. These components are discrete and unrelated, failing to optimize the overall structure. Furthermore, vehicle body structure optimization involves multiple disciplines such as mechanics, materials science, NVH (noise, vibration, and harshness), and structural safety. These disciplines are interconnected, making it difficult for traditional single-objective optimization methods to achieve a balance between various performance aspects, resulting in limited optimization effects. Multi-objective optimization can solve optimization problems with contradictory sub-objectives. However, multi-objective optimization is prone to problems such as multivariate complexity and combinatorial explosion when dealing with large-scale optimization problems with high-dimensional decision variables. Traditional multi-objective optimization methods often employ problem reconstruction to transform high-dimensional decision variable problems into low-dimensional optimization problems before solving them using multi-objective optimization algorithms. This method is prone to getting trapped in local optima and fails to reflect the actual situation of the problem.

[0006] Traditional optimization algorithms are generally suitable for linear and simple systems, but they exhibit limitations when applied to multi-objective optimization problems. Solving multi-objective optimization problems is both time-consuming and difficult, and the global or near-global optimal solution may be overlooked within the feasible region. For example, exhaustive search algorithms can be applied to simple structures with a small number and range of adjustable variables, but as the number and range of adjustable variables increase, the entire search space may become excessively large, making it difficult to find the optimal solution. Intelligent optimization algorithms have advantages over traditional optimization algorithms in solving multi-objective optimization problems, as they are better able to find the global optimum and avoid local optima, but they suffer from problems such as high time consumption and premature convergence. For example, the Bacterial Foraging Optimization (BFO) algorithm is a relatively new intelligent optimization algorithm that finds the global optimum through trending operations, effectively avoiding local optima, but its convergence time is long and the search step size is difficult to determine. Hybrid optimization algorithms combine the advantages of traditional and intelligent optimization algorithms, resulting in faster solution speeds and higher reliability of global optima when dealing with multi-objective optimization problems.

[0007] Therefore, how to provide a lightweight design and optimization method for composite materials used in automotive flooring has become a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0008] To overcome the shortcomings of the prior art, this invention provides a lightweight design and optimization method for composite materials used in automotive flooring. This invention comprehensively considers the balance between materials, structure, and process when optimizing automotive flooring, proposing a composite material flooring layup design and optimization method that integrates conceptual design and multi-objective optimization design. This achieves integrated collaborative optimization design of automotive flooring materials, structure, and process, ensuring vehicle performance while completing the lightweight design of the flooring.

[0009] To achieve the aforementioned inventive objectives, the present invention employs the following technical solution:

[0010] A lightweight design and optimization method for composite materials used in automotive flooring, the design and optimization method specifically includes the following steps:

[0011] The first step is to establish a parametric model of the body-in-white using implicit parametric modeling methods;

[0012] The second step is to establish a finite element model of the body-in-white based on the parametric model of the body-in-white, and verify the accuracy of the finite element model of the body-in-white through bending, torsion and modal analysis.

[0013] The third step is to establish a finite element model of the electric vehicle and, based on the vehicle crash test methods and modeling process in the C-NCAP Management Rules, establish frontal and side impact models of the vehicle. The reliability of the finite element model is verified by the energy changes and vehicle deformation modes during the collision, and the correctness of the model is verified by the safety performance indicators of acceleration and intrusion.

[0014] The fourth step is to select composite materials, including fiber-reinforced materials and matrix materials, and use the Tsai-Wu strength theory as the failure criterion for composite laminates. The mechanical properties of the composite materials are then tested to obtain the mechanical property parameters of the composite materials.

[0015] Step 5: Using the selected composite material, the front, middle, and rear floors of the car are designed as a whole. The designed composite floor first undergoes free-size optimization to obtain the composite ply thickness and ply block shape. Then, the shape of each ply block is cut to ensure that the ply blocks are more regular and easier to process. Finally, the initial number of layers for each ply block is determined by size optimization. The number of layers for each ply block is rounded down to obtain the initial number of layers for each ply block. Through the above optimization, the initial model of the composite floor ply is obtained, including the continuous ply thickness results of the front, middle, and rear blocks.

[0016] Step 6: Optimize the layup sequence of the composite flooring based on engineering and manufacturing constraints, specifically including the following steps:

[0017] (1) Use the continuous variable discretization rounding strategy to process the initial ply number after downward rounding, while ensuring that the in-plane stiffness of the front, middle and rear blocks is not reduced, so as to obtain the specific ply number of the front, middle and rear blocks. On this basis, ply with corresponding angles is added to ensure balanced symmetry constraints.

[0018] (2) Using the continuous fiber layup strategy, the distribution results of global shared layup, subdomain shared layup and independent layup are obtained;

[0019] (3) The improved swarm intelligence optimization (GA-VNS) algorithm is used to optimize the piling order of the global shared piling;

[0020] (4) Using the optimized global shared ply as the initial ply base, insert the subdomain shared ply and independent ply into the global shared ply according to the interpolation strategy to form a ply sequence scheme.

[0021] (5) Optimize the mathematical model based on the ply sequence, determine whether the scheme is feasible, and output the optimal ply sequence scheme and the corresponding target value;

[0022] Step 7: Compare the floor performance before and after optimization to determine whether the various performance characteristics of the body-in-white have been improved after the floor layering sequence optimization.

[0023] Step 8: To improve the overall performance of the body-in-white and maximize the advantages of the composite material floor, a multi-objective optimization and lightweight design of the body-in-white is required, which includes the following steps:

[0024] (1) Select appropriate materials for each component on the white car body according to its function;

[0025] (2) Based on the implicit parameterized body-in-white model, find the design variables of each component on the body-in-white and collect them into a design variable library;

[0026] (3) Based on the importance of the components during the collision, select the key design variables from the design variable library;

[0027] (4) Based on the performance analysis results of the body-in-white, set the objective function and constraints for the multi-objective optimization problem of lightweight body-in-white;

[0028] (5) Construct an approximate model based on the mathematical model of the multi-objective optimization problem of lightweight body-in-white;

[0029] (6) Using the approximate model as the optimization object, the body-in-white is designed with multiple objectives by the MNSGA-II algorithm to obtain the Pareto solution set, where each solution represents an optimization scheme.

[0030] (7) Use the entropy weight grey relational analysis method to determine the optimal solution in the Pareto solution set, and use it as the best optimization scheme for the lightweight design of the body-in-white in multi-objective optimization.

[0031] Step 9: By comparing the performance before and after multi-objective optimization, verify whether the multi-objective optimization scheme is feasible, and further verify the feasibility of the lightweight design and optimization method for automotive composite floor.

[0032] The aforementioned lightweight design and optimization method for composite materials used in automotive floors, in the second step, involves constraining the translational degrees of freedom in the XY axis direction of the front suspension assembly center of the body-in-white, and constraining the translational degrees of freedom in the XYZ axis direction of the rear suspension assembly center of the body-in-white; the load is uniformly applied to the R points of the four front and rear passenger seats, and the Z-direction displacements at various locations on the front longitudinal beam, sill beam, and rear longitudinal beam of the body-in-white are measured. The static bending stiffness of the body-in-white is then calculated using the following formula:

[0033]

[0034] In the formula, K b D is the static bending stiffness; F is the resultant load; Lmax D represents the maximum Z-axis displacement of the measuring point on the left. Rmax This represents the maximum Z-axis displacement of the measuring point on the right.

[0035] The aforementioned lightweight design and optimization method for composite materials used in automotive flooring, in the second step, involves constraining the translational degree of freedom along the Y-axis at the midpoint of the front crossbeam of the body-in-white, and constraining the translational and rotational degrees of freedom along the XYZ axes at the assembly center of the rear suspension of the body-in-white. A torque is applied to the rotation centers of the two front suspensions of the body-in-white, equivalent to applying forces in opposite directions to the two front suspension assembly centers. The Z-direction displacement at the same position as the bending stiffness is measured, and the static torsional stiffness of the body-in-white is calculated using the following formula:

[0036]

[0037] In the formula, K t M is the torsional stiffness; θ is the applied torque; D is the torsional angle; L D represents the Z-direction displacement of the left loading point; R L represents the Z-direction displacement of the right loading point; L is the distance between the left and right front overhang loading points.

[0038] The lightweight design and optimization method for composite materials used in automotive flooring, in the second step, modal analysis is performed without applying any load to the body-in-white or constraining any of the six degrees of freedom; only the modal characteristics of the body-in-white in its free state are analyzed.

[0039] In the aforementioned lightweight design and optimization method for composite materials used in automotive flooring, the acceleration measurement point in the third step of the frontal collision performance index is the bottom of the left and right B-pillars; the intrusion measurement points are the footrest of the front bulkhead, the center of the steering column, and the four positions of the upper left, lower left, upper right, and lower right of the door frame.

[0040] In the aforementioned lightweight design and optimization method for composite materials used in automotive flooring, the third step of the side impact performance index includes acceleration measurement points at four locations on the B-pillar corresponding to a person's head, chest, abdomen, and H-point, as well as the center of the battery box; intrusion measurement points are at four locations on the B-pillar corresponding to a person's head, chest, abdomen, and H-point, as well as the center of the battery box and the sill beam closest to the battery box.

[0041] The lightweight design and optimization method for composite materials used in automotive flooring, in the fourth step, uses the Tsai-Wu strength theory as follows:

[0042] The failure criteria are mainly based on the allowable stress and allowable strain of composite materials. The failure index of the material is obtained by calculating the ply and matrix. When the failure index is less than "1", the stress or strain is within the allowable range; while when the failure index is greater than "1", the stress or strain exceeds the allowable range.

[0043] Based on a synthesis of multiple intensity criteria, Tsai-Wu proposed the energy polynomial intensity criterion:

[0044] F i σ i +F ij σ i σ j +F ijk σ i σ j F ij σ i σ k +LL=1

[0045] In engineering applications, only the first two items are usually considered:

[0046] F i σ i +F ij σ i σ j =1(i,j=1,2,3,4,5,6)

[0047] In the formula: F i These are the strength parameters of the material;

[0048] For a two-dimensional plane stress problem, the above equation simplifies to:

[0049]

[0050] The strength parameters in the formula are as follows:

[0051]

[0052] Among them, X t Longitudinal tensile strength; X cY represents longitudinal compressive strength. t Y represents the transverse tensile strength. c S is the transverse compressive strength; S is the plane shear strength.

[0053] The above failure criterion states that the material will fail if the maximum stress or strain of the material or a certain ply exceeds the allowable value of the material.

[0054] The lightweight design and optimization method for composite materials used in automotive flooring includes, in the fourth step, composite material mechanical property testing, which includes uniaxial tensile tests at 0° and 90° layup, uniaxial compression tests at 0° and 90° layup, and in-lay shear tests at ±45° layup.

