Energy storage battery multi-scale mechanical evolution mechanism and failure path prediction method and system

Abstract

The application discloses a method and system for predicting multi-scale mechanical evolution mechanism and failure path of energy storage batteries, and belongs to the technical field of batteries, comprising the following steps: constructing a geometric model containing a plurality of randomly oriented primary particles; coupling and solving lithium ion diffusion equations, linear elastic mechanics equations and phase field damage evolution equations, calculating diffusion-induced stress through anisotropic lattice strain input by X-ray diffraction data; performing constant current charge and discharge cycle simulation, outputting first principal stress distribution, phase field damage variable distribution and crack evolution path; comparing stress and damage distribution under different charge and discharge rates, primary particle morphological configurations or material mechanical parameters to evaluate the influence degree of each parameter on the failure path; the method and system for predicting multi-scale mechanical evolution mechanism and failure path of energy storage batteries can accurately, efficiently and parameterizedly predict the chemical-mechanical failure path of the positive electrode material.

Description

Technical Field

[0001] This invention relates to the field of battery technology, and in particular to a method and system for predicting the multi-scale mechanical evolution mechanism and failure path of energy storage batteries. Background Technology

[0002] Energy storage batteries (such as lithium-ion batteries) have become the preferred power technology for electric vehicles and large-scale energy storage due to their high energy density, long cycle life, and low self-discharge rate. As market demands for longer driving ranges and faster charging speeds continue to increase, high-nickel layered cathode materials (such as LiNi) are becoming increasingly important. x CoMnzO2 (NCM) has become a mainstream cathode material due to its high specific capacity and cost advantages. However, high-nickel cathodes face serious multi-scale mechanical evolution problems during cycling, mainly manifested as chemical-mechanical degradation from the lattice scale to the particle scale, ultimately leading to the formation and propagation of intergranular cracks. These cracks not only cause rapid capacity decay but may also trigger safety hazards such as thermal runaway. Therefore, accurately predicting its mechanical evolution mechanism and failure path is crucial for the life assessment and safety design of energy storage batteries.

[0003] Numerous studies have shown that the mechanical failure of NCM cathodes stems from the multi-scale mechanical response induced by their polycrystalline aggregate structure. At the lattice scale, layered oxide crystals exhibit significant anisotropy: the diffusion rates along the a-axis and c-axis differ during lithium-ion insertion / extraction, and the lattice constant undergoes a non-monotonic evolution with lithium concentration—experiencing a sudden contraction of approximately 5% along the c-axis during deep delithiation. At the particle scale, NCM secondary particles are composed of dozens of randomly oriented primary particles aggregated. Each primary particle undergoes anisotropic volume changes due to differences in crystal orientation, resulting in mismatched deformation between adjacent particles and the accumulation of enormous diffusion-induced stress and cohesive stress at grain boundaries. Once the stress exceeds the fracture strength of the grain boundaries, intergranular cracks initiate and gradually propagate along the grain boundaries, ultimately leading to particle fragmentation and battery failure. This multi-scale mechanical evolution process, from lattice strain to intergranular cracks, constitutes the intrinsic mechanism of energy storage battery material failure.

[0004] Current technologies are unable to accurately, efficiently, and parametrically predict the multi-scale mechanical evolution mechanism and failure path of energy storage batteries. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for predicting the multi-scale mechanical evolution mechanism and failure path of energy storage batteries. It can continuously describe the entire process of stress accumulation, damage initiation and crack propagation from the perspective of microscopic lattice behavior, and provide a theoretical basis for material optimization and life assessment of energy storage batteries.

[0006] To achieve the above objectives, this invention provides a method for predicting the multi-scale mechanical evolution mechanism and failure path of energy storage batteries, comprising the following steps: S1. Construct a geometric model containing multiple randomly oriented primary particles based on focused ion beam scanning electron microscope images. The grain boundaries of the primary particles are generated by the Voronoi algorithm, and each primary particle is assigned a random crystal orientation angle. S2. Coupled solution of lithium-ion diffusion equation, linear elasticity equation and phase field damage evolution equation, wherein the lithium-ion diffusion coefficient is set anisotropically along the a-axis and c-axis of the primary particle, and the diffusion-induced stress is calculated by anisotropic lattice strain input by X-ray diffraction data. S3. Perform constant current charge-discharge cycle simulation and output the first principal stress distribution, phase field damage variable distribution and crack evolution path; S4. By comparing the stress and damage distribution under different charge / discharge rates, primary particle morphology configurations, or material mechanical parameters, assess the degree of influence of each parameter on the failure path.

