Method and system for obtaining crack propagation through parallel operation under stress state judgment criterion

Abstract

The present invention provides a method and system for obtaining crack propagation through parallel computing under a stress state judgment criterion. Aiming at the diversity of fracture modes of materials under complex external force constraints, based on the crack propagation algorithm of EigenErosion with energy as the failure criterion, a pre-judgment of the stress state is introduced. By judging the stress state of material points under a specific criterion, the accuracy and robustness of the EigenErosion algorithm with a single strain energy magnitude as the failure criterion for crack propagation are enhanced, enabling the EigenErosion operation method to adapt to various external force constraint conditions. The parallel computer and parallel strategy are successfully utilized to achieve parallelization in the homogenization calculation of the energy release rate.

Description

Technical Field

[0001] The present invention relates to the technical field of dynamic crack propagation in medium mechanics, and more particularly to a method for obtaining crack propagation through parallel computing under a stress state judgment criterion. It is a dynamic crack propagation method using the equivalent energy release rate as the failure condition. Based on the EigenErosion algorithm, it adds a stress state criterion judgment criterion and realizes its use in distributed multi-process parallelization to efficiently and accurately solve dynamic crack propagation in continuum mechanics problems. Background Art

[0002] The initiation and propagation of cracks until complete fracture are continuous characteristics of materials and are extremely complex processes, involving the combination and competition of various energy propagation and dissipation methods such as brittle and ductile fractures. According to different stress states, fracture can be divided into three modes: opening, splitting, and tearing. Accurately describing these characteristics requires a physics-based crack propagation algorithm. However, existing crack propagation algorithms have problems such as non-convergence in theory and grid or discretization-related issues. Their errors do not decrease with the refinement of the grid or discretization, and they do not converge to the correct result when the element size approaches zero.

[0003] Take the element erosion method for example. It lacks a physics-based element deletion criterion. After deleting elements, mass conservation is not satisfied, and all physical quantities are also not conserved, resulting in overestimated failure results. At the same time, the results obtained for different grids vary significantly. In the cohesive zone model, cracks can only propagate along the surface of elements. If a crack needs to pass through the interior of an element, re-meshing is required. However, no matter how finely / remeshed the grid is; in the damage algorithms operation, physical relevance is not satisfied, and different discrete point distribution methods will result in different crack distributions, and the error of the calculation result will be larger when the discrete points are denser. The results obtained by existing methods will not converge, and their crack prediction results are not based on physics but determined by the grid.

[0004] The crack propagation algorithm can be divided into two categories according to different failure criteria: stress-based and energy-based. The model based on stress as the failure criterion lacks physical theory support and cannot describe clearly the entire strongly coupled energy dissipation process under loading rate, pressure, and temperature conditions. Its parameters are generally obtained from experiments and are also powerless for predictions when the calculation conditions exceed the test range. The model based on energy as the failure criterion has a certain theoretical basis and can allocate energy more reasonably during the energy dissipation process. Among them, the EigenErosion intrinsic material point deletion algorithm based on energy and variational principle can strictly prove the convergence of calculation results mathematically, that is, the finer the discretization, the more accurate the result. The only parameter that needs to be determined is the intrinsic energy release rate of the material. Without any other test models, it can predict the failure behavior of materials under different working conditions and the internal crack propagation process. By averaging energy to define the local energy release rate, it avoids the grid or discretization correlation of the crack propagation path. And an intrinsic deformation parameter (Eigenstrain) representing the local crack generation ratio is introduced into the EigenErosion framework. It is equivalent to the failure degree in the failure model, but it is a homogenized description of the internal crack size calculated based on the physical mechanism of variational crack propagation and free surface generation. There is no need to explicitly describe the direction and size of the crack, and it has a very simple processing method for three-dimensional geometry and topology structures without reconstructing complex crack surfaces. It is also the only physically based, convergent, and discretization-independent crack propagation algorithm in the meshless method. EigenErosion can be applied to problems such as simulation of brittle and ductile fractures, small and large deformations, and static and dynamic crack propagations.

