A method for topology optimization and fiber path design in continuous fiber 3D printing

By using finite element software and non-equidistant fiber path planning, the problem of incomplete optimization scale in continuous fiber 3D printing was solved, and the simultaneous optimization of fiber content, orientation and topology was achieved, generating manufacturable fiber paths and improving the accuracy and efficiency of optimization.

CN115310154BActive Publication Date: 2026-11-17HUAZHONG UNIV OF SCI & TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210809160.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-07-11
Publication Date
2026-11-17
Estimated Expiration
2042-07-11

AI Technical Summary

Technical Problem

Existing technologies in continuous fiber 3D printing fail to fully consider the optimization scale, neglect fiber spacing optimization, fail to achieve actual manufacturing, and do not consider composite material failure constraints, resulting in optimization results that cannot be directly used for printing.

Method used

A model was established using finite element software. Through simultaneous optimization of fiber content, fiber orientation, and topology, combined with non-equidistant fiber path planning, a manufacturable continuous fiber trajectory was generated, taking into account polymerization failure constraints and composite material mixing ratio criteria.

Benefits of technology

It achieves full-scale optimization, improves the accuracy and efficiency of optimization, and generates fiber paths that can be directly used for continuous fiber 3D printing, thus possessing manufacturability.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115310154B_ABST
    Figure CN115310154B_ABST
Patent Text Reader

Abstract

This invention relates to a topology optimization and fiber path design method for continuous fiber 3D printing. Based on continuous fiber 3D printing technology, this method considers failure constraints and aims at structural stiffness. It performs integrated, multi-scale, parallel optimization design of fiber layout, printing spacing, and fiber orientation. A two-way progressive structural optimization method is used to discretize the element density, and a penalty-free variable density method is used to continuously optimize the fiber content within the element. Simultaneously, the principal stress direction is used as the continuous fiber orientation. Furthermore, the failure of composite material structures is considered, and the Tsai-Hill criterion for polymeric P-norm integration is used to approximate the maximum failure point. The failure constraints are supplemented by introducing Lagrange multipliers as a design objective. Finally, based on the optimization scheme, a method for planning continuous fiber paths is proposed, and the resulting finished product is printed.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of continuous fiber 3D printing technology, and in particular to a topology optimization and fiber path design method for continuous fiber 3D printing. Background Technology

[0002] Continuous fiber reinforced composites possess a series of excellent characteristics, including high specific strength, high specific modulus, and strong designability of material properties, leading to their wide application in the aerospace, shipbuilding, and automotive fields. Due to limitations in traditional manufacturing processes, the traditional structural optimization problem for continuous fiber reinforced composites involves optimizing the stacking sequence of laminates under current loading conditions, assuming uniform fiber distribution and constant fiber orientation in each layer, to achieve maximum mechanical properties. However, advancements in modern manufacturing processes have led to sophisticated composite material manufacturing techniques. For example, continuous fiber-based 3D printing technology has made significant progress. This technology designs based on fiber angles in localized areas, offering greater design freedom. By constructing a fiber orientation optimization problem for composite structures according to the manufacturing constraints of the designed part, it can more effectively distribute the load across the entire structure, maximizing structural stiffness.

[0003] Topology optimization is a type of structural design. Its core idea is to discretize the design domain into several elements, with each element represented by a pseudo-density between 0 and 1. The presence or absence of an element in the structure is indicated by an interpolation relationship between the pseudo-density and material properties such as volume and elastic modulus. Although the performance of continuous fiber 3D printed prototypes is not yet directly usable, the combination of advanced manufacturing processes and structural design methods often yields complementary results. Isotropic materials have produced many high-performance, lightweight structures by combining topology optimization with 3D printing. However, research on topology optimization for composite materials is still in its developmental stage. Topology optimization for composite materials needs to consider the influence of fiber orientation angles. Existing research on composite material topology optimization shows that the choice of fiber angle has a significant impact on topological morphology and mechanical properties; therefore, combining topology optimization with fiber orientation design is a promising design method.

[0004] In addition, considering the degrees of freedom in the 3D printing process, the printing spacing (which can be equated to the fiber volume fraction in that region) can be used as a design variable during printing. Theoretical analysis was also conducted to improve the stress state. The study on printing spacing fully utilizes the degrees of freedom in the topology optimization of 3D printed composite materials, demonstrating how incorporating printing spacing as a variable helps the optimization results, thus opening up new avenues for composite material topology optimization.

[0005] The joint optimization study on fiber angles and topological configurations revealed the performance improvement effect of the optimized structure. Dani El Peeters et al. used the variable density method to solve the composite material topology optimization problem, used the lamination parameters to indirectly represent the fiber angles and performed joint optimization, and obtained the fiber path distribution through some post-processing, achieving good results; Haoqing Ding et al. used the normal distribution function to form a material interpolation scheme, realized continuous fiber angle design, and established a parallel topology optimization model; Xiaolei Yan et al. established a parallel optimization design method for structure / material topology and material orientation based on the bidirectional progressive topology optimization (BESO) method, based on uniform porous microstructures.

