Long fiber composite material performance prediction method based on discrete entity unit
By establishing microscopic and mesoscopic RVE models based on discrete solid elements and combining them with finite element analysis, the problem of describing the microstructure of long fiber composite materials was solved, and high-precision macroscopic performance prediction was achieved.
Patent Information
- Application Number
- CN202511071511.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-08-01
- Publication Date
- 2025-11-21
AI Technical Summary
Existing modeling and analysis methods are insufficient to accurately describe the microstructure of long fiber composites, making it difficult to accurately predict their properties.
A discrete solid element-based approach is adopted to establish micro and mesoscopic RVE models by collecting model parameters. Combined with finite element analysis, the macroscopic performance parameters of long fiber composites are predicted. A microscopic model with high fiber volume content is generated by using unit cell array and fiber repositioning technology, and a mesoscopic model is generated by randomly placing prepreg sheets.
It achieves high-precision prediction of the macroscopic properties of long fiber composite materials, taking into account microstructural features such as fiber diameter, length, and orientation tensor, reducing computational resource requirements and providing an efficient numerical analysis method.
Smart Images
Figure CN120998367A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The application belongs to the field of composite material simulation calculation, and relates to a long fiber composite material performance prediction method based on discrete entity elements. BACKGROUND
[0002] In recent years, discontinuous long fiber reinforced composites have been paid more and more attention and applied in modern industries such as aerospace, automobile, national defense, etc. It is easy to form complex components, and at the same time, it can realize very high fiber volume fraction (50%-60%), so it has very good mechanical properties. This characteristic makes up for the shortcomings of continuous fiber composites and short fiber composites. It is of great significance to establish a model capable of accurately predicting the performance of this material to promote its engineering application. Long fiber composites are formed by cutting pre-impregnated materials into long rectangular sheets and then molding. Therefore, at the mesoscale, long fiber composites are composed of individual pre-impregnated material sheets. And at the microscale, the pre-impregnated material sheet is composed of fibers and a resin matrix.
[0003] However, long fiber composites have a very complex microstructure, including fiber diameter and length, random orientation of fibers, fiber volume fraction, size of pre-impregnated material sheets, etc. Therefore, how to accurately describe its microstructure is the key to accurately predicting its performance, which poses a challenge to existing modeling and analysis methods. SUMMARY
[0004] In view of the above defects or improvement needs of the prior art, the application provides a long fiber composite material performance prediction method based on discrete entity elements. The method only needs to determine the material parameters of the fiber and the resin matrix, the volume fraction of the fiber, the size of the pre-impregnated material sheet, the orientation tensor of the fiber, etc. Physical quantities, based on the sequential multiscale method, the macroscopic performance parameters of long fiber composites with high fiber volume content can be obtained by performing finite element progressive damage analysis. At the same time, the application adopts standard hexahedral elements for the mesoscopic RVE model, reduces the number of node degrees of freedom, and can realize high-efficiency and high-precision macroscopic performance prediction of long fiber composites.
[0005] The technical scheme adopted by the application is as follows:
[0006] A long fiber composite material performance prediction method based on discrete entity elements, comprising: first, collecting model parameters; then, based on the cell array and fiber relocation, establishing a micro-representative volume element (RVE) model of long fiber composites, and predicting the equivalent performance parameters of the pre-impregnated material sheet; finally, by simulating the random placement of the pre-impregnated material sheet, establishing a mesoscopic RVE model, and combining the equivalent performance parameters of the pre-impregnated material sheet, predicting the equivalent macroscopic performance parameters of the long fiber composite material. Specifically, the steps include:
[0007] Step 1, collect model parameters, including micro-RVE model size, meso-RVE model size (length, width and height), fiber diameter d f , fiber volume fraction V f .
[0008] Step 2, based on the unit cell array and fiber relocation, establish a micro-RVE model.
[0009] According to the fiber diameter d f and the fiber volume fraction V f , the size a of a single square unit cell is calculated as shown in equation (1):
[0010]
[0011] Randomly place fibers within a single unit cell, and array the fiber-placed unit cell within the entire micro-RVE model area.
