Battery casing thermal runaway safety prediction method and system

By combining thermal cycling load experiments and thermo-mechanical coupling simulation models of battery casings with yield strength parameters, the safety of battery casings under thermal runaway is predicted. This solves the problem of insufficient assessment of material performance degradation due to long-term thermal cycling in existing technologies, and realizes structural safety assessment throughout the entire life cycle.

CN122864049BActive Publication Date: 2026-11-17CHANGSHU INSTITUTE OF TECHNOLOGY
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202611356360.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2026-09-03
Publication Date
2026-11-17
Estimated Expiration
2046-09-03

Smart Images

  • Figure CN122864049B_ABST
    Figure CN122864049B_ABST
Patent Text Reader

Abstract

This invention provides a method and system for predicting the safety of battery casings under thermal runaway, belonging to the field of battery safety assessment technology. The invention obtains the degradation curve of the casing material's yield strength with the number of cycles through cyclic thermal loading experiments; establishes a thermo-mechanical coupling simulation model of the casing, assigning the yield strength corresponding to each evaluation node to the model to obtain simulation models under different aging states; applies thermal runaway loads to each aging model, and solves for the mechanical response results of the casing under each aging state; compares the mechanical response results with a preset threshold, calculates the safety margin, and plots a safety evaluation curve that changes with the number of cycles. This invention correlates the degradation of the casing material caused by long-term thermal cycling with the structural response under thermal runaway impact, realizing the safety prediction of the battery casing throughout its entire life cycle.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of battery safety assessment technology, specifically to a method and system for predicting the safety of thermal runaway of battery casings. Background Technology

[0002] During the charging and discharging process, lithium-ion batteries generate heat internally, causing the battery casing temperature to rise. During the resting or cooling phase, the casing temperature drops back to ambient temperature. Each charge-discharge cycle involves a "heating-cooling" thermal cycle in the casing. As the number of cycles accumulates, the casing material gradually degrades under repeated thermal expansion and contraction, primarily manifested as decreased yield strength and increased brittleness.

[0003] When a battery experiences thermal runaway, high-temperature gases and high-pressure ejecta are rapidly generated inside, and the casing is subjected to severe thermal and pressure shocks in a very short time. Whether the casing can maintain its structural integrity without cracking or excessive deformation under the impact of thermal runaway directly determines the severity of the thermal runaway accident.

[0004] Currently, there are two main technical directions in the field of battery safety assessment: one focuses on early warning of thermal runaway inside the cell, and determines whether thermal runaway will occur by monitoring electrical parameters such as voltage, current and temperature; the other focuses on the instantaneous structural response of the casing under thermal runaway impact, and calculates the deformation and stress distribution of the casing under high temperature and high pressure through simulation.

[0005] However, the aforementioned existing technologies all have the following shortcomings: First, existing thermal runaway early warning technologies focus on the electrochemical state inside the battery cell, without addressing the structural and mechanical properties of the battery casing, thus failing to answer the structural safety question of whether the casing can withstand thermal runaway. Second, the material parameters used in the existing thermal runaway simulation technology for the casing are all fresh material parameters that have not undergone cyclic aging. They do not take into account the degradation of casing material performance caused by repeated thermal cycling during long-term battery service, and their evaluation results are biased towards danger. Third, there is a lack of a full life-cycle shell safety prediction method that correlates long-term thermal cycling aging with short-term thermal runaway impact; Therefore, it is necessary to develop a predictive method that can take into account the degradation of shell materials caused by cyclic thermal loads and, based on this, assess the safety of shell structures under thermal runaway shocks.

[0006] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0007] The purpose of this invention is to provide a method and system for predicting the safety of thermal runaway of battery casings, so as to solve the problems mentioned in the background art.

[0008] To achieve the above objectives, the present invention provides the following technical solution: The method for predicting the safety of battery casing thermal runaway includes the following steps: Step 1: Obtain the thermal cycling load spectrum of the battery casing material to be evaluated, conduct cyclic thermal loading experiments on the battery casing material according to the thermal cycling load spectrum, and obtain the change data of yield strength with the number of cycles by performing mechanical property tests at each preset cycle number node, and fit the change data into a performance degradation curve; Step 2: Establish a thermo-mechanical coupling simulation model of the battery casing to be evaluated. Select several cycle number nodes as evaluation nodes from the cycle number node range of the battery casing in the cyclic thermal loading experiment. Extract the corresponding yield strength parameter from the performance degradation curve according to each evaluation node, and input the yield strength parameter as a material property into the thermo-mechanical coupling simulation model to obtain the aging state simulation model corresponding to each evaluation node. Step 3: Input the preset thermal runaway load into the aging state simulation model corresponding to each evaluation node, and solve for the mechanical response results of the battery casing under each evaluation node. The thermal runaway load includes the temperature evolution curve and internal pressure evolution curve that change with time. Step 4: Set a critical deformation threshold for the battery casing, compare the thermal runaway mechanical response results at each evaluation node with the critical deformation threshold to obtain the safety margin corresponding to each evaluation node, and establish a safety evaluation curve that evolves with the number of cycles based on the safety margin.

[0009] Furthermore, the thermal cycling load spectrum of the battery casing material to be evaluated is obtained, and cyclic thermal loading experiments are performed on the battery casing material to obtain the performance degradation curve. The specific steps are as follows: Obtain the charge / discharge rate, heat generation parameters, and service environment information of the battery cells to be evaluated under preset operating conditions; Based on the heat generation parameters and service environment information, determine the highest service temperature, lowest service temperature, and heating rate to reach the highest service temperature of the battery casing in a single cycle. Based on the charge / discharge rate, determine the duration of a single continuous eruption of the battery casing at the highest service temperature, and set the duration of the single continuous eruption as the high-temperature holding time of the thermal cycling load spectrum; Based on the highest service temperature, lowest service temperature, heating rate, and high temperature holding time, an alternating heat load curve with time as the horizontal axis and temperature as the vertical axis is constructed as the thermal cycle load spectrum. Prepare multiple sets of standard mechanical specimens made of the same material as the battery casing to be evaluated, and perform reciprocating thermal cycling loading according to the thermal cycling load spectrum described above; At each preset cycle number node, standard mechanical specimens of the corresponding group are taken out for mechanical property testing to obtain data on the change of yield strength with the number of cycles. Using the number of cycles as the independent variable and the corresponding yield strength parameter as the dependent variable, a performance degradation curve characterizing the relationship between the mechanical properties of the material and the number of cycles is obtained by fitting.

[0010] Furthermore, a thermo-mechanical coupling simulation model of the battery casing to be evaluated is established. The specific steps are as follows: Obtain the geometric dimensions of the battery casing to construct a three-dimensional geometric model, and then divide the three-dimensional geometric model into several finite mesh elements using the finite element method. Define the basic material properties of the battery casing and assign the basic material properties to each mesh cell. The basic material properties include: thermal conductivity, specific heat capacity, elastic modulus, Poisson's ratio and coefficient of thermal expansion. A material parameter interface is set, wherein the yield strength parameter is set as an evolution variable to receive yield strength parameter data corresponding to different evaluation nodes; The boundary condition setting interface includes: defining the thermal load boundary and mechanical load boundary for inputting the thermal runaway load; Thermal boundary conditions are set, including: applying a transient heat flux density load triggered by thermal runaway to the inner surface of the three-dimensional geometric model, and applying the convective heat transfer coefficient and the radiative heat dissipation coefficient with the ambient air to the outer surface. The transient temperature field distribution of the inner wall of the battery casing is calculated by solving the three-dimensional heat conduction equation based on the transient heat flux density load, the convective heat transfer coefficient and the radiative heat dissipation coefficient, combined with the thermal conductivity and specific heat capacity. Set mechanical boundary conditions, which include: setting fixed constraints as displacement boundary conditions at the connection parts of the three-dimensional geometric model, and applying the internal expansion pressure load generated by thermal runaway as an external load input to the inner surface mesh elements of the three-dimensional geometric model. Based on the thermal expansion coefficient of the battery casing material, the transient temperature field distribution is transformed into the thermal strain of each grid element, and combined with the elastic modulus and Poisson's ratio, it is mapped to the corresponding internal thermal stress. The internal thermal stress and the internal expansion pressure load are vector-superimposed as the total load for structural analysis and applied to the three-dimensional geometric model to construct a thermo-mechanical coupling simulation model.

[0011] Furthermore, the yield strength parameter is input as a material property into the thermo-mechanical coupling simulation model to obtain the aging state simulation model corresponding to each evaluation node. The specific steps are as follows: For each evaluation node, the yield strength parameter at the corresponding cycle number node is extracted from the performance degradation curve; The extracted yield strength parameter is assigned to the battery casing material properties in the thermo-mechanical coupling simulation model; Keeping the geometric dimensions and displacement boundary conditions in the thermo-mechanical coupling simulation model unchanged, only changing the yield strength parameter value in the battery casing material properties, an aging state simulation model corresponding to each evaluation node is generated.

[0012] Furthermore, the thermal runaway load is obtained in the following manner: Obtain the characteristic parameters of the battery cell to be evaluated under thermal runaway triggered state, including the instantaneous maximum temperature of the ejected material, the peak pressure, and the duration of continuous ejection; Based on the aforementioned characteristic parameters, a numerical distribution function is defined to construct temperature evolution curves and internal pressure evolution curves that change over time. The temperature evolution curve and the internal pressure evolution curve are respectively loaded into the aging state simulation model as input excitations for the thermal boundary conditions and mechanical boundary conditions.

[0013] Furthermore, loading the temperature evolution curve and internal pressure evolution curve into each of the aging state simulation models specifically involves: The heat transfer hysteresis parameter is set according to the distance from the point of action of the ejected object to various parts of the battery casing, and the temperature evolution curve is mapped to the simulation model of each aging state in the form of a non-uniform temperature field. The internal pressure evolution curve is applied to the inner surface of each aging state simulation model with a dynamic surface load that varies with time.

[0014] Furthermore, the preset thermal runaway load is input into the aging state simulation model corresponding to each evaluation node, and the mechanical response results of the battery casing under each evaluation node are obtained by solving the problem. The specific steps are as follows: The aging state simulation model is solved nonlinearly and incrementally using a finite element solver. At each time step, the structural equilibrium equations are solved simultaneously, and the displacement vectors of each evaluation node of the battery casing are obtained synchronously. Based on the displacement vector, the geometric deformation ratio at various points of the battery casing is calculated. Combined with the material parameters of the pre-assigned aging state simulation model, the stress distribution inside the battery casing is calculated through the stress-strain conversion relationship. The stress distribution includes: a transient stress field reflecting the instantaneous stress intensity, and a plastic strain field reflecting whether the material has suffered permanent damage. From the displacement vector, transient stress field and plastic strain field mentioned above, the maximum values ​​at each evaluation node are extracted and used as the highest displacement, maximum equivalent stress and maximum plastic strain at that evaluation node, respectively.