[0006] In addition, some studies have considered the degrees of freedom in the 3D printing process, such as using the printing spacing (which can be equated to the fiber volume fraction in that region) as a design variable during printing, and conducting theoretical analysis to improve the stress state. AVMalakhov et al. defined the fiber as a continuous curve distributed along the principal stress trajectory. Based on this, they studied the influence of changes in fiber orientation and fiber distance on the stress concentration factor using the finite element method. The results showed that this optimization method significantly reduced the stress concentration factor. Subsequently, they conducted molding experiments on printed specimens, and the results showed that the optimized strength and stiffness were also significantly improved. Zhanghao Hou et al. designed based on the stress gradient distribution, so that the fiber content corresponds to the stress distribution and the fiber direction corresponds to the maximum principal stress distribution, obtaining trajectory lines with unequal fiber spacing. The test results also showed good results. Hang Li et al. proposed a full-scale optimization scheme, which simultaneously optimizes the topology, fiber orientation, and fiber spacing of the structure. This scheme is based on the dual-material element density optimization framework. By constraining the local fiber volume fraction, the fiber spacing is constrained, so that the fiber spacing can achieve global equidistant variation, thus realizing full-scale optimization design to a certain extent. The study on printing spacing makes full use of the degrees of freedom in the topology optimization of 3D printed composite materials, and demonstrates how adding printing spacing as a variable helps the optimization results, thus opening up new ideas for the topology optimization of composite materials.

[0007] The existing topology optimization has the following defects: (1) The optimization scale is not comprehensive; there are many optimization designs for fiber angle, and some consider fiber angle and topology optimization at the same time, but few consider fiber spacing for optimization, and there is a lack of a full-scale optimization framework for continuous fiber printed structures; (2) The optimization process does not consider manufacturing issues. Most current optimization methods only obtain some structures in theory, without actual printing and manufacturing, and do not consider manufacturing constraints such as fiber continuity and printing feasibility in the printing process; (3) The optimization results cannot be directly used for printing; (4) The failure constraints of composite materials are not considered in the optimization. Summary of the Invention

[0008] The technical problem to be solved by this invention is to provide a topology optimization and fiber path design method for continuous fiber 3D printing. This method is based on the fiber mixing ratio criterion of composite materials and performs simultaneous optimization of fiber content, fiber orientation and topology under the constraint of polymerization failure. The optimization scale is comprehensive, and further, a non-equidistant fiber path planning method is used to generate continuous fiber trajectories to achieve the actual printing and manufacturing results.

[0009] The technical solution adopted in this invention is a topology optimization and fiber path design method for continuous fiber 3D printing, the method comprising the following steps:

[0010] (1) A continuous fiber layout optimization model is established using finite element software. The design domain of the continuous fiber layout optimization model is discretized into several finite element elements, each of which includes four nodes.

[0011] (2) Initialize the design variables of the continuous fiber layout optimization model using the discretized design domain from step (1). The design variables include the element density ρ. B The orientation angle θ of the unit fiber and the unit fiber content index ρ S The finite element calculation file is output from the continuous fiber layout optimization model;

[0012] (3) Establish a continuous fiber material interpolation model based on the composite material mixing ratio criterion, and output the material card of the continuous fiber material from the continuous fiber material interpolation model; define optimization parameters in MATLAB, and read the finite element calculation file output by the continuous fiber layout optimization model from MATLAB, and obtain the element stress, element displacement and element principal stress angle of each finite element in the continuous fiber material; each element stress includes four nodal stresses and each element displacement includes four nodal displacements;

[0013] (4) The objective function for optimizing continuous fiber composite materials under polymerization failure constraint is set as follows:

[0014]

[0015] Where F represents the optimization objective function of the continuous fiber composite material under the constraint of polymerization failure, C represents the overall flexibility of the continuous fiber composite structure, λ is the Lagrange multiplier, and FI max FI min FI represents the maximum and minimum values ​​of the failure index, respectively. pn The global aggregation of failure indices is represented by U, F, and K, which represent the global displacement, global load vector, and global stiffness matrix of the continuous fiber composite structure, respectively. e and k e ρ represents the element displacement vector and element stiffness matrix, respectively. B ρ represents the element density. S V(x) represents the fiber content index of a unit, θ represents the fiber orientation angle of the unit; V(x) and V * Represent the total volume of the structure and the target volume, respectively; vf and vf * These are the structural fiber content and the target fiber content, respectively.

[0016] (5) Calculate the element compliance of each finite element based on the nodal stress and nodal displacement of each finite element obtained in step (3), and adjust the element compliance of each finite element according to the corresponding element density ρ. B and unit fiber content index ρ S The element density ρ of each finite element is obtained by differentiating the derivatives. B The sensitivity and the element fiber content index ρ of each finite element element S The sensitivity is assessed by performing finite element analysis on the element stiffness matrix and nodal displacements of each finite element to obtain the stress-associated variable values ​​for each finite element. Based on these stress-associated variable values, the element density ρ under the aggregate failure constraint is then calculated. B Sensitivity and unit fiber content index ρ S Sensitivity;

[0017] (6) Perform polymerization on the optimization objective function of the continuous fiber composite material under the polymerization failure constraint set in step (4). The specific process is as follows: for the unit density ρ B The objective function F is iterated several times, and in each iteration, the element density ρ under the aggregation failure constraint calculated in step (5) is adjusted. B The sensitivity was filtered, and the element density ρ in the optimization objective function F of the continuous fiber composite was updated using the BESO method. B In each iteration, the unit fiber content index ρ SThe element fiber orientation angle θ remains unchanged and is updated with the element principal stress angle in each iteration. A finite element calculation file is exported for each iteration. The stress results are analyzed based on the exported finite element calculation file. The iteration stops when the stress results meet the failure index constraint or the iteration reaches the maximum cycle value.

