High-precision structure damage prediction method

Abstract

The invention discloses a high-precision structural damage prediction method, which comprises the following steps of: establishing a finite element structural part model for any actual structural part, and constructing a discrete element structural part model in a discrete element space through a double-ball modeling method; calculating a displacement field uF of the finite element structural member model, and calculating a displacement field uD of the discrete element structural member model; comparing the difference between the displacement field uF and the displacement field uD, and iteratively adjusting the microscopic parameters of the discrete element structural member model based on the difference until the difference meets a specified standard; and performing iterative calculation on the adjusted discrete element structural member model through a discrete element method to complete fracture prediction of the structural member.

Description

Technical Field

[0001] The invention relates to the technical field of fracture structure simulation prediction, and in particular to a high-precision structural damage prediction method. Background Art

[0002] Fracture is critical to the safety, reliability and performance of materials, engineering structures and equipment. In order to understand the fracture mechanism and phenomenon, researchers have adopted a variety of methods to study. Classical continuum mechanics is widely used to study the structural mechanical response behavior, but it faces some challenges when dealing with fracture problems. This is because the classical continuum mechanics governing equations involve derivatives, which do not exist at discontinuities (fractures). Phase field theory uses continuous damage functions to estimate the existence of discontinuous surfaces, but it essentially operates in a continuous field framework and does not simulate real fractures. Extended finite element methods enhance traditional finite element methods by introducing a function to incorporate discontinuities into the displacement field. However, when faced with complex crack branching scenarios, a large number of parameters need to be set, and extended finite element methods may encounter challenges. Peridynamics is an alternative to traditional differential equations, using integral equations to circumvent the singularity problem at the crack tip, but problems related to reduced stiffness near the material boundary may occur in peridynamic simulations. Discrete element methods treat materials as discrete entities, where individual blocks or particles follow Newton's second law. By enabling displacement, rotation, sliding, and even separation, the method is able to realistically simulate a variety of phenomena including fracture and large-scale deformations.

[0003] However, the discrete element method still faces some challenges that need to be overcome. The first is the modeling challenge, because there are large errors when using the random particle model to calculate the displacement field. Using a simple uniaxial test to compare the displacement field, such as Figure 1 It can be observed that the displacement field of the general random model is more chaotic than the finite element displacement field. This problem is related to the particle size and modeling method. This is because the discrete element is a numerical method based on the interaction between particles. The larger the particle size, the larger the range of interaction between particles, which leads to a large deviation between the simulation results and the actual situation. However, if the simulated particle size is reduced to the size of a molecule or to the order of 100 million, in this case, the range of interaction between particles will be significantly reduced, making the discrete element simulation closer to the actual situation, and the accuracy of the displacement field will be significantly improved. However, to achieve such a fine-grained simulation, it usually takes an inestimable amount of computing resources and time. Therefore, it is not feasible to reduce the calculation error from the perspective of reducing the particle size and increasing the number of particles.

[0004] The second is the challenge of micro-parameter calibration. Since the micro-parameters of discrete elements, such as stiffness ratio, are different from the macro-parameter Poisson's ratio, and the micro-parameters cannot be obtained through experiments, calibration is required. Calibration initially uses trial and error or empirical methods to calibrate force-displacement curves and micro-parameters, but there are many blind operations and it is not efficient enough. Then the experimental design method can achieve calibration more quickly by solving equations. Compared with the trial and error method and the empirical method, it is more scientific, but the experimental design method is weak in handling nonlinearity. Then machine learning calibration appeared. Although machine learning calibration has strong nonlinear processing capabilities, it requires a large amount of annotated data, which wastes a lot of manpower and time. In addition, the machine learning calibration method focuses on the correspondence of force-displacement curves, but this does not mean the accurate correspondence of displacement fields. The problem of displacement field mismatch is not only closely related to geometric modeling, but also closely related to micro-parameters. Summary of the invention

[0005] In view of the above-mentioned prior art, the present invention aims to provide a high-precision structural damage prediction method, which mainly solves the technical problems existing in the above-mentioned background technology.