[0015] Furthermore, the specific steps to obtain the safety margin corresponding to each evaluation node are as follows: Obtain the assembly gap between the battery casing and internal components, and set the critical deformation threshold of the battery casing based on the assembly gap; The ratio of the highest displacement to the critical deformation threshold in the thermal runaway mechanical response results at each evaluation node is calculated to obtain the deformation safety margin. The strength safety margin is calculated based on the yield strength parameters corresponding to each evaluation node and the maximum equivalent stress in the thermal runaway mechanical response results under each evaluation node. A failure strain threshold is preset for the battery casing material to be evaluated. The strain safety margin is obtained by comparing the maximum plastic strain in the thermal runaway mechanical response results at each evaluation node with the failure strain threshold. For each assessment node, the corresponding deformation safety margin, strength safety margin, and strain safety margin are compared, and the minimum value is selected as the comprehensive safety margin index for that assessment node.

[0016] Furthermore, a safety evaluation curve evolving with the number of cycles is established based on the aforementioned safety margin. The specific steps are as follows: Obtain the comprehensive safety margin index under each evaluation node, and plot the safety evaluation curve with each evaluation node as the horizontal axis and the comprehensive safety margin index as the vertical axis. Identify the abrupt change in the slope of the safety evaluation curve, and determine the evaluation node corresponding to the abrupt change in slope as the safety failure node of the battery casing.

[0017] The present invention also provides a battery casing thermal runaway safety prediction system, the prediction system being used to execute the above-described battery casing thermal runaway safety prediction method, comprising: The performance degradation acquisition module is used to acquire the thermal cycling load spectrum of the battery casing material to be evaluated, conduct cyclic thermal loading experiments on the battery casing material according to the thermal cycling load spectrum, and obtain the change data of yield strength with the number of cycles by performing mechanical property tests at each preset cycle number node, and fit the change data into a performance degradation curve. The simulation modeling module is used to establish a thermo-mechanical coupling simulation model of the battery casing to be evaluated. From the range of cycle number nodes of the battery casing in the cyclic thermal loading experiment, several cycle number nodes are selected as evaluation nodes. The corresponding yield strength parameter is extracted from the performance degradation curve according to each evaluation node, and the yield strength parameter is input as a material property into the thermo-mechanical coupling simulation model to obtain the aging state simulation model corresponding to each evaluation node. The response calculation module is used to input the preset thermal runaway load into the aging state simulation model corresponding to each evaluation node, and solve for the mechanical response results of the battery casing under each evaluation node. The thermal runaway load includes the temperature evolution curve and internal pressure evolution curve that change with time. The margin evaluation module is used to preset the critical deformation threshold of the battery casing, compare the thermal runaway mechanical response results at each evaluation node with the critical deformation threshold, obtain the safety margin corresponding to each evaluation node, and establish a safety evaluation curve that evolves with the number of cycles based on the safety margin.

[0018] Compared with the prior art, the beneficial effects of the present invention are: This invention combines material mechanical property degradation experiments with thermo-mechanical coupling simulation. The performance degradation curve of the shell material with the number of cycles is obtained through experiments. Then, the degraded material parameters are input into the simulation model to calculate the mechanical response under thermal runaway shock. This realizes the correlation assessment between "long-term thermal cycling aging" and "short-term thermal runaway shock". At the same time, by calculating the deformation safety margin, strength safety margin and strain safety margin and taking the minimum value as the comprehensive safety margin index, a safety evaluation curve of the safety margin changing with the number of cycles is plotted. This provides an intuitive and quantitative basis for the design optimization of battery shell and the determination of replacement cycle. This invention also obtains the degradation curve of the yield strength of the shell material with the number of cycles through cyclic thermal loading experiments, and uses this degradation data as input to the simulation model to realize the structural safety assessment of the battery shell at different aging stages throughout its entire life cycle. This solves the problem that the existing technology only focuses on the instantaneous deformation of the shell in a single thermal runaway event, while ignoring the degradation of the shell material performance caused by long-term thermal cycling. Attached Figure Description

[0019] Figure 1 This is a flowchart of the steps of the present invention; Figure 2 This is a graph showing the performance degradation of the yield strength of the present invention with the number of cycles; Figure 3 This is a functional block diagram of the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0021] It should be noted that, unless otherwise defined, the technical or scientific terms used in this invention should have the ordinary meaning understood by one of ordinary skill in the art to which this invention pertains. The terms "first," "second," and similar terms used in this invention do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects. Terms such as "connected" or "linked" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. Terms such as "upper," "lower," "left," and "right" are used only to indicate relative positional relationships; when the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0022] Example: Please see Figure 1-2 The present invention provides a technical solution: The method for predicting the safety of battery casing thermal runaway includes the following steps: Step 1: Obtain the thermal cycling load spectrum of the battery casing material to be evaluated. Conduct cyclic thermal loading experiments on the battery casing material according to the thermal cycling load spectrum. By performing mechanical property tests at each preset cycle number node, obtain the change data of yield strength with the number of cycles, and fit the change data into a performance degradation curve.

[0023] Specifically, the thermal cycling load spectrum is used to simulate the repeated temperature changes experienced by the battery casing due to battery heat generation and ambient temperature variations during actual charge and discharge operations. The accuracy of its construction directly determines the degree to which subsequent aging experiments reproduce real service conditions. First, the charge / discharge rate, heat generation parameters, and service environment information of the battery cell to be evaluated under preset conditions are obtained.

[0024] The charge / discharge rate refers to the current rate of a single battery cell during charging or discharging, measured in C. 1C represents the current required to fully charge or discharge the battery in one hour. The charge / discharge rate characterizes the speed of battery charging and discharging, directly affecting the heat generation power and total heat output during the process, and is a fundamental input for determining the casing temperature level. For example, in fast charging scenarios, the charge / discharge rate can be set to 1C; in standard charging and discharging scenarios, it can be set to 0.3C or 0.5C.

[0025] Heat generation parameters include the heat generated by a single battery cell during charging and discharging due to internal resistance, polarization effects, and electrochemical reactions. These parameters can be obtained through charge-discharge testing of the battery or from battery cell specifications or publicly available literature. This embodiment uses the Bernardi heat generation model to characterize the heat generation rate.

[0026] in, The heat generation rate of a single battery cell, expressed in W. This is the charging and discharging current, measured in amperes (A). This is the open-circuit voltage, measured in volts (V). This is the terminal voltage, in volts (V). This is the absolute temperature of the battery, expressed in Kelvin (K). The temperature coefficient of the open-circuit voltage, i.e., the entropy coefficient, is expressed in V / K. The first term on the right-hand side of the formula represents irreversible heat, caused by ohmic resistance and polarization effects; its value equals the product of current and overpotential, and is always positive. The second term represents reversible entropy heat, caused by the entropy change of the electrochemical reaction; its sign depends on the sign of the entropy coefficient, characterizing the endothermic or exothermic contribution of the electrochemical reaction. This model was chosen because it decomposes heat generation into separately calibrable irreversible and reversible terms based on energy conservation. Both terms can be obtained from voltage, current, and temperature data in charge-discharge tests, resulting in higher calculation accuracy than the simplified approach that approximates heat generation using only ohmic heat. This provides a reliable quantitative basis for subsequently determining the maximum service temperature and heating rate of the casing.

[0027] Service environment information refers to the external ambient temperature conditions under which the battery operates during actual use, including the highest ambient temperature, the lowest ambient temperature, and the range of ambient temperature variation. These values ​​depend on the specific application scenario of the battery. For example, for batteries used in electric vehicles, the service environment temperature range can be set to -20℃ to 45℃; for batteries used in energy storage power stations, the service environment temperature range can be set to -10℃ to 40℃; and for batteries used in outdoor energy storage devices in tropical regions, the service environment temperature range can be set to 0℃ to 50℃. These values ​​are illustrative examples and can be adjusted according to specific usage scenarios and regional climate conditions in actual applications.

[0028] Based on heat generation parameters and service environment information, the highest and lowest service temperatures of the battery casing, as well as the rate of temperature rise to reach the highest service temperature, are determined during a single charge-discharge cycle. The heat generated by the battery during charging and discharging is transferred to the battery casing through heat conduction, convection, and radiation, causing the casing temperature to rise. Simultaneously, heat exchange occurs between the casing and the external environment through natural convection and radiation, causing the casing temperature to decrease. Under the combined effect of heat generation during charging and discharging and heat dissipation from the external environment, the casing temperature exhibits an alternating pattern of first rising and then falling within a charge-discharge cycle. The highest service temperature refers to the peak temperature reached by the casing during charging and discharging, determined jointly by heat generation parameters and service environment information. For example, when a single battery cell is continuously discharged at a 1C rate and the ambient temperature is 25°C, the highest service temperature of the casing can reach 55°C to 65°C. The lowest service temperature refers to the lowest temperature the casing temperature drops to when the battery is at rest, primarily determined by service environment information. For example, when the ambient temperature is -10°C, the lowest service temperature of the casing can approach -5°C to 0°C. The heating rate refers to the rate at which the casing temperature rises from the minimum service temperature to the maximum service temperature. It is influenced by both the battery's heat generation power and the casing's heat capacity; the greater the heat generation power and the smaller the casing's heat capacity, the faster the heating rate. For example, under 1C discharge conditions, the casing's heating rate can reach 2°C / min to 5°C / min.

[0029] The duration of a single sustained burst of heat from the battery casing at the highest service temperature is determined based on the charge / discharge rate, and this duration is set as the high-temperature hold time in the thermal cycling load spectrum. The high-temperature hold time refers to the length of time the casing temperature remains near the highest service temperature during a complete charge / discharge cycle. It should be noted that the duration of a single sustained burst of heat from the battery casing, in the context of thermal runaway loads, describes the duration of the transient thermal runaway process. However, in the context of the cyclic aging test in this step, it characterizes the duration of a single sustained burst of heat from the casing at the highest service temperature. Both correspond to the thermal damage accumulation process of the casing material undergoing short-term high-temperature exposure, and therefore the same parameter name is used for unified characterization. The high-temperature hold time and the charge / discharge rate are inversely proportional.