[0055] The lightweight design and optimization method for composite materials used in automotive flooring, in the fifth step, free dimension optimization uses 0°, 90°, and ±45° as ply angles. To simplify the model, ply blocks with the same angle are grouped together, with the ply thickness at each angle as the design variable and the lightweight coefficient of the body-in-white as the optimization objective. The formula for calculating the lightweight coefficient of the body-in-white is as follows:

[0056]

[0057] In the formula, L is the lightweighting coefficient; M is the body-in-white mass (kg); K t Let A be the torsional stiffness of the body-in-white, N*m / (deg); and let A be the projected footprint area of ​​the product of the wheelbase and track width in the Z direction, m. 2 ;

[0058] Using the original body-in-white bending stiffness, first-order bending frequency, and first-order torsional frequency as performance constraints, and ±45° ply balance and symmetry, each of the four angle plies accounting for no less than 10%, and the ply shape and thickness being symmetrical about the neutral plane as manufacturing constraints, a free-size optimization mathematical model is constructed as follows:

[0059]

[0060] Where: L is the lightweight coefficient of the body-in-white; BS T BF T and TF T These represent the bending stiffness, first-order bending frequency, and first-order torsional frequency of the body-in-white, respectively; BS0, BF0, and TF0 are initial values; C2, C4, and C5 are manufacturing constraints.

[0061] The lightweight design and optimization method for composite materials used in automotive flooring, in the fifth step, uses the thickness of the ply block as the design variable for dimensional optimization, the floor mass as the optimization objective, and the bending stiffness, torsional stiffness, first-order bending frequency, and first-order torsional frequency of the body-in-white as performance constraints. When designing the number of half-thickness ply layers, the Tsai-Wu strength theory is introduced, and the constructed optimization mathematical model is as follows:

[0062]

[0063] In the formula: M is the mass of the floor; BS(T) i ), TS(T i ), BF(T i ) and TF(T i The values ​​are the bending stiffness, torsional stiffness, first-order bending frequency, and first-order torsional frequency of the body-in-white, respectively. BS0, TS0, BF0, and TF0 are initial values; Tsai-Wu is the failure criterion.

[0064] The lightweight design and optimization method for composite materials used in automotive floors, wherein the engineering manufacturing constraints in step six are:

[0065] S1: The surface ply of the structure consists of a pair of ±45° plies, and the surface ply is continuous;

[0066] S2: The angle difference between two adjacent angled plies shall not exceed 45°;

[0067] S3: The proportion of each ply at each angle to the total number of ply layers shall not be less than 10%;

[0068] S4: The maximum number of consecutive layers at the same angle is 3;

[0069] S5: All plies are symmetrical about the center plane, and the number of plies at ±45° is the same;

[0070] S6: The tilt angle of the transition zone between different thicknesses should not exceed 7° and the increase in length should be greater than 8 times the decrease in thickness;

[0071] S7: The thinnest module in each block of flooring of different thicknesses should maximize the number of shared layers across the entire area and minimize the number of missing layers;

[0072] S8: The maximum number of layers lost in the same cross-sectional area between adjacent blocks shall not exceed 4.

[0073] The lightweight design and optimization method for composite materials used in automotive flooring, in step six, the continuous variable discretization rounding strategy can eliminate the situation where the number of ply layers is reduced and the stiffness of the laminate is reduced due to downward rounding. To ensure the load-bearing performance of the floor, the laminate needs to be compensated by adding layers, while also meeting the lightweight requirements. The number of added layers should be minimized. Since there are four ply angles: 0°, 45°, -45°, and 90°, the maximum number of layers lost during downward rounding is 4. The combinations of the number of added layers for 0° and 90° ply angles are 0, 1, 2, 3, and 4, respectively. The number of ply angles for 45° and -45° ply angles appears in pairs due to symmetrical equilibrium constraints. Therefore, the combinations of the number of added layers for ±45° are 0, 1, and 2.

[0074] The aforementioned lightweight design and optimization method for composite materials used in automotive flooring, in its sixth step, employs a continuous fiber layup strategy to ensure maximum continuity in the transition region, preventing discontinuity in the force transmission path due to layer loss and thus protecting the overall mechanical properties of the flooring. To facilitate the description of the continuous fiber layup strategy, the following definition is used: Where θ represents the pavement orientation angle; l, m, n, o, p, q represent the number of pavement layers corresponding to the pavement orientation angle θ; A, B, C correspond to independent blocks; (ABC) is a whole block formed by connecting three blocks, indicating that the entire domain shares pavement; (AB) represents the block where blocks A and B are connected; (BC) represents the block where blocks B and C are connected, indicating that the subdomains share pavement.

[0075] The lightweight design and optimization method for composite materials used in automotive flooring, wherein the improved swarm intelligence optimization algorithm (GA-VNS) in the sixth step is obtained by adding the traveling salesman problem, memory bank checking strategy and variable neighborhood search algorithm to the traditional genetic algorithm. It has high search efficiency and high computational accuracy when dealing with the ply sequence optimization problem, and can avoid getting trapped in local optima and premature convergence.

[0076] The specific implementation steps of the algorithm are as follows:

[0077] S01: Initialize the population P, set the maximum number of iterations I. max Maximum number of iterations without improvement (NI) max Parameters such as population size N;

[0078] S02: Initialize the current optimal solution S best ;

[0079] S03: The individual optimal solution S best Perform N perturbation operations;

[0080] S04: Search each individual in population P using the local search operator, ATSP operator, and crossover operator (where local search includes five neighborhood search operators: subtree shift operator, subtree exchange operator, node shift operator, node exchange operator, and mixed shift operator);

[0081] S05: Update the current solution S current and the global optimal solution S best ;

[0082] S06: Determine the updated global optimal solution S based on the memory bank checking strategy. best If the data matches the data stored in the memory bank, continue updating the global optimal solution S. best If there is no overlap, the result is updated.

[0083] S07: Apply a perturbation operator to the population P to perform a certain perturbation;

[0084] S08: Determine the termination condition of the optimization iteration. If the optimization reaches the maximum number of iterations or the convergence tolerance, the optimization process ends and the optimization result is output.

[0085] The lightweight design and optimization method for composite materials used in automotive flooring, in step six, uses an intercalation strategy to determine the sequence of global shared ply, subdomain shared ply, and independent ply. If engineering constraints are not met during execution, the ply is fine-tuned using an intercalation strategy. The steps of the intercalation strategy are as follows:

[0086] S01: Sort the shared and independent pavers of the subdomain according to the order of pavers area from largest to smallest;

[0087] S02: Adjust the order according to engineering constraints;

[0088] S03: Insert the plywood with the adjusted order into the global shared plywood from the outside to the middle surface to complete the sorting of each plywood.

[0089] The aforementioned lightweight design and optimization method for composite materials used in automotive flooring, in step six, uses the layup sequence optimization mathematical model with the lightweight coefficient and bending stiffness of the body-in-white as optimization objectives, the first-order bending frequency and the first-order torsional frequency as performance constraints, and process constraints, layup sequence constraints, and continuous layup constraints as engineering constraints, to perform multi-objective optimization design of the vehicle floor structure. The optimization mathematical model is as follows:

[0090]

[0091] In the formula: x i1 x i2 L Lx in Different arrangements are design variables; L is the lightweight coefficient; BS(X)i BF(X) represents the vehicle body bending stiffness; i ), TF(X) i These are the first-order bending frequency and first-order torsional frequency of the white body, respectively, with BF0 and TF0 as initial values; C i For engineering constraints.

[0092] The aforementioned lightweight design and optimization method for composite materials used in automotive flooring includes, in step seven, a performance comparison before and after optimization of the body-in-white flooring layup sequence. This includes determining whether the failure index of the composite material flooring under bending and torsional conditions is less than 1; determining whether the stress of the composite material flooring under bending and torsional conditions is less than the transverse tensile strength of the composite material; and comparing the bending stiffness, torsional stiffness, and low-order modes of the body-in-white before and after optimization.

[0093] The aforementioned lightweight design and optimization method for composite materials used in automotive floors, in step eight, uses the objective function of the multi-objective optimization mathematical model, which comprehensively considers the performance and lightweight design of the body-in-white, with the minimum lightweight coefficient (L) of the body-in-white as the objective function; considering occupant safety in a frontal collision, the average acceleration of the left and right B-pillars is used. Minimize as the objective function; considering occupant safety during a side impact, use the B-pillar head intrusion amount (x) as the objective function. H Minimize it as the objective function;

[0094] The constraints of the multi-objective optimization mathematical model, considering the static performance indicators of the body-in-white, require that the optimized body-in-white bending stiffness (BS), first-order bending modal frequency (BF), and first-order torsional modal frequency (TF) be no lower than those before optimization; considering occupant safety in a frontal collision, the optimized intrusion acceleration (α) at ​​the rear end of the front longitudinal beam after a frontal collision must be guaranteed. L ), anterior intrusion amount (x P The acceleration (a) should not exceed the pre-optimization value; considering occupant safety during a side impact, the optimized B-pillar head and chest intrusion acceleration (a) after the side impact must be guaranteed. H a C No higher than before optimization;

[0095] The mathematical model for the multi-objective optimization problem is shown below:

[0096]

[0097] In the formula: X represents the key design variables selected from the design variable library; L is the lightweight coefficient; x represents the average acceleration of the left and right B pillars in a head-on collision. H The maximum intrusion of the B-pillar head in a side impact; BS, TS, BF, and TF are the bending stiffness, torsional stiffness, first-order bending frequency, and first-order torsional frequency of the body-in-white, respectively; BS0, TS0, BF0, and TF0 are initial values; aL For the maximum intrusion acceleration at the rear end of the front longitudinal beam in a head-on collision, x P For the maximum intrusion amount in a head-on collision with the frontal enclosure, a H a C The maximum intrusion acceleration for the head and chest upon side impact with the B-pillar; α L0 x P0 α H0 and α C0 This is the initial value.

[0098] The lightweight design and optimization method for composite materials used in automotive flooring, specifically the improved non-dominated sorting genetic algorithm (MNSGA-II) in step eight, introduces a fast non-dominated sorting method, an elite preservation strategy, and a crowding comparison method based on the genetic algorithm. The specific implementation steps of the MNSGA-II algorithm are as follows:

[0099] S01: Generate a parent population P with a population size of N. t And perform a non-dominated sort on all of them;

[0100] S02: Select, crossover, and mutate the sorted individuals to generate the next generation population Q. t ;

[0101] S03: Merge parent populations P t and offspring population Q t A mixed population R with a population size of 2N was obtained. t And for the mixed population R t Individuals perform non-dominated ranking;

[0102] S04: Based on the non-dominated ordination criteria, for population R t Perform non-dominated sorting operations and calculate the crowding distance for each individual within each level;

[0103] S05: Select N outstanding individuals from the lowest to the highest level to form a new parent population P. t+1 When selecting individuals, priority should be given to those at lower levels, followed by those at the same level with greater crowding distance.

[0104] S06: If the termination condition is met, the iteration ends; otherwise, repeat the optimization from step 2 until the non-dominated solution meets the termination condition.

