A method for predicting fatigue life of additive manufacturing material under random load spectrum based on continuous damage mechanics

Abstract

The application provides a kind of fatigue life prediction method of additive manufacturing material under random load spectrum based on continuous damage mechanics, comprising: constructing the constitutive equation coupled with damage and the damage evolution equation considering additive manufacturing process;Adopt particle swarm algorithm to calibrate the parameters of constitutive equation and damage evolution equation;Convert the random load spectrum of finite length from time domain to cycle domain using rainflow counting method, obtain the cycle domain load information sequence;Based on the calibrated constitutive equation and damage evolution equation, using finite element-based numerical calculation method, combined with the damage evolution law coupled with damage, construct the finite element model of additive manufacturing material specimen;Based on the finite element model of additive manufacturing material specimen, the cycle domain load information sequence is simulated and calculated, and the fatigue damage calculation and life prediction of additive manufacturing material specimen under random load spectrum are completed.The application provides technical support for reliability design and life management of additive manufacturing material structure.

Description

Technical Field

[0001] This invention belongs to the field of fatigue damage analysis technology for additive manufacturing materials, specifically relating to a method for predicting the fatigue life of additive manufacturing materials under a random load spectrum based on continuous damage mechanics. Background Technology

[0002] With the continuous improvement of additive manufacturing technology, materials with different properties and applications can be manufactured by controlling processing parameters. Furthermore, it can rapidly manufacture complex structures. Therefore, the application scenarios for additive manufacturing materials are gradually increasing, including in the aerospace field where they are used in many parts. Continuous damage mechanics uses damage variables to describe the degree of material degradation and has been widely used in damage analysis and fatigue damage analysis due to its good physical interpretability. For additive manufacturing materials, some damage theory models consider the influence of additive manufacturing process parameters on the fatigue life of materials during the evolution of damage variables. For example, the authors' previous work considered the influence of additive manufacturing process parameters such as laser power, laser scanning rate, powder bed thickness, and scanning interval on the performance of aluminum alloy materials manufactured by selective laser melting (SLM), defining a fatigue damage evolution model coupled with laser volume energy density, and establishing a fatigue damage analysis method for aluminum alloy specimens manufactured by SLM.

[0003] However, given the practical engineering challenge of random load spectra faced by additively manufactured materials in complex service environments such as aerospace, conventional damage evolution models are only applicable to constant-amplitude spectra. Fatigue damage analysis under random spectra requires new methods. Currently, common fatigue damage analysis methods under random spectra primarily rely on rainflow counting and nominal stress methods for linear damage accumulation analysis, while damage evolution often resembles nonlinearity. Furthermore, additive manufacturing process parameters have been shown to influence the evolution of damage variables.

[0004] Therefore, it is necessary to develop nonlinear fatigue damage analysis and corresponding life prediction methods for additive manufacturing materials under random load spectra based on continuous damage mechanics and considering process parameters. Summary of the Invention

[0005] The purpose of this invention is to propose a fatigue damage analysis and life prediction method for additive manufacturing materials under random load spectra with finite lengths, based on continuous damage mechanics. This method establishes a nonlinear damage evolution and accumulation mechanism applicable to random spectrum loading within the framework of continuous damage mechanics. Furthermore, by combining this technical method with known additive manufacturing process parameters, it enables quantitative analysis of fatigue damage and life prediction of additive manufacturing materials under finite length and repeated random spectrum loading. This improves the physical consistency and engineering applicability of life assessment under random spectrum conditions, providing technical support for the reliability design and life management of additive manufacturing material structures.

[0006] To achieve the above objectives, the present invention provides the following solution: A method for predicting the fatigue life of additively manufactured materials under random load spectra based on continuous damage mechanics includes: Based on the assumptions of continuous damage mechanics and strain equivalence, a damage variable to describe material degradation is introduced to construct a constitutive equation for coupled damage; the laser volumetric energy density is defined according to the additive manufacturing process parameters, and an energy density influence factor is introduced to construct a damage evolution equation considering the additive manufacturing process. Using uniaxial tensile test results and constant amplitude fatigue test data, the parameters of the constitutive equation and the damage evolution equation are calibrated using the particle swarm optimization algorithm. The rainflow counting method is used to convert the finite-length random load spectrum from the time domain to the cyclic domain to obtain the cyclic domain load information sequence; Based on the calibrated constitutive equation and damage evolution equation, a finite element model of the additive manufacturing material specimen is constructed by using a finite element-based numerical calculation method and combining the damage evolution law of coupled damage. Based on the finite element model of the additive manufacturing material specimen, the cyclic domain load information sequence is simulated and calculated to complete the fatigue damage calculation and life prediction of the additive manufacturing material specimen under the random load spectrum.