[0012] Relocate each fiber in the micro-RVE model area and iteratively optimize the placement position to avoid interference to achieve random distribution of fibers. Specifically, randomly generate new position coordinates for each fiber in the micro-RVE model area, determine whether fiber interference occurs between any two fibers at the new position coordinates, if fiber interference occurs, randomly generate new position coordinates for the interfering fibers again and determine whether fiber interference occurs again, if no fiber interference occurs, place the corresponding fiber at the new position coordinates, and iterate and optimize in this way until all fibers in the micro-RVE model area do not interfere; if a certain fiber cannot find new position coordinates after reaching a given maximum number of iterations, the fiber remains in the original position.
[0013] Determine the failure criteria and stiffness damage factors of fibers and resin matrix respectively:
[0014] 1) For fibers, since they are generally brittle materials, the maximum stress criterion is used as the failure criterion, and the initial failure of the fiber occurs when the axial stress of the fiber reaches the corresponding strength.
[0015] The expression of the fiber failure criterion is:
[0016] -X fc <σ x <X ft (2)
[0017] In the formula: σ x is the axial stress of the fiber, X ft and X fc correspond to the tensile strength and compressive strength of the fiber respectively.
[0018] The stiffness damage factor of the fiber is defined as follows:
[0019]
[0020] When the stiffness damage factor of the fiber d1 = 1, it represents that the corresponding fiber unit is damaged, and the corresponding fiber unit is deleted.
[0021] 2) For the resin matrix, its failure mode is closely related to the load it bears. When subjected to tensile load, the resin matrix usually exhibits brittle failure, at which time its failure criterion and stiffness damage factor definition are the same as those of the fiber; when subjected to compression or shear load, the resin matrix usually exhibits elastic-plastic failure, at which time it needs to be considered as an isotropic elastic-plastic material.
[0022] The yield criterion of the resin matrix when subjected to compression or shear load is defined by using the generalized Mises yield criterion (Mises criterion), as shown in formula (4):
[0023]
[0024] In the formula: is the Mises equivalent stress, I1≡tr(σ) is the first invariant of the stress tensor, S≡σ-I1 / 3I is the deviatoric stress tensor, σ is the stress tensor, and I is the Kronecker constant, and are the tensile strength constant and the compression strength constant of the resin matrix material.
[0025] When subjected to compression or shear load, the resin matrix does not fail immediately after reaching the yield criterion, but enters the yield state, thereby causing redistribution of stress, and then the stiffness damage factor d m is defined according to formula (5):
[0026]
[0027] Periodic boundary conditions and corresponding displacement loads are applied to the micro RVE model, the failure criteria and stiffness damage factors of the fiber and the resin matrix are set, the finite element analysis of progressive failure is carried out in Abaqus, the equivalent stress-strain curve is obtained, and thus the equivalent performance parameters of the prepreg sheet can be obtained.
[0028] Step 3, a micro RVE model is established by simulating the random placement of the prepreg sheet.
[0029] Step 3.1, a coordinate system is established with the length direction of the mesoscopic RVE model as the X axis, the width direction as the Y axis, and the height direction as the Z axis, and an initial RVE model is generated in the coordinate system according to the given size of the mesoscopic RVE model; the mesoscopic RVE model is partitioned according to the set size of the hexahedral element, with the hexahedral element as the partition unit.
[0030] Step 3.2, the position and orientation angle θ of the prepreg sheet are randomly generated, and the prepreg sheet is placed in the mesoscopic RVE model area in layers, wherein the orientation angle θ is the layup angle of the prepreg sheet, which obeys the second-order orientation tensor A:
[0031] A = diag(t, 1-t, 0) (6)
[0032] Where t represents the probability of the orientation angle θ along the X axis.
[0033] Step 3.3, it is judged whether there is a center point of the partition inside the prepreg sheet, if there is, the number of plies of the partition is increased by one, and the layup angle is the orientation angle θ of the prepreg sheet; repeat step 3.2 to place the prepreg sheet until the average number of plies of all partitions meets the requirements.