[0105] The lightweight design and optimization method for composite materials used in automotive flooring, specifically the step eight involving entropy-weighted grey relational analysis to determine the optimal solution, is as follows:

[0106] S01: Each performance response needs to be normalized to dimensionless data between 0 and 1 to facilitate quantitative analysis. If the objective function has the characteristic of "the larger the better," then the normalization formula is:

[0107]

[0108] If the objective function has the property of "the smaller the better", then the normalization formula is:

[0109]

[0110] If the objective function has the property that "the closer it is to a specific value T, the better its performance", then the normalization formula is:

[0111]

[0112] In the formula: x is the normalized value of the i-th response in the k-th objective function. i (k) represents the initial value of the objective function; max k x i (k) and min k x i (k) represent the maximum and minimum values ​​of the k-th objective function, respectively; T is a specific value;

[0113] Due to the lightweight coefficient (L) of the body-in-white and the average acceleration of the B-pillars on both sides... and the amount of B-pillar head intrusion (x) H All three objective functions are to be as small as possible, therefore the choice formula is... The three objective functions are normalized.

[0114] S02: Calculate the gray correlation coefficient γ for each design scheme, as shown in the following formula:

[0115]

[0116] In the formula: The normalized value of the i-th response in the k-th objective function is the ideal experimental design. Δ is the normalized value of the i-th response in the k-th objective function of the designed experimental scheme. 0i (k) is and The absolute difference between them; Δ max With Δ min Δ 0i The maximum and minimum values ​​of (k); ζ is the discrimination coefficient, ζ∈[0,1];

[0117] S03: Calculate the lightweight coefficient (L) of the body-in-white and the average acceleration of the left and right B-pillars using the entropy weight method. and the amount of B-pillar head intrusion (x) HThe weight coefficients of the three objective functions can be calculated using the information entropy, which represents the degree of uncertainty of the random variables. The information entropy of the k-th objective function is:

[0118]

[0119] In the formula: i = 1, 2, 3...L...m, where m is the number of responses; k = 1, 2, 3...L...n, where n is the number of objective functions; x... ik This is the standardized value of the i-th response in the k-th objective function;

[0120] The weight coefficients of the objective function can be calculated as follows:

[0121]

[0122] In the formula: d k d represents the degree of bias of the k-th objective function. k =1-e k ;

[0123] From the above formula, we can obtain L, x H The weight coefficients of the three objective functions;

[0124] S04: The grey correlation coefficient of each optimization scheme is calculated using the grey correlation coefficient and the weight coefficient of each objective function. The larger the value, the better the scheme. The grey correlation coefficient is calculated as follows:

[0125]

[0126] In the formula: n is the number of objective functions; ω k The weight coefficients of the k-th objective function are...

[0127] The lightweight design and optimization method for composite materials used in automotive floors, in the ninth step, includes comparing the performance of the body-in-white before and after multi-objective optimization, including comparing the bending stiffness, torsional stiffness, and low-order modes of the body-in-white before and after optimization; comparing the frontal collision safety performance of the body-in-white before and after optimization; and comparing the side collision safety performance of the body-in-white before and after optimization.

[0128] By employing the technical solution described above, the present invention has the following advantages:

[0129] This invention comprehensively considers multiple aspects such as materials, structure, and process, reflecting the integration of multiple disciplines. Through implicit parameter modeling, it can accurately and efficiently establish a white body model; it integrates the front, middle, and rear floors of the vehicle into a single unit, using composite materials and molds for integrated design, simplifying the processing technology, reducing mold development costs, simplifying floor assembly difficulty, and improving floor rigidity performance; while adhering to manufacturing processes, it determines the shape of the ply blocks, the thickness of the ply blocks, the number of ply layers, and the ply sequence through free dimension optimization, size optimization, and ply sequence optimization; in the ply sequence optimization, a continuous variable discretization rounding strategy is used to process the initial number of layers of the ply blocks after downward rounding, avoiding changes in the structural rigidity of the floor. To mitigate performance loss, a continuous fiber layup strategy is adopted to reduce layer loss and ensure the continuity of force transmission in the composite floor. An intercalation strategy is employed to rationally allocate the layup sequence, guaranteeing the structural performance of the floor. An improved swarm intelligence optimization algorithm is proposed, capable of optimizing discrete problems. It boasts high search efficiency and computational accuracy when handling layup sequence optimization, while avoiding getting trapped in local optima and premature convergence. Multi-objective optimization is used to optimize the design of other components in the vehicle body, maximizing the advantages of the composite floor. The proposed entropy-weighted grey relational analysis method objectively identifies the optimal solution among numerous multi-objective optimization schemes. This invention achieves integrated collaborative optimization design of automotive floor materials, structure, and processes while ensuring floor performance, making it suitable for widespread promotion and application. Attached Figure Description

[0130] Figure 1 An implicit parameterized model for the body-in-white;

[0131] Figure 2 A comparison diagram of the displacement of the measured points under the bending condition in simulation and experiment;

[0132] Figure 3 A comparison diagram of the displacement of the measured points under torsional conditions between simulation and experiment;

[0133] Figure 4 This is a comparison diagram of the simulated and experimental low-order mode shapes during modal analysis;

[0134] Figure 5 These are frontal and side impact model diagrams of the entire vehicle.

[0135] Figure 6 This is a graph showing the energy changes during a head-on collision with a complete vehicle.

[0136] Figure 7 A comparison diagram of the vehicle's deformation modes in simulation and testing after a frontal collision;

[0137] Figure 8 This is a graph showing the acceleration change of column B during a head-on collision.

[0138] Figure 9 This diagram shows the location of the front enclosure intrusion measurement points during a head-on collision.

[0139] Figure 10 Diagram showing the location of the steering column intrusion measurement point during a head-on collision;

[0140] Figure 11 This is a diagram showing the locations of the measurement points for door deformation during a head-on collision.

[0141] Figure 12 This is a cloud map showing the deformation of the front section after a head-on collision.

[0142] Figure 13 This is a graph showing the deformation of the car door over time during a head-on collision.

[0143] Figure 14 This is a graph showing the energy changes during a side impact of a vehicle.

[0144] Figure 15 A comparison diagram of the deformation modes of the whole vehicle after side impact simulation and test;

[0145] Figure 16 This diagram shows the locations of the measurement points for the acceleration and displacement of the B-pillar during a side impact.

[0146] Figure 17 Diagram showing the location of the intrusion measurement points on the threshold beam;

[0147] Figure 18 Diagram showing the location of acceleration measurement points in the battery box;

[0148] Figure 19 The graph shows the acceleration of each measuring point on column B as a function of time.

[0149] Figure 20 The graph shows the change in intrusion amount at the measuring point on column B over time.

[0150] Figure 21 These are the basic performance parameters for carbon fiber and epoxy resin;

[0151] Figure 22 These are the mechanical property parameters of carbon fiber composite materials;

[0152] Figure 23 This is a structural diagram of an integrated composite floor.

[0153] Figure 24 Initial model for composite floor layup:

[0154] Figure 25 To simplify the analytical diagram of the plywood model;

[0155] Figure 26 Thickness contour plots of plywood at various angles after free-size optimization;

[0156] Figure 27 The shapes of the four ply blocks for a 45° ply in the front, middle and rear sections;

[0157] Figure 28 A comparison image of the shape of the plywood before and after cutting;

[0158] Figure 29 Results of half-thickness layup for the front, middle, and rear portions of the composite flooring:

[0159] Figure 30 Flowchart for optimizing the layup sequence of composite flooring;

[0160] Figure 31 A rounding strategy for discretizing continuous variables;

[0161] Figure 32 Pseudocode for the memory checking strategy;

[0162] Figure 33 Flowchart of the improved swarm intelligence optimization algorithm (GA-VNS);

[0163] Figure 34 This is a half-thickness rounded ply result;

[0164] Figure 35 This represents the result of continuous fiber layup distribution;

[0165] Figure 36 To establish a shared initial layering order across the entire domain;

[0166] Figure 37 For composite flooring half-thickness layup sequence;

[0167] Figure 38 This is a distribution chart of the floor failure index.

[0168] Figure 39 This is a diagram showing the stress distribution in the floor.

[0169] Figure 40 This is a comparison chart of the bending stiffness of the composite floor and the original floor;

[0170] Figure 41 This is a comparison chart of the torsional stiffness of the composite floor and the original floor;

[0171] Figure 42 This is a comparison diagram of the first-order bending modes of the composite floor and the original floor;

[0172] Figure 43 This is a comparison diagram of the first-order torsional modes of the composite floor and the original floor;

[0173] Figure 44 Optimize the lightweight design process for the body-in-white with multiple objectives;

[0174] Figure 45 A variable library for multi-objective optimization design of body-in-white;

[0175] Figure 46 The Pareto solution set optimized for multiple objectives;

[0176] Figure 47 Schematic diagram of the improved non-dominated sorting genetic algorithm (MNSGA-II);

[0177] Figure 48 The grey relational degree values ​​of the Pareto solution set obtained by the entropy-weighted grey relational degree analysis method;

[0178] Figure 49 To optimize the comparison of bending and torsional stiffness of the front and rear body-in-white;

[0179] Figure 50 To optimize the low-order modal comparison of the body-in-white before and after;

[0180] Figure 51 Comparison of intrusion amount before and after optimization of the front bulkhead;

[0181] Figure 52 Comparison of acceleration changes at various positions before and after optimization of the B-pillar. Detailed Implementation

[0182] The present invention can be explained in more detail through the following embodiments, but the present invention is not limited to the following embodiments;

[0183] The present invention discloses a lightweight design and optimization method for composite materials used in automotive flooring, the design and optimization method specifically including the following steps:

[0184] I. Using the implicit parametric modeling software SFE Concept, models of the left side of modules such as the front bumper beam, front longitudinal beam, front bulkhead, floor, side panels, roof, and battery box are created using base points, baselines, and cross-sections. The accuracy of each module model is ensured by adjusting the base point positions, controlling the baseline curvature, and optimizing the curvature of the reference cross-sections. Then, MAP mapping relationships are used to connect and encapsulate each module. The left-side model is copied to the right side via mirroring, and local asymmetric features are modified. Finally, the parametric model of the body-in-white is assembled using connectors. The implicit parametric model of the body-in-white is shown below. Figure 1 As shown.

[0185] Second, the parametric model of the body-in-white was simplified and geometrically cleaned in the finite element software. The mesh of each component was divided by extracting the mid-surface. The parts on the body-in-white were assigned materials and properties. Penetration and interference analysis were performed on the body-in-white. Finally, the connection between the components was established to obtain the finite element model of the body-in-white. The accuracy of the finite element model of the body-in-white was verified by bending, torsion and modal analysis. The analysis showed that the established finite element model of the body-in-white had good consistency with the actual body-in-white.

[0186] Furthermore, during the bending condition analysis and verification of the body-in-white model, the translational degrees of freedom in the XY axis direction of the front suspension assembly center and the translational degrees of freedom in the XYZ axis direction of the rear suspension assembly center were constrained. A 6000N load was uniformly applied to the R points of the four front and rear passenger seats. Ten measuring points were selected on the left and right sides of the front longitudinal beam, sill beam, and rear longitudinal beam of the body-in-white, and the Z-axis displacement of the measuring points was measured. The simulation and experimental comparison diagram is shown below. Figure 2 As shown. The formula for calculating the static bending stiffness of the body-in-white is as follows:

[0187]

[0188] In the formula, K b D is the static bending stiffness; F is the resultant load; Lmax D represents the maximum Z-axis displacement of the measuring point on the left. Rmax This represents the maximum Z-axis displacement of the measuring point on the right.