[0007] Preferably, the constitutive equation is expressed as follows: , in, A , B and n For the constitutive parameters of the material, D Represents a one-dimensional damage variable. It represents plastic strain.

[0008] Preferably, the method for constructing the damage evolution equation includes: The laser volumetric energy density is calculated based on the laser power, laser scanning rate, powder bed thickness, and scanning interval of the additive manufacturing process. The set of process parameters with the longest fatigue test life in the additive manufacturing sample control group was taken as the reference set for the laser volume energy density during the analysis process to obtain the volume energy density. Based on the laser volume energy density and the volume energy density, calculate the energy density influence factor; Based on the energy density influence factor and combined with the preset material damage evolution equation parameters, the damage evolution equation is constructed.

[0009] Preferably, methods for calibrating the parameters of constitutive equations include: Based on the stress-strain curve data obtained from uniaxial tensile tests, an objective function for constitutive equation parameter calibration is constructed. Set the population size, number of iterations, learning factor, and inertia weight of the particle swarm optimization algorithm, and initialize the position and velocity of the particles, where the position vector of the particles corresponds to the calibration parameters of the constitutive equation. The velocity and position of the particles are iteratively updated, the fitness value of each particle is calculated, and the optimal position of the individual and the optimal position of the group are updated until the preset convergence condition is met. The optimal parameters are then output as the calibration result of the constitutive equation.

[0010] Preferably, the method for obtaining the cyclic domain load information sequence includes: Using the rainflow counting rule, complete stress cycles are sequentially identified and extracted from the acquired original random load time series, and the stress amplitude, stress mean, and cycle count of each cycle are recorded. Based on the start and end position indices of each stress cycle in the original random load time series, the smaller of the two values ​​is taken as the position marker of the stress cycle. Based on the location markers, all extracted stress cycles are sorted to generate a cyclic domain load information sequence that retains the original loading order, thus completing the conversion from the time domain to the cyclic domain.

[0011] Preferred methods for extracting complete stress cycles include: Four consecutive peak and valley points are selected sequentially from the original random load time series. The amplitudes A1 of the first and second points, A2 of the second and third points, and A3 of the third and fourth points are calculated. Determine whether A1≥A2 and A3≥A2 are satisfied. If satisfied, determine that the second and third points form a complete independent loop, record the amplitude A2 and mean of the independent loop, and remove these two points from the original data sequence. If not satisfied, move the selection window one point backward and repeat the judgment process.

[0012] Preferably, the method for constructing a finite element model of an additive manufacturing material specimen includes: Based on the calibrated constitutive equation and damage evolution equation, the step-by-step method is used to calculate damage accumulation and update damage. Based on the calculation results of the damage accumulation and the damage update, the damage increment of all materials after cycles in the cyclic domain load information sequence is calculated. Based on the damage increment, the material parameters at the integration point are updated to the effective stress space, and based on the updated material parameters, the numerical calculation of the stress-strain field and damage accumulation is completed. Repeat the above process until the simulation calculation of all cyclic loads in the cyclic domain load information sequence is completed, and the finite element model of the additive manufacturing material specimen is obtained.

[0013] Preferably, the damage variable at the integration point of the finite element model of the additive manufacturing material specimen is updated until the damage variable reaches 1, at which point the calculation stops and the fatigue life prediction result is obtained.