[0018] (7) Based on step (6), the optimization objective function of the continuous fiber composite material under the polymerization failure constraint set in step (4) is polymerized again. The specific process is as follows: for the unit fiber content index ρ S The objective function λ is iterated several times. In each iteration, the fiber content index ρ of the unit cell under the polymerization failure constraint obtained in step (5) is adjusted. S The sensitivity was filtered, and the unit fiber content index ρ in the optimization objective function of continuous fiber composites was updated using the SIMP method. S In each iteration, the unit density ρ B The element fiber orientation angle θ remains unchanged, and is updated with the element principal stress angle in each iteration. The stress results are analyzed based on the finite element calculation file exported in each iteration. The iteration stops when the stress results meet the failure index constraint or the iteration reaches the maximum cycle value.

[0019] (8) After the iteration is completed, output the finite element calculation file;

[0020] (9) Based on the finite element calculation file output in step (7), a continuous fiber trajectory is generated by using a non-equidistant fiber path planning method.

[0021] Preferably, in step (8), the specific process of generating a continuous fiber trajectory using a non-equidistant fiber path planning method based on the finite element calculation file output in step (9) includes the following steps:

[0022] First, the element density ρ of each finite element in the finite element calculation file output in step (8). B All are discrete variables, taking values ​​of 0 or 1, with the element density ρ being the most significant. B The topological configuration is constructed using finite element units of type 1.

[0023] Second, sample points are evenly distributed on the long scale of the topological configuration constructed in step one, and the topological configuration constructed in step one is divided into multiple regions with different numbers of cross-sectional branches according to the cross-sectional size of the sample points; the fiber content index ρ of all finite element elements in the finite element calculation file output in step (8) is then used to calculate the fiber content index ρ of all finite element elements. S The fiber content v of the finite element within the topological configuration constructed in step one is converted into the fiber content v. f The unit fiber content index ρS It is a continuous variable, and its value can take any value between 0 and 1; the fiber content v of the finite element within the topological configuration constructed in step one. f The expression is:

[0024] Third, adjacent regions with the same number of cross-sectional branches are merged to obtain several merged branch regions, which then form a new topological configuration. The fiber content (v) of different branches within each merged branch region is then considered. f The fiber count is allocated to obtain the expression for the fiber count in each of the merged branch regions: Among them, R i This represents the number of fibers within any branch region; sum(ρ) Si ∩ρ B =1) represents the element density ρ within the branch region. B The sum of unit fiber content with a value of 1; R represents the total fiber content before branching. total This indicates the total number of fibers before branching;

[0025] Fourth, after the fiber counts of different branches within each merged branch region have been allocated, the fiber points on the sample points are allocated. The specific process is as follows: based on the obtained fiber counts in each branch region and the unit fiber content index ρ before branching... S And the fiber content v of the finite element within the topological configuration obtained in step two. f The fiber points on the sample points are assigned based on the fiber content v within the finite element elements of the new topology obtained in step three, which is present between different fiber points within that branch. f Equal; due to the fiber content v of each finite element within the topological configuration. f They are not equal, therefore non-equidistant fiber points will be obtained;

[0026] Fifth, for a complete fiber, all fiber points on the fiber are filtered according to distance weights, and the final unit fiber orientation angle θ generated in step (7) is obtained to obtain the smoothed fiber direction; the expression for filtering the final unit fiber orientation angle θ is: in, This represents the sum of all distances within the filter radius. Represents the product of distance and fiber angle; θ i θ represents the angle of the filtered fibers. j cel indicates the fiber angle within the filtration range. The distance, or weighting factor, is represented by r. sen The filter radius is represented by the value; finally, based on the Hermite interpolation method, the fiber points are used as the coordinates of the sample points, and the smoothed fiber direction is used as the first derivative to obtain the fiber distribution curve under the minimum printing spacing constraint.

[0027] Preferably, in step (3), the optimization parameters defined in MATLAB include: design variables, maximum cycle value, target volume, failure index constraint, and element density ρ. B Filtration radius and unit fiber content index ρ S Filter radius;

[0028] Preferably, in step (3), the specific process of establishing a continuous fiber material interpolation model based on the composite material mixing ratio criterion and outputting the material card from the continuous fiber material interpolation model includes the following steps:

[0029] (3.1) The fiber content value of continuous composite materials in the composite structure is expressed by the unit fiber content index ρ. S Interpolation is performed to obtain v based on different maximum and minimum print spacings dmax and dmin, respectively, for a given print layer thickness. fmax and v fmin v fmax v represents the maximum fiber content of a continuous composite material within a topological configuration. fmin The minimum fiber content of a continuous composite material within a topological configuration is represented by the fiber content v of the finite element of the continuous composite material within the topological configuration. f For linear interpolation, the expression is: In the formula This represents the unit fiber content index of the i-th unit;

[0030] (3.2) Based on the modulus of the composite material structure and the modulus of the fiber material, and then using the composite material mixing ratio formula, the ρ value is calculated. s The interpolated model for continuous fiber materials is expressed as follows:

[0031] E1 = v f E f +(1-v f E m ,

[0032]

[0033]

[0034] v12 = v f v f +(1-v f )vm

[0035] Among them, E m G m v m These are the elastic modulus, shear modulus, and Poisson's ratio of the matrix material constituting the composite material, respectively; E m G m v m E1, E2, and G represent the elastic modulus, shear modulus, and Poisson's ratio of the fiber material. 12 v 12 These represent the longitudinal elastic modulus, transverse elastic modulus, in-plane shear modulus, and Poisson's ratio of the composite fiber after mixing, respectively.