[0189] Furthermore, during the torsional condition analysis and verification of the body-in-white model, the translational degree of freedom along the Y-axis of the middle of the front crossbeam of the body-in-white was constrained, and the translational and rotational degrees of freedom along the XYZ axes of the rear suspension assembly center of the body-in-white were also constrained. A torque of 2000N was applied to the rotation centers of the two front suspensions of the body-in-white, which is equivalent to applying forces of 1724N in opposite directions to the two front suspension assembly centers respectively. The selection of measuring points was the same as that for the bending condition, and the simulation and experimental comparison of the Z-axis displacement of the measuring points is shown in the figure below. Figure 3 As shown. The formula for calculating the static torsional stiffness of the body-in-white is as follows:

[0190]

[0191] In the formula, K t M is the torsional stiffness; θ is the applied torque; D is the torsional angle; L D represents the Z-direction displacement of the left loading point; R L represents the Z-direction displacement of the right loading point; L is the distance between the left and right front overhang loading points.

[0192] Furthermore, during modal analysis verification of the body-in-white model, the boundary conditions for the modal analysis of the body-in-white were not subject to any loads, nor were any of the six degrees of freedom constrained; only the modal characteristics of the body-in-white in its free state were analyzed. Considering that under real road conditions, the body-in-white is mainly subjected to low-frequency excitation, the low-order modes of the body-in-white were analyzed, with a frequency bandwidth of 0–100 Hz. Simulation and experimental results for the first-order torsional mode and the first-order bending mode were obtained. The comparison diagram of the mode shapes of the first-order torsional mode and the first-order bending mode is shown below. Figure 4 As shown.

[0193] III. To further verify the accuracy of the established finite element model, it is necessary to combine the finite element model of the body-in-white with those of the powertrain, chassis assembly, battery, doors, and windows to form a complete electric vehicle finite element model. Based on the vehicle crash test methods and modeling procedures in the 2021 edition of the *C-NCAP Management Rules*, frontal and side impact models of the vehicle were established. The frontal impact model uses a vertically fixed rigid plane to simulate a rigid wall and a horizontally fixed plane to simulate the road surface, causing the vehicle model to collide head-on with the rigid wall at a speed of 50 km / h. The side impact model uses a horizontally fixed plane to simulate the road surface. A movable barrier is modeled using finite element analysis according to the dimensions in C-NCAP. The X-axis coordinate of the barrier's center point is the same as the R-axis coordinate of the driver's seat in the vehicle. The movable barrier model moves towards the vehicle model at a speed of 50 km / h and collides with it. The frontal and side impact vehicle crash models are as follows: Figure 5 As shown. The reliability of the finite element model is often verified by the energy changes and vehicle deformation modes during a collision, while the correctness of the model is verified by safety performance indicators such as acceleration and intrusion amount. The collision test results show that the constructed vehicle model meets the accuracy requirements for modeling a real vehicle.

[0194] Furthermore, under the condition of a full-vehicle head-on collision, the energy change diagram during the collision is as follows: Figure 6 As shown, the deformation pattern of the entire vehicle after the collision is as follows: Figure 7 As shown in the diagram. Measuring the acceleration at the bottom of the left and right B-pillars represents the change in passenger compartment acceleration during the collision. The B-pillar acceleration change diagram is shown in the figure below. Figure 8 As shown. In a head-on collision, the intrusion of the front bumper and steering column reflects the survival space of the front occupants. The deformation of the front doors determines whether the front occupants can exit the vehicle in time for rescue. The intrusion measurement points are selected at four locations: the footrest area of ​​the front bumper, the center point of the steering column, and the upper left (E1), lower left (E2), upper right (E3), and lower right (E4) positions on the door frame. The front bumper intrusion measurement point locations are as follows. Figure 9 As shown, the location of the measuring point on the steering wheel column is as follows: Figure 10 As shown, the locations of the door deformation measurement points are as follows: Figure 11 As shown in the figure. The deformation contour diagram of the front bulge after a head-on collision, as simulated by finite element method. Figure 12 As shown, the point of maximum deformation is located at the right-side footrest. The finite element simulation curve of the door deformation versus time during a frontal collision is shown below. Figure 13 As shown, the point of maximum deformation is the upper right measuring point of the car door. The changes in acceleration and intrusion at each measuring point in the simulation and experiment are not significant, verifying the correctness of the frontal collision model.

[0195] Furthermore, under the condition of a full-vehicle side impact, the energy change diagram during the collision is as follows: Figure 14 As shown, the deformation pattern of the entire vehicle after the collision is as follows: Figure 15As shown in the diagram. The acceleration and intrusion of the B-pillar after impact represent the compression and impact experienced by the occupants during the collision. The location of the measurement points for the B-pillar acceleration and intrusion is shown in the diagram below. Figure 16 As shown. The amount of intrusion into the sill beam after an impact affects the battery's safety performance. The measurement points for the sill beam intrusion are shown in the figure. Figure 17 As shown. The acceleration of the battery pack reflects the state of the battery during the collision. The location of the acceleration measurement point of the battery pack is shown in the figure. Figure 18 As shown in the figure. After the side impact, the acceleration curves of each measuring point on the B-pillar as a function of time are as follows. Figure 19 As shown in the figure. The curve of the intrusion amount at the measuring point on the B-pillar during a side impact simulation as a function of time is shown in the figure. Figure 20 As shown, the measurement point at the abdominal location represents the point of maximum intrusion. The changes in acceleration and intrusion at each measurement point were small in both the simulation and the experiment, verifying the correctness of the side-impact model.

[0196] IV. Composite materials, needing to withstand combined loads such as tension, compression, bending, and shear, also require certain stiffness and strength. Carbon fiber is selected as the fiber reinforcement material, and resin as the matrix material to form a composite material. The basic performance parameters of carbon fiber and resin are as follows: Figure 21 As shown. Using the Tsai-Wu strength theory as the failure criterion for the composite laminate, mechanical property tests were conducted on the composite material, including uniaxial tensile tests at 0° and 90° layups, uniaxial compression tests at 0° and 90° layups, and in-lay shear tests at ±45° layups. The measured mechanical property parameters of the carbon fiber composite material are as follows: Figure 22 As shown.

[0197] Furthermore, the Tsai-Wu strength theory is as follows:

[0198] Failure criteria are primarily based on the allowable stress and strain of composite materials, and the failure index of the material is obtained by calculating the ply and matrix. When the failure index is less than "1", the stress or strain is within the allowable range; while when the failure index is greater than "1", the stress or strain exceeds the allowable range.

[0199] Based on a synthesis of multiple intensity criteria, Tsai-Wu proposed the energy polynomial intensity criterion:

[0200] F i σ i +F ij σ i σ j +F ijk σ i σ j F ij σ i σ k +LL=1

[0201] In engineering applications, only the first two items are usually considered:

[0202] F i σ i +F ij σ i σ j =1(i,j=1,2,3,4,5,6)

[0203] In the formula: F i This refers to the strength parameters of the material.

[0204] For the two-dimensional plane stress problem, equation (4) can be simplified to:

[0205]

[0206] The strength parameters in the formula are as follows:

[0207]

[0208] Among them, X t Longitudinal tensile strength; X c Y represents longitudinal compressive strength. t Y represents the transverse tensile strength. c S represents the transverse compressive strength; S represents the planar shear strength.

[0209] The above failure criterion states that the material will fail if the maximum stress or strain of the material or a certain ply exceeds the allowable value of the material.

[0210] Fifth, integrating the front, middle, and rear floors of a car into a single unit using composite materials and molds simplifies the manufacturing process, reduces mold development costs, simplifies floor assembly, improves floor rigidity, and better aligns with the integration of composite material structure and function. The resulting integrated floor structure is as follows: Figure 23 As shown. Next, the composite flooring structure needs to be designed with layup specifications, including obtaining the composite layup thickness and layup block shape using free-size optimization, cutting each layup block shape, determining the initial number of layers for each layup block using size optimization, and rounding down the number of layers for each layup block. Finally, an initial model of the composite flooring layup is created in Fibersim software, as shown. Figure 24 As shown.

[0211] Furthermore, when designing ply thickness and ply block shape, grouping ply blocks with the same angle together (each ply block is manufactured with the same thickness) can simplify the model, such as... Figure 25As shown. In composite materials, a 90° ply can withstand lateral loads, while 0° and 45° plies can withstand axial and shear loads, respectively. Therefore, 0°, 90°, and ±45° are used as ply angles. Free-size optimization is used to obtain the ply thickness at each angle and the shape of each ply block. Using the ply thickness at each angle as the design variable and the body-in-white lightweight coefficient as the optimization objective, the formula for calculating the body-in-white lightweight coefficient is as follows:

[0212]

[0213] In the formula, L is the lightweighting coefficient; M is the body-in-white mass (kg); K t Let A be the torsional stiffness of the body-in-white, N*m / (deg); and let A be the projected footprint area of ​​the product of the wheelbase and track width in the Z direction, m. 2 The performance constraints are the original body-in-white bending stiffness, first-order bending frequency, and first-order torsional frequency. Manufacturing constraints include ±45° ply balance and symmetry, each of the four ply angles accounting for no less than 10%, and ply shape and thickness being symmetrical about the neutral plane. The mathematical model is constructed as follows:

[0214]

[0215] Where: L is the lightweight coefficient of the body-in-white; BS T BF T and TF T These represent the bending stiffness, first-order bending frequency, and first-order torsional frequency of the body-in-white, respectively; BS0, BF0, and TF0 are initial values; C2, C4, and C5 are manufacturing constraints.

[0216] During optimization, ensure that the thickness design variable has sufficient margin. Figure 26 The thickness of the ply at each angle is optimized for free-size applications. After optimization, each angle ply has 4 ply blocks, for a total of 16 ply blocks. Figure 27 The shapes of the four ply blocks with a 45° layup after optimization for free dimensions at the front, middle and rear positions.

[0217] Furthermore, trimming the shape of each ply block is to make the ply blocks more regular, thereby simplifying the composite ply fabrication process. To distinguish each ply block, they are numbered 111-114. The first digit, ranging from 1 to 3, represents the front, middle, and back flooring layers; the second digit, ranging from 1 to 4, represents the four angles of the ply: 0°, 45°, -45°, and 90°; and the third digit, ranging from 1 to 4, represents the four ply blocks for each angle. For example, number 323 represents ply block 3 corresponding to the 45° ply of the back flooring. A comparison of the ply block shapes before and after trimming is provided. Figure 28 As shown.