[0014] Compared with existing technologies, the beneficial effects of this invention are as follows: For additive manufacturing materials subjected to finite-length and repetitive random load spectra, a fatigue damage analysis and life prediction method for additive manufacturing materials under random load spectra based on continuous damage mechanics is proposed. Within the framework of continuous damage mechanics, a nonlinear damage evolution and accumulation mechanism suitable for random spectrum loading is established. Furthermore, based on this technical method and combined with known additive manufacturing process parameters, quantitative analysis of fatigue damage and life prediction of additive manufacturing materials under finite-length and repetitive random spectrum loading can be achieved. This improves the physical consistency and engineering applicability of life assessment under random spectrum conditions, providing technical support for the reliability design and life management of additive manufacturing material structures. Attached Figure Description

[0015] To more clearly illustrate the technical solution of the present invention, the drawings used in the embodiments are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a flowchart of the fatigue life prediction method for additive manufacturing materials under random load spectrum according to an embodiment of the present invention; Figure 2 A schematic diagram showing the shape and loading of a smooth test specimen for additive manufacturing materials; Figure 3 A schematic curve illustrating the application of material damage evolution parameter calibration results to the lifetime prediction integral formula; Figure 4 This is a schematic curve illustrating the constitutive properties of the material. Figure 5 This is a schematic diagram of the load sequence after the random load spectrum has been processed by the rainflow counting method. Figure 6 This is a schematic diagram of the damage evolution curve under random load spectrum. Detailed Implementation

[0017] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0018] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0019] Example 1: like Figure 1 As shown, a method for predicting the fatigue life of additive manufacturing materials under a random load spectrum based on continuous damage mechanics includes: S1: Based on the assumptions of continuous damage mechanics and strain equivalence, a damage variable is introduced to describe material degradation, and a constitutive equation for coupled damage is constructed. The energy density of the laser body is defined according to the additive manufacturing process parameters, and an energy density influence factor is introduced to construct a damage evolution equation that considers the additive manufacturing process.

[0020] In step S1, based on continuous damage mechanics, damage variables are introduced to describe the damage in the material. Considering the convenience of engineering applications, a one-dimensional damage variable is introduced in a representative volume element. D The initial damage value is 0, when D When = 1, it signifies complete material damage at the representative volume element. Based on the strain equivalence assumption and damage coupling based on the JC constitutive model, we can obtain: , in, A , B and n For the constitutive parameters of the material, D Represents a one-dimensional damage variable. It represents plastic strain.

[0021] A further implementation method involves constructing the damage evolution equation, including: Laser power based on additive manufacturing process P Laser scanning rate v Powder bed thickness t and scan interval h Calculate the volumetric energy density of the laser. E d : , The set of process parameters with the longest fatigue test life in the additive manufacturing sample control group was selected as the reference set for the laser volumetric energy density during the analysis, and the volumetric energy density was obtained. It is used for subsequent fatigue damage analysis and life prediction calculations.

[0022] Based on the laser volumetric energy density and energy density, an energy density influence factor is calculated. Specifically, based on the definition of laser volumetric energy density, an energy density influence factor can be further defined to describe the impact of different additive manufacturing parameters on the fatigue properties of manufactured materials, and further reflect the influence of different process parameters on fatigue damage evolution. Considering the ease of application and the effectiveness of the parameter process, the energy density influence factor Δ is defined as follows: in, q These are parameters that control the manufacturing effect of different additive manufacturing processes.

[0023] Based on the energy density influence factor and combined with preset material damage evolution equation parameters, a damage evolution equation is constructed. Specifically, based on the energy density influence factor, the existing damage evolution equation can be modified to reflect the influence of different additive manufacturing process parameters on damage evolution. Therefore, a damage evolution equation considering parameterized processes is established: in, , , m and b These are the parameters of the material damage evolution equation, which can be calibrated using the least squares method or particle swarm optimization algorithm based on fatigue test data of the material. By combining the constitutive equation of coupled damage and the damage evolution equation, the fatigue life of specimens under different additive manufacturing process parameters can be calculated using the damage mechanics-finite element method.

[0024] S2: Using uniaxial tensile test results and constant amplitude fatigue test data, the parameters of the constitutive equation and damage evolution equation are calibrated using the particle swarm optimization (PSO) algorithm. In step S2, the parameters are calibrated by finding the minimum value of the objective function using the PSO algorithm. The PSO algorithm uses the idea of ​​information transfer and sharing among the population to find the global optimal solution to the problem. The particle swarm is a group, and each particle is an individual. Velocity and position are two attributes describing the behavior of an individual. The evolution of an individual is guided by its own historical best position and the historical best position of the population until the objective function reaches its optimum.