[0030] in, This refers to the time the temperature is maintained at high temperature, expressed in minutes. This refers to the charge / discharge rate, expressed in C. This is a calibration coefficient related to battery thermal capacity and heat dissipation conditions, expressed in min·C. The physical basis for this inverse relationship is that the higher the charge / discharge rate, the shorter the time required for the battery to complete one charge / discharge cycle, and the shorter the time the casing remains near its peak temperature; conversely, the lower the rate, the longer the high-temperature retention time. The calibration coefficient k is determined through preliminary experiments. The specific calibration process is as follows: a reference rate is selected, exemplarily 1C, and a complete charge / discharge cycle is performed on a single battery cell at this rate. The duration for which the casing temperature remains near the maximum service temperature is measured. ,Depend on The calibration coefficients are calculated, where This is a reference rate. For example, when the charge / discharge rate is 1C, the high-temperature holding time can be set to 15 minutes; when the rate is increased to 2C, the high-temperature holding time is correspondingly shortened to 7 to 8 minutes. By setting the high-temperature holding time through this relationship, the cumulative amount of thermal damage under different rate conditions is comparable, ensuring that the cyclic aging test is an equivalent reproduction of the actual service thermal exposure process.

[0031] Based on the maximum service temperature, minimum service temperature, heating rate, and high-temperature holding time, an alternating heat load curve with time on the x-axis and temperature on the y-axis is constructed as the thermal cycle load spectrum. A complete cycle of the alternating heat load curve includes the following sequential stages: starting from the minimum service temperature, heating up to the maximum service temperature at a set heating rate; maintaining the high temperature at the maximum service temperature for the specified time; then cooling down from the maximum service temperature to the minimum service temperature; and finally returning to the cycle starting point after maintaining the minimum service temperature for a period of time. For example, a complete thermal cycle can be set as follows: starting from 25°C, heating up to 60°C at 3°C / min, holding at 60°C for 15 minutes, then cooling down to 0°C at 2°C / min, holding at 0°C for 10 minutes, and finally heating back to 25°C at 2°C / min.

[0032] After obtaining the thermal cycling load spectrum, it is used as the temperature control program for the environmental chamber to perform cyclic thermal loading experiments. Specifically, the time-temperature data points in the load spectrum are input into the environmental chamber's control system as piecewise linear curves. The environmental chamber automatically performs heating, holding, and cooling operations according to this program, without manual intervention. The temperature control accuracy of the environmental chamber is ±1℃ to ±2℃. A temperature sensor inside the environmental chamber monitors the air temperature near the sample in real time. When the measured temperature deviates from the set value, the control system automatically adjusts the heating or cooling power to compensate, keeping the deviation between the actual temperature curve and the load spectrum within the allowable range, thus ensuring the accuracy and repeatability of the cyclic thermal loading process.

[0033] The standard mechanical specimens of the battery casing material are used in the cyclic heat loading experiment. Multiple sets of standard mechanical specimens made of the same material as the battery casing to be evaluated are prepared. These specimens are cut from the battery casing material or prepared using the same grade and state of material as the battery casing. The shape and size of the specimens conform to the tensile testing standards for metallic materials. The specimens are placed on a specimen rack in an environmental chamber, ensuring air circulation around them to prevent uneven temperature distribution caused by localized overheating or undercooling. The total number of cycles is set to a preset value, determined based on the battery's designed service life. Specifically, the total number of cycles can be the product of the designed service life and the average number of charge-discharge cycles per year. In this embodiment, the total number of cycles is set to 1500. Several cycle count nodes are preset within the total number of cycles as sampling points. In this embodiment, these cycle count nodes are set to 0, 100, 200, 300, 500, 700, 800, 1000, 1200, and 1500 cycles. Cycle count 0 represents the baseline node before cyclic heat loading and is used to characterize the initial properties of the material. The distribution of nodes for each cycle number follows the pattern of being denser in the early stage and sparser in the later stage. This is because the performance degradation of the shell material usually changes rapidly in the early stage and then gradually becomes more gradual. Setting denser sampling points in the early stage can capture the rapid change stage of material performance, while setting sparser sampling points in the later stage can reflect the stable degradation trend of material performance. This reduces unnecessary experimental and testing workload while ensuring the accuracy of the degradation curve.

[0034] When the environmental chamber reaches the preset cycle count nodes, the experiment is paused, and the corresponding group of standard mechanical specimens are removed for mechanical property testing to obtain the yield strength variation data with the number of cycles. The removed specimens are immediately marked with the cycle count and number and stored in a dry environment to avoid moisture or oxidation affecting the test results. In this embodiment, the mechanical property test is a room temperature tensile test, conducted according to the standard method for room temperature tensile testing of metallic materials: the specimen is mounted between the upper and lower clamps of the universal testing machine, aligning the specimen axis with the loading direction to avoid eccentric loading that could lead to test result deviations; the testing machine applies a tensile load at a constant displacement rate until the specimen breaks. During the test, the testing machine automatically records the load-displacement curve or stress-strain curve, and the yield strength value is read from this curve. For materials without a clear yield plateau, the specified non-proportional elongation strength is taken as the yield strength value. For example, for aluminum alloy shell materials, the tensile rate is set to 1 mm / min to 5 mm / min. This rate range falls within the quasi-static loading range, which avoids the loading rate affecting the yield strength test results. At least three parallel specimens should be tested at the same cycle number node, and the average yield strength of the specimens should be taken as the representative yield strength value at that node. The purpose of taking the average value is to eliminate the material's own inhomogeneity and experimental operation errors, and to ensure the representativeness and reliability of the data.

[0035] After obtaining the yield strength data at each cycle number node, the discrete data points are fitted into a continuous performance degradation curve using the cycle number node as the independent variable and the corresponding yield strength parameter as the dependent variable. Specifically, the average yield strength corresponding to each cycle number node is plotted on a coordinate system with the cycle number as the abscissa and the yield strength as the ordinate. Considering the above three-stage degradation law, this embodiment adopts an exponential decay model as the fitting function form:

[0036] in, The number of loops is The yield strength at which the pressure is measured in MPa. This represents the number of loop iterations. The initial strength attenuation amplitude is expressed in MPa, reflecting the potential for performance loss of the material in the early stages of aging. The decay rate constant, with dimensions of the reciprocal of the number of cycles, characterizes the sensitivity of a material to thermal cycling loads. A higher value indicates faster performance degradation; This represents the residual strength asymptote, measured in MPa, and signifies the theoretically lowest strength level a material can retain after undergoing sufficient cycles. The rationale for choosing this function form is that the exponential decay function exhibits the mathematical characteristic of rapid initial change followed by a gradual approach to the asymptote, corresponding to the physical mechanism of rapid evolution of dislocation density followed by saturation during the thermal fatigue process of aluminum alloys. This allows for the simultaneous characterization of the three-stage degradation features. Furthermore, this function... The place is given It directly corresponds to the initial yield strength measured at the reference node, making it easy to constrain the parameters using experimental data.

[0037] The yield strength value corresponding to each cycle number node is obtained through the above method.

[0038] Table 1 shows the yield strength test data at different cycle counts. Based on Table 1 and... Figure 2 It can be seen that the yield strength of the shell material gradually decreases with the increase of the number of cycles. Specifically, in the 0 to 300 cycle stage, the yield strength decreases from 130 MPa to 120 MPa, a decrease of 10 MPa, indicating relatively slow material performance degradation; in the 300 to 1000 cycle stage, the yield strength decreases from 120 MPa to 85 MPa, a decrease of 35 MPa, indicating significantly faster material performance degradation; in the 1000 to 1500 cycle stage, the yield strength decreases from 85 MPa to 62 MPa, a decrease of 23 MPa, indicating a slowdown in the degradation rate. These trends indicate that the performance degradation of the shell material under cyclic thermal loading follows a pattern of "slow degradation first, then accelerated degradation, and finally leveling off."

[0039] Table 1: Yield strength test data at different cycle counts

[0040] After obtaining the yield strength data at each cycle number node, these discrete data points are fitted into a continuous performance degradation curve.

[0041] The fitting process uses the least squares method, determining the undetermined coefficients by minimizing the sum of squared residuals between all data points and the fitted curve. , , :

[0042] in, It is the sum of squared residuals; The total number of nodes in the loop; For the first The average yield strength measured at each cycle number node; This represents the calculated value of the fitted function at this node. The parameter solution is performed in two steps: first, the initial values ​​are determined, where... The average yield strength at later nodes (exemplarily 1200 to 1500 cycles) is used as the initial estimate. From the initial yield strength and The difference is determined, that is , Estimated from the initial degradation rate, and using the fitting function in The derivative at point is equal to This relationship is obtained by approximating the derivative using the yield strength difference between the first two nodes. The initial value is then determined; then, starting from this initial value, the Levenberg-Marquardt nonlinear least squares algorithm is used for iterative optimization. , , The process continues until the sum of squared residuals no longer decreases, yielding the final fitting parameters.

[0043] After fitting is complete, the coefficient of determination of the fitted curve is calculated to evaluate the fitting effect:

[0044] in, This is the sum of squared residuals, which is the sum of the squares of the differences between each measured value and the fitted value. The coefficient of determination (COP) is the sum of squares, which is the sum of the squares of the differences between each measured value and the mean of all measured values. The closer the COP is to 1, the higher the interpretability of the fitted curve for the original data points. This embodiment requires a COP of 0.95 or higher. This threshold is set because: if the COP is below 0.95, the deviation between the fitted curve and the measured data points increases, which will distort the yield strength value after being substituted into the simulation, thus affecting the reliability of the safety margin evaluation; 0.95 is the minimum fitting accuracy to ensure the reliability of the evaluation results. If the COP does not meet the requirement, check for outliers in the experimental data and verify them, or adjust the form of the fitting function and refit until the COP meets the requirement. After fitting a continuous function, substituting any number of iterations into the function will yield the corresponding estimated yield strength value. This continuous curve is the performance degradation curve, providing material constitutive input for subsequent simulations.