[0005] EigenErosion defines the local energy release rate by averaging energy. However, as a scalar, energy does not have directionality. Under complex external force constraints, the isotropic homogenization process adopted by the algorithm still has limitations, affecting the accuracy of predicting the speed and direction of crack propagation. At the same time, to obtain the energy information of adjacent material points for the homogenization process of a certain material point. In parallel computing, how to establish the neighborhood of material points between divided regions and the dynamic update of the neighborhood during the calculation process are problems that must be solved for EigenErosion to achieve parallelization. Summary of the Invention

[0006] The first aspect of the present invention provides a method for obtaining crack propagation through parallel computing under a stress state judgment criterion, including the following steps:

[0007] Step 1: Discretize the continuous medium problem domain and initialize the operation parameters at t kAt the initial moment when t = 0, divide the material points in the continuous medium problem domain according to the total number of processors. Divide each material point in the initialized set of material points into multiple subsets equal to the total number of processors according to the coordinate region and send them to each corresponding processor;

[0008] Step 2, start step-by-step operations from the moment when t k = 1 until the preset total number of calculation steps is completed. During the operation process at each moment, calculate and update the neighborhood of each effective material point on the i-th processor, and obtain the set of neighborhoods of each material point within the subset of material points on each processor. The effective material point refers to the material point that has not failed due to fracture;

[0009] Step 3, based on the set of neighborhoods of each material point within the subset of material points on each processor, determine the range interval of the material points on each processor, that is, the maximum and minimum values of the material points within the neighborhoods of all material points on each processor in each dimension coordinate. Update from the maximum and minimum values to obtain the subset of dynamic material points included in the processor at time t k moment;

[0010] Step 4, according to the range interval of the material points on each processor, define the material points belonging to the subsets of dynamic material points of two or more processors as ghost material points, divide them into the ghost interval, and store them in the shared data exchange table;

[0011] Step 5, calculate and update the strain energy of each material point in the global domain at time t k moment;

[0012] Step 6, detect the stress state of each effective material point one by one, calculate the stress state of the effective material point and determine whether the effective material point can fail based on the stress state criterion: if the stress state of the detected material point meets the selected failure criterion, calculate the equivalent energy release rate of the effective material point, otherwise do not calculate the equivalent energy release rate of the effective material point;

[0013] Step 7, compare the calculated equivalent energy release rate of a certain effective material point with the preset critical energy release rate. If it is greater than the critical value, it is determined that the effective material point fails and will no longer participate in subsequent deformation and operations related to crack propagation. At the same time, the strain energy of the material points within the domain of the effective material point is released, that is, the strain energy is cleared to zero, and reset and store them in the shared data exchange table;

[0014] Step 8, repeat steps 6 - 7 until all effective material points have been detected, and implement the operation and solution process at the current moment.

[0015] The second aspect of the present invention also proposes a computer system, including:

[0016] One or more processors;

[0017] A memory storing instructions executable to cause the one or more processors to perform operations including performing the processes of the foregoing method when executed by the one or more processors.

[0018] A third aspect of the present invention further provides a computer-readable medium storing software, the software including instructions executable by one or more computers, and the instructions perform the processes of the foregoing method when executed by the one or more computers.

[0019] Thus, the method for obtaining crack propagation by combining parallel operation of the EigenErosion algorithm under the stress state judgment criterion proposed by the present invention (parallel EigenErosion with Stress State Criterion, pEESSC), on the basis of the crack propagation algorithm of EigenErosion with energy as the failure criterion, introduces a pre-judgment of the stress state, enables the EigenErosion operation method to adapt to various external force constraint conditions, and successfully applies parallel computers and parallel strategies to achieve parallelization in the calculation of the homogenization of the energy release rate.