[0006] To achieve the above object, the technical solution of the embodiment of the present invention is implemented as follows:

[0007] A high-precision structural damage prediction method, the method comprising the following steps:

[0008] For any actual structural part, a finite element structural part model is established, and a discrete element structural part model is constructed in discrete element space by using a double sphere modeling method;

[0009] Calculate the displacement field u of the finite element structural model F , calculate the displacement field u of the discrete element structural component model D ;

[0010] Compare displacement fields u F and the displacement field u D and iteratively adjusting microscopic parameters of the discrete element structural component model based on the difference until the difference meets a specified standard;

[0011] The adjusted discrete element structural component model is iteratively calculated using the discrete element method to complete the fracture prediction of the structural component.

[0012] Optionally, a discrete element structural component model is constructed by a double-sphere modeling method, specifically including:

[0013] A regional space is enclosed in the discrete element space, and large particles are first formed in the regional space by a circulation function, and then small particles are formed by the circulation function, and a quadrilateral structural model is formed by arranging the large particles and the small particles in the regional space;

[0014] The quadrilateral structural model is used as a structural component embryo, and microscopic parameters are set;

[0015] The contact model is selected for the structural component embryo, and the corresponding contact model microscopic parameters are set to obtain a complete discrete element structural component model file.

[0016] Optionally, the large particles and the small particles satisfy the following relationship: (2) 2 =(R+r) 2 +(+r) 2 ,in Where R is the radius of the large particle and r is the radius of the small particle.

[0017] Optionally, contact between the large particles occurs at a lateral angle of 45°.

[0018] Optionally, based on the difference, iteratively adjusting microscopic parameters of the discrete element structural component model specifically includes:

[0019] Firstly, n groups of microscopic parameters and error limit e are given;

[0020] Calculate the displacement field u of the finite element structural model F and the displacement field u of the discrete element structural component model D The difference between

[0021] If the difference exceeds the specified error limit e, the iterative adjustment phase is entered, during which the micro parameters are updated.

[0022] Optionally, the error limit e is calculated by: Where λ is the coefficient and T is the transposed matrix.

[0023] Optionally, find the minimum value of the error in the displacement field of the structural component between n sets of discrete elements and finite elements, denoted by e min , when e min When it is greater than the error limit e, it enters the iterative adjustment stage.

[0024] Optionally, an update algorithm is used to iteratively adjust until e min Less than the error limit e.

[0025] Optionally, an update algorithm is used for iterative adjustment, specifically comprising the following steps:

[0026] For n sets of microscopic parameter vectors β, the change of the microscopic parameter vector is set to Δβ i , and set the macroscopic parameter coordinate vector B obtained by simulation for each initial microscopic parameter vector β n , its Δβ i The calculation method is as follows:

[0027] Δβ i =[v1v2……v n ]

[0028] Among them, v n is the amount of change of each microscopic parameter;

[0029] Since there is no evolutionary experience at the beginning, all these variables are random, so the new microscopic parameter vector update formula is as follows:

[0030] Z i =β i +Δβ i =[z1z2……z n ]

[0031] where z n is the new evolutionary microscopic parameter, for each z n The macroscopic parameter coordinate vector C is obtained through simulation. n ;

[0032] Use the macro parameter coordinate vector C n and B n Carry out numerical simulations respectively to obtain corresponding multiple result quantities;

[0033] Based on the result quantity, it is judged whether it meets the requirements. If it does not meet the requirements, the macro parameter coordinate vector C is selected. n , B n The microscopic parameter vector corresponding to the minimum value obtained after numerical simulation is taken as the optimal microscopic parameter P gbest , based on the optimal microscopic parameter P gbest Update Z again i ;

[0034] Numerical simulation is performed based on the second update results and the first update results, and the above process is repeated to obtain the optimal microscopic parameters again until the obtained result quantity meets the requirements.

[0035] The beneficial effects of the present invention are as follows: a high-precision structural damage prediction method provided by an embodiment of the present invention, the double-sphere model adopted by the method can effectively simulate the fracture process of the structural member, from the initiation of cracks to their expansion and finally to destruction, and exhibits high accuracy in simulating the deformation of the structural member in the elastic stage, as well as high precision in the elastic stage. The calibration method iteratively adjusts the microscopic parameters so that the displacement field of the discrete element model matches the displacement field of the finite element model as much as possible, thereby improving efficiency and reducing manual intervention. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 A schematic diagram of a flow chart of a high-precision structural damage prediction method in an embodiment of the present application;

[0037] Figure 2 Schematic diagram of the core formula of the evolutionary algorithm in the embodiment of this application