[0007] Preferably, step S1 specifically includes: Focused ion beam scanning electron microscopy was used to perform continuous cross-sectional imaging of secondary particles of cathode material to obtain a series of cross-sectional images. Image recognition software was used to extract the boundaries of each cross-sectional image to obtain the morphological parameters of each primary particle. Based on the extracted morphological parameters, a two-dimensional polygon aggregation model is generated using the Voronoi algorithm; Each primary particle is assigned a random crystal orientation angle. , defined as the angle between the crystallographic a-axis and c-axis of the primary particle and the x-axis of the global coordinate system, where the a-axis direction angle of the primary particle is . The c-axis direction angle of this particle is ; Apply constrained mechanical boundary conditions to the outer surface of the secondary particles: The normal displacement is fixed at 0 to simulate the compaction state of the electrode material in the battery. Tangential displacement is free, allowing secondary particles to slip due to expansion / contraction within the plane; The internal primary grain boundaries are set to mutual support conditions: normal separation is allowed between adjacent primary grains, but normal penetration is prohibited; relative sliding is allowed in the tangential direction.

[0008] Preferably, based on the extracted morphological parameters, a two-dimensional polygon aggregation model is generated using the Voronoi algorithm, specifically as follows: Randomly generated within the computational domain Each seed point represents the number of primary particles; Lloyd relaxation iteration is performed on the seed points to make the polygon size distribution more uniform, which is used to avoid polygons that are too small or distorted. Construct a Voronoi diagram, where each Voronoi unit is a primary grain and the common edge between Voronoi units is a grain boundary; The generated polygons are smoothed to approximate the irregular shapes of real particles.

[0009] Preferably, the lithium-ion diffusion equation in step S2 adopts a modified Fick's second law: ; in, Lithium concentration, For time, It is the second-order diffusion coefficient tensor. This represents the partial molar volume of lithium ions. Let be the ideal gas constant. Absolute temperature It is the hydrostatic pressure; The equations of linear elasticity include: Equilibrium equations: ; Constitutive equation: ; in, For the isotropic elastic stiffness tensor, For stress tensor, For the total strain tensor, For lithium insertion / extraction induced inelastic strain, , It is the displacement vector; Introducing phase field variables ,in, Indicates that the material is in good condition. Indicating complete cracking, the phase field evolution follows Allen-Cahn type equations: ; Total Energy Function Including elastic strain energy and fracture surface energy: ; in, For elastic strain energy density, The inherent fracture energy release rate, The phase field length dimension is used to control the width of the crack dispersion region. Phase mobility; Substituting the total energy functional into the evolution equation, we obtain the explicit form: .

[0010] Preferably, in step S2, the diffusion-induced stress is calculated using anisotropic lattice strain, specifically as follows: The lattice constant of the cathode material under different battery charge states was measured by X-ray diffraction experiments. Along axis a: The lattice constant decreases monotonically with decreasing lithium concentration, where, The initial value of the a-axis lattice constant is given. The magnitude of the change in lattice constant. This represents the maximum lithium concentration. The lattice constant along the c-axis first increases and then decreases with decreasing lithium concentration, and undergoes a sudden contraction in the phase transition region; The volumetric strain exhibits a non-linear relationship with the normalized lithium concentration, and the volumetric strain is: ; For each primary particle, based on its crystal orientation angle Transform the strain in the crystal coordinate system to the global coordinate system: ; in, Represents the strain tensor in crystal coordinates. This represents the strain tensor in the global coordinate system. This represents the initial value of the lithium concentration.

[0011] Preferably, the phase field variable in step S2 represents the continuous transition of the material from intact to cracked and is bidirectionally coupled with the stress field and concentration field, specifically: Phase field → Stress field: Effective stress tensor This indicates that the stiffness of the cracked region has degraded; Phase field → Concentration field: Effective diffusion coefficient This indicates that lithium transport is blocked in the crack region; Stress field / concentration field → Phase field: Elastic strain energy Directly drives crack propagation; concentration field influences This will affect .

[0012] Preferably, the charge / discharge rate in step S3 ranges from 0.5C to 4C; the material mechanical parameters include elastic modulus, Poisson's ratio, and inherent fracture energy release rate; the output first principal stress distribution, phase field damage variable distribution, and crack evolution path are used to identify the most dangerous stress concentration area and determine the location and propagation trend of cracks.

[0013] Preferably, the evaluation of the influence of each parameter on the failure path in step S4 specifically involves: The single-parameter scanning method was used to record the peak value of the first principal stress, the average damage variable, and the percentage of intergranular crack area under different parameter combinations. Calculate the local sensitivity coefficient for the parameter The local sensitivity coefficient is defined as: ; in, The output value is the value under the reference parameter value. As the baseline parameter value, This represents the change in the output value after the parameter is changed. For parameter changes; when When this happens, the parameter is considered a sensitive parameter.

[0014] This invention also provides a multi-scale mechanical evolution mechanism and failure path prediction system for energy storage batteries, including: Model building module: used to build a geometric model containing multiple randomly oriented primary particles based on focused ion beam scanning electron microscope images, wherein the grain boundaries of the primary particles are generated by the Voronoi algorithm, and each primary particle is assigned a random crystal orientation angle; Coupled solution module: used to couple the solution of lithium-ion diffusion equation, linear elasticity equation and phase field damage evolution equation, where the lithium-ion diffusion coefficient is set anisotropically along the a-axis and c-axis, diffusion-induced stress is calculated through anisotropic lattice strain, and phase field variables are bidirectionally coupled with stress field and concentration field; Path prediction module: used to perform constant current charge-discharge cycle simulation, outputting the first principal stress distribution, phase field damage variable distribution, and crack evolution path; Parameter evaluation module: Used to evaluate the influence of each parameter on the chemical-mechanical failure path by comparing stress and damage distribution under different parameter combinations.