[0036] (3.3) Introducing the unit fiber orientation angle θ, let m = cosθ, n = sinθ, then the off-axis rotation matrix of the composite material is:

[0037] (3.4) Introducing the element density ρ B , ρ B ∈{0,1}, representing the presence or absence of material, let ρ B Substituting the off-axis rotation matrix of the composite material, we obtain the interpolated continuous fiber material structure model, whose expression is:

[0038]

[0039] Preferably, in step (5), the corresponding element density ρ is adjusted according to the element compliance of each element. B Differentiation yields the element density ρ B The sensitivity is obtained using the adjoint method, and its expression is: in, This represents the stiffness matrix of the i-th element; based on the element flexibility of each element, the corresponding element fiber content ρ is... S Differentiation yields the unit fiber content index ρ S The sensitivity is obtained using the adjoint method, and its expression is: in, B i T i D i Let represent the stiffness matrix, strain-displacement matrix, off-axis matrix, and constitutive matrix of the i-th element, respectively.

[0040] Preferably, in step (5), the corresponding stress-associated variable values ​​are calculated based on the element stiffness matrix and nodal displacements of each element, and then the stress-associated variable values ​​are calculated accordingly. value Obtain the unit density ρ under the constraint of polymerization failure.B and unit fiber content index ρ S The specific process for determining sensitivity includes the following steps:

[0041] (5.1) Obtain the global aggregate FI of the failure index PN For ρ Bi The sensitivity of the α is expressed as: in, V is the failure index matrix, σ i Let X be the element stress vector, where X represents the tensile strength along the fiber direction and Y represents the tensile strength perpendicular to the fiber direction. Indicates in-plane shear strength;

[0042] (5.2) Differentiating the failure index of the i-th element with respect to stress, its expression is:

[0043] (5.3) σ i =D0B i u i Substituting the expression from step (4.2), we get: Among them, L i Satisfy u i =L i U, u i B represents the element displacement vector. i D represents the shape function. i Denotes the constitutive matrix;

[0044] (5.4) Based on the overall equilibrium equation of the finite element method, the expression is obtained as follows:

[0045] (5.5) Substituting the expression in step (5.3) into the expression in step (5.4), we get:

[0046] (5.6) By introducing the adjoint variable μ into the expression in step (5.5) Bi The expression obtained is: Where K represents the overall structural stiffness matrix;

[0047] (5.7) Based on the expression in step (5.6), the sensitivity of the P-norm aggregation failure constraint is simplified to: because: Finally, the P-norm aggregation failure index constraint on the design variable ρ is obtained. Bi The expression for the sensitivity is: Similarly, the P-norm aggregation failure index can be obtained for the design variable ρ. Si The expression for the sensitivity is: in Let be the vector of the associated node values ​​of the i-th element with respect to the unit fiber content index;

[0048] (5.8) Combining the sensitivity of the aggregation failure constraint to the design variables obtained in (5.7) with the sensitivity of the P-norm aggregation failure index to the design variables, the final sensitivity of the objective function to the element density is obtained as follows: The final sensitivity of the objective function to the unit fiber content index is:

[0049] The beneficial effects of this invention are as follows: The topology optimization process for continuous fiber 3D printing described above is based on the fiber mixing ratio criterion of the composite material. Under the constraint of polymerization failure, it simultaneously optimizes fiber content, fiber orientation, and topological configuration, comprehensively considering the optimization scale. This process boasts high accuracy and efficiency. Furthermore, based on this topology-optimized structure, a non-equidistant fiber path planning method is used to generate continuous fiber trajectories, ensuring manufacturability. Attached Figure Description

[0050] Figure 1 This is a flowchart of a topology optimization and fiber path design method for continuous fiber 3D printing according to the present invention.

[0051] Figure 2 This is an abstract schematic diagram of the continuous fiber layout optimization model in this invention;

[0052] Figure 3 for Figure 2 Enlarged view of the fiber phase angle at point A;

[0053] Figure 4 for Figure 3 Enlarged view of the non-equidistant fibers at point B;

[0054] Figure 5 This is a schematic diagram of the printing path planning process for the cantilever beam in this invention;

[0055] 1. Substrate; 2. Fiber; 3. Blank area; 4. Fiber framing angle. Detailed Implementation

[0056] The invention will be further described below with reference to the accompanying drawings and specific embodiments, so that those skilled in the art can implement it based on the description. The scope of protection of the invention is not limited to these specific embodiments.

[0057] This invention provides a method for topology optimization and fiber path design for continuous fiber 3D printing, such as... Figure 1 As shown, the method includes the following steps:

[0058] (1) A continuous fiber layout optimization model is established using finite element software. The design domain of the continuous fiber layout optimization model is discretized into several finite element elements, each of which includes four nodes.