[0038] Figure 3 This is a schematic diagram of the relationship between large particles and small particles in the embodiments of the present application;

[0039] Figure 4 This is a flow chart of the calibration process in the embodiment of the present application;

[0040] Figure 5 Schematic diagram of the x-direction displacement field of the finite element method and the discrete element method in the embodiment of the present application;

[0041] Figure 6 Schematic diagram of the displacement field in the y direction of the finite element method and the discrete element method in the embodiment of the present application;

[0042] Figure 7 This is a schematic diagram of the displacement of cutting line 1 and cutting line 2 in the embodiment of the present application;

[0043] Figure 8 This is a schematic diagram of the displacement of the cutting line 3 in the embodiment of the present application;

[0044] Fig. 9 This is the force chain diagram of the ball model in the embodiment of this application;

[0045] Fig.10 Schematic diagram of the evolution process of the evolutionary algorithm in the embodiment of the present application. DETAILED DESCRIPTION

[0046] The technical solution of the present invention is further elaborated in detail below in conjunction with the drawings and specific embodiments of the specification. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by technicians in the technical field of the present invention. The terms used in the specification of the present invention herein are only for the purpose of describing specific embodiments and are not intended to limit the present invention. In the following description, the expression "some embodiments" is related to a subset of all possible embodiments, but it should be understood that "some embodiments" can be the same subset or different subsets of all possible embodiments, and can be combined with each other without conflict.

[0047] In the following description, a large number of specific details are provided to provide a more thorough understanding of the present invention. However, it is apparent to those skilled in the art that the present invention can be implemented without one or more of these details. In other examples, in order to avoid confusion with the present invention, some technical features well known in the art are not described.

[0048] It should be understood that the present invention can be implemented in different forms and should not be interpreted as being limited to the embodiments proposed herein. On the contrary, providing these embodiments will make the disclosure thorough and complete, and the scope of the present invention will be fully conveyed to those skilled in the art. And the purpose of the terms used herein is only to describe specific embodiments and is not intended to be a limitation of the present invention. When used herein, the singular forms of "one", "one" and "said / the" are also intended to include plural forms, unless the context clearly indicates another way. It should also be understood that the terms "compose" and / or "include" when used in this specification determine the presence of the features, integers, steps, operations, elements and / or parts, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, parts and / or groups. When used herein, the term "and / or" includes any and all combinations of the relevant listed items.

[0049] It should also be noted that when an element is referred to as being "fixed to" another element, it may be directly on the other element or there may be a central element. When an element is considered to be "connected to" another element, it may be directly connected to the other element or there may be a central element at the same time. The terms "vertical", "horizontal", "inside", "outside", "left", "right" and similar expressions used herein are for illustrative purposes only and are not intended to be the only implementation method.

[0050] In order to fully understand the present invention, a detailed structure will be proposed in the following description to illustrate the technical solution proposed by the present invention. The optional embodiments of the present invention are described in detail as follows, but in addition to these detailed descriptions, the present invention may also have other implementations.

[0051] Please refer to the attached Figures 1 to 4 The present application provides a high-precision structural damage prediction method, the method comprising the following steps:

[0052] S1. For any actual structural component, a finite element structural component model is established, and a discrete element structural component model is constructed in a discrete element space by a double sphere modeling method;

[0053] S2. Calculate the displacement field u of the finite element structural model F , calculate the displacement field u of the discrete element structural component model D ;

[0054] S3. Comparison of displacement field u F and the displacement field u D and iteratively adjusting microscopic parameters of the discrete element structural component model based on the difference until the difference meets a specified standard;

[0055] S4. The adjusted discrete element structural component model is iteratively calculated by the discrete element method to complete the fracture prediction of the structural component.

[0056] In a possible implementation, a discrete element structural component model is constructed by a double-sphere modeling method. This method can achieve relatively high-precision calculations using relatively few particles, and can simulate the initiation and propagation of cracks in a structural component under load until the structural component is destroyed. Specifically, it includes:

[0057] A regional space is enclosed in the discrete element space, and large particles are first formed in the regional space by a circulation function, and then small particles are formed by the circulation function, and a quadrilateral structural model is formed by arranging the large particles and the small particles in the regional space;

[0058] The quadrilateral structural model is used as a structural component embryo, and microscopic parameters are set;

[0059] The contact model is selected for the structural component embryo, and the corresponding contact model microscopic parameters are set to obtain a complete discrete element structural component model file.