[0218] Furthermore, the initial number of ply layers for each ply block is obtained through size optimization. Using the thickness of the ply block as the design variable, floor quality as the optimization objective, and the bending stiffness, torsional stiffness, first-order bending frequency, and first-order torsional frequency of the body-in-white as performance constraints, the Tsai-Wu strength theory is introduced when designing the number of ply layers for half-thickness. The resulting optimization mathematical model is as follows:

[0219]

[0220] In the formula: M is the mass of the floor; BS(T) i ), TS(T i ), BF(T i ) and TF(T i The values ​​are the bending stiffness, torsional stiffness, first-order bending frequency, and first-order torsional frequency of the body-in-white, respectively. BS0, TS0, BF0, and TF0 are initial values; Tsai-Wu is the failure criterion.

[0221] After optimization, the optimal ply thickness T for each ply block was obtained. i Taking a single-layer board of thickness k as the manufacturing unit, the specific number of ply layers N = T in each ply block of the half-thickness ply of composite flooring. i / k. Due to the influence of symmetry and balance constraints, the floor ply is balanced and symmetrical, therefore the actual number of ply layers in each ply block is 2N. Figure 29 The image shows the half-thickness layup results for the front, middle, and rear sections of the composite flooring. The actual layup thickness and number of layers are twice the data shown in the image.

[0222] VI. Optimize the layup sequence of composite material layers, taking into account engineering and manufacturing constraints. The specific process for layup sequence optimization is as follows: Figure 30 As shown, the ply sequence optimization process is as follows:

[0223] (1) The number of ply blocks after downward rounding is processed by the continuous variable discretization rounding strategy to obtain the specific number of ply blocks in the front, middle and back floor blocks. On this basis, ply blocks with corresponding angles are added to ensure the balance and symmetry of the ply blocks.

[0224] (2) Using the continuous fiber layup strategy, the distribution results of global shared layup, subdomain shared layup and independent layup are obtained;

[0225] (3) The GA-VNS optimization algorithm is used to optimize the piling order of the global shared piling while ensuring the engineering constraints of the piling, so as to obtain the optimized piling order of the global shared piling.

[0226] (4) Using the optimized global shared ply as the initial ply base, insert the subdomain shared ply and independent ply into the global shared ply according to the interpolation strategy to form a ply sequence scheme.

[0227] (5) Optimize the mathematical model based on the ply sequence, determine whether the scheme is feasible, and output the optimal ply sequence scheme and the corresponding target value.

[0228] Furthermore, the process constraints considered in the optimization of the layup sequence design include:

[0229] S1: The surface ply of the structure consists of a pair of ±45° plies, and the surface ply is continuous;

[0230] S2: The angle difference between two adjacent angled plies shall not exceed 45°;

[0231] S3: The proportion of each ply at each angle to the total number of ply layers shall not be less than 10%;

[0232] S4: The maximum number of consecutive layers at the same angle is 3;

[0233] S5: All plies are symmetrical about the center plane, and the number of plies at ±45° is the same;

[0234] S6: The tilt angle of the transition zone between different thicknesses should not exceed 7° and the increase in length should be greater than 8 times the decrease in thickness;

[0235] S7: The thinnest module in each block of flooring of different thicknesses should maximize the number of shared layers across the entire area and minimize the number of missing layers;

[0236] S8: The maximum number of layers lost in the same cross-sectional area between adjacent blocks shall not exceed 4.

[0237] Furthermore, the continuous variable discretization rounding strategy aims to eliminate the reduction in the number of ply layers and the decrease in laminate stiffness caused by downward rounding. To ensure the load-bearing performance of the floor, additional layers are added to compensate for the loss of ply thickness, while also meeting the lightweight requirement by minimizing the number of additional layers. Since there are four ply angles: 0°, 45°, -45°, and 90°, downward rounding results in a maximum of four dropped layers. Specifically, the combinations of additional layers for 0° and 90° ply angles are 0, 1, 2, 3, and 4, respectively. Due to symmetrical equilibrium constraints, the ply angles for 45° and -45° ply angles appear in pairs, therefore the combinations of additional layers for ±45° are 0, 1, and 2. The specific process of the continuous variable discretization rounding strategy is as follows... Figure 31 As shown.

[0238] Furthermore, the continuous fiber layup strategy aims to ensure maximum continuity in the transition area, preventing discontinuity in the force transmission path due to layer loss and thus protecting the overall mechanical properties of the flooring. To facilitate the description of the continuous fiber layup strategy, we define... Where θ represents the paving angle; l, m, n, o, p, and q represent the number of ply layers corresponding to the paving angle θ; A, B, and C each correspond to an independent block; (ABC) is a whole block formed by connecting the three blocks, representing a shared ply across the entire area; (AB) represents a block where blocks A and B are connected; (BC) represents a block where blocks B and C are connected, representing a shared ply across sub-domains. For example, if the front, middle, and rear flooring are divided into three independent modules A, B, and C, and the number of ply layers at 0°, ±45°, and 90° for the front flooring module is 7, 3, 3, and 5 respectively, the number of ply layers at each angle for the middle flooring module is 5, 4, 4, and 3 respectively, and the number of ply layers at each angle for the rear flooring module is 5, 5, 5, and 2 respectively, the number of ply layers at each angle can be expressed by the following formula:

[0239]

[0240] After processing with the continuous fiber layup strategy, the following formulas are obtained for half-thickness global shared layup, subdomain shared layup, and independent layup:

[0241]

[0242] After processing, the number of layers for the global shared ply at 0°, ±45°, and 90° are 5, 3, and 2, respectively. The sub-domain shared ply at ±45° is shared by the middle and rear floor blocks with 1 layer, and the sub-domain shared ply at 90° is shared by the front and middle floor blocks with 1 layer. The independent ply at 0° has 2 layers in the front floor block, the independent ply at ±45° has 1 layer in the rear floor block, and the independent ply at 90° has 2 layers in the rear floor block.

[0243] Furthermore, the improvement process of the swarm intelligence optimization algorithm (GA-VNS) is as follows:

[0244] Traditional genetic algorithms (GA) create an initial population through encoding, evaluate the fitness of each individual in the initial population using a genetic algorithm fitness function, and perform genetic operations based on the fitness value to achieve population iteration. During population evolution, better individuals have a higher probability of being retained for iteration; better individuals may also mutate to produce even better individuals, resulting in differences among the better individuals in the iterated population; different individuals may also crossover to produce new offspring. Through selection, crossover, and mutation, evolution continues generation by generation until the goal is achieved.

[0245] For the discrete optimization problem of the layup sequence of composite flooring, the solution to the Traveling Salesman Problem (TSP) can be used as a reference to analogize the layup sequence problem to the problem of visiting the layup angles in sequence. The order of visits is equivalent to the sorting of the flooring layups, and the path length is equivalent to the optimization objective. Therefore, a genetic algorithm can be used to handle the TSP problem.

[0246] However, in the TSP problem, each city is unique, and regardless of the selection, the paths between any two cities are different, preventing repetition. But fiber-reinforced composite laminates have four layup angle specifications: -45°, 0°, 45°, and 90°. In TSP, invalid swaps of layup orders with the same angle can occur. To filter out valid laminate orderings, a memory-checking strategy is proposed. During optimization, different layup order schemes are stored in a memory. Before each new ordering scheme is calculated, it is compared with existing schemes in the memory. If the layup order is the same, the scheme is invalid, and the existing calculation result is directly output; otherwise, it is a valid scheme, and the corresponding calculation is performed. Qualified layup orders and calculation results are then stored in the memory. The pseudocode for the memory-checking strategy is as follows: Figure 32 As shown.

[0247] Traditional genetic algorithms are prone to getting trapped in local optima and exhibiting premature convergence. The Variable Neighborhood Search (VNS) algorithm possesses superior neighborhood search capabilities. Its basic process is as follows: An initial solution is randomly generated within the first neighborhood. The algorithm begins searching for the best solution within this neighborhood. When a new solution is found, the current best solution is updated, and the algorithm continues to search for the best solution within the new neighborhood. If no better solution is found in the current neighborhood, the algorithm searches for the best solution within neighborhoods with different structures. This process is repeated until the algorithm reaches its termination condition and outputs the search result. During neighborhood transformations, the structure of the neighboring neighborhoods in the VNS algorithm is not fixed but rather a combination of structures with different patterns. This advantage allows the VNS algorithm to continuously change the neighborhood structure and transform the search space, thus avoiding getting trapped in local optima.

[0248] The flowchart of the improved swarm intelligence optimization algorithm (GA-VNS) is as follows: Figure 33 As shown, the specific implementation steps are as follows:

[0249] Step 1: Initialize the population P and set the maximum number of iterations I. max Maximum number of iterations without improvement (NI) max Parameters such as population size N;

[0250] Step 2: Initialize the current optimal solution S best ;

[0251] Step 3: Calculate the individual optimal solution S best Perform N perturbation operations;

[0252] Step 4: Search each individual in population P using the local search operator, ATSP operator, and crossover operator (where local search includes five neighborhood search operators: subtree shift operator, subtree exchange operator, node shift operator, node exchange operator, and mixed shift operator);

[0253] Step 5: Update the current solution S current and the global optimal solution S best ;

[0254] Step 6: Determine the updated global optimal solution S based on the memory bank checking strategy. best If the data matches the data stored in the memory bank, continue updating the global optimal solution S. best If there is no overlap, the result is updated.

[0255] Step 7: Apply a perturbation operator to the population P to perform a certain perturbation;

[0256] Step 8: Determine the termination condition of the optimization iteration. If the optimization reaches the maximum number of iterations or the convergence tolerance is reached, the optimization process ends and the optimization result is output.

[0257] Furthermore, the intercalation strategy is used to determine the order of global shared ply, subdomain shared ply, and independent ply. If engineering constraints are not met during execution, the ply is fine-tuned using the intercalation strategy. The steps of the intercalation strategy are as follows:

[0258] Step 1: Sort the shared and independent pavers of the subdomains in descending order of ply area;

[0259] Step 2: Adjust the order according to engineering constraints;

[0260] Step 3: Insert the plywood with the adjusted order from the outside to the middle into the global shared plywood to complete the sorting of each plywood.

[0261] Furthermore, the layup sequence optimization mathematical model uses the lightweight coefficient and bending stiffness of the body-in-white as optimization objectives, the first-order bending frequency and the first-order torsional frequency as performance constraints, and process constraints, layup sequence constraints, and continuous layup constraints as engineering constraints to perform multi-objective optimization design of the body floor structure. The optimization mathematical model is as follows:

[0262]

[0263] In the formula: x i1 x i2 L Lx in Different arrangements are design variables; L is the lightweight coefficient; BS(X) i BF(X) represents the vehicle body bending stiffness; i ), TF(X) i These are the first-order bending frequency and first-order torsional frequency of the white body, respectively, with BF0 and TF0 as initial values; C i For engineering constraints.

[0264] VII. When optimizing the floor layer laying sequence design, the data processed by the discretization rounding strategy is as follows: Figure 29 The data in the image, after processing, yields the half-thickness rounded ply result as follows: Figure 34 As shown; continuous fiber layup strategy processing Figure 34 The results of the half-thickness global shared ply, subdomain shared ply, and independent ply distribution obtained after data processing are as follows: Figure 35 As shown; Improved swarm intelligence optimization algorithm processing Figure 35 The optimal initial sorting of globally shared plying is obtained by using the global shared plying method. Figure 36 As shown; after inserting the subdomain shared ply and independent ply into the global shared ply using the intercalation strategy, the final half-thickness ply sequence of the composite flooring is as follows. Figure 37 As shown.