[0045] Please see Figure 2 It can be seen that the yield strength of the shell material gradually decreases with the increase of the number of cycles. Specifically, in the 0 to 300 cycle stage, the yield strength decreases from 130 MPa to 120 MPa, a decrease of 10 MPa, and the curve is relatively flat, indicating that the material performance degradation is relatively slow. In the 300 to 1000 cycle stage, the yield strength decreases from 120 MPa to 85 MPa, a decrease of 35 MPa, and the slope of the curve increases significantly, indicating that the material performance degradation has entered an accelerated stage. In the 1000 to 1500 cycle stage, the yield strength decreases from 85 MPa to 62 MPa, a decrease of 23 MPa, and the curve gradually flattens out, indicating that the material performance degradation rate has slowed down. Overall, the degradation of yield strength shows a pattern of "first a slow decrease, then an accelerated decrease, and then a gradual flattening out," which is consistent with the fatigue degradation characteristics of aluminum alloy materials under cyclic thermal loading.

[0046] Step 2: Establish a thermo-mechanical coupling simulation model of the battery casing to be evaluated. Select several cycle number nodes as evaluation nodes from the cycle number node range of the battery casing in the cyclic thermal loading experiment. Extract the corresponding yield strength parameter from the performance degradation curve according to each evaluation node, and input the yield strength parameter as a material property into the thermo-mechanical coupling simulation model to obtain the aging state simulation model corresponding to each evaluation node.

[0047] Specifically, the construction of the thermo-mechanical coupling simulation model serves as a bridge connecting the material aging state and the structural mechanical response. First, the geometric dimensions of the battery casing are obtained to construct a three-dimensional geometric model. Based on the actual design drawings or measured dimensions of the battery casing, a three-dimensional solid model of the casing is constructed in finite element preprocessing software. The geometric model includes structural features such as the main walls, bottom, top flange, pressure relief valve installation area, and connecting lugs. Subsequently, the three-dimensional geometric model is meshed using finite element methods, discretizing the continuous three-dimensional geometric model into a finite number of non-overlapping mesh elements. Since the wall thickness of the battery casing is usually much smaller than its length and width, shell elements are used for meshing the main body area to reduce the computational scale while ensuring computational accuracy. For areas with larger wall thickness or larger stress gradients, solid elements are used for meshing to capture more accurate stress distribution. The mesh density directly affects computational accuracy and efficiency: a denser mesh yields more accurate results but takes longer and consumes more computational resources; a coarser mesh results in higher computational efficiency but may lose detailed information about local stress concentrations. In this embodiment, the mesh size for the main region is set to 1mm to 3mm, while the mesh size for stress concentration areas such as the area around the pressure relief valve, near the connecting lugs, and at the corners of the shell is locally refined to 0.5mm to 1mm. This size range is determined based on mesh sensitivity analysis: when the mesh size is less than 1mm, the computation time increases exponentially but the result changes by less than 2%; when the mesh size is greater than 3mm, it is impossible to accurately capture the local plastic deformation zone under thermal runaway impact. The above mesh scheme achieves a balance between computational accuracy and computational efficiency, and after meshing, a finite element calculation model for subsequent thermo-mechanical coupling analysis is generated.

[0048] After mesh generation, the basic material properties of the battery casing are defined and assigned to each mesh cell. In this embodiment, the battery casing material is aluminum alloy, exemplarily using 3003 aluminum alloy or 5052 aluminum alloy. The basic material properties include two categories: thermal parameters and mechanical parameters. Thermal conductivity characterizes the material's ability to conduct heat, measured in W / (m·K). A higher thermal conductivity results in faster heat transfer within the casing. Specific heat capacity characterizes the amount of heat required to raise the material's temperature by one unit, measured in J / (kg·K). A higher specific heat capacity results in a slower temperature rise in the casing. Elastic modulus characterizes the material's ability to resist elastic deformation, measured in MPa. A higher elastic modulus results in smaller deformation under the same load. Poisson's ratio is a dimensionless quantity that characterizes the ratio of lateral contraction to longitudinal elongation when the material is under tension. The coefficient of thermal expansion characterizes the material's ability to change length with temperature changes, measured in degrees Celsius. A higher coefficient of thermal expansion results in greater thermal strain under the same temperature difference. For example, the density of 3003 aluminum alloy can be set to 2700 kg / m³ to 2740 kg / m³, the thermal conductivity can be set to 150 W / (m·K) to 200 W / (m·K), the specific heat capacity can be set to 880 J / (kg·K) to 920 J / (kg·K), the elastic modulus can be set to 68000 MPa to 71000 MPa, the Poisson's ratio can be set to 0.33, and the coefficient of thermal expansion can be set to 23 × 10⁻⁶. -6 Up to 24 × 10 per degree Celsius -6 Per degree Celsius. The above basic material properties can be obtained from material handbooks or standard specifications. Different grades of aluminum alloys should use the corresponding standard values, not limited to the specific values ​​listed in this embodiment. The basic material properties are fixed parameters in the model because: the above parameters have low sensitivity to thermal cycling aging, while the yield strength changes significantly with aging. Using the yield strength as the only evolution variable reduces the dimensionality of the problem and ensures that the simulation models for different aging states are completely consistent except for the yield strength, so that the response differences of each model can be uniquely attributed to material degradation.

[0049] Based on the fundamental material properties, a material parameter interface is set within the model, with the yield strength parameter designated as an evolutionary variable. This interface receives yield strength parameter data corresponding to different evaluation nodes. The material parameter interface connects to the elastoplastic constitutive model within the model. When the yield strength parameter is input through this interface, the constitutive model uses this yield strength as the critical threshold for determining whether the material has entered the plastic stage. The yield strength parameter, along with the elastic modulus and Poisson's ratio, constitutes the complete mechanical constitutive model input. In terms of implementation, this interface is implemented through parametric scripts or subroutines in finite element software, exemplarily using APDL scripts or Python subroutines. The yield strength is defined as an assignable variable, and batch values ​​are assigned to different evaluation nodes using loop statements, eliminating the need for repetitive modeling for each evaluation node.

[0050] Thermal boundary conditions are set to simulate the transient thermal shock process experienced by the battery casing after thermal runaway is triggered. A transient heat flux density load triggered by thermal runaway is applied to the inner surface of the 3D geometric model. This load originates from the heating effect of the high-temperature gas and electrolyte vapor ejected from the battery during thermal runaway on the inner wall of the casing. The temperature of the ejected material is much higher than the casing temperature, and forced convection heat transfer occurs between the high-temperature gas and the inner wall of the casing, transferring a large amount of heat to the casing in a very short time. "Transient" refers to the heat flux density changing over time rather than being constant. Convective heat transfer coefficients and radiative heat dissipation coefficients with ambient air are applied to the outer surface of the 3D geometric model. Convective heat transfer is calculated according to Newton's law of cooling, and radiative heat dissipation is calculated according to the Stefan-Boltzmann law.

[0051] in, This is the heat flux density for convective heat transfer, expressed in W / m². The convective heat transfer coefficient, expressed in W / (m²·K), is used to simulate the natural or forced convection heat transfer process between the outside of the shell and the surrounding air. This refers to the local temperature on the outer surface of the casing, expressed in Kelvin (K). The ambient temperature is expressed in Kelvin (K). This is the heat flux density for radiative heat dissipation, expressed in W / m². The radiation heat dissipation coefficient, i.e., surface emissivity, is dimensionless and is used to simulate the thermal radiation heat dissipation process between the exterior of the shell and the environment. This is the Stefan-Boltzmann constant, with a value of 5.67 × 10⁻⁶. -8 W / (m²·K 4 Convection heat transfer coefficient The values ​​are determined based on the empirical range of air convection heat transfer and in conjunction with preliminary experimental calibration: the natural convection heat transfer coefficient is set to 5 W / (m²·K) to 15 W / (m²·K), and the forced convection heat transfer coefficient is set to 20 W / (m²·K) to 50 W / (m²·K). Radiation heat dissipation coefficient. Depending on the oxidation state and coating characteristics of the shell surface, the value is set to 0.2 to 0.8. For example, 0.2 to 0.4 is used for naturally oxidized aluminum alloy surfaces, and 0.6 to 0.8 is used for coated or deeply oxidized surfaces. Accurate setting of the above parameters is a prerequisite for ensuring the accuracy of temperature field calculations.

[0052] The transient temperature field distribution on the inner wall of the battery casing is calculated by solving the three-dimensional heat conduction equation using transient heat flux density load, convective heat transfer coefficient, and radiative heat dissipation coefficient, combined with thermal conductivity and specific heat capacity.

[0053] in, This refers to the material density, expressed in kg / m³. Specific heat capacity, expressed in J / (kg·K); Temperature, in Kelvin (K). Time, in seconds; is the thermal conductivity, with units of W / (m·K); The term represents the internal heat source, with units of W / m³. The physical meaning of this equation is energy conservation: the left-hand side represents the rate of change of internal energy per unit volume of material over time; the first term on the right-hand side represents the net heat transfer into the unit volume via heat conduction, and the second term represents the heat contribution from the internal heat source. There is no internal heat source within the shell wall region. Taking zero, the heating effect of the ejected material on the shell is applied in the form of heat flux density at the inner surface boundary. The transient temperature field distribution refers to the temperature value at each location of the shell at each time point during the thermal runaway eruption. This temperature field is the physical basis for the thermal expansion of the shell material, which in turn generates thermal stress.

[0054] Mechanical boundary conditions are set to simulate the mechanical constraints and pressure loads experienced by the battery casing during thermal runaway. Fixed constraints are set as displacement boundary conditions at the connection points of the 3D geometric model. In actual use, the battery casing is fixed to the battery module frame by bolts, welding, or clips. These connection points will not move freely during thermal runaway. By applying fixed constraints, the displacement response of the model is made close to the actual working condition, preventing the casing from translating or rotating as a whole under load, thus obtaining effective deformation and stress calculation results. The internal expansion pressure load generated by thermal runaway is used as an external load input and applied to the mesh elements on the inner surface of the 3D geometric model. The internal expansion pressure load is a transient pressure curve that varies with time to simulate the pressure fluctuation process of gas generation, accumulation, and ejection through the pressure relief valve during thermal runaway. Its direction is the same as the transient heat flux density load, both pointing from the inside of the casing to the outside.