[0020] The pEESSC method proposed by the present invention, in view of the diversity of fracture modes of materials under complex external force constraints, enhances the accuracy and robustness of the EigenErosion algorithm with a single strain energy magnitude as the failure criterion for crack propagation by judging the stress state of material points under specific criteria. At the same time, a set of shadow point calculation data partitioning schemes is developed at the MPI layer to achieve distributed parallelization, enabling the algorithm to be effectively utilized in parallel operations.

[0021] It should be understood that all combinations of the foregoing concepts and additional concepts described in greater detail below are considered to be part of the inventive subject matter of the present disclosure as long as such concepts are not mutually inconsistent. Additionally, all combinations of the claimed subject matter are considered to be part of the inventive subject matter of the present disclosure.

[0022] The foregoing and other aspects, embodiments, and features of the teachings of the present invention can be more fully understood from the following description in conjunction with the accompanying drawings. Other additional aspects of the present invention, such as features and / or beneficial effects of exemplary embodiments, will be apparent from the following description or will be learned through practice of specific embodiments according to the teachings of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The accompanying drawings are not intended to be drawn to scale. In the accompanying drawings, each identical or nearly identical component shown in each figure may be denoted by the same reference numeral. For the sake of clarity, not every component is labeled in each figure. Now, embodiments of various aspects of the present invention will be described by way of example and with reference to the accompanying drawings, in which:

[0024] Figure 1 is a schematic flow diagram of a method for obtaining crack propagation through parallel operation under a stress state judgment criterion in an embodiment of the present invention.

[0025] Figure 2 is a schematic diagram showing different crack propagation results obtained by using different stress state criteria when simulating the Brazilian test on concrete materials.

[0026] Figure 3 is a schematic diagram of the failure surface of each stress state criterion in a three-dimensional stress space. Detailed implementation manners

[0027] For a better understanding of the technical content of the present invention, specific embodiments are hereby given and described in conjunction with the accompanying drawings as follows.

[0028] In the present disclosure, various aspects of the present invention are described with reference to the accompanying drawings, in which many illustrative embodiments are shown. The embodiments of the present disclosure are not necessarily intended to include all aspects of the present invention. It should be understood that the various concepts and embodiments introduced above, as well as those described in more detail below, can be implemented in any of many ways, because the concepts and embodiments disclosed in the present invention are not limited to any implementation manner. Additionally, some aspects of the present invention can be used alone, or in any suitable combination with other aspects of the present invention.

[0029] Combined with Figure 1 the method for obtaining crack propagation through parallel operation under a stress state judgment criterion in the exemplary embodiment of the present invention shown in, includes the following steps:

[0030] Step 1, discretize the continuous medium problem domain and initialize the operation parameters. At the initial moment of t k = 0, divide the material points in the continuous medium problem domain according to the total number of processors, divide each material point in the initialized set of material points into multiple subsets equal to the total number of processors according to the coordinate region, and send them to each corresponding processor;

[0031] Step 2, from t kStarting from the moment of =1, perform step-by-step operations until the preset total number of calculation steps is completed. During the operation at each moment, calculate and update the neighborhood of each valid material point on the i-th processor, and obtain the set of neighborhoods of each material point within the subset of material points on each processor. A valid material point refers to a material point that has not failed due to fracture.

[0032] Step 3: Based on the set of neighborhoods of each material point within the subset of material points on each processor, determine the range interval of the material points on each processor, that is, the maximum and minimum values of the material points within the neighborhoods of all material points on each processor in each dimension coordinate. Update the dynamic subset of material points included in the processor at time t k from the maximum and minimum values.

[0033] Step 4: According to the range interval of the material points on each processor, define the material points belonging to the dynamic subset of material points of two or more processors as ghost material points, divide them into the ghost interval, and store them in the shared data exchange table.

[0034] Step 5: Calculate and update the strain energy of each material point in the global domain at time t k from the maximum and minimum values.

[0035] Step 6: Detect the stress state of each valid material point one by one, calculate the stress state of the valid material point, and determine whether the valid material point can fail based on the stress state criterion: If the stress state of the detected material point meets the selected failure criterion, calculate the equivalent energy release rate of the valid material point; otherwise, do not calculate the equivalent energy release rate of the valid material point.