[0060] Furthermore, the large particles and the small particles satisfy the following relationship: (2) 2 =(+r) 2 +(R+r) 2 ,in Where R is the radius of the large particle and r is the radius of the small particle. The contact between the large particles occurs at a lateral angle of 45°. The small particles are surrounded by the large particles and there is no direct contact between the small particles.

[0061] For example, when constructing, first set the effective space, then set four walls to form a square structure, and then arrange the positions of large particles in the walls. Since the double-ball model has a specific positional relationship, large particles can be quickly formed through a loop function. Similarly, small particles are generated in a loop, and finally a quadrilateral structural model formed by the arrangement of double balls is obtained. For example, a square, the square structural model is the embryo of the structural part, and at this time, the microscopic parameters such as stiffness ratio and effective modulus are set for the embryo of the structural part;

[0062] After obtaining the square structural part embryo, it is necessary to pre-press it so that the particles keep in contact with each other, and then select the contact model for the structural part embryo, such as the parallel bonding contact model, and then set the microscopic parameters of the parallel bonding contact model. Then delete the wall, and then you can carve the structural part embryo. The shape that can be carved is such as a hollow plate structural part or a Brazilian disc with prefabricated cracks. You only need to delete the redundant particles using the command, and the structural part model constructed by the double ball arrangement is now built.

[0063] After solving the structural component model problem, the next step is to calibrate the microscopic parameters of the structural component. When looking at the specific implementation of microscopic parameter calibration, a finite element structural component is established, and then the macroscopic parameters of the calibration target such as elastic modulus and Poisson's ratio are given, and then the displacement field of the finite element structural component model is simulated and calculated. For example, when the structural component is a hollow plate, the displacement field can be a cutting line at one-quarter or one-half of the boundary of the hollow plate, recorded as u F .

[0064] Then, the microscopic parameters of the discrete element are randomly selected and the displacement field of the same-position cutting line of the discrete element structural component model is calculated, denoted as u D Next, the difference between the displacement field of the finite element structural component model and the displacement field of the discrete element structural component model is compared. If the requirements are not met, the microscopic parameters of the structural component will be iteratively adjusted until the error meets the specified standard. The entire process uses evolutionary algorithms to achieve adaptive calibration. Due to full automation, manual work is reduced and efficiency is improved.

[0065] In a possible implementation, based on the difference, iteratively adjusting the microscopic parameters of the discrete element structural component model specifically includes:

[0066] First, given n sets of microscopic parameters and the error limit e, the error limit e is calculated by the following formula:

[0067]

[0068] Where λ is a coefficient, which ranges from 0 to 1, T is a transposed matrix, (u F -λu F ) represents the vector of the displacement field of the cutting line of the finite element structural component minus a coefficient multiplied by the displacement field of the cutting line of the finite element structural component. The meaning of this vector is to assume that the error of all calculated displacement fields of the structural component is (1-λ). Then, multiply this vector by its transpose and take the square root to get the second-order norm of the error vector of the assumed structural component displacement field. If the error is large, λ is small; if the error is small, λ is large.

[0069] Then, under n sets of microscopic parameters, the displacement field u of the finite element structural model is calculated. F and the displacement field u of the discrete element structural component model D The difference between the n sets of discrete elements and the finite element displacement field of the structural component is found and the minimum value among the n sets of errors is recorded as e min , when e min When it is greater than the error limit e, it enters the iterative adjustment stage, and the calculation formula of the difference is as follows;

[0070]

[0071] If the difference exceeds the specified error limit e, enter the iterative adjustment stage, and update the microscopic parameters during the iterative adjustment process, where the evolutionary algorithm is used for iterative adjustment until e min is less than the error limit e.

[0072] In the above formula, (u F - u D ) represents the error vector obtained by subtracting the displacement field of the discrete element structural member from the displacement field of the finite element structural member. The errors of each displacement field in this vector are likely to be different. To facilitate comparison and selection of the optimal microscopic parameter set, it is necessary to multiply this vector by its transpose and then take the square root to obtain the second-order norm of the displacement field error vector of the discrete element structural member. This second-order norm represents a criterion for evaluating the calculation accuracy of the structural member displacement field and can be compared with the error limit e. If e min < e, it represents successful calibration. If not, record this optimal e min . This is the parameter set that calculates the structural member displacement field closest to the finite element structural member displacement field among the n sets of microscopic parameters and can be used as a reference for overall learning of microscopic parameters, and then continue with the next step.