[0265] After optimizing the floor layer sequence, the total mass of the car floor was reduced. By comparing and analyzing the performance of the optimized floor, it was determined that the various performance characteristics of the car body were improved after the floor layer sequence was optimized.

[0266] Furthermore, it was determined whether the failure index of the composite flooring under bending and torsional conditions was less than 1. Figure 38 This is a distribution chart of the floor failure index.

[0267] Furthermore, it was determined whether the stress in the composite flooring under bending and torsional conditions was less than the transverse tensile strength of the composite material. Figure 39 This is a diagram showing the stress distribution on the floor.

[0268] Furthermore, the bending stiffness of the composite material body-in-white and the original body-in-white were compared, and the results are as follows: Figure 40 As shown.

[0269] Furthermore, the torsional stiffness of the composite material body-in-white and the original body-in-white were compared, and the results are as follows: Figure 41 As shown.

[0270] Furthermore, the first-order bending modes of the composite material body-in-white and the original body-in-white were compared, and the comparison results are as follows: Figure 42 As shown.

[0271] Furthermore, the first-order torsional modes of the composite material body-in-white and the original body-in-white were compared, and the comparison results are as follows: Figure 43 As shown.

[0272] 8. To improve the overall performance of the body-in-white and maximize the advantages of the composite material floor, a multi-objective optimization lightweight design of the body-in-white is also required. The multi-objective optimization lightweight design process for the body-in-white is as follows: Figure 44 As shown, the specific process of multi-objective optimization is as follows:

[0273] (1) Select appropriate materials for each component on the white car body according to its function;

[0274] (2) Based on the implicitly parametric body-in-white model, identify the design variables for each component on the body-in-white and compile them into a design variable library, such as... Figure 45 As shown;

[0275] (3) Based on the importance of the components during the collision, key design variables in the design variable library are selected;

[0276] (4) Based on the performance analysis results of the body-in-white, set the objective function and constraints for the multi-objective optimization problem of lightweight body-in-white;

[0277] (5) Construct an approximate model based on the mathematical model of the multi-objective optimization problem of lightweight body-in-white;

[0278] (6) Using the approximate model as the optimization object, the MNSGA-II algorithm is used to perform multi-objective optimization design on the body-in-white to obtain the Pareto solution set, such as... Figure 46 As shown, each solution represents an optimization scheme;

[0279] (7) The optimal solution in the Pareto solution set is determined by the entropy weight grey relational analysis method, and this solution is used as the best optimization scheme for the lightweight design of the body-in-white.

[0280] Furthermore, the objective function for multi-objective optimization is established as follows: taking into account both the performance design and lightweight design of the body-in-white, the objective function is to minimize the lightweight coefficient (L) of the body-in-white; considering occupant safety in a frontal collision, the objective function is the average acceleration of the left and right B-pillars. Minimize as the objective function; considering occupant safety during a side impact, use the B-pillar head intrusion amount (x) as the objective function. H Minimize it as the objective function.

[0281] Furthermore, the constraint conditions for multi-objective optimization are established as follows: Considering the static performance indicators of the body-in-white, it is necessary to ensure that the bending stiffness (BS), first-order bending modal frequency (BF), and first-order torsional modal frequency (TF) of the optimized body-in-white are not lower than those before optimization; considering occupant safety during a frontal collision, it is necessary to ensure that the optimized rear-end intrusion acceleration (α) of the front longitudinal beam after a frontal collision is... L ), anterior intrusion amount (x P The acceleration (a) should not exceed the pre-optimization value; considering occupant safety during a side impact, the optimized B-pillar head and chest intrusion acceleration (a) after the side impact must be guaranteed. H a C The value is not higher than before optimization.

[0282] Furthermore, the mathematical model for the multi-objective optimization problem is shown below:

[0283]

[0284] In the formula: X represents the key design variables selected from the design variable library; L is the lightweight coefficient; x represents the average acceleration of the left and right B pillars in a head-on collision. H The maximum intrusion of the B-pillar head in a side impact; BS, TS, BF, and TF are the bending stiffness, torsional stiffness, first-order bending frequency, and first-order torsional frequency of the body-in-white, respectively; BS0, TS0, BF0, and TF0 are initial values; a L For the maximum intrusion acceleration at the rear end of the front longitudinal beam in a head-on collision, x P For the maximum intrusion amount in a head-on collision with the frontal enclosure, a H a C The maximum intrusion acceleration for the head and chest upon side impact with the B-pillar; α L0 x P0 α H0 and α C0 This is the initial value.

[0285] Furthermore, the improved non-dominated sorting genetic algorithm (MNSGA-II) introduces a fast non-dominated sorting method, an elite preservation strategy, and a crowding comparison method on the basis of the genetic algorithm. The principle diagram of the MNSGA-II algorithm is shown below. Figure 47 As shown, the specific implementation steps of the algorithm are as follows:

[0286] Step 1: Generate a parent population P with a population size of N. t And perform a non-dominated sort on all of them;

[0287] Step 2: Select, crossover, and mutate the sorted individuals to generate the next generation population Q. t ;

[0288] Step 3: Merge parent populations P t and offspring population Q t A mixed population R with a population size of 2N was obtained. t And for the mixed population R t Individuals perform non-dominated ranking;

[0289] Step 4: Based on the non-dominated ordination criteria, sort the population R... t Perform non-dominated sorting operations and calculate the crowding distance for each individual within each level;

[0290] Step 5: Select N outstanding individuals from the lowest to the highest level to form a new parent population P. t+1 When selecting individuals, priority should be given to those at lower levels, followed by those at the same level with greater crowding distance.

[0291] Step 6: If the termination condition is met, the iteration ends; otherwise, repeat the optimization from step 2 until the non-dominated solution meets the termination condition.

[0292] Furthermore, the steps for selecting the optimal solution in the Pareto solution set using entropy-weighted grey relational analysis are as follows:

[0293] Step 1: The performance responses need to be normalized to dimensionless data between 0 and 1 to facilitate quantitative analysis. If the objective function has the property of "the larger the better," then the normalization formula is:

[0294]

[0295] If the objective function has the property of "the smaller the better", then the normalization formula is:

[0296]

[0297] If the objective function has the property that "the closer it is to a specific value T, the better its performance", then the normalization formula is:

[0298]

[0299] In the formula: x is the normalized value of the i-th response in the k-th objective function. i (k) represents the initial value of the objective function; max k x i (k) and min k x i (k) represents the maximum and minimum values ​​of the k-th objective function, respectively; T is a specific value.

[0300] Due to the lightweight coefficient (L) of the body-in-white and the average acceleration of the B-pillars on both sides... and the amount of B-pillar head intrusion (x) H Since all three objective functions should be as small as possible, we choose equation (60) to normalize the three objective functions.

[0301] Step 2: Calculate the grey correlation coefficient γ for each design scheme, as shown in the following formula:

[0302]

[0303] In the formula: The normalized value of the i-th response in the k-th objective function is the ideal experimental design. Δ is the normalized value of the i-th response in the k-th objective function of the designed experimental scheme. 0i (k) is and The absolute difference between them; Δ max With Δ min Δ 0iThe maximum and minimum values ​​of (k); ζ is the discrimination coefficient, ζ∈[0,1].

[0304] Step 3: Calculate the lightweight coefficient (L) of the body-in-white and the average acceleration of the left and right B-pillars using the entropy weight method. and the amount of B-pillar head intrusion (x) H The weight coefficients of the three objective functions can be calculated using the information entropy, which represents the degree of uncertainty of the random variables. The information entropy of the k-th objective function is:

[0305]

[0306] In the formula: i = 1, 2, 3...L...m, where m is the number of responses; k = 1, 2, 3...L...n, where n is the number of objective functions; x... ik This is the standardized value of the i-th response in the k-th objective function.

[0307] The weight coefficients of the objective function can be calculated as follows:

[0308]

[0309] In the formula: d k d represents the degree of bias of the k-th objective function. k =1-e k .

[0310] From the above formula, we can obtain L, x H The weighting coefficients of the three objective functions.

[0311] Step 4: Calculate the grey correlation degree value for each optimization scheme using the grey correlation coefficient and the weight coefficient of each objective function. A higher value indicates a better scheme. The grey correlation degree value is calculated as follows:

[0312]

[0313] In the formula: n is the number of objective functions; ω k The weight coefficients of the k-th objective function are...

[0314] The grey relational degree values ​​of each optimization scheme were calculated as follows: Figure 48 As shown, the optimization scheme with the highest grey relational value is the optimal optimization scheme.

[0315] 9. By comparing the performance before and after multi-objective optimization, the effectiveness of the multi-objective optimization scheme was verified, and the feasibility of the lightweight design and optimization method for automotive composite floor was further demonstrated.

[0316] Furthermore, the bending and torsional stiffness of the body-in-white before and after optimization were compared, such as... Figure 49As shown.

[0317] Furthermore, comparing the low-order modes of the body-in-white before and after optimization, such as... Figure 50 As shown.

[0318] Furthermore, comparing the frontal collision safety performance of the body-in-white before and after optimization involves comparing the changes in intrusion amount of the front bulkhead, the deformation amount of the front door, the acceleration of the B-pillar, and the intrusion amount of the steering column before and after optimization over time. For example, comparing the intrusion amount of the front bulkhead before and after optimization... Figure 51 As shown.

[0319] Furthermore, comparing the side-impact safety performance of the body-in-white before and after optimization involves comparing the changes in acceleration at various positions of the B-pillar before and after optimization, the changes in intrusion amount at various positions of the B-pillar before and after optimization over time, the changes in intrusion amount of the sill beam before and after optimization over time, and the changes in the acceleration of the battery pack before and after optimization over time. For example, the changes in acceleration at various positions of the B-pillar before and after optimization are compared to... Figure 52 As shown.

[0320] Compared with the prior art, the advantages of the present invention are as follows:

[0321] 1. Traditional methods for lightweighting automotive floors mostly focus on structural and material aspects, dividing the floor into three parts (front, middle, and rear) and optimizing them separately, resulting in less than ideal optimization effects. This invention proposes a design and optimization method for automotive composite material floors that comprehensively considers materials, structure, and processes. This method achieves integrated and collaborative optimization design of automotive floor materials, structure, and processes, resulting in significant lightweighting effects.

[0322] 2. Traditional composite material layup optimization only considers the performance constraints of the vehicle, neglecting the engineering constraints of the composite material itself. This hinders the engineering manufacturing of the composite material and affects the strength and force transmission of the floor. This invention proposes a carbon fiber floor layup design method that uses composite materials to integrate the front, middle, and rear floor sections into a single mold design, simplifying the processing technology, reducing costs, and improving floor performance. While adhering to manufacturing processes, the shape of the layup blocks, the thickness of the layup layers, the number of layers, and the layup sequence are optimized to ensure vehicle performance while achieving lightweight flooring.