[0036] Step 7: Compare the calculated equivalent energy release rate of a certain valid material point with the preset critical energy release rate. If it is greater than the critical value, it is determined that the valid material point fails and will no longer participate in subsequent deformation and operations related to crack propagation. At the same time, the strain energy of the material points within the domain of the valid material point is released, that is, the strain energy is cleared to zero, and they are reset and stored in the shared data exchange table.

[0037] Step 8: Repeat steps 6 - 7 until all valid material points have been detected, and the operation and solution process at the current moment is completed.

[0038] Among them, in step 1, discretizing the continuous medium problem domain and initializing the operation parameters include:

[0039] Step 1.1: Discretize the continuous medium problem domain Ω using a finite element mesh. For a two-dimensional continuous medium domain, use a first-order triangular element mesh; for a three-dimensional continuous medium domain, use a first-order tetrahedral element mesh.

[0040] Step 1.2: Using the finite element mesh obtained by discretization, initialize a set of material points {xp，0 where \(p = 1, 2, \cdots, M\) and a set of node sets \(\{x\) a，0 \(_{a}\), where \(a = 1, 2, \cdots, N\), and where \(x\) p，0 represents the \(p\)th material point at \(t\) k \( = 0\), \(k = 0, 1, \cdots, m\), \(p = 1, 2, \cdots, M\), \(x\) a，0 represents the \(a\)th node at \(t\) k \( = 0\), \(a = 1, 2, \cdots, N\), \(k = 0, 1, \cdots, n\), and \(n\) represents the total number of preset calculation steps during the initialization process; the node set is the node set of the finite element mesh, and the material point set is the set of the center points of each unit of each finite element mesh; then, for each unit of the finite element mesh, the centroid of each unit is taken as the material point, the four corner points of each unit are the nodes, and the nodes of each unit are the initial neighborhood of the material point;

[0041] Step 1.3, define a set of distributed processors \(\{P\) i \(_{i}\}\), \(i = 1, \cdots, I\), and \(I\) is the total number of processors

[0042] Thus, the initialization of the total number of calculation steps, the processor set, and the material point set and the node set is completed.

[0043] Among them, in Step 2, at each moment \(t\) k moment, calculate and update the neighborhood of each effective material point on the \(i\)th processor, \(i = 1, 2, \cdots, I\), thereby obtaining the set of neighborhoods of each material point within the subset of material points on each processor An effective material point refers to a material point that has not failed due to fracture;

[0044] Among them, for the material point \(p\) on the \(i\)th processor, retrieve each effective material point other than the material point \(p\) to obtain the \(\in\) p，k adjacent material points

[0045]

[0046] where \(q = 1, 2, \cdots, M\), \(q\neq p\);

[0047] \(\in\) p，k represents the dynamic \(\in\) neighborhood size of the material point \(x\) p，k \(_{p}\), \(\in\) p，k \(=\delta x\times h\) p，k where \(\delta x\) is an artificial coefficient; the value range of the artificial coefficient \(\delta x\) is \(1.5 - 3.0\); \(h\) p，k is the characteristic size of the material point \(p\) at \(t\) k moment.

[0048] As an optional embodiment, calculating the equivalent energy release rate of the effective material point includes:

[0049] Calculate the equivalent energy release rate according to the following formula:

[0050]

[0051] where α is the geometric correction parameter with a value of 1.5; V p，k represents the volume of material point p at time k, and W(x p，k ) represents the strain energy of material point x p，k ; the summation operation represents the strain energy stored in the material points near the crack tip

[0052] During the calculation of the equivalent energy release rate, the shared material point data in the shared data exchange table C k is used for data interaction between distributed processes through the information transfer interface MPI, and finally the equivalent energy release rate of the material point in the original continuous medium problem domain is obtained.