[0073] In the current calibration algorithm, the evolutionary algorithm is the most advanced calibration method. Its core content is to jointly learn and guide evolution globally and locally, and the core formula is as Figure 2 : In Figure 2 , the change in the microscopic parameter vector of the structural member is Δβ i , the individual historical best microscopic parameter P i,pbest , the overall historical best microscopic parameter P gbest , k is the iteration number (starting from 1), w is the inertia learning factor, c1 is the individual learning factor, c2 is the social learning factor, and r1 and r2 are random numbers.

[0074] The principle of calibration by the evolutionary algorithm is as follows: First, assume that there is a macroscopic parameter space, and its dimension is determined by the number of parameters in the macroscopic parameter vector α. For example, if there are two parameters in α, then the macroscopic parameter space will be two-dimensional. Each point in this space corresponds to a unique set of macroscopic parameters, which we call the target value.

[0075] Therefore, for the n sets of microscopic parameter vectors β, a set of macroscopic parameters B n can be obtained for each microscopic parameter in each β vector through numerical simulation experiments. B i to B n is equivalent to a coordinate in the macroscopic space of the macroscopic test calculated from the initial microscopic parameters. However, usually, the coordinates calculated from these n microscopic parameters do not directly match the target value. To solve this problem, it is necessary to update the microscopic parameter vector β.

[0076] Set the change of its microscopic parameter vector to Δβ i , its Δβ i The calculation method is as follows:

[0077] Δβ i =[v1 v2……v n ]

[0078] Among them, v n is the amount of change of each microscopic parameter;

[0079] Since there is no evolutionary experience at the beginning, all these variables are random, so the new microscopic parameter vector update formula is as follows:

[0080] Z i =β i +Δβ i =[z1 z2……z n ]

[0081] where z n is the new evolutionary microscopic parameter, for each z n The macroscopic parameter coordinate vector c is obtained by simulation. n ;

[0082] Use the macro parameter coordinate vector C n and B n Carry out numerical simulations respectively to obtain corresponding multiple result quantities;

[0083] Based on the result quantity, it is judged whether it meets the requirements. If it does not meet the requirements, the macro parameter coordinate vector C is selected. n , B n The microscopic parameter vector corresponding to the minimum value obtained after numerical simulation is taken as the optimal microscopic parameter P gbest , based on the optimal microscopic parameter P gbest Update Z again i ;

[0084] Among them, whether it meets the requirements is determined by the following formula:

[0085]

[0086] σ UCS is the uniaxial compressive strength measured in the laboratory. u is the macroscopic parameter coordinate vector C n , B n The uniaxial compressive strength obtained by numerical simulation. ex is the Young's modulus measured in the laboratory, and u d For the macroscopic parameter coordinate vector C n , B nThe Young's modulus obtained by numerical simulation, that is, the minimum value, must be smaller than the target value, and the setting of the target value is obtained by those skilled in the art based on experience.

[0087] Exemplarily, for the macro-parameter coordinate vectors B1 and C1, numerical simulations are performed on B1 and C1 respectively to obtain the corresponding uniaxial compressive strength and Young's modulus. If the above-mentioned result quantities obtained do not conform to the calculation results of the above formula, the result quantity obtained when the macro-parameter coordinate vector B1 is compared with the result quantity when the macro-parameter coordinate vector C1 is obtained. For example, if the result quantity corresponding to the macro-parameter coordinate vector C1 is smaller than the result quantity when the macro-parameter coordinate vector B1 is obtained, the micro-parameter corresponding to the macro-parameter coordinate vector C1 is used as the local optimal parameter P. i,pbest , P i,pbest =[p p1 p p2 ……p pn ];

[0088] Similarly, for the macroscopic parameter coordinate vectors B1, B2, B3 and C1, C2, C3, numerical simulations are performed on B1, B2, B3, C1, C2, C3 respectively to obtain multiple sets of uniaxial compressive strength and Young's modulus, which are calculated using the above formula. When the above results do not meet the above judgment criteria, the microscopic parameters corresponding to the minimum results among B1, B2, B3, C1, C2, C3 are selected as the global optimal parameters. gbest , P gbest =[p g1 p g2 ……p gn ],based on Figure 2 The formula in updates all Z changes again.