[0055] The essence of thermo-mechanical coupling lies in the fact that temperature changes cause materials to expand and contract, generating thermal strain. When this thermal strain is constrained by surrounding materials, thermal stress is generated within the material. Therefore, the temperature field results obtained from thermal boundary condition calculations are transformed into input loads for mechanical analysis through physical equations. First, based on the thermal expansion coefficient of the battery casing material, the transient temperature field distribution is transformed into the thermal strain of each mesh element:

[0056] in, Thermal strain, dimensionless; This is the coefficient of thermal expansion, expressed in degrees Celsius. This represents the difference between the current temperature of the mesh element and the reference temperature, in degrees Celsius. The reference temperature is taken as the temperature when the shell is in a stress-free state, exemplarily set as the assembly ambient temperature of 25°C. Since the temperature varies at different locations on the shell, the thermal strain also varies at each location. Then, the thermal strain, combined with the elastic modulus and Poisson's ratio, is mapped to the corresponding internal thermal stress using the material's elastic constitutive equation:

[0057] in, For stress; This is the elasticity matrix, composed of the elastic modulus and Poisson's ratio; For total strain; Let represent thermal strain. The physical meaning of this formula is: after deducting the freely occurring thermal strain from the total strain, the remaining part is the elastic strain that generates stress. When the shell expands freely, the total strain equals the thermal strain, and the stress is zero; when the thermal strain is constrained, the elastic strain is not zero, thus generating thermal stress. The battery shell is a thin-walled structure, and the main region is in a plane stress state. The elastic matrix is ​​represented as:

[0058] in, For elastic modulus, Poisson's ratio. The elastic modulus determines the magnitude of stress; the larger the elastic modulus, the greater the thermal stress generated under the same constrained strain. Poisson's ratio describes the coupling effect between strains in different directions, so that when thermal expansion in one direction is constrained, additional stress will be induced in the transverse direction.

[0059] Finally, the internal thermal stress and internal expansion pressure load are vector-superimposed as the total load for structural analysis and applied to the three-dimensional geometric model to construct a thermo-mechanical coupling simulation model. Vector superposition refers to merging the contributions of thermal stress and pressure load into a resultant force at the same location and in the same direction, which is represented as the superposition of equivalent force vectors at the nodes in the finite element implementation. Both types of loads act simultaneously during thermal runaway. The vector superposition follows the principle of mechanical equilibrium, ensuring that the thermal load and mechanical load are correctly synthesized in the same coordinate system, jointly driving the shell to produce deformation and stress response.

[0060] A boundary condition interface is defined within the model, including thermal load boundaries and mechanical load boundaries for inputting thermal runaway loads. Thermal load boundaries are preset on the inner surface of the 3D geometric model to receive temperature evolution curves; mechanical load boundaries are preset on the inner surface mesh elements to receive internal pressure evolution curves. The boundary condition interface is connected to the aforementioned thermal and mechanical boundary conditions. When a thermal runaway load is input through this interface, the corresponding boundary conditions are activated and participate in the solution process.

[0061] After the above steps, the constructed thermo-mechanical coupling simulation model contains two variable inputs: the first is the material evolution interface, used to input the yield strength parameter. This parameter controls when the shell material transitions from elastic deformation to plastic deformation under stress, and is a core variable characterizing the aging degree of the shell material. It is assigned a value in step 2 and replaced with the corresponding yield strength value at different evaluation nodes. The second is the load excitation interface, used to input the thermal runaway load, which includes the temperature evolution curve and the internal pressure evolution curve, serving as the input excitations for the model's thermal boundary conditions and mechanical boundary conditions, respectively. It is assigned a value in step 3. The two inputs are independent of each other, representing the current aging state of the shell material and the external thermal runaway impact conditions, respectively, and jointly determining the mechanical response calculation results of the shell. The model's own computational framework, including geometric dimensions, mesh generation, basic material properties, and boundary condition settings, is fixed before the two inputs are assigned values ​​and does not change with the evaluation nodes. This interface separation design allows the model to be reused in batches; calculations at different evaluation nodes only require replacing parameter values, avoiding redundant modeling.

[0062] In step 1, the performance degradation curve of yield strength with the number of cycles was obtained. In this step, several cycle number nodes are selected from the range of cycle number nodes in the cyclic thermal loading experiment of the battery casing as evaluation nodes. The evaluation nodes are used to represent the aging state of the battery casing at different stages of use, and each evaluation node corresponds to a specific cycle number value. In this embodiment, from ten experimental nodes of 0, 100, 200, 300, 500, 700, 800, 1000, 1200, and 1500 cycles, 200, 500, 800, 1000, and 1200 cycles are selected as evaluation nodes. The selection covers the early, middle, and late stages of degradation, so that the safety evaluation curve can reflect the complete trend of safety margin change. The number of evaluation nodes is set according to actual needs: the more evaluation nodes selected, the smoother the safety evaluation curve, and the more accurately the trend of safety margin change can be captured; however, too many evaluation nodes will increase the computational workload.

[0063] For each evaluation node, the yield strength parameter corresponding to the number of cycles is extracted from the performance degradation curve. Since the performance degradation curve is a continuous function obtained by fitting experimental data, the cycle number value of the evaluation node is substituted into the fitting function during extraction to obtain the yield strength parameter corresponding to that evaluation node. For example, the yield strength parameters corresponding to evaluation nodes of 200, 500, 800, 1000, and 1200 cycles are 124 MPa, 112 MPa, 96 MPa, 85 MPa, and 75 MPa, respectively. The extracted yield strength parameter is assigned to the battery casing material properties in the thermo-mechanical coupling simulation model, specifically as the threshold parameter determining the plasticity initiation in the material properties. In this process, the geometric dimensions and displacement boundary conditions in the thermo-mechanical coupling simulation model are kept unchanged, that is, the models corresponding to all evaluation nodes have the same geometry, mesh generation, basic material properties, and boundary conditions. Only the yield strength parameter value in the battery casing material properties is changed, thereby generating an aging state simulation model corresponding to each evaluation node. This single-variable substitution strategy ensures model consistency, attributing response differences at different evaluation nodes entirely to yield strength degradation, while significantly reducing the workload of repetitive modeling. The physical significance of this assignment process lies in the fact that as usage time increases, the shell material gradually degrades due to repeated thermal cycling, resulting in a gradual decrease in yield strength and a weakening of the shell's ability to resist thermal runaway shocks. By establishing simulation models under different aging conditions, the safety of the shell at various stages throughout its entire lifespan can be systematically evaluated.

[0064] Step 3: Input the preset thermal runaway load into the aging state simulation model corresponding to each evaluation node, and solve for the mechanical response results of the battery casing under each evaluation node. The thermal runaway load includes the temperature evolution curve and internal pressure evolution curve that change with time.

[0065] Specifically, the accuracy of the thermal runaway load directly determines the reliability of the prediction results. First, the characteristic parameters of the battery cell to be evaluated under thermal runaway triggering conditions are obtained. These parameters include the instantaneous maximum temperature of the ejected material, the peak pressure, and the duration of the ejection. These parameters can be obtained from publicly available literature, battery datasheets, or through thermal runaway triggering experiments. For example, for a certain type of ternary lithium-ion battery, under thermal runaway triggering conditions, the instantaneous maximum temperature of the ejected material is approximately 800°C to 1000°C, the peak pressure is approximately 1.5 MPa to 3 MPa, and the duration of the ejection is approximately 0.5 seconds to 2 seconds. The instantaneous maximum temperature determines the magnitude of the temperature load, the peak pressure determines the magnitude of the pressure load, and the duration of the ejection determines the duration of the load effect; these three together constitute a complete description of the thermal runaway load.

[0066] Based on characteristic parameters, a numerical distribution function is defined to construct temperature evolution curves and internal pressure evolution curves over time. The temperature evolution curves describe the entire process from the moment thermal runaway is triggered, as the ejecta temperature rises from its initial value to its peak value, and then decays back to ambient temperature. Because the energy release of the thermal runaway reaction is concentrated in a very short time, the temperature rises rapidly, while the heat dissipation after the eruption ends is relatively slow, resulting in an asymmetrical characteristic of rapid rise and relatively slow fall in the temperature evolution curves. This embodiment uses a piecewise function to construct the temperature evolution curves:

[0067] in, for The temperature of the ejected material at all times; The initial temperature at the trigger moment is taken as the current temperature of the casing; The instantaneous maximum temperature of the ejected material is given by characteristic parameters; The duration of the heating phase is set to 20% to 30% of the continuous eruption duration, corresponding to the stage in which the temperature rapidly rises from the initial value to the instantaneous maximum temperature; The ambient temperature; The thermal decay time constant characterizes the rate at which the temperature decays from its peak value. It is preferentially calibrated by fitting the measured temperature curve of the thermal runaway triggering experiment. When measured data is lacking, it is set to 0.5 to 1 times the duration of continuous eruption.

[0068] The internal pressure evolution curve describes the entire process from the moment thermal runaway is triggered, as the internal pressure rises from its initial value to its peak value and then falls back to atmospheric pressure over time. The rapid accumulation of gas generated by the thermal runaway reaction causes a rapid pressure increase, and the opening of the pressure relief valve causes a sudden pressure drop due to the ejection of gas; therefore, both the rise and fall rates of internal pressure are relatively fast. This embodiment also uses a piecewise function to construct the internal pressure evolution curve:

[0069] in, for Constant internal pressure of the casing; The initial pressure is taken as atmospheric pressure; The peak pressure is given by the characteristic parameter; The duration of the boost phase is set to 10% to 20% of the continuous eruption duration; Assuming the final downward pressure, normal pressure will be used. The pressure decay time constant is related to the flow area of ​​the pressure relief valve and the gas generation rate. It is preferentially calibrated by fitting the measured internal pressure curve. When measured data is lacking, it is set according to the design parameters of the pressure relief valve. The specific functional form of the above evolution curve can be determined by fitting actual experimental data. In the case of lack of experimental data, a piecewise linear function or exponential function can be used for approximate description. Those skilled in the art can select appropriate functional forms and parameter values ​​to construct evolution curves based on the obtained characteristic parameters and specific battery types.

[0070] Temperature and internal pressure evolution curves are loaded into simulation models for each aging state, serving as input excitations for thermal and mechanical boundary conditions, respectively. The different application methods of temperature and pressure loads reflect the different physical mechanisms of heat conduction and mechanical loading. The temperature evolution curve, as the input excitation for thermal boundary conditions, is converted into transient heat flux density or transient temperature boundary applied to the inner surface of the shell, driving the thermal analysis calculations in the model; the internal pressure evolution curve, as the input excitation for mechanical boundary conditions, is converted into dynamic pressure load applied to the inner surface of the shell, driving the structural analysis calculations in the model.