[0059] (2) Initialize the design variables of the continuous fiber layout optimization model using the discretized design domain from step (1). The design variables include the element density ρ. B The orientation angle θ of the unit fiber and the unit fiber content index ρ S The finite element calculation file is output from the continuous fiber layout optimization model;

[0060] (3) Establish a continuous fiber material interpolation model based on the composite material mixing ratio criterion, and output the material card of the continuous fiber material from the continuous fiber material interpolation model; define optimization parameters in MATLAB, and read the finite element calculation file output by the continuous fiber layout optimization model from MATLAB, and obtain the element stress, element displacement and element principal stress angle of each finite element in the continuous fiber material; each element stress includes four nodal stresses and each element displacement includes four nodal displacements;

[0061] (4) The objective function for optimizing continuous fiber composite materials under polymerization failure constraint is set as follows:

[0062]

[0063] Where F represents the optimization objective function of the continuous fiber composite material under the constraint of polymerization failure, C represents the overall flexibility of the continuous fiber composite structure, λ is the Lagrange multiplier, and FI max FI min FI represents the maximum and minimum values ​​of the failure index, respectively. pn The global aggregation of failure indices is represented by U, F, and K, which represent the global displacement, global load vector, and global stiffness matrix of the continuous fiber composite structure, respectively. e and k e ρ represents the element displacement vector and element stiffness matrix, respectively. B ρ represents the element density. S V(x) represents the fiber content index of a unit, θ represents the fiber orientation angle of the unit; V(x) and V * Represent the total volume of the structure and the target volume, respectively; vf and vf *These are the structural fiber content and the target fiber content, respectively.

[0064] (5) Calculate the element compliance of each finite element based on the nodal stress and nodal displacement of each finite element obtained in step (3), and adjust the element compliance of each finite element according to the corresponding element density ρ. B and unit fiber content index ρ S The element density ρ of each finite element is obtained by differentiating the derivatives. B The sensitivity and the element fiber content index ρ of each finite element element S The sensitivity is assessed by performing finite element analysis on the element stiffness matrix and nodal displacements of each finite element to obtain the stress-associated variable values ​​for each finite element. Based on these stress-associated variable values, the element density ρ under the aggregate failure constraint is then calculated. B Sensitivity and unit fiber content index ρ S Sensitivity;

[0065] (6) Perform polymerization on the optimization objective function of the continuous fiber composite material under the polymerization failure constraint set in step (4). The specific process is as follows: for the unit density ρ B The objective function F is iterated several times, and in each iteration, the element density ρ under the aggregation failure constraint calculated in step (5) is adjusted. B The sensitivity was filtered, and the element density ρ in the optimization objective function F of the continuous fiber composite was updated using the BESO method. B In each iteration, the unit fiber content index ρ S The element fiber orientation angle θ remains unchanged and is updated with the element principal stress angle in each iteration. A finite element calculation file is exported for each iteration. The stress results are analyzed based on the exported finite element calculation file. The iteration stops when the stress results meet the failure index constraint or the iteration reaches the maximum cycle value.

[0066] (7) Based on step (6), the optimization objective function of the continuous fiber composite material under the polymerization failure constraint set in step (4) is polymerized again. The specific process is as follows: for the unit fiber content index ρ S The objective function λ is iterated several times. The λ mentioned in (6) is unrelated to the λ in this iteration. In each iteration, the unit fiber content index ρ under the polymerization failure constraint obtained in step (5) is adjusted. S The sensitivity was filtered, and the unit fiber content index ρ in the optimization objective function of continuous fiber composites was updated using the SIMP method. S In each iteration, the unit density ρ BThe element fiber orientation angle θ remains unchanged, and is updated with the element principal stress angle in each iteration. The stress results are analyzed based on the finite element calculation file exported in each iteration. The iteration stops when the stress results meet the failure index constraint or the iteration reaches the maximum cycle value.

[0067] (8) After the iteration is completed, output the finite element calculation file;

[0068] (9) Based on the finite element calculation file output in step (7), a continuous fiber trajectory is generated by using a non-equidistant fiber path planning method.

[0069] The above-mentioned topology optimization and fiber path design method for continuous fiber 3D printing is adopted. The topology optimization in this method is based on the fiber mixing ratio criterion of composite materials. Under the constraint of polymerization failure, the fiber content, fiber orientation and topology configuration are simultaneously optimized, which comprehensively considers the optimization scale and has high accuracy. On the basis of the topology optimization structure, a non-equidistant fiber path planning method is further used to generate continuous fiber trajectories, which has manufacturability.

[0070] Steps (1) to (8) describe a topology optimization framework for continuous fiber 3D printing. This framework is proposed by combining the bidirectional progressive optimization method (BESO) and the variable density method (SIMP) in isotropic topology optimization and extending it to orthotropic anisotropic materials. An abstract schematic diagram of this design framework is shown in the figure. The design domain is discretized using a four-node finite element mesh. The parallel optimization framework simultaneously optimizes multiple design variables under stress constraints. The discrete element density obtained by the BESO method is used as the material distribution, and the continuous fiber content distribution is obtained by the non-penalized SIMP method. The principal stress direction is used as the fiber direction. Next, a multi-design variable material interpolation model is introduced to represent the relationship between the objective function and the variables. Based on maximizing the stiffness of the objective function, a cohesive constraint is added using Lagrange multipliers to prevent failure. The sensitivity derivation of the design variables is given to determine the update scheme of the variables. The optimal structure is found in the given design domain according to the above optimization model. In order to further meet the manufacturing requirements, the printing path of the obtained optimized structure is planned and designed, and printing tests are carried out.

[0071] In the process of topology optimization for continuous fiber 3D printing, the inconsistent fiber density in local areas necessitates careful consideration to avoid structural fracture and material failure caused by high stress values ​​or stress concentration in sparse fiber regions, which severely impacts test results. Therefore, the method described in this invention is of great significance in studying composite material failure constraints. Stress constraints require the structure to satisfy them at all locations, making them highly sensitive to changes in local parameters. Discretized numerical optimization methods require satisfying a large number of constraints, which obviously increases the difficulty of optimization and computational cost. To reduce the number of constraints, these numerous local constraints can be written into a global constraint form, expressed as: Where p represents stress concentration; FI i According to the Tsai-Hill criteria get It should be noted that a larger p can increase the weight of the maximum stress and reduce the chance of local stress exceeding the limit, but it will also make iterative oscillations more likely; σ i Let be the stress vector of the element. In isotropic materials, von Mises stress is usually used as the criterion, while the Tsai-hill criterion is widely used in strength calculations for anisotropic materials. Therefore, the Tsai-hill criterion is used as the criterion in the failure constraints of composite materials in this paper.