[0089] When the minimum value is less than the target value, the calibration will stop; otherwise, the iteration will continue. Fig.10 As shown, the point marked with an asterisk indicates the location of the target value. i Represents the macro parameters of the initial calculation, the solid circle C i Represents the macro parameters that changed after the first evolution. The solid arrow indicates the direction of the first evolution, while the dotted arrow indicates the final direction reached after several steps of movement, which is also the process of evolution. The overall idea of ​​the evolutionary algorithm is to learn and share the whole and the local together until the target value is reached.

[0090] During the calibration process, the initial values are set as i = 0 and j = 0. Considering there are n groups of microscopic parameters, when i < n, the minimum structural member displacement field error value in the history of the i-th group of microscopic parameters is compared. After each calculation, i is incremented by 1, and then the microscopic parameters in the history of the i-th group that are closest to the finite element structural member displacement field are stored until i = n. The purpose of this process is to obtain the optimal value of each group of microscopic parameters, that is, the individual optimal solution. Similarly, when updating the n groups of microscopic parameters, when j < n, we calculate the j-th group of microscopic parameters and update them. This update process needs to consider both the global optimal solution and the individual optimal solution. After each calculation, j is incremented by 1, and then the updated microscopic parameters are stored until j = n.

[0091] After updating the new microscopic parameters, return to the step of calculating the displacement field of the structural members of the n groups of microscopic parameters. At this time, m = 2, and m is incremented by 1 each time a calculation is performed. Repeat this process until e min <e, at this time, the microscopic parameters are output, indicating the end of calibration, and finally a complete discrete element structural member model is obtained.

[0092] Furthermore, in step S4, the adjusted discrete element structural member model is iteratively calculated by the discrete element method to complete the fracture prediction of the structural members. Substantially, the iterative calculation of the adjusted discrete element structural member model by the discrete element method is to perform numerical simulation of the structural members using discrete element calculation software. In the software, the boundary conditions of the structural members can be set. For example, in the case of a hollow plate structural member for a uniaxial test. At this moment, the top and bottom particles of the hollow plate structural member need to be separately selected, and then the velocities of these particles are fixed. It is necessary to ensure that the velocity is not too fast, otherwise there will be a large error. Then, the termination condition of the structural member simulation needs to be set. For example, the simulation stops when certain particles reach a certain position. If it is desired to better observe the fracture phenomenon of the structural member during the simulation, a fracture file can be introduced. This file is recorded in the discrete element operation manual. To better process the data, an observation circle can be set within the structural member. The observation circle can know the stress and strain changes within this circle. After handling these contents, the structural member simulation can be carried out. After the simulation is completed, some data such as the magnitude of particle displacement or force can be set to output files to form txt files for output, facilitating researchers to extract and analyze the data.

[0093] To verify the accuracy of the double-sphere model. The present invention verifies from two aspects: elastic behavior and fracture behavior. For the verification in the elastic stage, a hollow plate structural member is used for simulation, and three cutting lines are taken in the model. The results are as Figure 5 and Figure 6 shown. Compared with the finite element analysis, the displacement fields of the two structural members of the double-sphere model are very similar. The comparison of the cutting line displacement fields of the structural members is as Figure 7 and Figure 8As shown in the figure, the maximum relative errors are 11.5%, 4.6%, 4.1% and 0.1% respectively. The average maximum relative error of the displacement of the three structural cutting lines is 5.1%. The double-sphere model provides a more accurate and controllable displacement field of the structural part. This verification step proves the high accuracy of the double-sphere model in simulating the deformation of the structural part in the elastic stage.

[0094] Then the present invention applies the double-ball model to the fracture simulation of structural parts. For example, Brazilian disc test, three-point groove test, etc. The simulation results show that: (i) the double-ball model can effectively simulate the entire fracture process of these structural parts, from the initiation of cracks to the expansion and damage. The cracks of the simulated structural parts are very similar to the cracks observed in the experimental test. (ii) The data of the simulated structural parts, such as the maximum tensile force and displacement during loading, are very consistent with the experimental results. Therefore, whether it is elastic behavior or fracture behavior, the calculation results of the double-ball model simulating structural parts are relatively accurate.