[0071] The temperature evolution curve is mapped onto the simulation model of each aging state as a non-uniform temperature field. The thermal runaway ejecta does not simultaneously contact all locations on the inner wall of the casing: after the ejecta is released from an initial location inside the battery, the high-temperature gas diffuses in all directions. In areas closer to the initial contact point of the ejecta, the gas arrives quickly and the temperature rises rapidly; in areas farther away, the gas arrives slowly and the temperature rises slowly. The difference in heat transfer distance between the ejecta and the casing determines the temporal difference in heating in different areas of the casing, resulting in different temperatures at different locations on the casing at the same time, forming a non-uniform temperature field distribution. To quantitatively characterize this difference, heat transfer hysteresis parameters are set according to the distance from the ejecta's contact point to various locations on the battery casing.

[0072] in, This is the heat transfer hysteresis parameter, in seconds (s). This represents the distance from the point of impact of the ejected material to the target mesh cell, in meters (m). The thermal diffusivity of the material, expressed in m² / s, is calculated from the basic material properties. It characterizes the rate at which temperature changes propagate through a material. This formula originates from the derivation of the characteristic timescale of one-dimensional unsteady-state heat conduction: the penetration depth of unsteady-state heat conduction satisfies a certain relationship with time. The relationship allows for deeper penetration. equal to distance That is, the propagation distance of the thermal disturbance is obtained. Required feature time ; equivalently from Fourier number From this angle, the formula corresponds to the critical state where the Fourier number is 1 / 4. In calibration, the thermal diffusivity... The thermal diffusivity of 3003 aluminum alloy is, for example, approximately 7 × 10⁻⁶, calculated directly from its thermal conductivity, density, and specific heat capacity. -5 m² / s to 8×10 -5 m² / s; The location of the ejected material is determined based on the installation location of the pressure relief valve and the measured results of the thermal runaway triggering experiment. Thermal runaway ejection usually occurs from the opening of the pressure relief valve. For example, for a square battery with a pressure relief valve on the top cover, the location of action is the area of ​​the top cover.

[0073] For shell regions near the ejecta's impact point, the heat transfer lag time is short, and the temperature rise begins earlier; for shell regions far from the ejecta's impact point, the heat transfer lag time is long, and the temperature rise begins later. During mapping, corresponding heat transfer lag parameters are set for each grid cell, transforming the overall temperature evolution curve into local temperature change curves at each cell location.

[0074] in, For a certain grid cell location at Local temperature at any given moment; This is the temperature evolution curve; This represents the heat transfer hysteresis parameter corresponding to the grid cell. For example, if the ejected material is located near the battery top cover, the heat transfer hysteresis time is shortest in the top cover region (approximately 0 seconds), followed by the sidewall region, and longest in the bottom region. This mapping accurately recreates the spatial temperature gradient under thermal runaway impact, avoiding the underestimation of the risk of localized overheating due to the uniform loading assumption.

[0075] The internal pressure evolution curve is applied as a time-varying dynamic surface load to the inner surface of the simulation model for each aging state. The pressure of the gas generated by thermal runaway is approximately equal at all locations on the inner surface of the shell. Considering the supersonic flow characteristics and pressure wave propagation velocity of the thermal runaway gas, the instantaneous internal pressure is uniformly distributed on the inner surface of the shell. Therefore, the pressure load is uniformly applied to all mesh elements on the inner surface of the shell in the form of a surface load. Since the internal pressure evolution curve changes with time, the amplitude of the surface load also changes with time: at each time step, the pressure value at the current moment is obtained from the internal pressure evolution curve, and this pressure value is applied to each mesh element on the inner surface of the shell. The above dynamic surface load is implemented using the pressure-time curve loading function in the finite element analysis software. That is, the discrete data points of the internal pressure evolution curve are input into the software, and the software automatically performs time interpolation during the calculation process to obtain the instantaneous pressure amplitude at each time step and apply it to the model.

[0076] In terms of the solution strategy, after applying the thermal runaway load to the aging state simulation model corresponding to each evaluation node, the finite element solver is used to perform nonlinear incremental iterative solution on the aging state simulation model. Nonlinearity refers to the presence of at least one nonlinear factor in the material constitutive relation, geometric deformation, or boundary conditions in the model. In this embodiment, the battery casing material may undergo plastic deformation under thermal runaway impact, and the deformation amount may be large. Both the material behavior and geometric behavior exhibit nonlinear characteristics. Using nonlinear solution can accurately describe the process of the casing transitioning from elastic deformation to plastic deformation. Incremental iterative solution refers to dividing the entire duration of the thermal runaway eruption into several time steps, and approximating the accurate solution of that time step through multiple iterations within each time step. The division of time steps is determined based on the load change rate: the duration of thermal runaway is 0.5 seconds to 2 seconds. In this embodiment, the time step is set to 1ms to 10ms, so that the load increment within a single time step is small enough to ensure iterative convergence and fully capture the pressure peak and temperature peak. If the time step is too large, peaks will be missed and iterative divergence will occur; if the time step is too small, the calculation time will be significantly increased.

[0077] At each time step, solve the structural equilibrium equations simultaneously:

[0078] in, For displacement-dependent tangent stiffness matrix; For displacement increment; This is the external load vector, which includes contributions from thermal stress and internal pressure. The internal resistance vector is composed of the accumulated internal forces under the current deformation state. Due to material and geometric nonlinearity, the stiffness matrix changes with displacement, and the equilibrium equations cannot be solved in one complete operation; they must be solved incrementally over time steps. Within each time step, the Newton-Raphson method is used for iteration: starting with the current displacement, the displacement increment is solved, and so on... After updating the displacements, reassembling the tangent stiffness matrix and internal resistance vectors, the solution is solved again until convergence. The Newton-Raphson method is chosen because it has a second-order convergence rate, requiring significantly fewer iterations than the initial stiffness method when approaching the exact solution. The convergence criterion is set as the residual force norm ratio being less than... or the ratio of displacement increment norm is less than The residual force norm ratio is the ratio of the unbalanced force norm to the external load norm, and the displacement increment norm ratio is the ratio of the displacement increment norm of the current iteration to the current displacement norm. This convergence level is set considering that overly loose criteria will cause the unbalanced force error to accumulate gradually, while overly strict criteria will significantly increase the number of iterations. The above-mentioned level allows the influence of the residual unbalanced force on the displacement results to be negligible, while controlling computational costs; it is a common convergence control level for engineering finite element analysis. Within each convergence time step, the displacement vectors of each node in the battery casing are acquired synchronously. The displacement vectors record the displacement of each mesh node in three spatial directions.

[0079] The geometric deformation ratio at various points on the battery casing is calculated based on displacement vectors. The geometric deformation ratio refers to the relative deformation of a mesh element after being subjected to force, including the degree of change in the element's length, area, or volume. It reflects the severity of deformation in local areas of the casing. Specifically, the strain tensor is obtained from the displacement vector through the strain-displacement relationship. Then, combining the material parameters of the pre-assigned aging state simulation model, the stress distribution inside the battery casing is calculated through the stress-strain conversion relationship. In the elastoplastic analysis, the total strain is decomposed into three parts: elastic strain, thermal strain, and plastic strain. The stress is obtained from the elastic strain part through mapping using the elastic matrix.

[0080] in, This represents plastic strain. Whether plastic flow occurs in the material is determined by the yield criterion; in this embodiment, the yield strength parameter of the current evaluation node is used as the threshold. ,when When the value is less than 0, the material is in the elastic stage. When the yield strength is 0, the material undergoes plastic flow. The increment of plastic strain accumulates along the flow direction of the yield surface. The accumulated plastic strain is obtained by summing the increments of plastic strain at each time step, i.e. The equivalent stress is calculated using the von Mises yield criterion, which combines the six stress components into a single equivalent scalar value.

[0081] in, , , There are three normal stress components. , , There are three shear stress components. The von Mises criterion is chosen because it is a distortion energy criterion, applicable to ductile metals such as aluminum alloys, and the equivalent stress is a scalar, which can be directly compared with the yield strength measured by a uniaxial tensile test, and is used to measure the comprehensive stress strength of a material under multiaxial stress.

[0082] The stress distribution includes transient stress field and plastic strain field. Transient stress field refers to the instantaneous stress value at each location and time point during the thermal runaway impact of the shell. The location with the highest stress at a given moment is the stress concentration point. The larger the value of the transient stress field, the higher the instantaneous stress intensity experienced by the shell at that location. Plastic strain field refers to the accumulated irreversible permanent deformation of the shell at each location and time point during the thermal runaway impact. When the stress exceeds the material's yield strength, the material undergoes plastic flow, and the resulting strain does not recover after the load is unloaded. The larger the value of the plastic strain field, the closer the material at that location is to a damaged state. From the displacement vector, transient stress field, and plastic strain field, the maximum values ​​at each evaluation node are extracted: For the simulation calculation results corresponding to a certain evaluation node, the maximum displacement value is found across all mesh nodes and all time steps as the highest displacement value, the maximum equivalent stress value is found as the maximum equivalent stress value, and the maximum plastic strain value is found as the maximum plastic strain value. The rationale for selecting the maximum value is that the overall safety of the shell is determined by the most dangerous location, and evaluating it based on the most unfavorable location aligns with conservative design principles. The maximum displacement characterizes the maximum deformation of the shell under this aging state, the maximum equivalent stress characterizes the maximum instantaneous stress intensity borne by the shell, and the maximum plastic strain characterizes the maximum degree of permanent damage to the shell. These three values ​​describe the mechanical response of the shell under thermal runaway impact from the dimensions of deformation, strength, and damage, respectively, and together constitute the thermal runaway mechanical response result at this evaluation node. The above operations are performed sequentially on different evaluation nodes to obtain the thermal runaway mechanical response result sequence corresponding to each evaluation node.

[0083] Step 4: Preset the critical deformation threshold of the battery casing, compare the thermal runaway mechanical response results at each evaluation node with the critical deformation threshold to obtain the safety margin corresponding to each evaluation node, and establish a safety evaluation curve that evolves with the number of cycles based on the safety margin.

[0084] Specifically, the safety margin evaluation employs a multi-dimensional indicator system to comprehensively cover different failure modes. First, the assembly gap between the battery casing and internal components is obtained, and a critical deformation threshold for the battery casing is set based on this gap. In the actual battery assembly structure, a certain spatial gap exists between the battery casing and internal components. This gap accommodates minor deformations of the casing caused by thermal expansion and contraction during normal service, preventing contact or compression between the casing and internal components. The specific value of the assembly gap is obtained from the battery's structural design drawings or design specifications. For example, for a certain type of square battery, the single-sided assembly gap between the casing and the internal cells is approximately 0.5mm to 2mm. The critical deformation threshold refers to the maximum allowable deformation of the battery casing under thermal runaway impact; exceeding this value indicates a safety risk. Its setting is based on the fact that when the casing deformation exceeds the assembly gap, the casing will contact or compress with the internal components, potentially causing damage to the internal components, short circuits, or casing rupture, leading to more serious safety accidents. Therefore, the critical deformation threshold is not greater than the assembly gap value, and a certain safety margin is reserved.