[0034] Step 3.4, the failure criterion of the mesoscopic RVE model is determined:
[0035] For the mesoscopic RVE model, each partition has different composite material plies, and the failure mode of this structure is similar to that of laminated composite materials, so the commonly used failure criterion for laminated composite materials is used as the failure criterion for the mesoscopic RVE model, as shown in equation (7):
[0036]
[0037] In the formula: σ 11 represents the stress in the X direction, σ 22 represents the stress in the Y direction, σ 33 represents the stress in the Z direction, σ 12 represents the shear stress in the XY plane, σ 13 represents the shear stress in the XZ plane, σ 23 represents the shear stress in the YZ plane, F ft represents the fiber tensile damage coefficient, F fc represents the fiber compression damage coefficient, F mt represents the resin matrix tensile damage coefficient, F mc represents the resin matrix compression damage coefficient, X T represents the tensile strength of the prepreg sheet in the X direction, X C represents the compression strength of the prepreg sheet in the X direction, Y T represents the tensile strength of the prepreg sheet in the Y direction, Y Crepresents the compressive strength of the prepreg ply in the Y direction, S L represents the shear strength of the prepreg ply in the longitudinal direction (i.e. the direction perpendicular to the plane of the prepreg), S T represents the shear strength of the prepreg ply in the transverse direction (i.e. the direction parallel to the plane of the prepreg).
[0038] Step 3.5, periodic boundary conditions and corresponding displacement loads are applied to the meso-RVE model, the meso-RVE model failure criterion is set, based on the equivalent performance parameters of the prepreg ply obtained in step 1, the finite element analysis of progressive failure is carried out in Abaqus, and the equivalent stress-strain curve is obtained, so that the equivalent macroscopic performance parameters of the long fiber composite material can be obtained.
[0039] Advantages of the present application:
[0040] 1. The present application can generate a micro-RVE model of long fiber composite material with a maximum fiber volume content of 75% by the method of unit cell array and fiber relocation.
[0041] 2. The present application can generate a long fiber composite material meso-RVE model with the same volume fraction as the prepreg by randomly placing prepreg plies to give different layer properties to each partition.
[0042] 3. The multiscale model proposed in the present application can fully consider various microstructure characteristics of long fiber composites when predicting the macroscopic performance parameters of long fiber composites, taking the diameter and length of the fiber, the fiber orientation tensor, the fiber volume fraction, the size of the prepreg ply, the strength of the fiber and the strength of the matrix, etc. as input characteristics, so as to accurately predict different failure modes of the microstructure of long fiber composites, thereby realizing high-precision prediction of the macroscopic performance parameters of long fiber composites.
[0043] 4. The size of the RVE model created by the present application can be freely defined, and standard hexahedral elements are used, reducing the number of node degrees of freedom and saving computing resources, providing an efficient numerical analysis means for performance prediction of long fiber composites. BRIEF DESCRIPTION OF DRAWINGS
[0044] Figure 1 Modeling algorithm for micro-RVE model.
[0045] Figure 2 Unit cell array for micro-RVE model.
[0046] Figure 3 Partitioning diagram for meso-RVE model.
[0047] Figure 4The X-direction stress-strain curve obtained by micro RVE model finite element analysis, wherein (a) is the tensile stress-strain curve, and (b) is the compression stress-strain curve.
[0048] Figure 5 The longitudinal tensile stress-strain curve obtained by micro RVE model finite element analysis. DETAILED DESCRIPTION
[0049] The technical solutions of the present application are further described below in conjunction with examples, but are not limited thereto, and any modification or equivalent replacement of the technical solutions of the present application without departing from the spirit and scope of the technical solutions of the present application shall be encompassed in the protection scope of the present application.
[0050] The present application is further described below in conjunction with an example of a certain long fiber composite material, comprising the following steps:
[0051] Step 1, collect model parameters, the long fiber composite material is cut into a 12.5x25x0.1mm rectangular sheet from T1100 carbon fiber / HJ220 epoxy resin prepreg, the size of the micro RVE model is 70x70x14μm, the size of the micro RVE model is 100x50x3mm, the diameter of T1100 carbon fiber is 7μm, the fiber volume fraction is 40%, and the material properties of the fiber and the resin matrix are shown in Table 1.