[0095] The reason why the double-sphere model is so accurate is that the forces between particles in the double-sphere model can propagate in multiple directions such as up, down, left, and right, which can make the boundary forces be transmitted to the interior of the structure more regularly, thus realizing more complex simulations. The force diagram of the double-sphere model is shown in Fig. 9 shown.

[0096] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. The protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A high-precision structural damage prediction method, characterized in that: The method comprises the following steps: For any actual structural part, a finite element structural part model is established, and a discrete element structural part model is constructed in discrete element space by using a double sphere modeling method; Calculate the displacement field u of the finite element structural model F , calculate the displacement field u of the discrete element structural model D ; Compare displacement fields u F and the displacement field u D and iteratively adjusting microscopic parameters of the discrete element structural component model based on the difference until the difference meets a specified standard; The adjusted discrete element structural component model is iteratively calculated using the discrete element method to complete the fracture prediction of the structural component.

2. A high-precision structural damage prediction method according to claim 1, characterized in that: The discrete element structural component model is constructed by double-sphere modeling method, including: A regional space is enclosed in the discrete element space, and large particles are first formed in the regional space by a circulation function, and then small particles are formed by the circulation function, and a quadrilateral structural model is formed by arranging the large particles and the small particles in the regional space; The quadrilateral structural model is used as a structural component embryo, and microscopic parameters are set; The contact model is selected for the structural component embryo, and the corresponding contact model microscopic parameters are set to obtain a complete discrete element structural component model file.

3. A high-precision structural damage prediction method according to claim 2, characterized in that: The large particles and the small particles satisfy the following relationship: (2R) 2 =(R+r) 2 +(R+r) 2 ,in Where R is the radius of the large particle and r is the radius of the small particle.

4. A high-precision structural damage prediction method according to claim 3, characterized in that: Contact between large particles occurs at a lateral angle of 45°.

5. A high-precision structural damage prediction method according to claim 4, characterized in that: Based on the differences, the microscopic parameters of the discrete element structural component model are iteratively adjusted, specifically including: Firstly, n groups of microscopic parameters and error limit e are given; Calculate the displacement field u of the finite element structural model F and the displacement field u of the discrete element structural component model D The difference between: If the difference exceeds the specified error limit e, the iterative adjustment phase is entered, during which the micro parameters are updated.

6. A high-precision structural damage prediction method according to claim 5, characterized in that: The error limit e is calculated by the following formula: Where λ is the coefficient and T is the transposed matrix.

7. A high-precision structural damage prediction method according to claim 6, characterized in that: The error of the displacement field of the structural part between n sets of discrete elements and finite elements is calculated and the minimum value among the n sets of errors is found, which is recorded as c min , when e min When it is greater than the error limit e, it enters the iterative adjustment stage.

8. A high-precision structural damage prediction method according to claim 7, characterized in that: e min The calculation formula is as follows; 9. A high-precision structural damage prediction method according to claim 8, characterized in that: Adopt the update algorithm to iteratively adjust until e min Less than the error limit e.

10. A high-precision structural damage prediction method according to claim 9, characterized in that: The update algorithm is used for iterative adjustment, which specifically includes the following steps: For n sets of microscopic parameter vectors β, the change of the microscopic parameter vector is set to Δβ i , and set the macroscopic parameter coordinate vector B obtained by simulation for each initial microscopic parameter vector β n , its Δβ i The calculation method is as follows: Db i =[v1 v2 …… v n ] Among them, v n is the amount of change of each microscopic parameter; Since there is no evolutionary experience at the beginning, all these variables are random, so the new microscopic parameter vector update formula is as follows: Z i =b i +Δβ i =[z1 z2 …… z n ] where z n is the new evolutionary microscopic parameter, for each z n The macroscopic parameter coordinate vector C is obtained through simulation. n ; Use the macro parameter coordinate vector C n and B n Carry out numerical simulations respectively to obtain corresponding multiple result quantities; Based on the result quantity, it is judged whether it meets the requirements. If it does not meet the requirements, the macro parameter coordinate vector C is selected. n , B n The microscopic parameter vector corresponding to the minimum value obtained after numerical simulation is taken as the optimal microscopic parameter P gbest , based on the optimal microscopic parameter P gbest Update Z again i ; Numerical simulation is performed based on the second update results and the first update results, and the above process is repeated to obtain the optimal microscopic parameters again until the obtained result quantity meets the requirements.

Images