[0072] To ensure the manufacturability of the obtained topology optimization results, this invention studies the path planning problem. Current optimization methods often rarely involve the planning and processing of later paths. This invention introduces a fiber filling mode suitable for this topology optimization method to solve the fiber discontinuity after optimization and ensure manufacturability.

[0073] In step (9), the specific process of generating a continuous fiber trajectory using a non-equidistant fiber path planning method based on the finite element calculation file output in step (8) includes the following steps:

[0074] First, the element density ρ of each finite element in the finite element calculation file output in step (8). B All are discrete variables, taking values ​​of 0 or 1, with the element density ρ being the most significant. B The topological configuration is constructed using finite element units of type 1.

[0075] Second, sample points are evenly distributed on the long scale of the topological configuration constructed in step one, and the topological configuration constructed in step one is divided into multiple regions with different numbers of cross-sectional branches according to the cross-sectional size of the sample points; the fiber content index ρ of all finite element elements in the finite element calculation file output in step (8) is then used to calculate the fiber content index ρ of all finite element elements. SThe fiber content v of the finite element within the topological configuration constructed in step one is converted into the fiber content v. f The unit fiber content index ρ S It is a continuous variable, and its value can take any value between 0 and 1; the fiber content v of the finite element within the topological configuration constructed in step one. f The expression is:

[0076] Third, adjacent regions with the same number of cross-sectional branches are merged to obtain several merged branch regions. These merged branch regions then form a new topological configuration, where each merged branch region contains an equal number of cross-sectional branches. The fiber content v of different branches within each merged branch region is then considered. f The fiber count is allocated to obtain the expression for the fiber count in each of the merged branch regions: Among them, R i This represents the number of fibers within any branch region; sum(ρ) Si ∩ρ B =1) represents the element density ρ within the branch region. B The sum of unit fiber content with a value of 1; R represents the total fiber content before branching. total This indicates the total number of fibers before branching;

[0077] Fourth, after the fiber counts of different branches within each merged branch region have been allocated, the fiber points on the sample points are allocated. The specific process is as follows: based on the obtained fiber counts in each branch region and the unit fiber content index ρ before branching... S And the fiber content v of the finite element within the topological configuration obtained in step two. f The fiber points on the sample points are assigned based on the fiber content v within the finite element elements of the new topology obtained in step three, which is present between different fiber points within that branch. f Equal; due to the fiber content v of each finite element within the topological configuration. f They are not equal, therefore non-equidistant fiber points will be obtained;

[0078] Fifth, for a complete fiber, all fiber points on the fiber are filtered according to distance weights, and the final unit fiber orientation angle θ generated in step (7) is obtained to obtain the smoothed fiber direction; the expression for filtering the final unit fiber orientation angle θ is: in, This represents the sum of all distances within the filter radius. Represents the product of distance and fiber angle; θi θ represents the angle of the filtered fibers. j cel indicates the fiber angle within the filtration range. The distance, or weighting factor, is represented by r. sen The filter radius is represented by the value; finally, based on the Hermite interpolation method, the fiber points are used as the coordinates of the sample points, and the smoothed fiber direction is used as the first derivative to obtain the fiber distribution curve under the minimum printing spacing constraint.

[0079] In step (3), the optimization parameters defined in MATLAB include: design variables, maximum cycle value, target volume, failure exponent constraint, and element density ρ. B Filtration radius and unit fiber content index ρ S Filter radius;

[0080] In step (3), the specific process of establishing a continuous fiber material interpolation model based on the composite material mixing ratio criterion and outputting material cards from the continuous fiber material interpolation model includes the following steps:

[0081] (3.1) The fiber content value of continuous composite materials in the composite structure is expressed by the unit fiber content index ρ. S Interpolation is performed to obtain v based on different maximum and minimum print spacings dmax and dmin, respectively, for a given print layer thickness. fmax and v fmin v fmax v represents the maximum fiber content of a continuous composite material within a topological configuration. fmin The minimum fiber content of a continuous composite material within a topological configuration is represented by the fiber content v of the finite element of the continuous composite material within the topological configuration. f For linear interpolation, the expression is: In the formula This represents the unit fiber content index of the i-th unit;

[0082] (3.2) Based on the modulus of the composite material structure and the modulus of the fiber material, and then using the composite material mixing ratio formula, the ρ value is calculated. s The interpolated model for continuous fiber materials is expressed as follows:

[0083] E1 = v f E f +(1-v f E m ,

[0084]

[0085]

[0086] v12 = v f v f +(1-v f )v m

[0087] Among them, E m G m v m These are the elastic modulus, shear modulus, and Poisson's ratio of the matrix material constituting the composite material, respectively; E m G m v m E1, E2, and G represent the elastic modulus, shear modulus, and Poisson's ratio of the fiber material. 12 v 12 These represent the longitudinal elastic modulus, transverse elastic modulus, in-plane shear modulus, and Poisson's ratio of the composite fiber after mixing, respectively.