[0052] Table 1 Properties of component materials
[0053] Parameter T1100 Carbon fiber HJ220 Epoxy resin Elastic modulus E (GPa) 330 4.0 Poisson's ratio v 0.3 0.35 Xt (MPa) 5700 75 Xc (MPa) 2990 150 S (MPa) - 70
[0054] Step 2, establish a micro RVE model based on a unit cell array and fiber relocation.
[0055] As Figure 1 , the size a of a single square unit cell is calculated according to the fiber diameter d f and the fiber volume fraction V f , as shown in formula (1). The fiber is randomly placed in the single unit cell, and the unit cell with the placed fiber is arrayed in the entire micro RVE model area, as Figure 2 .
[0056] The fibers in the micro RVE model region are relocated, and the placement position is iteratively optimized to avoid interference to realize the random distribution of the fibers. Specifically, new position coordinates are randomly generated for each fiber in the micro RVE model region, it is judged whether fiber interference occurs between any two fibers at the new position coordinates, if fiber interference occurs, new position coordinates are randomly generated again for the interfering fibers and it is judged again whether fiber interference occurs, if no fiber interference occurs, the corresponding fiber is placed at the new position coordinates, and the iteration optimization is performed in this way until all the fibers in the micro RVE model region do not interfere; if a certain fiber still cannot find new position coordinates after reaching a given maximum number of iterations, the fiber remains in the original position.
[0057] The failure criteria and stiffness damage factors of the fibers and the resin matrix are determined respectively:
[0058] 1) For the fibers, since they are generally brittle materials, the maximum stress criterion is adopted as the failure criterion, and the initial failure of the fibers occurs when the axial stress of the fibers reaches the corresponding strength. The expression of the fiber failure criterion is as formula (2). The stiffness damage factor of the fiber is as formula (3).
[0059] 2) For the resin matrix, its failure mode is closely related to the load it bears. When subjected to tensile load, the resin matrix usually exhibits brittle failure, at this time, its failure criterion and stiffness damage factor are defined the same as the fibers; when subjected to compression or shear load, the resin matrix usually exhibits elastic-plastic failure, at this time, it needs to be considered as an isotropic elastic-plastic material.
[0060] The yield criterion of the resin matrix when subjected to compression or shear load is shown as formula (4).
[0061] When subjected to compression or shear load, the resin matrix does not fail immediately after reaching the yield criterion, but enters the yield state, thereby causing stress redistribution, and the stiffness damage factor d m is defined according to formula (5).
[0062] The micro RVE model is subjected to periodic boundary conditions and corresponding displacement load, the failure criteria and stiffness damage factors of the fibers and the resin matrix are set, and the finite element analysis of progressive failure is performed in Abaqus to obtain the equivalent stress-strain curve (such as Figure 4 ), so that the equivalent performance parameters of the prepreg sheet can be obtained, as shown in Table 2:
[0063] Table 2 Equivalent performance parameters of the prepreg sheet
[0064]
[0065]
[0066] Step 3, a meso RVE model is established by simulating the random placement of prepreg sheets.
[0067] Step 3.1, a coordinate system is established with the length direction of the meso RVE model as the X axis, the width direction as the Y axis, and the height direction as the Z axis. An initial RVE model is generated in the coordinate system according to the given size of the meso RVE model. The meso RVE model is partitioned with a hexahedral element as the partition unit, and the number of partitions is 800, as shown in Figure 3 .
[0068] Step 3.2, the position and orientation angle θ of the prepreg sheet are randomly generated and placed in the meso RVE model area. The orientation angle θ is the layup angle of the prepreg sheet, which obeys the second-order orientation tensor A = diag(0.55, 0.45, 0).
[0069] Step 3.3, it is judged whether there is a center point of the partition inside the prepreg sheet. If there is, the number of plies of the partition is increased by one. Repeat step 3.2 to place the prepreg sheet until the average number of plies of all partitions meets the requirements.
[0070] Step 3.4, the failure criterion of the meso RVE model is determined:
[0071] For the meso RVE model, each partition has different composite layups. The failure mode of this structure is similar to that of laminated composites, so the commonly used failure criterion for laminated composites is used as the failure criterion for the meso RVE model, as shown in equation (7).