[0053] As an optional example, the stress state criteria in step 6 include:

[0054] (1) Positive pressure criterion: where σ1, σ2, and σ3 are the principal stresses of the material point;

[0055] (2) Positive minimum principal stress criterion: σ min > 0, where σ min is the minimum principal stress of the material point;

[0056] (3) Positive maximum principal stress criterion: σ max > 0, where σ max is the maximum principal stress of the material point;

[0057] (4) Stress triaxiality criterion: where p is the pressure of the material point, σ v is the von Mises equivalent stress of the material point, and m is an empirical parameter measured experimentally with a default value of -1;

[0058] (5) Mohr-Coulomb criterion: where p is the pressure of the material point, τ max is the maximum shear stress of the material point, and m is an empirical parameter measured experimentally with a default value of -1.

[0059] Among them, during the initialization of operation parameters in step 1, it also includes selecting the stress state criterion for subsequent stress state detection according to the force conditions of the problem domain.

[0060] Among them, at t kAt a moment, the strain energies of the detected material points obtained through deformation calculation are used to calculate the relationship of stress components and determine whether they meet the determined stress state criterion.

[0061] Next, we will combine Figure 1 the following example of the process to describe the implementation process of the pEESSC method of the above embodiments of the present invention.

[0062] Let Ω represent the d-dimensional continuous medium problem domain, which is discretized into a set of material point sets {x p，k} and a set of node sets {x a，k}, where p = 1, 2,..., M; k = 0, 1,..., m, a = 1, 2,..., N; k = 0, 1,..., n, and x p，k represents the p-th material point at the k-th moment, and x a，k represents the a-th node at the k-th moment.

[0063] When using the pEESSC method (parallel EigenErosion with Stress State Criterion, a parallel algorithm of the eigen deletion algorithm based on stress state determination) to perform arithmetic processing on the material fracture and crack propagation problems in the continuous medium domain, the process includes:

[0064] The first step: Select the stress state criterion used in the fixed crack propagation algorithm (one can be selected from five criteria according to the force conditions of the problem domain), define the total number of calculation steps n, define the distributed processor set {P i}, where i = 1,..., I, and I is the total number of processors, and initialize the material point set and the node set.

[0065] At t k = 0 moment, according to the number of processors I, divide the material points in the continuous medium problem domain, which specifically includes:

[0066] Step 1.1: Discretize the continuous medium problem domain Ω with a finite element mesh. For a two-dimensional continuous medium domain, use a first-order triangular element mesh; for a three-dimensional continuous medium domain, use a first-order tetrahedral element mesh.

[0067] Step 1.2: Use the obtained finite element mesh to initialize a set of material point sets {x p，0 , p = 1, 2,..., M} and a set of node sets {x a，0 , a = 1, 2,..., N}, where the node set is the node set of the finite element mesh, and the material point set is the center point set of each unit of each finite element mesh. Then, for each unit of a finite element mesh, the centroid of each unit is taken as a material point, the four corner points of each unit are nodes, and the nodes of each unit are the initial neighborhood of the material point.

[0068] Then, each material point in the initialized material point set is divided into I subsets according to the coordinate region and sent to the corresponding processors.

[0069] Step 2: Starting from the initial time t k = 0, calculate and update the ∈ k neighborhoods of each valid material point, that is, the non-fractured and non-failed material points, on the i-th processor at time t p，k ; taking the material point p on the i-th processor as an example, retrieve each valid material point (including all material points outside processor i) except this material point to obtain the ∈ p，k adjacent material points

[0070] q = 1, 2,..., M and q ≠ p

[0071] where ∈ p，k is the dynamic ∈ neighborhood size of the material point x p，k , ∈ p，k = δx × h p，k , δx is an artificial coefficient, and h p，k is the characteristic size of the material point p at time t k .

[0072] Step 3: Determine the range interval (Range Box) of the material points on each processor according to the set of neighborhoods of each material point in the material point subsets on each processor obtained in Step 2 , that is, the maximum and minimum values of the material points within the neighborhoods of all material points on processor i in each coordinate dimension, and update from the RangeBox to obtain the dynamic material point subset contained in processor i at time t ; k

[0073] Step 4: According to the Range Box determined in Step 3, define the material points belonging to the dynamic subsets of two or more processors as ghost material points (Shadow material point), divide them into the ghost interval (Shadow Box), and store them in the shared data exchange table C k .