[0025] A further implementation method involves calibrating the parameters of the constitutive equation by: constructing an objective function for calibrating the constitutive equation parameters based on stress-strain curve data obtained from uniaxial tensile tests; setting the population size, number of iterations, learning factor, and inertia weight of the particle swarm optimization algorithm, and initializing the positions and velocities of the particles, wherein the position vectors of the particles correspond to the parameters to be calibrated in the constitutive equation; iteratively updating the velocities and positions of the particles, calculating the fitness value of each particle, updating the optimal position of the individual and the optimal position of the population, until the preset convergence condition is met, and outputting the optimal parameters as the calibration result of the constitutive equation.

[0026] Specifically, in the process of calibrating the constitutive parameters, the objective function can be taken as: , in, and This represents the stress and plastic strain values ​​corresponding to the data points on the experimental curve. k The table shows the number of test points taken on the curve (total number is...). N ).

[0027] In the calibration process of the damage evolution equation parameters, the incremental damage equation is first calibrated. from arrive Integrating, we can obtain the integral formula for predicting the life of smooth parts: , Therefore, the objective function can be taken as: in, , and Represents experimental data, k Represents the number of additive manufacturing and fatigue test specimens (total number of...). M ), and Further numerical calculations using trial and error are needed to find the most suitable value, which can also make... The formula is simplified.

[0028] Based on the particle swarm optimization algorithm, the evolution of individual activities is described by updating particle velocity and position. The velocity update formula can be viewed as consisting of an inertial component, a cognitive component, and a social component. in, N For particle swarm scale, i For particle serial numbers, N D In particle dimension, d The particle dimension number. k For the number of iterations, w For inertial weights, c 1 represents the individual learning factor. c 2 represents the group learning factor. r 1 and r 2 is a random number within the interval [0,1] to increase the randomness of the search. For particles i In the k In the nth iterationd 3D velocity vector particle i In the k In the nth iteration d A dimensional position vector, particle i In the k In the nth iteration d The historical optimal position of dimension, that is, the position at the dimensional level. k After the nth iteration, the th i The optimal solution obtained by searching for individual particles (individuals) The group in the k In the nth iteration d The historical optimal position of dimension, that is, the position at the dimensional level. k The optimal solution in the entire particle swarm after the iteration.

[0029] The aforementioned particle swarm optimization algorithm can be used to calibrate both the material constitutive parameters and the damage evolution equation parameters. During the calibration process using the particle swarm algorithm, the positions of the particles represent the material parameters to be calibrated, with dimensions corresponding to the number of constitutive parameters and damage evolution equation parameters, respectively. The particle swarm calibration method obtains suitable material parameters by updating the position states of a large number of particles.

[0030] S3: The finite-length random load spectrum is converted from the time domain to the cyclic domain using the rainflow counting method to obtain the cyclic domain load information sequence; a further implementation method for obtaining the cyclic domain load information sequence includes: Using rainflow counting rules, complete stress cycles are sequentially identified and extracted from the acquired raw random load time series, and the stress amplitude, mean stress, and cycle count for each cycle are recorded. A further implementation method for extracting complete stress cycles includes: Four consecutive peaks and valleys are selected sequentially from the original random load time series. The amplitudes A1 of the first and second points, A2 of the second and third points, and A3 of the third and fourth points are calculated. It is determined whether A1≥A2 and A3≥A2 are satisfied. If they are satisfied, the second and third points are determined to form a complete independent cycle. The amplitude A2 and mean of the independent cycle are recorded, and these two points are removed from the original data series. If they are not satisfied, the selection window is moved one point forward, and the judgment process is repeated.

[0031] Based on the start and end indexes of each stress cycle in the original random load time series, the smaller of the two indices is taken as the position marker of the stress cycle. All extracted stress cycles are then sorted based on these position markers to generate a cyclic domain load information sequence that retains the original loading order, completing the conversion from the time domain to the cyclic domain. Specifically, this iterative process is performed on the entire load sequence, repeating the above steps until all data points have been processed, resulting in a conversion from the frequency domain to the cyclic domain. Further, based on the indices of the cycle start and end points in the original sequence, a correspondence is established between the cycle and the time position of the original load time history. The smaller of the cycle start and end indices is taken as the position marker for that cycle. All cycles are then reordered to obtain a cyclic domain load spectrum that better matches the original random load spectrum loading order.