[0072] Step 3.5, periodic boundary conditions and corresponding displacement loads are applied to the meso RVE model, and the meso RVE model failure criterion is set. Based on the equivalent performance parameters of the prepreg sheet obtained in step 1, the finite element analysis of progressive failure is carried out in Abaqus, and the equivalent stress-strain curve is obtained, as shown in Figure 5 , and the equivalent macroscopic performance parameters of long fiber composites can be obtained, as shown in Table 3.
[0073] Table 3 Equivalent macroscopic performance parameters of long fiber composites
[0074] Parameter Value Parameter Value Elastic modulus in X direction (GPa) 53.4 Tensile strength in X direction / MPa 268 Elastic modulus in Y direction (GPa) 43.6 Compressive strength in X direction / MPa 215 Elastic modulus in Z direction (GPa) 12.60 Tensile strength in Y direction / MPa 182 Poisson's ratio in XY direction 0.362 Compressive strength in Y direction / MPa 156 Poisson's ratio in XZ direction 0.283 Shear strength in XY direction / MPa 141 Poisson's ratio in YZ direction 0.308 Shear strength in YZ direction / MPa 10 Shear modulus in XY direction (GPa) 17.57 Shear modulus in XZ direction (GPa) 3.70 Shear modulus in YZ direction (GPa) 3.67
Claims
1. A method for predicting the properties of long fiber composites based on discrete entity elements, characterized by, The application relates to a method for predicting the equivalent macroscopic performance parameters of long-fiber composites. Step 1, collect model parameters, including micro RVE model size, meso RVE model size, fiber diameter d f , fiber volume fraction V f , prepreg ply size; The method comprises the following steps: Step 2, establishing a micro RVE model of long-fiber composites based on a cell array and fiber repositioning, and predicting equivalent performance parameters of prepreg sheets; 2. The method of claim 1, wherein, Step 3, establishing a micro RVE model by simulating random placement of the prepreg sheets, combining the equivalent performance parameters of the prepreg sheets, and predicting equivalent macroscopic performance parameters of long-fiber composites. The step 2 is specifically as follows: The size a of a single square cell is calculated; fibers are randomly placed in the single cell, the cell with the placed fibers is arrayed in the whole micro RVE model area, fibers in the micro RVE model area are repositioned, and the placement position is iteratively optimized to avoid interference; Failure criteria and stiffness damage factors of the fibers and the resin matrix are respectively determined; 3. The method of claim 2, wherein the long fiber composite material performance prediction method is based on discrete entity units. According to the fiber diameter d f and the fiber volume fraction V f The size a of the individual square cells is calculated as shown in equation (1):
4. The method of claim 2, wherein the long fiber composite material performance prediction method based on discrete entity units is characterized by, Periodic boundary conditions and corresponding displacement loads are applied to the micro RVE model, the failure criteria and the stiffness damage factors of the fibers and the resin matrix are set, progressive failure finite element analysis is carried out, equivalent stress-strain curves are obtained, and thus the equivalent performance parameters of the prepreg sheets can be obtained.
5. The method of claim 4, wherein the long fiber composite material performance prediction method based on discrete entity units is characterized by, The fibers in the micro RVE model area are repositioned, and the placement position is iteratively optimized to avoid interference, and the specific process is as follows: new position coordinates of each fiber in the micro RVE model area are randomly generated, whether fiber interference occurs between any two fibers in the new position coordinates is judged, if fiber interference occurs, new position coordinates of the fibers with interference are randomly generated again and whether fiber interference occurs is judged again, if no fiber interference occurs, the corresponding fibers are placed in the new position coordinates, and the iterative optimization is continued until all the fibers in the micro RVE model area do not interfere.