[0085] in, This is the critical deformation threshold, in mm; Assembly clearance, in mm; Safety margin factor, dimensionless. For example, it is set to 0.75 to 0.8. The reason for leaving a safety margin is that there are uncertainties in manufacturing tolerances and thermal expansion. This margin is set according to the 3σ principle in engineering statistics to ensure that the shell will not interfere within 99.7% of the manufacturing deviation range. For example, if the assembly gap is 1mm, take... The critical deformation threshold is set to 0.8 mm; if the assembly gap is 2 mm, the critical deformation threshold can be set to 1.5 mm. The above ratios can be adjusted according to the material properties, structural form, and safety level requirements of the casing; the higher the safety level requirement, the lower the ratio should be. If the battery casing design has specific engineering allowable deformation requirements, then that allowable deformation should be directly used as the critical deformation threshold.

[0086] Based on this, the safety margins in the three dimensions are calculated respectively.

[0087] The deformation safety margin is obtained by calculating the ratio of the highest displacement to the critical deformation threshold in the thermal runaway mechanical response results at each evaluation node.

[0088] in, For deformation safety margin, dimensionless; This is the critical deformation threshold, i.e., the maximum allowable deformation. This represents the highest displacement at this evaluation node, i.e., the actual maximum deformation under thermal runaway impact. The purpose of using a ratio is that the ratio is dimensionless and directly reflects the multiple of the safety margin, facilitating horizontal comparisons between different evaluation nodes. The deformation safety margin assesses the shell's safety from a deformation perspective: the actual deformation of the shell under thermal runaway impact must not exceed the maximum allowable deformation. If the deformation safety margin is greater than 1, it indicates that the actual response of the shell in terms of deformation is less than the allowable limit, and it is safe; if the deformation safety margin is equal to 1, it indicates that the shell deformation has reached the allowable limit and is in a critical state; if the deformation safety margin is less than 1, it indicates that the shell deformation has exceeded the allowable limit, and it is judged as a failure.

[0089] The strength safety margin is calculated based on the yield strength parameters corresponding to each evaluation node and the maximum equivalent stress in the thermal runaway mechanical response results at each evaluation node:

[0090] in, For strength safety margin, dimensionless; The yield strength parameter corresponding to this evaluation node is obtained from the performance degradation curve; This represents the maximum equivalent stress at this evaluation node. The numerator uses the aged yield strength of the current evaluation node instead of the initial yield strength. This is the core difference between this embodiment and the static evaluation method: static evaluation uses the initial yield strength as a benchmark, which overestimates the load-bearing capacity of the shell after aging; while using the yield strength after real-time degradation as a benchmark allows the strength evaluation to be dynamically updated with service aging. The strength safety margin assesses the shell's safety from a stress strength perspective: the maximum equivalent stress generated inside the shell under thermal runaway impact cannot exceed the material's yield strength. When the equivalent stress exceeds the yield strength, the shell material begins to undergo permanent deformation, and even if the load disappears, the deformation will not fully recover. If the strength safety margin is greater than 1, it indicates that the shell is mainly in the elastic deformation stage and is safe; if the strength safety margin is equal to or less than 1, it indicates that the shell has entered or exceeded the plastic stage and is deemed unsafe.

[0091] A failure strain threshold is preset for the battery casing material to be evaluated. The strain safety margin is obtained by comparing the maximum plastic strain in the thermal runaway mechanical response results at each evaluation node with the failure strain threshold. The failure strain threshold refers to the maximum amount of plastic deformation that the casing material can withstand before ductile fracture. It is usually determined by the elongation after fracture or the ultimate plastic strain of the material, and can be obtained through standard tensile testing or by consulting a material handbook. For example, the elongation after fracture of 3003 aluminum alloy varies with different heat treatment states; the specific value is based on the actual results of standard tensile testing. It should be noted that elastic strain can be fully recovered after load unloading, while plastic strain is permanently retained after unloading, reflecting irreversible microstructural damage that has already occurred inside the material. The formula for calculating the strain safety margin is:

[0092] in, For strain safety margin, dimensionless; The failure strain threshold; This represents the maximum plastic strain at this evaluation node. The physical meaning of the strain safety margin is: after the shell has entered the plastic stage, it determines whether the plastic deformation has approached the material's fracture limit. This indicator complements the strength safety margin, jointly assessing the shell's remaining load-bearing capacity after permanent deformation. If the maximum plastic strain is 0, meaning the shell is in the elastic stage, the strain safety margin is assigned a preset maximum value much greater than 1, exemplarily set to 100. The reason for this setting is that there is no plastic damage in the elastic stage, and this indicator should not constrain the comprehensive evaluation. Assigning a maximum value can avoid mistakenly triggering the strain criterion in the elastic stage when taking the minimum value later. If the strain safety margin is greater than 1, it indicates that the cumulative damage of the shell in terms of plastic deformation is less than the allowable limit, and it is safe; if the strain safety margin is equal to or less than 1, it indicates that the plastic deformation of the shell has approached or reached the fracture critical value, and it is judged as failure.

[0093] For each assessment node, the corresponding deformation safety margin, strength safety margin, and strain safety margin are compared, and the minimum value is selected as the comprehensive safety margin index for that assessment node.

[0094] in, This is a comprehensive safety margin indicator. The minimum value is chosen based on the weakest link principle: battery casing failure is triggered by the factor that first reaches its limit. Regardless of the margins of other indicators, failure is declared as soon as any indicator exceeds its limit. Therefore, taking the minimum value is the most conservative and logically sound evaluation method, simultaneously covering three failure modes: casing structural integrity, material load-bearing capacity, and damage tolerance, avoiding the one-sidedness of evaluation based on a single indicator.

[0095] The comprehensive safety margin index at each evaluation node is obtained. A safety evaluation curve is plotted with each evaluation node as the x-axis and the comprehensive safety margin index as the y-axis. The x-axis represents the number of cycles corresponding to each evaluation node, indicating the battery's lifespan; the y-axis represents the comprehensive safety margin, a dimensionless ratio. As the number of cycles increases, the casing material gradually degrades due to repeated thermal cycling, the yield strength gradually decreases, and the casing's ability to resist thermal runaway impact weakens accordingly. The safety evaluation curve shows a gradually decreasing overall trend. When the safety margin is greater than 1, the casing is in a safe state; when the safety margin equals 1, the casing reaches the critical state between safety and unsafety; when the safety margin is less than 1, the casing is unsafe.

[0096] Identify abrupt changes in the slope of the security evaluation curve. First, calculate the slope of the line connecting adjacent evaluation nodes:

[0097] in, For the first The evaluation node and the first The slope of the line connecting the evaluation nodes; and These are the comprehensive safety margin indicators for two adjacent evaluation nodes; and These represent the number of iterations for two adjacent evaluation nodes. The slope reflects the rate at which the safety margin decreases with increasing iteration count. Then, a difference comparison is performed on adjacent slopes to determine abrupt slope changes.

[0098] When the absolute value of the difference between adjacent slopes is greater than the threshold At that time, the judgment of the first Each evaluation node is a slope abrupt change point. Threshold Based on the overall trend of the curve, this embodiment uses twice the standard deviation of the slope sequence, i.e. ,in This represents the standard deviation of the slopes of all adjacent nodes. The basis for this setting is that using statistical significance as the criterion for abrupt changes can eliminate random fluctuations in slopes caused by numerical calculation noise and data dispersion. Two standard deviations correspond to a confidence level of approximately 95%. Slope changes exceeding this range can be judged as significant changes rather than random fluctuations, thereby accurately capturing the real degradation mechanism transformation.

[0099] The physical significance of slope abrupt change points lies in characterizing a qualitative change in the material degradation mechanism. For example, it represents a shift from slow degradation dominated by dislocation movement and recovery to rapid degradation dominated by microcrack initiation and propagation, resulting in a sudden acceleration in the rate of safety margin decline and a sharp deterioration in safety performance. In practical engineering, battery casing performance often does not abruptly jump from safe to unsafe at a certain point, but rather begins to deteriorate rapidly at a certain stage. Slope abrupt change points are used to capture the starting point of this rapid deterioration. Determining the evaluation node corresponding to the slope abrupt change point as the safety failure node of the battery casing can provide a clear time window for preventive maintenance. Furthermore, when a data point on the safety evaluation curve first reaches 1 or is less than 1, the corresponding evaluation node is the position where the casing first reaches a critical safety state. Slope abrupt change points usually appear earlier than this position, serving as an early warning.

[0100] Please see Figure 3 The present invention also provides a battery casing thermal runaway safety prediction system, the prediction system being used to execute the above-described battery casing thermal runaway safety prediction method, comprising: The performance degradation acquisition module is used to acquire the thermal cycling load spectrum of the battery casing material to be evaluated, conduct cyclic thermal loading experiments on the battery casing material according to the thermal cycling load spectrum, and obtain the change data of yield strength with the number of cycles by performing mechanical property tests at each preset cycle number node, and fit the change data into a performance degradation curve. The simulation modeling module is used to establish a thermo-mechanical coupling simulation model of the battery casing to be evaluated. From the range of cycle number nodes of the battery casing in the cyclic thermal loading experiment, several cycle number nodes are selected as evaluation nodes. The corresponding yield strength parameter is extracted from the performance degradation curve according to each evaluation node, and the yield strength parameter is input as a material property into the thermo-mechanical coupling simulation model to obtain the aging state simulation model corresponding to each evaluation node. The response calculation module is used to input the preset thermal runaway load into the aging state simulation model corresponding to each evaluation node, and solve for the mechanical response results of the battery casing under each evaluation node. The thermal runaway load includes the temperature evolution curve and internal pressure evolution curve that change with time. The margin evaluation module is used to preset the critical deformation threshold of the battery casing, compare the thermal runaway mechanical response results at each evaluation node with the critical deformation threshold, obtain the safety margin corresponding to each evaluation node, and establish a safety evaluation curve that evolves with the number of cycles based on the safety margin.

[0101] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0102] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented in software, the above embodiments can be implemented, in whole or in part, as a computer program product. Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution.