[0074] Step 5: Calculate and update the strain energy W(x k ) of each material point in the global domain at time t p，k , p = 1, 2,..., M.

[0075] Step 6: Detect the stress state of each effective material point one by one. If the stress state of the detected material point meets a selected criterion, calculate the equivalent energy release rate of this material point:

[0076]

[0077] where α is the geometric correction parameter, generally taking a value of 1.5, and V p，k represent the volume of the material point p at time k respectively, and the summation term represents the strain energy stored in the material points near the crack tip.

[0078] During the calculation of the equivalent energy release rate of this material point, data interaction between distributed processes for the shared material point data in C k is carried out through MPI, and finally the equivalent energy release rate of this material point in the original problem domain is obtained.

[0079] If the stress state of the detected material point does not meet the selected failure criterion, there is no need to calculate the equivalent energy release rate of this material point, and the detection of the next material point is entered.

[0080] Step 7: According to the equivalent energy release rate of a certain effective material point calculated in Step 6, compare it with the preset critical energy release rate, G c (this value can be obtained from the fracture toughness measured by experiments or molecular dynamics calculations). If it is greater than the critical value, that is, G(x p，k ) > G c , then this material point fails (fractures), and will no longer participate in subsequent deformation and calculations related to crack propagation. At the same time, the strain energy of the material points in the p，k domain is released, that is, the strain energy is cleared, reset and store them in the shared data exchange table C k .

[0081] Step 8: Repeat Steps 6 and 7 until the detection of all effective material points is completed.

[0082] Step 9: Repeat Steps 2 to 8 until all calculation steps are completed.

[0083] In the embodiment of the present invention, according to the different external force constraints in the problem domain, the conditions and modes of material fracture will vary, resulting in changes in the direction and rate of crack propagation. The concept of the stress state criterion is similar to the yield criterion of plastic deformation.

[0084] When the material points in the problem domain are in a multi-axial stress state, all stress components must be considered simultaneously. Under certain deformation conditions (such as deformation temperature, deformation speed, etc.), only when a certain relationship exists between the stress components will the material points fail and cracks will occur.

[0085] Therefore, in the pEESSC method proposed in the embodiments of the present invention, before comparing the equivalent energy release rate of the effective material point with the critical value in the EigenErosion algorithm, the stress state of the material point is calculated and it is determined whether it can fail.

[0086] The stress state criteria include:

[0087] (1) Positive pressure criterion: where σ1, σ2, and σ3 are the principal stresses of the material point;

[0088] (2) Minimum principal stress is positive criterion: σ min > 0, where σ min is the minimum principal stress of the material point;

[0089] (3) Maximum principal stress is positive criterion: σ max > 0, where σ max is the maximum principal stress of the material point;

[0090] (4) Stress triaxiality criterion: where p is the pressure of the material point, σ v is the von Mises equivalent stress of the material point, and m is an empirical parameter measured experimentally, with a default value of -1;

[0091] (5) Mohr-Coulomb criterion: where p is the pressure of the material point, τ max is the maximum shear stress of the material point, and m is an empirical parameter measured experimentally, with a default value of -1.

[0092] In the calculation initialization (the first step of the parallelization process), a stress state criterion to be used in the subsequent crack propagation calculation needs to be defined, that is, one of the above criteria. In the sixth step of the parallelization process, the stress state of the material point is determined. At time t k , the stresses of the detected material point can be obtained through deformation calculation, and thus the relationship of the current stress components can be calculated and it can be determined whether it satisfies the determined stress state criterion.

[0093] Figure 3 Exemplarily shows a schematic of the failure surface of each stress state criterion in the three-dimensional stress space.