[0323] 3. This invention uses a continuous variable discretization rounding strategy to process the number of ply layers obtained by downward rounding, thus avoiding the loss of structural stiffness of the floor; it uses a continuous fiber plying strategy to reduce the phenomenon of missing layers and ensure the continuity of force transmission in composite flooring; and it uses an intercalation strategy to rationally allocate the plying sequence, thus ensuring the structural performance of the flooring.

[0324] 4. Traditional optimization algorithms are mostly designed for continuous optimization problems and cannot handle discrete optimization problems such as ply order optimization. This invention proposes an improved swarm intelligence optimization algorithm that, based on the genetic algorithm, uses an analogy to the traveling salesman problem, introduces a memory bank checking strategy, and incorporates a variable neighborhood search algorithm to optimize discrete problems. This algorithm boasts high search efficiency and computational accuracy, while avoiding getting trapped in local optima and premature convergence, effectively solving the ply order optimization problem for flooring.

[0325] 5. The multi-objective optimization method proposed in this invention improves various performance aspects of the vehicle by optimizing key components on the vehicle body, maximizing the advantages of composite material flooring. The entropy-weighted grey relational analysis method used in the optimization can objectively identify the optimal solution among many multi-objective optimization schemes, avoiding subjective factors such as human preference, and is flexible in operation, suitable for handling decision-making problems in engineering.

[0326] The parts of this invention not described in detail are prior art.

[0327] The embodiments selected herein for the purpose of disclosing the inventive objectives are currently considered suitable; however, it should be understood that the invention is intended to include all variations and modifications of the embodiments that fall within the scope of this concept and invention.

Claims

1. A lightweight design and optimization method for composite materials used in automotive flooring, characterized by: The design and optimization method specifically includes the following steps: The first step is to establish a parametric model of the body-in-white using implicit parametric modeling methods; The second step is to establish a finite element model of the body-in-white based on the parametric model of the body-in-white, and verify the accuracy of the finite element model of the body-in-white through bending, torsion and modal analysis. The third step is to establish a finite element model of the electric vehicle and, based on the vehicle crash test methods and modeling process in the C-NCAP Management Rules, establish frontal and side impact models of the vehicle. The reliability of the finite element model is verified by the energy changes and vehicle deformation modes during the collision, and the correctness of the model is verified by the safety performance indicators of acceleration and intrusion. The fourth step is to select composite materials, including fiber-reinforced materials and matrix materials, and use the Tsai-Wu strength theory as the failure criterion for composite laminates. The mechanical properties of the composite materials are then tested to obtain the mechanical property parameters of the composite materials. Step 5: Using the selected composite material, the front, middle, and rear floors of the car are designed as a whole. The designed composite floor first undergoes free-size optimization to obtain the composite ply thickness and ply block shape. Then, the shape of each ply block is cut to ensure that the ply blocks are more regular and easier to process. Finally, the initial number of layers for each ply block is determined by size optimization. The number of layers for each ply block is rounded down to obtain the initial number of layers for each ply block. Through the above optimization, the initial model of the composite floor ply is obtained, including the continuous ply thickness results of the front, middle, and rear blocks. Step 6: Optimize the layup sequence of the composite flooring based on engineering and manufacturing constraints, specifically including the following steps: (1) Use the continuous variable discretization rounding strategy to process the initial ply number after downward rounding, while ensuring that the in-plane stiffness of the front, middle and rear blocks is not reduced, so as to obtain the specific ply number of the front, middle and rear blocks. On this basis, ply with corresponding angles is added to ensure balanced symmetry constraints. (2) Using the continuous fiber layup strategy, the distribution results of global shared layup, subdomain shared layup and independent layup are obtained; (3) The improved swarm intelligence optimization GA-VNS algorithm is used to optimize the piling order of the global shared piling; (4) Using the optimized global shared ply as the initial ply base, insert the subdomain shared ply and independent ply into the global shared ply according to the intercalation strategy to form a ply sequence scheme; (5) Optimize the mathematical model based on the ply sequence, determine whether the scheme is feasible, and output the optimal ply sequence scheme and the corresponding target value; Step 7: Compare the floor performance before and after optimization to determine whether the various performance characteristics of the body-in-white have been improved after the floor layering sequence optimization. Step 8: To improve the overall performance of the body-in-white and maximize the advantages of the composite material floor, a multi-objective optimization and lightweight design of the body-in-white is required, which includes the following steps: (1) Select appropriate materials for each part of the white car body according to its function; (2) Based on the implicit parameterized body-in-white model, identify the design variables of each component on the body-in-white and compile them into a design variable library; (3) Based on the importance of the components during the collision, select the key design variables from the design variable library; (4) Based on the performance analysis results of the body-in-white, set the objective function and constraints for the multi-objective optimization problem of lightweight body-in-white; (5) Construct an approximate model based on the mathematical model of the multi-objective optimization problem of lightweight body-in-white; (6) Using the approximate model as the optimization object, the body-in-white is designed with multiple objectives by the MNSGA-Ⅱ algorithm to obtain the Pareto solution set, where each solution represents an optimization scheme; (7) Use the entropy weight grey relational analysis method to determine the optimal solution in the Pareto solution set, and use it as the best optimization scheme for the lightweight design of the body-in-white in multi-objective optimization. Step 9: By comparing the performance before and after multi-objective optimization, verify whether the multi-objective optimization scheme is feasible, and further verify the feasibility of the lightweight design and optimization method for automotive composite floor.

2. The lightweight design and optimization method for composite materials used in automotive flooring according to claim 1, characterized in that: In the second step, the bending condition constrains the translational degrees of freedom in the XY axis direction of the front suspension assembly center of the body-in-white, and constrains the translational degrees of freedom in the XYZ axis direction of the rear suspension assembly center of the body-in-white. The load is evenly applied to the R points of the four front and rear passenger seats, and the Z-direction displacements on the front longitudinal beam, sill beam, and rear longitudinal beam of the body-in-white are measured. The static bending stiffness of the body-in-white is then calculated using the following formula: In the formula, Static bending stiffness; The resultant load; This represents the maximum Z-axis displacement of the measuring point on the left. This represents the maximum Z-axis displacement of the measuring point on the right.

3. The lightweight design and optimization method for composite materials used in automotive flooring according to claim 1, characterized in that: In the second step, the torsional condition constrains the translational degree of freedom of the middle of the front crossbeam of the body-in-white along the Y-axis, and the translational and rotational degrees of freedom of the rear suspension assembly center of the body-in-white along the XYZ axes. A torque is applied to the rotation centers of the two front suspensions of the body-in-white, which is equivalent to applying forces in opposite directions to the two front suspension assembly centers respectively. The Z-direction displacement at the same position as the bending stiffness is measured, and the static torsional stiffness of the body-in-white is calculated using the following formula: In the formula, For torsional stiffness; The applied torque; It is the angle of twist; This represents the Z-axis displacement of the left loading point; This represents the Z-axis displacement of the right loading point; This represents the distance between the loading points of the left and right front suspensions.

4. The lightweight design and optimization method for composite materials used in automotive flooring according to claim 1, characterized in that: In the second step, the modal analysis is performed without applying any load to the body-in-white or constraining any of the six degrees of freedom; only the modal characteristics of the body-in-white in its free state are analyzed.

5. The lightweight design and optimization method for composite materials used in automotive flooring according to claim 1, characterized in that: In the third step of the frontal collision performance index, the acceleration measurement point is the bottom of the left and right B-pillars; the intrusion measurement points are the footrest of the front bumper, the center of the steering column, and the four positions of the upper left, lower left, upper right, and lower right of the door frame.

6. The lightweight design and optimization method for composite materials used in automotive flooring according to claim 1, characterized in that: In the third step, the side impact performance indicators include acceleration measurement points at four locations on the B-pillar corresponding to the person's head, chest, abdomen, and H-point, as well as the center of the battery box; intrusion measurement points are at four locations on the B-pillar corresponding to the person's head, chest, abdomen, and H-point, as well as the center of the battery box and the sill beam closest to the battery box.

7. The lightweight design and optimization method for composite materials used in automotive flooring according to claim 1, characterized in that: The Tsai-Wu strength theory in the fourth step is as follows: The failure criterion is based on the allowable stress and allowable strain of composite materials. The failure index of the material is obtained by calculating the ply and matrix. When the failure index is less than "1", the stress or strain is within the allowable range; while when the failure index is greater than "1", the stress or strain exceeds the allowable range. Based on a synthesis of multiple intensity criteria, Tsai-Wu proposed the energy polynomial intensity criterion: In engineering applications, only the first two items are usually considered: In the formula: F i These are the strength parameters of the material; For a two-dimensional plane stress problem, the above equation simplifies to: The strength parameters in the formula are as follows: in, Longitudinal tensile strength; Longitudinal compressive strength; It is the transverse tensile strength; Transverse compressive strength; It is the plane shear strength; The above failure criterion states that the material will fail if the maximum stress or strain of the material or a certain ply exceeds the allowable value of the material.

8. The lightweight design and optimization method for composite materials used in automotive flooring according to claim 1, characterized in that: The mechanical property testing of composite materials in the fourth step includes uniaxial tensile tests of 0° and 90° ply, uniaxial compression tests of 0° and 90° ply, and in-ply shear tests of ±45° ply.

9. The lightweight design and optimization method for composite materials used in automotive flooring according to claim 1, characterized in that: In the fifth step, the free dimension optimization uses 0°, 90°, and ±45° as ply angles. To simplify the model, ply blocks with the same angle are grouped together, with the ply thickness at each angle as the design variable and the body-in-white lightweight coefficient as the optimization target. The formula for calculating the body-in-white lightweight coefficient is as follows: In the formula, The weight reduction factor; The mass of the body-in-white is in kg; For the torsional stiffness of the body-in-white, ; The area of ​​the projected footprint of the product of the vehicle's wheelbase and track width in the Z direction. ; Using the original body-in-white bending stiffness, first-order bending frequency, and first-order torsional frequency as performance constraints, and ±45° ply balance and symmetry, each of the four angle plies accounting for no less than 10%, and the shape and thickness of each ply being symmetrical about the neutral plane as manufacturing constraints, a free-size optimization mathematical model is constructed as follows: In the formula: The lightweight coefficient of the body-in-white; , and These are the bending stiffness of the body-in-white, the first-order bending frequency, and the first-order torsional frequency, respectively. , and Initial value; , and To create constraints.

10. The lightweight design and optimization method for composite materials used in automotive flooring according to claim 1, characterized in that: In the fifth step, the size optimization uses the thickness of the ply block as the design variable, the floor quality as the optimization objective, and the bending stiffness, torsional stiffness, first-order bending frequency, and first-order torsional frequency of the body-in-white as performance constraints. When designing the number of half-thickness ply layers, the Tsai-Wu strength theory is introduced, and the optimization mathematical model is as follows: In the formula: For floor quality; , , and These are the bending stiffness, torsional stiffness, first-order bending frequency, and first-order torsional frequency of the body-in-white. , , and The initial value is Tsai-Wu, which is the failure criterion.