6. The method of claim 2, wherein, If a fiber still cannot find new position coordinates after reaching a given maximum number of iterations, the fiber remains in the original position. The failure criteria and the stiffness damage factors of the fibers are as follows: The maximum stress criterion is adopted as the failure criterion of the fibers, and the initial failure of the fibers occurs when the axial stress of the fibers reaches the corresponding strength; X fc σ x X ft (2) where: σ x is the axial stress of the fiber, X ft and X fc correspond to the tensile and compressive strength of the fiber, respectively. The expression of the fiber failure criterion is as follows: The stiffness damage factor of the fibers is defined as follows:
7. The method of claim 2, wherein the long fiber composite material performance prediction method based on discrete entity units is characterized by, When the stiffness damage factor d1 of the fibers is 1, it represents that the corresponding fiber unit is damaged, and the corresponding fiber unit is deleted. The failure criteria and the stiffness damage factors of the resin matrix are as follows: The failure criteria and the stiffness damage factors of the resin matrix are as follows: wherein: is the Mises equivalent stress, I1≡tr(σ) is the first invariant of the stress tensor, S≡σ-I1 / 3I is the deviatoric stress tensor, σ is the stress tensor, I is the Kronecker constant, and are the tensile and compressive strength constants of the resin matrix material; When subjected to compression or shear loading, the resin matrix does not fail immediately upon reaching the yield criterion, but rather enters a yield state, resulting in a redistribution of stresses, and the stiffness damage factor d m is defined according to equation (5):
8. The method of claim 2, wherein the long fiber composite material performance prediction method is based on discrete entity units. The failure criteria and the stiffness damage factors of the resin matrix are as follows: The failure criteria and the stiffness damage factors of the resin matrix are as follows: The yield criterion of the resin matrix under compression or shear load is defined by using the generalized Mises yield criterion, as shown in formula (4). The step 3 is specifically as follows: Step 3.1, a coordinate system is established with the length direction of the mesoscopic RVE model as the X axis, the width direction as the Y axis and the height direction as the Z axis, and an initial RVE model is generated in the coordinate system according to the given size of the mesoscopic RVE model; and the mesoscopic RVE model is partitioned according to the set size of the hexahedral element, with the hexahedral element as the partition unit; Step 3.2, the position and orientation angle θ of the prepreg sheet are randomly generated and the prepreg sheet is placed in the mesoscopic RVE model area in layers; Step 3.3, it is judged whether there is a center point of the partition inside the prepreg sheet, if yes, the number of plies of the partition is increased by one and the ply angle is the orientation angle θ of the prepreg sheet; the prepreg sheet is placed repeatedly until the average number of plies of all partitions meets the requirement; Step 3.4, the failure criterion of the mesoscopic RVE model is determined; Step 3.5, the periodic boundary condition and the corresponding displacement load are applied to the mesoscopic RVE model, the failure criterion of the mesoscopic RVE model is set, the equivalent performance parameters of the prepreg sheet obtained in step 1 are used for the finite element analysis of progressive failure, and the equivalent stress-strain curve is obtained, so that the equivalent macroscopic performance parameters of the long-fiber composite material can be obtained.
9. The method of claim 2, wherein the long fiber composite material performance prediction method based on discrete entity units is characterized by, In 3.2, the orientation angle θ is subject to a second-order orientation tensor A; A = diag(t, 1-t, 0) (6) Wherein, t represents the probability of the orientation angle θ along the X axis direction.
10. The method of claim 2, wherein the long fiber composite material performance prediction method based on discrete entity units is characterized by, The step 3.4 is specifically: The failure mode of the mesoscopic RVE model is similar to that of the laminated composite material, so the commonly used failure criterion of the laminated composite material is used as the failure criterion of the mesoscopic RVE model, as shown in formula (7): wherein: σ 11 represents stress in the X direction, σ 22 represents stress in the Y direction, σ 33 represents stress in the Z direction, σ 12 represents shear stress in the XY plane, σ 13 represents shear stress in the XZ plane, σ 23 represents shear stress in the YZ plane, F ft represents fiber tensile damage coefficient, F fc represents fiber compressive damage coefficient, F mt represents resin matrix tensile damage coefficient, F mc represents resin matrix compressive damage coefficient, X T represents tensile strength of the prepreg sheet in the X direction, X C represents compressive strength of the prepreg sheet in the X direction, Y T represents tensile strength of the prepreg sheet in the Y direction, Y C represents compressive strength of the prepreg sheet in the Y direction, S L represents shear strength of the prepreg sheet in the longitudinal direction, S T represents shear strength of the prepreg sheet in the transverse direction.