[0015] Therefore, the present invention, employing the above-mentioned multi-scale mechanical evolution mechanism and failure path prediction method and system for energy storage batteries, has the following beneficial effects: (1) The model generated by focusing ion beam scanning electron microscopy images and Voronoi algorithm can accurately reflect the morphology, size distribution and random orientation of primary particles in NCM secondary particles, providing a real structural basis for stress-damage analysis. (2) Simultaneously considering lithium diffusion anisotropy (different diffusion coefficients along the a / c axis) and lattice strain anisotropy (different variation of the a / c axis with lithium concentration), and through experimental data input, the calculation of diffusion-induced stress is made more accurate; (3) The phase field damage variable is bidirectionally coupled with the stress field and concentration field. The elastic strain energy serves as the crack driving force, which can naturally simulate the crack initiation, intergranular propagation and possible transgranular transformation without pre-setting the crack path. (4) It can simulate different primary particle arrangements such as random orientation, radial orientation, and core-shell configuration, and quantitatively evaluate the effect of each configuration on crack suppression. (5) Through the calculation of local sensitivity coefficient, it was revealed for the first time that Poisson's ratio has the most significant effect on stress level, providing a clear direction for the optimization of material mechanical properties; (6) It can systematically study the differences in failure paths under different C ratios, and find that concentration polarization is intensified, the area of ​​intergranular cracks increases, and damage extends to the particle core under high C ratios, providing a basis for safety assessment of fast charging conditions.

[0016] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0017] Figure 1 This is a flowchart of the multi-scale mechanical evolution mechanism and failure path prediction method for energy storage batteries of the present invention; Figure 2 The area distribution of the primary particles in this embodiment of the invention roughly follows a log-normal distribution. Figure 3 The lattice changes with SOC in an embodiment of the present invention, where (a) is the a-axis and (b) is the c-axis; Figure 4 This is an overall flowchart of the multi-scale mechanical evolution mechanism and failure path prediction system for energy storage batteries of the present invention. Detailed Implementation

[0018] The following detailed description of embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0019] Please see Figure 1This paper presents a multi-scale mechanical evolution mechanism and failure path prediction method for energy storage batteries. "Multi-scale" refers to describing the structural characteristics and physicochemical behavior of battery electrode materials across multiple levels, from the lattice scale (atomic / molecular level) to the particle scale (micrometer level). This embodiment constructs a geometric model (particle scale) containing multiple primary particles based on FIB-SEM images, while simultaneously inputting anisotropic lattice strain (lattice scale) using XRD data, enabling the transfer and coupling of mechanical and chemical information between the two scales. The mechanical evolution mechanism refers to the dynamic changes in stress generation, accumulation, release, and impact on material structure caused by lithium-ion insertion / extraction during charge-discharge cycles. This embodiment couples and solves the lithium-ion diffusion equation, linear elasticity equation, and phase-field damage evolution equation, outputting the first principal stress distribution and phase-field damage variables as a function of cycle time, revealing the dynamic feedback mechanism between stress, concentration, and damage. The failure path refers to the crack initiation location, propagation direction, propagation rate, and final crack network morphology during the process of the battery electrode material progressing from an intact state to complete cracking. This embodiment performs constant current charge-discharge cycle simulation, outputs the crack evolution path, and compares the damage distribution differences under different charge-discharge rates, primary particle morphology, and material mechanical parameters. This allows for a quantitative assessment of the influence of each parameter on the failure path. Specifically, it includes the following: S1. Construct a geometric model containing multiple randomly oriented primary particles based on focused ion beam scanning electron microscope images. The geometric model can be a two-dimensional or three-dimensional geometric model. The grain boundaries of the primary particles are generated by the Voronoi algorithm, and each primary particle is assigned a random crystal orientation angle. Primary particle morphology configurations may include at least one of the following: Random orientation configuration: The orientation angles of all primary grain crystals are randomly distributed; Radial orientation configuration: Primary particles elongate radially along the secondary particles, with the c-axis pointing towards the particle center; Core-shell configuration: internal primary particles are radially oriented, while external primary particles are randomly oriented.

[0020] The radial orientation configuration can be configured to shorten the lithium diffusion path in the radial direction, make the concentration gradient distribution more uniform, and reduce the stress level and damage at the grain boundaries compared to the random orientation configuration.