11. The lightweight design and optimization method for composite materials used in automotive flooring according to claim 1, characterized in that: The engineering manufacturing constraints in step six are as follows: S 1: The surface ply of the structure consists of a pair of ±45° plies, and the surface ply is continuous; S 2: The angle difference between two adjacent angled plies shall not exceed 45°; S 3: The proportion of each ply at each angle to the total number of ply layers shall not be less than 10%; S 4: The maximum number of consecutive layers at the same angle is 3; S 5: All plies are symmetrical about the center plane, and the number of plies at ±45° is the same; S 6: The tilt angle of the transition zone between different thicknesses should not exceed 7° and the increase in length should be greater than 8 times the decrease in thickness; S 7: For flooring of different thicknesses, the thinnest module in each section should maximize the number of shared layers across the entire area and minimize the number of missing layers; S 8: The maximum number of layers lost in the same cross-sectional area between adjacent blocks shall not exceed 4.

12. The lightweight design and optimization method for composite materials used in automotive flooring according to claim 1, characterized in that: The continuous variable discretization rounding strategy in step six can eliminate the situation where the number of ply layers is reduced and the stiffness of the laminate is reduced due to downward rounding. In order to ensure the load-bearing performance of the floor, the laminate needs to be compensated by adding layers. At the same time, the lightweight requirements must be met, and the number of added layers should be minimized. Since there are four ply types: 0°, 45°, -45°, and 90°, the maximum number of layers lost by downward rounding is 4. Among them, the combination of the number of added layers for 0° and 90° ply types are 0, 1, 2, 3, and 4, respectively. The number of ply types for 45° and -45° ply types appears in pairs due to symmetrical equilibrium constraints. Therefore, the combination of the number of added layers for ±45° is 0, 1, and 2.

13. The lightweight design and optimization method for composite materials used in automotive flooring according to claim 1, characterized in that: The continuous fiber layup strategy in step six ensures maximum continuity in the transition area, preventing discontinuity in the force transmission path due to layer loss and thus protecting the overall mechanical properties of the flooring. To facilitate the description of the continuous fiber layup strategy, we define... , , , , , ,in θ Indicates the direction of the paving; l , m , n , o , p , q express θ The number of ply layers corresponding to the ply angle; A , B , C Each corresponds to an independent block; ABC () represents a unified block consisting of three connected blocks, indicating a globally shared layer; AB ) represents a block A and B Connected blocks, ( BC ) represents a block B and C Connected blocks indicate shared subdomain layers.

14. The lightweight design and optimization method for composite materials used in automotive flooring according to claim 1, characterized in that: The improved swarm intelligence optimization algorithm GA-VNS in the sixth step is obtained by adding the traveling salesman problem, memory bank checking strategy and variable neighborhood search algorithm to the traditional genetic algorithm. It has high search efficiency and high computational accuracy when dealing with the layer order optimization problem, and can avoid getting trapped in local optima and premature convergence. The specific implementation steps of the algorithm are as follows: S01: Initialize the population P Set the maximum number of iterations. I max Maximum number of iterations without improvement NI max Population size N parameter; S02: Initialize the current optimal solution S best ; S 03: Find the individual optimal solution S best implement N Secondary disturbance operation; S04: For the population P Each individual uses local search operators, ATSP operators, and crossover operators to search. Local search includes five neighborhood search operators: subtree shift operator, subtree exchange operator, node shift operator, node exchange operator, and hybrid shift operator. S05: Update the current solution S current and the global optimal solution S best ; S06: Determine the updated global optimal solution based on the memory bank checking strategy. S best If the data matches the data stored in the memory bank, continue updating the global optimal solution. S best If there is no overlap, the result is updated. S07: For the population P Use a perturbation operator to perform a certain perturbation; S08: Determine the termination condition of the optimization iteration. If the optimization reaches the maximum number of iterations or the convergence tolerance, the optimization process ends and the optimization result is output.

15. The lightweight design and optimization method for composite materials used in automotive flooring according to claim 1, characterized in that: The interpolation strategy in step six is ​​used to determine the order of global shared ply, subdomain shared ply, and independent ply. If engineering constraints are not met during execution, the ply is fine-tuned through the layering strategy. The steps of the interpolation strategy are as follows: S01: Sort the shared and independent pavers of the subdomain according to the order of pavers area from largest to smallest; S02: Adjust the order according to engineering constraints; S03: Insert the plywood with the adjusted order into the global shared plywood from the outside to the middle surface to complete the sorting of each plywood.

16. The lightweight design and optimization method for composite materials used in automotive flooring according to claim 1, characterized in that: In the sixth step, the mathematical model for optimizing the ply sequence uses the lightweight coefficient and bending stiffness of the body-in-white as optimization objectives, the first-order bending frequency and the first-order torsional frequency as performance constraints, and process constraints, ply sequence constraints, and continuous ply constraints as engineering constraints to perform multi-objective optimization design of the body floor structure. The optimization mathematical model is as follows: In the formula: The different arrangements of these variables are design variables; L The weight reduction factor; For vehicle body bending stiffness; These are the first-order bending frequency and the first-order torsional frequency of the body-in-white, respectively. Initial value; For engineering constraints.

17. The lightweight design and optimization method for composite materials used in automotive flooring according to claim 1, characterized in that: The seventh step, comparing the performance before and after optimization of the body-in-white floor layup sequence, includes determining whether the failure index of the composite floor under bending and torsional conditions is less than 1; determining whether the stress of the composite floor under bending and torsional conditions is less than the transverse tensile strength of the composite material; and comparing the bending stiffness, torsional stiffness, and low-order modes of the body-in-white before and after optimization.

18. The lightweight design and optimization method for composite materials used in automotive flooring according to claim 1, characterized in that: The objective function of the multi-objective optimization mathematical model in the eighth step comprehensively considers the performance design and lightweight design of the body-in-white, with the lightweight coefficient of the body-in-white as the key factor. L Minimize as the objective function; Considering occupant safety in a head-on collision, the average acceleration of the left and right B-pillars is taken as the average acceleration. Minimize as the objective function; considering occupant safety during a side impact, use the head intrusion amount at the B-pillar as the objective function. Minimize as the objective function; The constraints of the multi-objective optimization mathematical model, considering the static performance indicators of the body-in-white, require ensuring the bending stiffness of the optimized body-in-white. BS First-order bending mode frequency BF and first-order torsional mode frequency TF Not lower than before optimization; considering occupant safety in a head-on collision, the intrusion acceleration of the optimized rear end of the front longitudinal beam must be guaranteed after a head-on collision. Anterior intrusion amount The acceleration should not exceed that before optimization; considering occupant safety during a side impact, the optimized B-pillar head and chest intrusion acceleration must be maintained after the side impact. , No higher than before optimization; The mathematical model for the multi-objective optimization problem is shown below: In the formula: X These are the key design variables selected from the design variable library; L The weight reduction factor; The average acceleration of the left and right B pillars in a head-on collision; This represents the maximum intrusion depth of the B-pillar head in a side impact collision. , , and These are the bending stiffness, torsional stiffness, first-order bending frequency, and first-order torsional frequency of the body-in-white. , , and Initial value; This represents the maximum intrusion acceleration at the rear end of the front longitudinal beam upon impact. This represents the maximum intrusion amount in a head-on collision with the front bumper. , The maximum intrusion acceleration for the head and chest upon side impact with the B-pillar; , , and This is the initial value.

19. The lightweight design and optimization method for composite materials used in automotive flooring according to claim 1, characterized in that: The improved non-dominated sorting genetic algorithm MNSGA-II in step eight introduces a fast non-dominated sorting method, an elite preservation strategy, and a crowding comparison method on the basis of genetic algorithms. The specific implementation steps of the MNSGA-II algorithm are as follows: S01: Generate a population of size... N parental population And perform a non-dominated sort on all of them; S02: Select, crossover, and mutate the sorted individuals to generate the next generation population. ; S03: Merge parent populations and offspring population The population size was 2. N Mixed population and for mixed populations Individuals perform non-dominated ranking; S04: Based on the non-dominated ordination criteria, the population... Perform non-dominated sorting operations and calculate the crowding distance for each individual within each level; S05: Select from lower to higher levels sequentially. N Outstanding individuals constitute a new parental population. When selecting individuals, priority should be given to those at lower levels, followed by those at the same level with greater crowding distance. S06: If the termination condition is met, the iteration ends; otherwise, repeat the optimization from step 2 until the non-dominated solution meets the termination condition.

20. The lightweight design and optimization method for composite materials used in automotive flooring according to claim 1, characterized in that: The steps for determining the optimal solution using the entropy-weighted grey relational analysis method in step eight are as follows: S01: Each performance response needs to be normalized to dimensionless data between 0 and 1 to facilitate quantitative analysis. If the objective function has the characteristic of "the larger the better," then the normalization formula is: If the objective function has the property of "the smaller the better", then the normalization formula is: If the objective function has the property of "getting closer to a specific value" The characteristic of "better performance" is then normalized as follows: In the formula: For the first In the objective function, the th The normalized value of each response, These are the initial values ​​for the objective function; and The first The maximum and minimum values ​​of the objective function; For a specific value; Due to the lightweight coefficient of the body-in-white L Average acceleration of the B-pillars on both sides and the amount of B-pillar head intrusion All three objective functions are to be as small as possible, therefore the selection formula... The three objective functions are normalized. S02: Calculate the grey correlation coefficient for each design scheme. As shown in the following formula: In the formula: When the ideal experimental scheme is the first k In the objective function, the th The normalized value of each response; When designing the test scheme, the first k In the objective function, the th The normalized value of each response; for and The absolute difference between them; and They are respectively The maximum and minimum values; To distinguish the coefficients, ; S03: Calculation of the lightweight coefficient of the body-in-white using the entropy weight method L Average acceleration of the B-pillars on both sides and the amount of B-pillar head intrusion The weighting coefficients of the three objective functions can be calculated using the information entropy, which represents the degree of uncertainty of the random variables. The information entropy of each objective function is: In the formula: , The number of responses; , The number of objective functions; For the first In the objective function, the th The standardized value of a response; The weight coefficients of the objective function can be calculated as follows: In the formula: For the first The degree of deviation of each objective function ; From the above formula, we can obtain... L、 , The weight coefficients of the three objective functions; S04: The grey correlation coefficient of each optimization scheme is calculated using the grey correlation coefficient and the weight coefficient of each objective function. The larger the value, the better the scheme. The grey correlation coefficient is calculated as follows: In the formula: The number of objective functions; For the first The weight coefficients of each objective function. .

21. The lightweight design and optimization method for composite materials used in automotive flooring according to claim 1, characterized in that: The performance comparison of the body-in-white before and after multi-objective optimization in the ninth step includes comparing the bending stiffness, torsional stiffness, and low-order modes of the body-in-white before and after optimization; comparing the frontal collision safety performance of the body-in-white before and after optimization; and comparing the side collision safety performance of the body-in-white before and after optimization.

Images