[0032] S4: Based on the calibrated constitutive equation and damage evolution equation, a finite element method is used, combined with the damage evolution law of coupled damage, to construct a finite element model of the additive manufacturing material specimen. A further implementation method for constructing the finite element model of the additive manufacturing material specimen includes: calculating damage accumulation and updating damage using the step-by-step method based on the calibrated constitutive equation and damage evolution equation; calculating the damage increment of all cyclic loads in the cyclic domain load information sequence based on the damage accumulation calculation results and damage updates; updating the material parameters at the integration points to the effective stress space based on the damage increments, and completing the numerical calculation of the stress-strain field and damage accumulation based on the updated material parameters; repeating the above process until the simulation calculation of all cyclic loads in the cyclic domain load information sequence is completed, thus obtaining the finite element model of the additive manufacturing material specimen.

[0033] Specifically, in step S4, the skip-step method is used to perform damage accumulation calculation and damage update. Simultaneously, material parameters are updated through damage variables to achieve numerical calculation of the stress-strain field and damage accumulation. For a random load spectrum of finite length, it can be assumed that the damage accumulation value of the material within each cycle of the random load spectrum is small, meaning the difference in the initial damage value at the time of damage accumulation within each cycle is small. Therefore, it can be assumed that the initial damage value at the time of damage accumulation within each cycle of the random load spectrum is the same. Thus, the damage increment of the material after all cycles in the random load spectrum can be obtained, i.e.: Among them, subscript k and k +1 indicates the number of times the damage is updated in the cyclic load of the i-th random load spectrum. N `total` represents the total number of cycles in the random load spectrum. After the cumulative damage calculation for the cyclic load of the i-th random load spectrum is completed, the parameters at the integration point are updated to the effective stress space and used for subsequent stress-strain calculations.

[0034] S5: Based on the finite element model of additive manufacturing material specimens, the cyclic domain load information sequence is simulated and calculated to complete the fatigue damage calculation and life prediction of additive manufacturing material specimens under random load spectrum.

[0035] A further implementation involves updating the damage variable at the integration point of the finite element model of the additive manufacturing material specimen until the damage variable reaches 1, at which point the calculation stops and the fatigue life prediction result is obtained.

[0036] Specifically, in step S5, the material damage analysis and life prediction method can be implemented using the ABAQUS finite element calculation platform and its subroutine interface. By customizing the damage evolution program and numerically simulating the fatigue process, the fatigue damage calculation and life prediction results of the additive manufacturing material specimen under random load spectrum are obtained.

[0037] A finite element model for numerical simulation is established, and an appropriate mesh is generated based on the mesh size convergence analysis. Correct boundary conditions and loads conforming to experimental conditions are applied. Based on the interaction between the ABAQUS platform and subroutines, a user-defined UMAT subroutine is used to perform numerical simulation in conjunction with ABAQUS. Based on the physical quantities input from ABAQUS to UMAT, the stress and strain at each integration point of each element under cyclic loading are calculated. A stress-strain update algorithm for coupled damage is developed. Based on the physical quantities input from ABAQUS to UMAT, the stress and strain at each integration point of each element under cyclic loading are calculated, and a damage evolution equation is added to continuously update the damage variables until the damage degree at a certain element integration point reaches 1, at which point the calculation stops. This achieves the updating of material parameters and the backhaul of stress field and state variable physical quantities.

[0038] Example 2: The implementation of the above-mentioned method for predicting the fatigue life of additive manufacturing materials under random load spectrum based on continuous damage mechanics is explained.

[0039] Assuming there is an existing... Figure 2 The additively manufactured smooth test specimens shown under different processing techniques exhibit varying stress ratios. R Constant-amplitude fatigue tests were conducted under fatigue loads of -1 / 0.0 / 0.06, and the energy density influence factor composed of additive manufacturing process parameters was quantitatively calculated. Based on the above method, the fatigue damage and life prediction of additively manufactured materials under random loads were analyzed and calculated. Specifically, this included: (1) Conduct multiple fatigue tests under different stress ratios, peak stresses and additive energy density influence factors, and organize the data according to the fatigue test conditions and test results; (2) Take the group with the longest fatigue test life and record it as the reference group. At the same time, calculate its Calculate Δ for each set of process parameters; (3) Based on the uniaxial tensile test result curve, the material constitutive model parameters are calibrated according to the particle swarm calibration algorithm: A , B and n ; (4) Based on the fatigue test life, the parameters of the damage evolution equation are calibrated according to the particle swarm calibration algorithm: m , b、 and ; (5) Use the rainflow counting method to process the random load spectrum into the load spectrum under the cyclic domain. Based on the index of the cycle start and end inflection point in the original sequence, establish the correspondence between the cycle and the time position of the original load time history. Take the small value between the cycle start index and the end index to mark the cycle position. Reorder all cycles to obtain the cyclic domain load list. (6) In ABAQUS software, a finite element calculation model of the specimen is established to carry out numerical simulation of the fatigue loading process. The load spectrum under the cyclic domain is input, and the damage variable at the unit integration point is continuously updated with the help of the UMAT subroutine. (7) The calculation stops when the damage variable at a certain unit integration point reaches 1, and the fatigue life prediction result is obtained.