[0088] (3.3) Introducing the unit fiber orientation angle θ, let m = cosθ, n = sinθ, then the off-axis rotation matrix of the composite material is:

[0089] (3.4) Introducing the element density ρ B , ρ B ∈{0,1}, representing the presence or absence of material, let ρ B Substituting the off-axis rotation matrix of the composite material, we obtain the interpolated continuous fiber material structure model, whose expression is:

[0090]

[0091] When ρ B When ρ is 0, the stress-strain matrix is ​​empty, and at this point, it is a void element with no stiffness contribution; when ρ B When ρ is 1, the material's stiffness matrix depends on the values ​​of other design variables. s Unpenalized interpolation of fiber fraction affects material stiffness, ρ s The punishment is actually caused by ρ B To achieve this, for regions without material, ρ s The change in ρ will not cause a change in the material's stiffness because the stress-strain matrix is ​​empty at this time, so ρ s For ρ BThe element sensitivity of θ approaches 0, thus implementing a penalty process. θ affects the coordinate transformation matrix T, which in turn affects the element stiffness matrix of the composite material. Design methods with completely free angles can lead to too many variables, slow calculation speed, and even failure to obtain calculation results. Therefore, considering that the results of topology optimization typically represent truss structures, the stress situation of branch members along the principal stress is more reasonable. Therefore, during the calculation process, the fiber orientation θ is set to the principal stress direction of the previous calculation result after each iteration.

[0092] In step (5), the corresponding element density ρ is determined according to the element compliance. B Differentiation yields the element density ρ B The sensitivity is obtained using the adjoint method, and its expression is: in, This represents the stiffness matrix of the i-th element; based on the element flexibility, the corresponding element fiber content ρ... S Differentiation yields the unit fiber content ρ S The sensitivity is obtained using the adjoint method, and its expression is: in, B i T i D i Let represent the stiffness matrix, strain-displacement matrix, off-axis matrix, and constitutive matrix of the i-th element, respectively.

[0093] In step (5), the corresponding stress-associated variable values ​​are calculated based on the element stiffness matrix and nodal displacements of each element, and then the stress-associated variable values ​​are... value Obtain the unit density ρ under the constraint of polymerization failure. B and unit fiber content ρ S The specific process for determining sensitivity includes the following steps:

[0094] (5.1) Obtain the global aggregate FI of the failure index PN For ρ Bi The sensitivity of the α is expressed as: in, V is the failure index matrix, σ i Let X be the element stress vector, where X represents the tensile strength along the fiber direction and Y represents the tensile strength perpendicular to the fiber direction. Indicates in-plane shear strength;

[0095] (5.2) Differentiating the failure index of the i-th element with respect to stress, its expression is:

[0096] (5.3) σi =D0B i u i Substituting the expression from step (4.2), we get: Among them, L i Satisfy u i =L i U, u i B represents the element displacement vector. i D represents the shape function. i Denotes the constitutive matrix;

[0097] (5.4) Based on the overall equilibrium equation of the finite element method, the expression is obtained as follows:

[0098] (5.5) Substituting the expression in step (5.3) into the expression in step (5.4), we get:

[0099] (5.6) By introducing the adjoint variable μ into the expression in step (5.5) Bi The expression obtained is: Where K represents the overall structural stiffness matrix;

[0100] (5.7) Finally, based on the expression in step (5.6), the P-norm polymerization stress sensitivity is simplified to: because: Finally, the P-norm aggregation failure constraint on the design variable ρ is obtained. Bi The expression for the sensitivity is: Similarly, the P-norm aggregation failure constraint on the design variable ρ can be obtained. Si The expression for the sensitivity is: in Let be the unit fiber content index associated node value vector of the i-th element;

[0101] (5.8) Combining the sensitivity of the aggregation failure constraint to the design variables obtained in (5.7) with the sensitivity of the P-norm aggregation failure index to the design variables, the final sensitivity of the objective function to the element density is obtained as follows: The final sensitivity of the objective function to the unit fiber content index is:

[0102] Generating continuous fiber trajectories using a non-equidistant fiber path planning method is based on ρ in the topology optimization results. S With ρ BBased on this, the results are divided into multiple branch regions, and continuous fiber paths within the optimized result range are created in an unequal-distance manner. In each structure, fiber paths can be automatically generated according to the number of input fibers and sample points. The fiber path arrangement of a cantilever beam is used as an example below.

[0103] Taking the optimization results of a cantilever beam as an example, such as Figure 3 As shown, its path planning was designed. The optimization result ρ... S With ρ B Based on this, continuous paths are placed along the maximum dimension of the structure (x-direction in this example); based on boundary conditions, starting from the final topological region, the topological configuration is divided into multiple branch regions according to the perpendicular lines of the sample points on the maximum dimension, and the fiber points are arranged along the cross-section of each sample point on the topological configuration according to ρ. S The values ​​are not evenly distributed. Finally, these points are smoothly connected along the direction of maximum dimension.

[0104] This method first divides the optimization results into several cross-sections by equidistantly distributing sample points on a long scale. Based on the unit density variable results fed back from the cross-sections, the optimized macroscopic geometric features are then divided into multiple regions according to the sample point cross-sections. The structural changes of the fiber are transformed into changes in the number of branches, and the fibers are branched and merged based on these changes.