[0103] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0104] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for predicting the safety of thermal runaway in battery casings, characterized in that, The specific steps include: Step 1: Obtain the thermal cycling load spectrum of the battery casing material to be evaluated, conduct cyclic thermal loading experiments on the battery casing material according to the thermal cycling load spectrum, and obtain the change data of yield strength with the number of cycles by performing mechanical property tests at each preset cycle number node, and fit the change data into a performance degradation curve; Step 2: Establish a thermo-mechanical coupling simulation model of the battery casing to be evaluated. Select several cycle number nodes as evaluation nodes from the cycle number node range of the battery casing in the cyclic thermal loading experiment. Extract the corresponding yield strength parameter from the performance degradation curve according to each evaluation node, and input the yield strength parameter as a material property into the thermo-mechanical coupling simulation model to obtain the aging state simulation model corresponding to each evaluation node. Step 3: Input the preset thermal runaway load into the aging state simulation model corresponding to each evaluation node, and solve for the mechanical response results of the battery casing under each evaluation node. The thermal runaway load includes the temperature evolution curve and internal pressure evolution curve that change with time. Step 4: Preset the critical deformation threshold of the battery casing, compare the thermal runaway mechanical response results at each evaluation node with the critical deformation threshold to obtain the safety margin corresponding to each evaluation node, and establish a safety evaluation curve that evolves with the number of cycles based on the safety margin. The thermal cycling load spectrum of the battery casing material to be evaluated is obtained, and cyclic thermal loading experiments are conducted on the battery casing material to obtain the performance degradation curve. The specific steps are as follows: Obtain the charge / discharge rate, heat generation parameters, and service environment information of the battery cells to be evaluated under preset operating conditions; Based on the heat generation parameters and service environment information, determine the highest service temperature, lowest service temperature, and heating rate to reach the highest service temperature of the battery casing in a single cycle. Based on the charge / discharge rate, determine the duration of a single continuous eruption of the battery casing at the highest service temperature, and set the duration of the single continuous eruption as the high-temperature holding time of the thermal cycling load spectrum; Based on the highest service temperature, lowest service temperature, heating rate, and high temperature holding time, an alternating heat load curve with time as the horizontal axis and temperature as the vertical axis is constructed as the thermal cycle load spectrum. Prepare multiple sets of standard mechanical specimens made of the same material as the battery casing to be evaluated, and perform reciprocating thermal cycling loading according to the thermal cycling load spectrum described above; At each preset cycle number node, standard mechanical specimens of the corresponding group are taken out for mechanical property testing to obtain data on the change of yield strength with the number of cycles. Using the number of cycles as the independent variable and the corresponding yield strength parameter as the dependent variable, a performance degradation curve characterizing the relationship between the mechanical properties of the material and the number of cycles is obtained by fitting. The yield strength parameter is input as a material property into the thermo-mechanical coupling simulation model to obtain the aging state simulation model corresponding to each evaluation node. The specific steps are as follows: For each evaluation node, the yield strength parameter at the corresponding cycle number node is extracted from the performance degradation curve; The extracted yield strength parameter is assigned to the battery casing material properties in the thermo-mechanical coupling simulation model; Keeping the geometric dimensions and displacement boundary conditions in the thermo-mechanical coupling simulation model unchanged, only changing the yield strength parameter value in the battery casing material properties, an aging state simulation model corresponding to each evaluation node is generated. The specific steps to obtain the safety margin corresponding to each evaluation node are as follows: Obtain the assembly gap between the battery casing and internal components, and set the critical deformation threshold of the battery casing based on the assembly gap; The ratio of the highest displacement to the critical deformation threshold in the thermal runaway mechanical response results at each evaluation node is calculated to obtain the deformation safety margin. The strength safety margin is calculated based on the yield strength parameters corresponding to each evaluation node and the maximum equivalent stress in the thermal runaway mechanical response results under each evaluation node. A failure strain threshold is preset for the battery casing material to be evaluated. The strain safety margin is obtained by comparing the maximum plastic strain in the thermal runaway mechanical response results at each evaluation node with the failure strain threshold. For each assessment node, the corresponding deformation safety margin, strength safety margin, and strain safety margin are compared, and the minimum value is selected as the comprehensive safety margin index for that assessment node.

2. The method for predicting the thermal runaway safety of a battery casing according to claim 1, characterized in that: The specific steps for establishing a thermo-mechanical coupling simulation model of the battery casing to be evaluated are as follows: Obtain the geometric dimensions of the battery casing to construct a three-dimensional geometric model, and then divide the three-dimensional geometric model into several finite mesh elements using the finite element method. Define the basic material properties of the battery casing and assign the basic material properties to each mesh cell. The basic material properties include: thermal conductivity, specific heat capacity, elastic modulus, Poisson's ratio and coefficient of thermal expansion. A material parameter interface is set, wherein the yield strength parameter is set as an evolution variable to receive yield strength parameter data corresponding to different evaluation nodes; The boundary condition setting interface includes: defining the thermal load boundary and mechanical load boundary for inputting the thermal runaway load; Thermal boundary conditions are set, including: applying a transient heat flux density load triggered by thermal runaway to the inner surface of the three-dimensional geometric model, and applying the convective heat transfer coefficient and the radiative heat dissipation coefficient with the ambient air to the outer surface. The transient temperature field distribution of the inner wall of the battery casing is calculated by solving the three-dimensional heat conduction equation based on the transient heat flux density load, the convective heat transfer coefficient and the radiative heat dissipation coefficient, combined with the thermal conductivity and specific heat capacity. Set mechanical boundary conditions, which include: setting fixed constraints as displacement boundary conditions at the connection parts of the three-dimensional geometric model, and applying the internal expansion pressure load generated by thermal runaway as an external load input to the inner surface mesh elements of the three-dimensional geometric model. Based on the thermal expansion coefficient of the battery casing material, the transient temperature field distribution is transformed into the thermal strain of each grid element, and combined with the elastic modulus and Poisson's ratio, it is mapped to the corresponding internal thermal stress. The internal thermal stress and the internal expansion pressure load are vector-superimposed as the total load for structural analysis and applied to the three-dimensional geometric model to construct a thermo-mechanical coupling simulation model.

3. The method for predicting the thermal runaway safety of a battery casing according to claim 2, characterized in that: The thermal runaway load was obtained in the following manner: Obtain the characteristic parameters of the battery cell to be evaluated under thermal runaway triggered state, including the instantaneous maximum temperature of the ejected material, the peak pressure, and the duration of continuous ejection; Based on the aforementioned characteristic parameters, a numerical distribution function is defined to construct temperature evolution curves and internal pressure evolution curves that change over time. The temperature evolution curve and the internal pressure evolution curve are respectively loaded into the aging state simulation model as input excitations for the thermal boundary conditions and mechanical boundary conditions.

4. The method for predicting the thermal runaway safety of a battery casing according to claim 3, characterized in that: Specifically, the temperature evolution curve and the internal pressure evolution curve are loaded into the aging state simulation model as follows: The heat transfer hysteresis parameter is set according to the distance from the point of action of the ejected object to various parts of the battery casing, and the temperature evolution curve is mapped to each of the aging state simulation models in the form of a non-uniform temperature field. The internal pressure evolution curve is applied to the inner surface of each of the aging state simulation models as a dynamic surface load that varies with time.

5. The method for predicting the thermal runaway safety of a battery casing according to claim 4, characterized in that: The preset thermal runaway load is input into the aging state simulation model corresponding to each evaluation node, and the mechanical response results of the battery casing under each evaluation node are obtained by solving the problem. The specific steps are as follows: The aging state simulation model is solved nonlinearly and incrementally using a finite element solver. At each time step, the structural equilibrium equations are solved simultaneously, and the displacement vectors of each evaluation node of the battery casing are obtained synchronously. Based on the displacement vector, the geometric deformation ratio at various points of the battery casing is calculated. Combined with the material parameters of the pre-assigned aging state simulation model, the stress distribution inside the battery casing is calculated through the stress-strain conversion relationship. The stress distribution includes: a transient stress field reflecting the instantaneous stress intensity, and a plastic strain field reflecting whether the material has suffered permanent damage. From the displacement vector, transient stress field and plastic strain field mentioned above, the maximum values ​​at each evaluation node are extracted and used as the highest displacement, maximum equivalent stress and maximum plastic strain at that evaluation node, respectively.

6. The method for predicting the thermal runaway safety of a battery casing according to claim 1, characterized in that: Based on the aforementioned safety margin, a safety evaluation curve evolving with the number of cycles is established. The specific steps are as follows: Obtain the comprehensive safety margin index under each evaluation node, and plot the safety evaluation curve with each evaluation node as the horizontal axis and the comprehensive safety margin index as the vertical axis. Identify the abrupt change in the slope of the safety evaluation curve, and determine the evaluation node corresponding to the abrupt change in slope as the safety failure node of the battery casing.

7. A battery casing thermal runaway safety prediction system, characterized in that: The prediction system is used to execute the battery casing thermal runaway safety prediction method according to any one of claims 1-6, including: The performance degradation acquisition module is used to acquire the thermal cycling load spectrum of the battery casing material to be evaluated, conduct cyclic thermal loading experiments on the battery casing material according to the thermal cycling load spectrum, and obtain the change data of yield strength with the number of cycles by performing mechanical property tests at each preset cycle number node, and fit the change data into a performance degradation curve. The simulation modeling module is used to establish a thermo-mechanical coupling simulation model of the battery casing to be evaluated. From the range of cycle number nodes of the battery casing in the cyclic thermal loading experiment, several cycle number nodes are selected as evaluation nodes. The corresponding yield strength parameter is extracted from the performance degradation curve according to each evaluation node, and the yield strength parameter is input as a material property into the thermo-mechanical coupling simulation model to obtain the aging state simulation model corresponding to each evaluation node. The response calculation module is used to input the preset thermal runaway load into the aging state simulation model corresponding to each evaluation node, and solve for the mechanical response results of the battery casing under each evaluation node. The thermal runaway load includes the temperature evolution curve and internal pressure evolution curve that change with time. The margin evaluation module is used to preset the critical deformation threshold of the battery casing, compare the thermal runaway mechanical response results at each evaluation node with the critical deformation threshold, obtain the safety margin corresponding to each evaluation node, and establish a safety evaluation curve that evolves with the number of cycles based on the safety margin.

Citation Information

Patent Citations

  • Model construction and simulation method and device of woven composite material and storage medium

    CN119538632A

  • Lithium manganese iron phosphate battery and shell forming method thereof

    CN119812600A