[0094] As Figure 2 shown in the example, when simulating the splitting test (Brazilian test) of concrete materials, the comparison of different crack propagation results obtained by using different stress state criteria is presented. Under the stress constraints of this test, the Mohr-Coulomb criterion can more accurately predict the fracture mode consistent with the real experiment.

[0095] According to the disclosure of the present invention, a computer system is further provided, including:

[0096] One or more processors;

[0097] A memory storing executable instructions, the instructions, when executed by one or more processors, cause the one or more processors to perform operations, the operations including performing the processes of the methods of any of the foregoing embodiments.

[0098] According to the disclosure of the present invention, a computer-readable medium storing software is further provided, the software including instructions executable by one or more computers, the instructions, when executed by the one or more computers, perform the processes of the methods of any of the foregoing embodiments.

[0099] Although the present invention has been disclosed above with preferred embodiments, it is not intended to limit the present invention. Those of ordinary skill in the technical field to which the present invention pertains can make various modifications and refinements without departing from the spirit and scope of the present invention. Therefore, the protection scope of the present invention shall be subject to what is defined by the claims.

Claims

1. A method for obtaining crack propagation through parallel operation under a stress state judgment criterion, characterized in that It includes the following steps: Step 1: Discretize the continuous medium problem domain and initialize the operation parameters. At the initial time t k = 0, divide the material points in the continuous medium problem domain according to the total number of processors. Divide each material point in the initialized material point set into multiple subsets equal to the total number of processors according to the coordinate region and send them to each corresponding processor; Step 2: Starting from the moment when t k = 1, perform step-by-step operations until the preset total number of calculation steps is completed. During the operation process at each moment, calculate and update the neighborhood of each effective material point on the i-th processor to obtain the set of neighborhoods of each material point within the subset of material points on each processor. The effective material point refers to the material point that has not broken and failed; Step 3: Based on the sets of neighborhoods of material points within the subsets of material points on each processor, determine the range intervals of the material points on each processor, that is, the maximum and minimum values of the material points within the neighborhoods of all material points on each processor in each dimension coordinate, and update from the maximum and minimum values to obtain the subset of dynamic material points included in the processor at time t k moment; Step 4: According to the range intervals of the material points on each processor, define the material points belonging to the dynamic material point subsets of two or more processors as ghost material points, divide them into ghost intervals, and store them in the shared data exchange table; Step 5, calculate and update t k The strain energy of each material point within the entire domain at the moment; Step 6: Detect the stress states of the effective material points one by one, calculate the stress states of the effective material points, and determine whether the effective material points can fail based on the stress state criterion: If the stress state of the detected material point meets the selected failure criterion, calculate the equivalent energy release rate of the effective material point; otherwise, there is no need to calculate the equivalent energy release rate of the effective material point; Step 7: Compare the calculated equivalent energy release rate of a certain effective material point with the preset critical energy release rate. If it is greater than the critical value, it is determined that the effective material point fails and will no longer participate in subsequent deformations and operations related to crack propagation. At the same time, the strain energy of the material points in the domain of the effective material point is released, that is, the strain energy is cleared to zero, and they are reset and stored in the shared data exchange table; Step 8: Repeat Steps 6-7 until all effective material points are detected, and the operation and solution process at the current moment is realized.

2. The method for obtaining crack propagation by parallel operation under the stress state judgment criterion according to claim 1, characterized in that In Step 1, the discretization of the continuous medium problem domain and the initialization of operation parameters include: Step 1.1: Discretize the continuous medium problem domain Ω with a finite element mesh. Among them, a first-order triangular element mesh is used for the two-dimensional continuous medium domain, and a first-order tetrahedral element mesh is used for the three-dimensional continuous medium domain; Step 1.