[0105] When branching, design variable ρ S Converted to fiber content v f The total number of fibers before branching and the fiber content v of each branch f The results were used to assign fiber numbers to the branches using fiber content as a weighting factor; the number of fibers in each branch was: Among them, R i R represents the number of fibers in this branch. total This represents the total number of fibers before branching. When merging, the total number of fibers after merging is determined based on the fibers of each branch to be merged. This method ensures that there is no cross-linking between fibers.

[0106] After calculating the total number of fibers to be allocated to each branch, based on the sample points of each section within the branch, proceed section by section according to the method mentioned above. With ρ S Calculate v f The fiber spacing within each branch is not evenly distributed, resulting in non-equidistant fiber sample points.

[0107] Then, a filter radius is applied around the assigned fiber point, and the fiber direction of this fiber point is taken as the average of the fiber directions of all units with a density of 1 within the filter radius, i.e.:

[0108] For a single fiber, which is composed of a series of points connected along its long dimension, to ensure a smooth result, similar to sensitivity filtering, the fiber orientation is filtered by measuring distance for all points on the fiber. The expression for filtering the optimized fiber angle θ is as follows: The purpose of this filtration is to counteract the effects of regions with complex fiber orientation variations, thereby obtaining a smooth curve, such as... Figure 4 As shown.

[0109] Finally, using the Hermite interpolation method, the fiber points are used as sample point coordinates, and the smoothed fiber direction is used as the first derivative to obtain the fiber distribution curve under the minimum printing spacing constraint.

Claims

1. A method for topology optimization and fiber path design for continuous fiber 3D printing, characterized in that: The method includes the following steps: (1) A continuous fiber layout optimization model is established using finite element software. The design domain of the continuous fiber layout optimization model is discretized into several finite element elements, each of which includes four nodes. (2) initializing design variables of the continuous fiber layout optimization model by using the discrete design domains in step (1), the design variables including cell density , cell fiber orientation angle , and cell fiber content index , outputting a finite element calculation file from the continuous fiber layout optimization model; (3) Establish a continuous fiber material interpolation model based on the composite material mixing ratio criterion, and output the material card of the continuous fiber material from the continuous fiber material interpolation model; define optimization parameters in MATLAB, and read the finite element calculation file output by the continuous fiber layout optimization model from MATLAB, and obtain the element stress, element displacement and element principal stress angle of each finite element in the continuous fiber material; each element stress includes four nodal stresses and each element displacement includes four nodal displacements; (4) The objective function for optimizing continuous fiber composite materials under the constraint of polymerization failure is set as follows: ; Where F represents the optimization objective function of the continuous fiber composite material under the constraint of polymerization failure, and C represents the overall flexibility of the continuous fiber composite material structure. For Lagrange multipliers, , These represent the maximum and minimum values ​​within the range of the failure index, respectively. The global aggregation of failure indices is represented by U, F, and K, which represent the overall displacement, overall load vector, and overall stiffness matrix of the continuous fiber composite structure, respectively. and These represent the element displacement vector and the element stiffness matrix, respectively. Represents unit density, Indicates the unit fiber content index. Indicates the fiber orientation angle of the unit; and These represent the total structural volume and the target volume, respectively. and These are the structural fiber content and the target fiber content, respectively. (5) Calculate the element compliance of each finite element based on the nodal stress and nodal displacement of each finite element obtained in step (3), and adjust the element density according to the obtained element compliance of each finite element. and unit fiber content index The element density of each finite element is obtained by taking the derivative separately. Sensitivity and fiber content index of each finite element element The sensitivity is assessed by performing finite element analysis on the element stiffness matrix and nodal displacements of each finite element to obtain the stress-associated variable values ​​for each element. Based on these stress-associated variable values, the element density under aggregate failure constraints is then calculated. Sensitivity and unit fiber content index Sensitivity; (6) Perform polymerization on the optimization objective function of the continuous fiber composite material under the polymerization failure constraint set in step (4). The specific process is as follows: for the unit density The objective function F is iterated several times, and in each iteration, the element density under the aggregation failure constraint calculated in step (5) is adjusted. The sensitivity was filtered, and the element density in the optimization objective function F of the continuous fiber composite was updated using the BESO method. The unit fiber content index during each iteration The unit fiber orientation angle remains unchanged. The element principal stress angle is updated during each iteration. A finite element calculation file is exported for each iteration. The stress results are analyzed based on the exported finite element calculation file. The iteration stops when the stress results meet the failure index constraint or the iteration reaches the maximum cycle value. (7) Based on step (6), the optimization objective function of the continuous fiber composite material under the polymerization failure constraint set in step (4) is polymerized again. The specific process is as follows: for the unit fiber content index The objective function λ is iterated several times. In each iteration, the fiber content index of the unit cell under the polymerization failure constraint obtained in step (5) is adjusted. The sensitivity was filtered, and the unit fiber content index in the optimization objective function of continuous fiber composites was updated using the SIMP method. Cell density in each iteration The unit fiber orientation angle remains unchanged. The stress is updated with the principal stress angle of the element during each iteration. The stress results are analyzed based on the finite element calculation file exported from each iteration. The iteration stops when the stress results meet the failure index constraint or the iteration reaches the maximum cycle value. (8) After the iteration is completed, output the finite element calculation file; (9) Based on the finite element calculation file output in step (7), a continuous fiber trajectory is generated by using a non-equidistant fiber path planning method.

2. The topology optimization and fiber path design method for continuous fiber 3D printing according to claim 1, characterized in that: In step (3), the optimization parameters defined in MATLAB include: design variables, maximum cycle value, target volume, failure index constraint, and element density. Filtration radius and unit fiber content index The filtration radius.