[0040] Based on the results of constant amplitude fatigue tests, data analysis was performed, and fatigue parameters were calibrated to obtain the material damage evolution parameter calibration results. These results were then applied to the life prediction integral formula, and the resulting fatigue load stress peak value-calculated life value schematic curve is shown below. Figure 3 As shown. Based on several experimental values ​​from the monotonic tensile curve, constitutive parameters were calibrated. These parameters were then applied to the JC constitutive equation to obtain a schematic stress-strain response curve for the additive manufacturing material, as shown. Figure 4 As shown.

[0041] The rainflow counting method is used to process finite-length random load spectra. Based on the indices of the cycle start and end points in the original sequence, a correspondence is established between the cycle and the time position of the original load time history. The smaller value between the cycle's initial and ending indices is used to mark the cycle position. All cycles are then reordered. The transformed load spectrum stress range and sequence are illustrated in the diagram below. Figure 5 As shown.

[0042] against Figure 5 The load spectrum shown is used to simulate the evolution of damage variables during fatigue under different additive manufacturing processes. The schematic diagram of the evolution of damage variables with the number of load cycles is shown below. Figure 6 As shown.

[0043] Example 3: This invention also provides a device for predicting the fatigue life of additive manufacturing materials under random load spectra, comprising: The first calculation module defines the laser energy density based on the additive manufacturing process parameters, inputs process parameters such as laser power, laser scanning rate, powder bed thickness, and scanning interval, as well as the reference value of laser energy density, calculates the energy density influence factor, and establishes a damage evolution equation that takes into account the additive manufacturing process. The second calculation module calibrates the parameters of the constitutive equation and the damage evolution equation using a particle swarm optimization algorithm based on the uniaxial tensile test results and fatigue test data. The third calculation module uses the rainflow counting method to process the finite-length random load spectrum, extracts the cyclic amplitude, mean, and sequence information, and reorders all cycles according to the small value in the initial and final indices of the cycles, thus transforming the time-domain problem into the cyclic domain. The fourth calculation module relies on the constitutive equation and damage evolution equation of coupled damage to establish a numerical calculation method based on the finite element method using large commercial software, and inputs the numerical calculation control parameters and algorithm file location. The fifth calculation module establishes a finite element model of the additive manufacturing material specimen, determines the mesh size based on mesh size convergence analysis, inputs the processed load information sequence for simulation calculation, and performs fatigue damage calculation and life prediction of the additive manufacturing material specimen under random load spectrum.

[0044] As one embodiment of the present invention, by inputting the fatigue test life and laser volumetric energy density reference values ​​of additive manufacturing materials under different processing techniques at stress ratios R=-1 / 0.0 / 0.06, the material damage evolution equation parameters and schematic curves (such as...) can be obtained. Figure 3 (As shown); by inputting several sets of monotonic loading test points for additive manufacturing materials under different processing techniques, the constitutive parameters and schematic curves of the material can be obtained (e.g., Figure 4 As shown); by inputting the random load spectrum, the processed cyclic domain load sequence can be obtained (e.g. Figure 5 As shown), further lifetime prediction under random spectral loading is completed, and damage evolution curves are obtained (as shown). Figure 6 (As shown).

[0045] The embodiments described above are merely preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Various modifications and improvements made to the technical solutions of the present invention by those skilled in the art without departing from the spirit of the present invention should fall within the protection scope defined by the claims of the present invention.