2. Initialize a set of material points {x p，0 , p = 1, 2, ..., M} and a set of nodes {x a，0 , a = 1, 2, ..., N} using the finite element mesh obtained by discretization, where x p，0 represents the p-th material point at t k = 0, k = 0, 1, … m, p = 1, 2, …, M, and x a，0 represents the a-th node at t k = 0, a = 1, 2, …, N, k = 0, 1, … n, and n represents the total number of preset calculation steps in the initialization process; the set of nodes is the set of nodes of the finite element mesh, and the set of material points is the set of the center points of each element of each finite element mesh; then, for each element of a finite element mesh, the centroid of each element is taken as a material point, the four corner points of each element are nodes, and the nodes of each element are the initial neighborhood of the material point. Step 1.

3. Define a set of distributed processors {P i}, where i = 1,..., I and I is the total number of processors So far, the initialization of the total number of calculation steps, the processor set, and the material point set and node set is completed.

3. The method for obtaining crack propagation by parallel operation under the stress state judgment criterion according to claim 2, wherein In step 2, at each moment t k calculate and update the neighborhood of each effective material point on the i-th processor, where i = 1, 2, ..., I, thereby obtaining a set of neighborhoods of each material point within the subset of material points on each processor The effective material point refers to a material point that has not failed due to fracture; Among them, for the material point p on the i-th processor, retrieve each effective material point other than the material point p to obtain the ∈ of the material point p p，k adjacent material point Where q = 1, 2,..., M, q ≠ p; ∈ p，k represents the dynamic ∈ neighborhood size of the material point x p，k , ∈ p，k = δx × h p，k , where δx is an artificial coefficient; h p，k is the characteristic size of the material point p at time t k .

4. The method for obtaining crack propagation through parallel operation under the stress state judgment criterion according to claim 3, characterized in that, The value range of the artificial coefficient δx is 1.5 - 3.

0.

5. The method for obtaining crack propagation by parallel operation under the stress state judgment criterion according to claim 3, characterized in that, Calculating the equivalent energy release rate of the effective material point includes: Calculating the equivalent energy release rate according to the following formula: where α is a geometric correction parameter with a value of 1.5; V p，k represents the volume of material point p at time k, and W(x p，k ) represents the strain energy of material point x p，k ; the summation operation represents the strain energy stored in material points near the crack tip In the process of calculating the equivalent energy release rate, the shared material point data in the shared data exchange table C is used for data interaction between distributed processes through the message passing interface MPI, and finally the equivalent energy release rate of the material point in the original continuum problem domain is obtained. k ​ 6. The method for obtaining crack propagation by parallel operation under the stress state judgment criterion according to claim 1, wherein The stress state criterion in Step 6 includes: (1) Positive pressure criterion: where σ1, σ2, and σ3 are the principal stresses of the material point; (2) Minimum principal stress positive criterion: σ min > 0, where σ min is the minimum principal stress of the material point; (3) Maximum principal stress positive criterion: σ max > 0, where σ max is the maximum principal stress of the material point; (4) Stress triaxiality criterion: where p is the pressure of the material point, and σ v is the von Mises equivalent stress of the material point, m is the empirical parameter measured experimentally, and the default value is -1; (5) Mohr-Columb criterion: where p is the pressure of the material point, and τ max is the maximum shear stress of the material point, and m is an empirical parameter measured experimentally with a default value of -1.

7. The method for obtaining crack propagation through parallel operation under the stress state judgment criterion according to claim 6, characterized in that, In the process of initializing the operation parameters in Step 1, it also includes selecting a stress state criterion for subsequent stress state detection according to the force conditions of the problem domain.

8. The method for obtaining crack propagation by parallel operation under the stress state judgment criterion according to claim 6, characterized in that, At time t k At this moment, the strain energies of the detected material points obtained through deformation calculation are used to calculate the relationship of stress components and determine whether they meet the determined stress state criterion.

9. A computer system, characterized in that, It includes: One or more processors; A memory that stores operable instructions, and the instructions, when executed by the one or more processors, cause the one or more processors to perform operations, and the operations include executing the process of the method described in any one of Claims 1-8.

10. A computer-readable medium storing software, characterized in that, The software includes instructions that can be executed by one or more computers, and the instructions, when executed by the one or more computers, execute the process of the method described in any one of Claims 1-8.

Images