[0021] In this embodiment, a focused ion beam scanning electron microscope (FBS) is used to perform sequential cross-sectional imaging of secondary particles of the cathode material (such as NCM), obtaining a series of cross-sectional images. Each slice has a thickness of 50–100 nm, and the image resolution is better than 10 nm / pixel. Image recognition software (such as AutoCAD, ImageJ, or Avizo) is used to extract the boundaries of each cross-sectional image, obtaining morphological parameters such as the contour, area, perimeter, and shape factor of each primary particle; for example... Figure 2As shown, the area distribution of the primary particles roughly follows a log-normal distribution, with an average area of ​​5–20 μm. 2 The horizontal axis (X-axis) represents the area of ​​the primary particles, expressed in square micrometers (μm). 2 The figure shows several area values: 0.18, 0.36, 0.54, 0.72, and 0.90 (all in μm). 2 ).

[0022] The vertical axis (Y-axis) corresponds to the particle count (Counts) of the area range, which is the number of primary particles with that area size in the NCM91 material. Overall, it shows a distribution characteristic of more particles in small areas and fewer particles in large areas.

[0023] Based on the extracted morphological parameters (number of primary particles, average area, and size distribution), a two-dimensional polygonal aggregation model is generated using the Voronoi algorithm; specifically: Randomly generated within the computational domain (typically a square region of 50μm × 50μm). Each seed point represents the number of primary particles; Perform Lloyd relaxation iterations (usually 3 to 5 times) on the seed points to make the polygon size distribution more uniform, which is used to avoid excessively small or distorted polygons. Construct a Voronoi diagram, where each Voronoi unit is a primary grain and the common edge between Voronoi units is a grain boundary; The generated polygons are smoothed (e.g., using Catmull-Rom splines) to approximate the irregular shapes of real particles.

[0024] Each primary particle is assigned a random crystal orientation angle. , defined as the angle between the crystallographic a-axis and c-axis of the primary particle and the x-axis of the global coordinate system, where the a-axis direction angle of the primary particle is . The c-axis direction angle of this particle is (Because the a-axis is perpendicular to the c-axis in layered crystals); the orientation angle directly affects the coordinate transformation of the anisotropic diffusion coefficient and lattice strain in subsequent calculations.

[0025] Apply constrained mechanical boundary conditions to the outer surface of the secondary particles: The normal displacement is fixed at 0 to simulate the compaction state of the electrode material in the battery (i.e., rigid constraint). Tangential displacement is free, allowing secondary particles to slip due to expansion / contraction within the plane; The internal primary grain boundaries are set to mutual support conditions: normal separation between adjacent primary grains is allowed (i.e., intergranular cracks can open), but normal penetration is prohibited; tangential relative sliding is allowed (the friction coefficient can be set to 0.2 to 0.5).

[0026] S2. Coupled solution of lithium-ion diffusion equation, linear elasticity equation, and phase-field damage evolution equation. The lithium-ion diffusion coefficient is anisotropically set along the a-axis and c-axis of the primary particle. Diffusion-induced stress is calculated using anisotropic lattice strain input from X-ray diffraction data. The above steps achieve strong bidirectional coupling of electrochemical, mechanical, and damage multiphysics fields. Anisotropic diffusion and lattice strain describe mass transport and volume changes during lithium insertion / extraction, and phase-field variables are introduced to continuously characterize crack initiation and propagation. This ensures accurate simulation of the interaction between stress, concentration, and damage. Specifically: The lithium-ion diffusion equation uses a modified Fick's second law: ; in, Lithium concentration (mol / m 3 ), initial value =38860, maximum value =46990, For time, Let be the second-order diffusion coefficient tensor, with diagonal elements in the crystal coordinate system (ac axis) as . =5×10⁻¹⁴m 2 / s, =1×10⁻¹⁴m 2 / s, with other components being 0. In the global coordinate system, this is achieved through a rotation matrix. The transformation yields: .in, For the Laplace operator, =3.497×10⁻⁶m 3 / mol represents the partial molar volume of lithium ions. 8.314 J / (mol·K) is the ideal gas constant. =293.15K is the absolute temperature. It is the hydrostatic pressure; For the concentration gradient driving term, This is a stress gradient driving term. Chemical potential; The equations of linear elasticity include: Equilibrium equations (ignoring inertia terms): ; Constitutive equation: ; in, For the isotropic elastic stiffness tensor, derived from the elastic modulus Compared to Poisson Calculate the baseline value of active particles. =93GPa, =0.3, For stress tensor, For the total strain tensor, This is the inelastic strain (anisotropic lattice strain) induced by lithium insertion / extraction. , It is the displacement vector; The diffusion-induced stress was calculated using anisotropic lattice strain, specifically as follows: The lattice constant of the cathode material under different states of charge (SOC, representing the percentage of the battery's current remaining capacity to its rated capacity) was measured using X-ray diffraction experiments. Figure 3 As shown: Along axis a: The lattice constant decreases monotonically with decreasing lithium concentration, where, =2.870Å is the initial value of the a-axis lattice constant. ≈0.04Å represents the variation range of the lattice constant. This represents the maximum lithium concentration. The lattice constant along the c-axis first increases and then decreases with decreasing lithium concentration. The time linearly increases to a maximum value (approximately 14.28 Å), at The situation remained generally stable. The phase transition region (H2→H3) shrinks sharply to approximately 13.5 Å. Piecewise linear interpolation tables were used for input.