Claims

1. A method for predicting the fatigue life of additively manufactured materials under random load spectra based on continuous damage mechanics, characterized in that, include: Based on the assumptions of continuous damage mechanics and strain equivalence, a damage variable to describe material degradation is introduced to construct a constitutive equation for coupled damage; the laser volumetric energy density is defined according to the additive manufacturing process parameters, and an energy density influence factor is introduced to construct a damage evolution equation considering the additive manufacturing process. Using uniaxial tensile test results and constant amplitude fatigue test data, the parameters of the constitutive equation and the damage evolution equation are calibrated using the particle swarm optimization algorithm. The rainflow counting method is used to convert the finite-length random load spectrum from the time domain to the cyclic domain to obtain the cyclic domain load information sequence; Based on the calibrated constitutive equation and damage evolution equation, a finite element model of the additive manufacturing material specimen is constructed by using a finite element-based numerical calculation method and combining the damage evolution law of coupled damage. Based on the finite element model of the additive manufacturing material specimen, the cyclic domain load information sequence is simulated and calculated to complete the fatigue damage calculation and life prediction of the additive manufacturing material specimen under the random load spectrum.

2. The method according to claim 1, characterized in that, The constitutive equation is expressed as follows: ， in, A , B and n For the constitutive parameters of the material, D Represents a one-dimensional damage variable. It represents plastic strain.

3. The method according to claim 1, characterized in that, The method for constructing the damage evolution equation includes: The laser volumetric energy density is calculated based on the laser power, laser scanning rate, powder bed thickness, and scanning interval of the additive manufacturing process. The set of process parameters with the longest fatigue test life in the additive manufacturing sample control group was taken as the reference set for the laser volume energy density during the analysis process to obtain the volume energy density. Based on the laser volume energy density and the volume energy density, calculate the energy density influence factor; Based on the energy density influence factor and combined with the preset material damage evolution equation parameters, the damage evolution equation is constructed.

4. The method according to claim 1, characterized in that, Methods for calibrating the parameters of constitutive equations include: Based on the stress-strain curve data obtained from uniaxial tensile tests, an objective function for constitutive equation parameter calibration is constructed. Set the population size, number of iterations, learning factor, and inertia weight of the particle swarm optimization algorithm, and initialize the position and velocity of the particles, where the position vector of the particles corresponds to the calibration parameters of the constitutive equation. The velocity and position of the particles are iteratively updated, the fitness value of each particle is calculated, and the optimal position of the individual and the optimal position of the group are updated until the preset convergence condition is met. The optimal parameters are then output as the calibration result of the constitutive equation.

5. The method according to claim 1, characterized in that, Methods for obtaining cyclic domain load information sequences include: Using the rainflow counting rule, complete stress cycles are sequentially identified and extracted from the acquired original random load time series, and the stress amplitude, stress mean, and cycle count of each cycle are recorded. Based on the start and end position indices of each stress cycle in the original random load time series, the smaller of the two values ​​is taken as the position marker of the stress cycle. Based on the location markers, all extracted stress cycles are sorted to generate a cyclic domain load information sequence that retains the original loading order, thus completing the conversion from the time domain to the cyclic domain.

6. The method according to claim 4, characterized in that, Methods for extracting complete stress cycles include: Four consecutive peak and valley points are selected sequentially from the original random load time series. The amplitudes A1 of the first and second points, A2 of the second and third points, and A3 of the third and fourth points are calculated. Determine whether A1≥A2 and A3≥A2 are satisfied. If satisfied, determine that the second and third points form a complete independent loop, record the amplitude A2 and mean of the independent loop, and remove these two points from the original data sequence. If not satisfied, move the selection window one point backward and repeat the judgment process.

7. The method according to claim 1, characterized in that, Methods for constructing finite element models of additive manufacturing material specimens include: Based on the calibrated constitutive equation and damage evolution equation, the step-by-step method is used to calculate damage accumulation and update damage. Based on the calculation results of the damage accumulation and the damage update, the damage increment of all materials after cycles in the cyclic domain load information sequence is calculated. Based on the damage increment, the material parameters at the integration point are updated to the effective stress space, and based on the updated material parameters, the numerical calculation of the stress-strain field and damage accumulation is completed. Repeat the above process until the simulation calculation of all cyclic loads in the cyclic domain load information sequence is completed, and the finite element model of the additive manufacturing material specimen is obtained.

8. The method according to claim 1, characterized in that, Update the damage variable at the element integration point of the finite element model of the additive manufacturing material specimen until the damage variable reaches 1, then stop the calculation and obtain the fatigue life prediction result.

Images