[0027] The volumetric strain exhibits a non-linear relationship with the normalized lithium concentration, and the volumetric strain is: ; For each primary particle, based on its crystal orientation angle Transform the strain in the crystal coordinate system to the global coordinate system: ; in, Represents the strain tensor in crystal coordinates. This represents the strain tensor in the global coordinate system. This represents the initial lithium concentration.

[0028] Introducing phase field variables ,in, Indicates that the material is in good condition. Indicating complete cracking, the phase field evolution follows Allen-Cahn type equations: ; Total Energy Function Including elastic strain energy and fracture surface energy: ; in, For elastic strain energy density, =10N / m is the inherent fracture energy release rate. =4.5×10 -3 μm is the phase field length dimension, used to control the width of the crack dispersion region. =1×10⁻⁶m 2 ·N -1 ·s -1 Phase mobility; Substituting the total energy functional into the evolution equation, we obtain the explicit form: .

[0029] Phase field variables represent the continuous transition of a material from intact to cracked and are bidirectionally coupled with the stress field and concentration field, specifically: Phase field → Stress field: Effective stress tensor This indicates that the stiffness of the cracked region has degraded; Phase field → Concentration field: Effective diffusion coefficient This indicates that lithium transport is blocked in the crack region; Stress field / concentration field → Phase field: Elastic strain energy Directly drives crack propagation; concentration field influences This will affect .

[0030] The strong coupling of the above equations is achieved using the "Solid Mechanics," "Dilute Mass Transport," and "Phase Field" interfaces of finite element software (such as COMSOL Multiphysics) combined with custom subroutines. A fully coupled solver (PARDISO or MUMPS) is used, with adaptive time step adjustment (initial step size 0.01s, minimum step size 0.001s). The iteration sequence within each time step is as follows: Based on the current concentration field Calculate anisotropic strain .

[0031] Solve the equations of linear elasticity to obtain the displacement. and stress .

[0032] Calculate elastic strain energy density .

[0033] Solving the phase field evolution equation, we obtain 1.

[0034] according to Correct the diffusion coefficient and stiffness, and update the concentration field. .

[0035] After convergence, proceed to the next time step.

[0036] S3. Perform constant current charge-discharge cycle simulation and output the first principal stress distribution, phase field damage variable distribution and crack evolution path. These outputs are used to identify the most dangerous stress concentration areas, determine the location and propagation trend of cracks, and provide basic data for subsequent parameter evaluation.

[0037] The charge / discharge rate ranges from 0.5C to 4C; material mechanical parameters include elastic modulus. (50-150 GPa), Poisson's ratio (0.2–0.4) and inherent fracture energy release rate (5~50N / m).

[0038] In this embodiment, a specific implementation method is given: (1) Charging and discharging conditions settings: Charge / discharge rate: 0.5C, 1C, 2C, 3C, and 4C are selectable. Taking 1C as an example, the nominal cycle time... =14400 seconds.

[0039] Charging method: Constant current (CC) charging, cutoff voltage 4.25V (relative to Li / Li). + ).

[0040] Discharge method: constant current discharge, cutoff voltage 3.6V.

[0041] Simulated temperature: constant at 293.15K.

[0042] Number of cycles: Usually simulates 1 to 200 cycles, and the maximum number of cycles can be set as needed (such as 500 cycles).

[0043] Normalized time: =0 charging starts, =0.5 Charging ends / Discharging begins =1 Discharge ends.

[0044] (2) Output content: First principal stress distribution: Output the first principal stress at each time step (or every 100 seconds). The entire field cloud map, as well as line distribution maps along representative grain boundaries and interfaces. The first principal stress is used to identify the most dangerous tensile regions, where cracks typically appear. At the peak.

[0045] Phase field damage variable distribution: Output phase field variables The full cloud map. The area with a value of 1 indicates that it has completely cracked; The region with a value of 0.5-0.9 indicates damage accumulation. The output frequency is at the end of each cycle or every 0.01 normalized time steps.

[0046] Crack evolution path: by tracing For continuous regions ≥0.9, the morphology of cracks is automatically identified. Output: The curve of crack area (or length) versus number of cycles; Location where the crack first appears (first time) =1 (element coordinates). The direction and speed of crack propagation (per unit time) =1 (extended distance of the region).

[0047] (3) Simulation convergence and termination: The simulation will be terminated prematurely if any of the following conditions are met: The maximum first principal stress exceeds the tensile strength of the material (>500 MPa); Phase field variables The total area of ​​the region with a value of 1 exceeds 30% of the total area of ​​the secondary particles; The capacity retention rate is less than 50%.

[0048] S4. By comparing the stress and damage distribution under different charge / discharge rates, primary particle morphology configurations, or material mechanical parameters, the influence of each parameter on the failure path is evaluated, providing quantitative guidance for material optimization (such as selecting radial orientation configurations and low Poisson's ratio materials) and process design.

[0049] The specific assessment of the impact of each parameter on the failure path is as follows: The single-parameter scanning method was used to record the peak value of the first principal stress, the average damage variable, and the percentage of intergranular crack area under different parameter combinations. Calculate the local sensitivity coefficient. The larger the absolute value of the sensitivity coefficient, the more significant the influence of this parameter on the failure path.

[0050] According to the results of local sensitivity analysis, Poisson's ratio is... The sensitivity coefficient to stress level is the highest, and the elastic modulus is the highest. Secondly, the inherent fracture energy release rate It has the lowest sensitivity coefficient.

[0051] It also includes a comparison of failure paths under different charge / discharge rates. For example, at low C rates (≤1C), cracks mainly propagate along grain boundaries, while at high C rates (≥2C), concentration polarization intensifies, stress levels increase, intergranular crack area increases, and damage propagates towards the grain core.

[0052] This embodiment is based on a specific implementation method: (1) Parameter classification and value range: Based on the research content, the parameters that need to be evaluated include: Charge / discharge rates: 0.5C, 1C, 2C, 3C, 4C (other parameters are fixed).

[0053] Primary particle morphological configurations: random orientation configuration (H), radial orientation configuration (R), and core-shell configuration (C). See above for the definitions and generation methods of these configurations.

[0054] Material mechanical parameters: elastic modulus (75–150 GPa), Poisson's ratio (0.2~0.4) fracture energy (5~50N / m).

[0055] (2) Single-parameter scanning method: Fix other parameters to the baseline value ( =93GPa, =0.3, =10N / m, 1C magnification, random orientation configuration). The target parameter is uniformly selected from 5 to 8 values ​​within its range (e.g., ...). =0.24, 0.27, 0.30, 0.33, 0.36), run a complete simulation (usually 100 laps) for each parameter value. Record the three failure indicators corresponding to each parameter value: Peak value of the first principal stress (The maximum value during the entire loop); Mean damage variable ; Percentage of intergranular crack area ( The proportion of the area of ​​the secondary particles to the total area of ​​the secondary particles with a density of ≥0.9.

[0056] (3) Calculation of local sensitivity coefficient: For parameters The local sensitivity coefficient is defined as: ; in, The output value is the value under the reference parameter value. As the baseline parameter value, This represents the change in the output value after the parameter is changed. This refers to the change in parameters; typically, the sensitivity between multiple adjacent points is calculated and then averaged. When this occurs, the parameter is considered a sensitive parameter. In this embodiment, The sensitivity is approximately 0.88. Approximately 0.52, Only 0.15.

[0057] (4) Comparison of failure paths at different charge / discharge rates: Following step S3, set different C-multipliers and compare the following characteristics: At low carbon ratios (≤1C), the concentration gradient is gentle, the stress peak is low (approximately 120 MPa at 1C), cracks mainly propagate along grain boundaries, and damage is concentrated at the outer edge of the grains (1C).

[0058] At high C ratios (≥2C), concentration polarization intensifies, the peak stress increases significantly (approximately 280 MPa at 4C), the area of ​​intergranular cracks increases (approximately 12% at 4C), damage extends from the particle surface to the core region, and transgranular cracks appear in some areas (4C).

[0059] (5) Comparison of different primary particle morphologies: Modeling the three configurations separately: Random orientation (H): The orientation angles of all primary particles are random.

[0060] Radial orientation (R): The c-axis points towards the particle center, and the a-axis is tangential. This can be achieved by assigning an orientation angle to each particle during Voronoi generation. To achieve, among which, , The coordinates of the secondary particle center are... , The coordinates are the seed point coordinates of the primary particle.

[0061] Core and shell (C): Internal particles (radius < 70% of particle radius) are radially oriented, while external particles are randomly oriented.

[0062] Simulation results show that the R configuration has the most uniform concentration distribution, the lowest central stress (approximately 80 MPa), and the least grain boundary damage (approximately 2% of the area); the H configuration has the highest stress (approximately 140 MPa) and the most severe damage (approximately 8% of the area); the C configuration is in the middle.

[0063] (6) Output format and optimization suggestions: Generate the following charts for user decision-making: Sensitivity histogram: The horizontal axis represents the parameter name ( , , The vertical axis is .

[0064] Parameter-failure index curves: such as the "Poisson's ratio-peak value of the first principal stress" curve and the "C ratio-intergranular crack area" curve.

[0065] Comprehensive recommendations: For example, to suppress chemical-mechanical failure, it is recommended to use a radial orientation configuration, a working rate ≤1C, a Poisson's ratio of active particles ≤0.3, and to control the modulus difference between active particles and electrolyte within 30GPa.

[0066] Step S4, by comparing stress and damage distribution under different parameters, assesses the influence of each parameter on the failure path, and may include the following steps: Comparison of failure paths under different charge / discharge rates.

[0067] Comparison of failure paths under different primary particle morphologies.

[0068] Sensitivity analysis under different material mechanical parameters.

[0069] Comprehensive evaluation and sensitivity coefficient calculation.

[0070] Through the specific implementation of the above four steps, this embodiment can accurately predict the chemical-mechanical failure path of NCM cathode materials under different operating conditions, providing a quantitative basis for crack-resistant material design and battery life assessment.

[0071] like Figure 4 As shown, the multi-scale mechanical evolution mechanism and failure path prediction system for energy storage batteries includes: Model building module: used to build a geometric model containing multiple randomly oriented primary particles based on focused ion beam scanning electron microscope images, wherein the grain boundaries of the primary particles are generated by the Voronoi algorithm, and each primary particle is assigned a random crystal orientation angle; Coupled solution module: used to couple the solution of lithium-ion diffusion equation, linear elasticity equation and phase field damage evolution equation, where the lithium-ion diffusion coefficient is set anisotropically along the a-axis and c-axis, diffusion-induced stress is calculated through anisotropic lattice strain, and phase field variables are bidirectionally coupled with stress field and concentration field; Path prediction module: used to perform constant current charge-discharge cycle simulation, outputting the first principal stress distribution, phase field damage variable distribution, and crack evolution path; Parameter evaluation module: Used to evaluate the influence of each parameter on the chemical-mechanical failure path by comparing stress and damage distribution under different parameter combinations.

[0072] The methods and systems provided above can be directly applied to the following scenarios: Rapidly screen primary particle orientation configurations (such as radial orientation) that have crack resistance capabilities. Determine the optimal charge / discharge rate window to avoid rapid failure at high rates; Guiding grain boundary engineering and microstructure design (such as controlling Poisson's ratio and modulus matching); Predict cycle life under given operating conditions (based on the correlation model between crack area and capacity decay).

[0073] Compared with the traditional trial-and-error method, this application can shorten the material development and verification cycle and significantly reduce R&D costs.

[0074] Therefore, the present invention adopts the above-mentioned multi-scale mechanical evolution mechanism and failure path prediction method and system of energy storage battery, which can accurately identify key failure influencing factors, guide the design of crack-resistant microstructure, and provide efficient and parameterized prediction tools for fast charging safety assessment, material optimization and full life cycle management of energy storage battery, significantly reducing R&D costs and shortening the development cycle.

[0075] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for predicting the multi-scale mechanical evolution mechanism and failure path of energy storage batteries, characterized in that, Includes the following steps: S1. Construct a geometric model containing multiple randomly oriented primary particles based on focused ion beam scanning electron microscope images. The grain boundaries of the primary particles are generated by the Voronoi algorithm, and each primary particle is assigned a random crystal orientation angle. S2. Coupled solution of lithium-ion diffusion equation, linear elasticity equation and phase field damage evolution equation, wherein the lithium-ion diffusion coefficient is set anisotropically along the a-axis and c-axis of the primary particle, and the diffusion-induced stress is calculated by anisotropic lattice strain input by X-ray diffraction data. S3. Perform constant current charge-discharge cycle simulation and output the first principal stress distribution, phase field damage variable distribution and crack evolution path; S4. By comparing the stress and damage distribution under different charge / discharge rates, primary particle morphology configurations, or material mechanical parameters, assess the degree of influence of each parameter on the failure path.

2. The method for predicting the multi-scale mechanical evolution mechanism and failure path of energy storage batteries according to claim 1, characterized in that, Step S1 is as follows: Focused ion beam scanning electron microscopy was used to perform continuous cross-sectional imaging of secondary particles of cathode material to obtain a series of cross-sectional images. Image recognition software was used to extract the boundaries of each cross-sectional image to obtain the morphological parameters of each primary particle. Based on the extracted morphological parameters, a two-dimensional polygon aggregation model is generated using the Voronoi algorithm; Each primary particle is assigned a random crystal orientation angle. , defined as the angle between the crystallographic a-axis and c-axis of the primary particle and the x-axis of the global coordinate system, where the a-axis direction angle of the primary particle is . The c-axis direction angle of this particle is ; Apply constrained mechanical boundary conditions to the outer surface of the secondary particles: The normal displacement is fixed at 0 to simulate the compaction state of the electrode material in the battery. Tangential displacement is free, allowing secondary particles to slip due to expansion / contraction within the plane; The internal primary grain boundaries are set to mutual support conditions: normal separation is allowed between adjacent primary grains, but normal penetration is prohibited; relative sliding is allowed in the tangential direction.

3. The method for predicting the multi-scale mechanical evolution mechanism and failure path of energy storage batteries according to claim 2, characterized in that, Based on the extracted morphological parameters, a two-dimensional polygon aggregation model is generated using the Voronoi algorithm, specifically: Randomly generated within the computational domain Each seed point represents the number of primary particles; Lloyd relaxation iteration is performed on the seed points to make the polygon size distribution more uniform, which is used to avoid polygons that are too small or distorted. Construct a Voronoi diagram, where each Voronoi unit is a primary grain and the common edge between Voronoi units is a grain boundary; The generated polygons are smoothed to approximate the irregular shapes of real particles.

4. The method for predicting the multi-scale mechanical evolution mechanism and failure path of energy storage batteries according to claim 1, characterized in that, The lithium-ion diffusion equation in step S2 adopts a modified Fick's second law: ； in, Lithium concentration, For time, It is the second-order diffusion coefficient tensor. This represents the partial molar volume of lithium ions. Let be the ideal gas constant. Absolute temperature It is the hydrostatic pressure; The equations of linear elasticity include: Equilibrium equations: ; Constitutive equation: ; in, For the isotropic elastic stiffness tensor, For stress tensor, For the total strain tensor, For lithium insertion / extraction induced inelastic strain, , It is the displacement vector; Introducing phase field variables ,in, Indicates that the material is in good condition. Indicating complete cracking, the phase field evolution follows Allen-Cahn type equations: ； Total Energy Function Including elastic strain energy and fracture surface energy: ； in, For elastic strain energy density, The inherent fracture energy release rate, The phase field length dimension is used to control the width of the crack dispersion region. Phase mobility; Substituting the total energy functional into the evolution equation, we obtain the explicit form: 。 5. The method for predicting the multi-scale mechanical evolution mechanism and failure path of energy storage batteries according to claim 4, characterized in that, In step S2, the diffusion-induced stress is calculated using anisotropic lattice strain, specifically as follows: The lattice constant of the cathode material under different battery charge states was measured by X-ray diffraction experiments. Along axis a: The lattice constant decreases monotonically with decreasing lithium concentration, where, The initial value of the a-axis lattice constant is given. The magnitude of the change in lattice constant. This represents the maximum lithium concentration. The lattice constant along the c-axis first increases and then decreases with decreasing lithium concentration, and undergoes a sudden contraction in the phase transition region; The volumetric strain exhibits a non-linear relationship with the normalized lithium concentration, and the volumetric strain is: ； For each primary particle, based on its crystal orientation angle Transform the strain in the crystal coordinate system to the global coordinate system: ； in, Represents the strain tensor in crystal coordinates. This represents the strain tensor in the global coordinate system. This represents the initial lithium concentration.

6. The method for predicting the multi-scale mechanical evolution mechanism and failure path of energy storage batteries according to claim 5, characterized in that, The phase field variables in step S2 represent the continuous transition of the material from intact to cracked and are bidirectionally coupled with the stress field and concentration field, specifically: Phase field → Stress field: Effective stress tensor This indicates that the stiffness of the cracked region has degraded; Phase field → Concentration field: Effective diffusion coefficient This indicates that lithium transport is blocked in the crack region; Stress field / concentration field → Phase field: Elastic strain energy Directly drives crack propagation; concentration field influences This will affect .

7. The method for predicting the multi-scale mechanical evolution mechanism and failure path of energy storage batteries according to claim 1, characterized in that, In step S3, the charge / discharge rate ranges from 0.5C to 4C; the material mechanical parameters include elastic modulus, Poisson's ratio, and inherent fracture energy release rate; the output first principal stress distribution, phase field damage variable distribution, and crack evolution path are used to identify the most dangerous stress concentration areas and determine the location and propagation trend of cracks.

8. The method for predicting the multi-scale mechanical evolution mechanism and failure path of energy storage batteries according to claim 1, characterized in that, Step S4 specifically assesses the impact of each parameter on the failure path as follows: The single-parameter scanning method was used to record the peak value of the first principal stress, the average damage variable, and the percentage of intergranular crack area under different parameter combinations. Calculate the local sensitivity coefficient for the parameter The local sensitivity coefficient is defined as: ； in, The output value is the value under the reference parameter value. As the baseline parameter value, This represents the change in the output value after the parameter is changed. For parameter changes; when When this happens, the parameter is considered a sensitive parameter.

9. A system for predicting the multi-scale mechanical evolution mechanism and failure path of energy storage batteries, employing the method for predicting the multi-scale mechanical evolution mechanism and failure path of energy storage batteries as described in any one of claims 1-8, characterized in that, include: Model building module: used to build a geometric model containing multiple randomly oriented primary particles based on focused ion beam scanning electron microscope images, wherein the grain boundaries of the primary particles are generated by the Voronoi algorithm, and each primary particle is assigned a random crystal orientation angle; Coupled solution module: used to couple the solution of lithium-ion diffusion equation, linear elasticity equation and phase field damage evolution equation, where the lithium-ion diffusion coefficient is set anisotropically along the a-axis and c-axis, diffusion-induced stress is calculated through anisotropic lattice strain, and phase field variables are bidirectionally coupled with stress field and concentration field; Path prediction module: used to perform constant current charge-discharge cycle simulation, outputting the first principal stress distribution, phase field damage variable distribution, and crack evolution path; Parameter evaluation module: Used to evaluate the influence of each parameter on the chemical-mechanical failure path by comparing stress and damage distribution under different parameter combinations.

Images