A short-circuit protection prediction method for new energy batteries with rubber casing

Abstract

This application provides a method for predicting short-circuit protection of new energy batteries with protective casings, comprising: characterizing and analyzing the surface and internal micromorphology of the casing in areas of residual stress concentration using microscopic observation to obtain the morphological characteristics of microcrack formation, and combining this with material mechanical property parameters, including fracture toughness and fracture strain, to analyze the mechanical properties of microcrack formation; if it is determined that the insulation of the constructed three-dimensional solid model of the new energy battery casing is reduced, then optimizing the injection molding process parameters, including injection temperature, holding time, and cooling rate, and obtaining the optimal combination of process parameters through orthogonal experimental design; constructing a reliability evaluation system for the injection molding quality of the casing, comprehensively considering residual stress, microcrack density, mechanical properties, and electrical properties, and using a fuzzy comprehensive evaluation method to quantitatively evaluate the casing quality, and screening and improving casings with defects.

Description

Technical Field

[0001] This invention relates to the field of information technology, and in particular to a method for predicting short-circuit protection of new energy batteries with rubber casings. Background Technology

[0002] During the injection molding process of new energy battery casings, improper injection process parameters, such as excessively high injection pressure and insufficient cooling time, can generate residual stress inside the casing. This residual stress gradually releases during use, causing changes in the casing's dimensions and reducing its mechanical strength. Simultaneously, the presence of residual stress can also induce and propagate microcracks in the casing's microstructure. These microcracks accumulate during battery charge-discharge cycles, eventually leading to casing cracking, electrolyte leakage, and safety issues. The presence of microcracks also affects the casing's insulation performance, increasing the risk of leakage and short circuits, and weakening the battery pack's short-circuit protection. Therefore, a thorough analysis of the residual stress generation mechanism during injection molding, optimization of process parameters, and minimizing residual stress to improve the mechanical strength and electrical safety of the casing are urgent technical problems that need to be addressed. Summary of the Invention

[0003] This invention provides a short-circuit protection prediction method for new energy batteries with rubber casings, mainly including:

[0004] Based on the constructed three-dimensional solid model of the new energy battery shell, the finite element simulation analysis method is used to input the mechanical property parameters of the material, the geometric dimensions of the parts and the injection molding process parameters to simulate the injection molding process and obtain the residual stress distribution cloud map inside the shell. The residual stress concentration area is determined by the stress distribution cloud map.

[0005] The surface and internal micromorphology of the shell in the residual stress concentration area were characterized by microscopic observation to obtain the morphological characteristics of microcrack formation. Combined with material mechanical property parameters, including fracture toughness and fracture strain, the mechanical properties of microcrack formation were analyzed.

[0006] Mechanical performance tests were conducted on the three-dimensional solid model of the constructed new energy battery shell. The residual stress level was determined based on the yield strength of the material. The influence of the residual stress level on the mechanical performance of the shell was analyzed. A residual stress control threshold was preset. When the residual stress exceeds the residual stress control threshold, the mechanical performance of the shell is determined to be unsatisfactory.

[0007] Electrical performance testing methods for dielectric constant and volume resistivity were used to evaluate the electrical insulation performance of the constructed three-dimensional solid model of the new energy battery shell, and to determine the microcrack density control range. If the microcrack density exceeded the range, it was determined that the insulation performance of the constructed three-dimensional solid model of the new energy battery shell was reduced.

[0008] If it is determined that the insulation of the constructed three-dimensional solid model of the new energy battery shell is reduced, the injection molding process parameters, including injection temperature, holding time, and cooling rate, are optimized, and the optimal combination of process parameters is obtained through orthogonal experimental design.

[0009] Based on the optimal combination of process parameters, injection molded shells are produced. Annealing is performed on the injection molded shells to promote the release of residual stress. Low-frequency vibration is applied to the shells to induce the redistribution of residual stress, reduce the degree of stress concentration, and inhibit the propagation of microcracks. Annealing and vibration aging treatments are performed to improve the residual stress state of the shells.

[0010] After annealing and vibration aging treatment, infrared thermal imaging is used to scan the surface of the shell to identify local heating areas caused by residual stress. Combined with the stress distribution cloud map in the three-dimensional solid model of the shell, the impact of microcrack propagation in the hot spot area on the insulation performance of the battery tab is predicted, the risk of battery short circuit in the hot spot area is assessed, and short circuit prevention design optimization suggestions are proposed for the shell in high-risk areas.

[0011] A reliability assessment system for injection molding quality of plastic shells is constructed. By comprehensively considering residual stress, microcrack density, mechanical properties, and electrical properties, a fuzzy comprehensive evaluation method is adopted to quantitatively assess the quality of the plastic shells and screen and improve those with defects.

[0012] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects:

[0013] This invention discloses a short-circuit protection prediction method for new energy battery casings. The method first analyzes the residual stress distribution during the injection molding process of the casing using finite element simulation, and then determines the microcrack formation characteristics through microscopic observation. Next, mechanical and electrical performance tests are conducted to establish a correlation model between residual stress, microcrack density, and casing performance, and control thresholds are set. When performance requirements are not met, the residual stress state is improved by optimizing injection molding process parameters, annealing treatment, and vibration aging. Infrared thermal imaging technology is used to identify hotspot areas, assess short-circuit risk, and propose short-circuit protection design optimization suggestions. Finally, a casing quality reliability assessment system is constructed, and a fuzzy comprehensive evaluation method is used to quantitatively assess and screen casing quality. This invention, through systematic analysis and control of residual stress, effectively suppresses microcrack propagation, improves the mechanical and electrical insulation performance of the casing, and significantly improves the overall quality reliability of new energy battery casings. Attached Figure Description

[0014] Figure 1 This is a flowchart of a short-circuit protection prediction method for a new energy battery with a rubber casing according to the present invention.

[0015] Figure 2 This is a schematic diagram of a short-circuit protection prediction method for a new energy battery with a rubber casing according to the present invention.

[0016] Figure 3 This is another schematic diagram of a short-circuit protection prediction method for a new energy battery with a rubber shell according to the present invention. Detailed Implementation

[0017] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.

[0018] like Figure 1 -3, This embodiment of a short-circuit protection prediction method for new energy batteries with rubber casings may specifically include:

[0019] S101. Based on the constructed three-dimensional solid model of the new energy battery shell, the finite element simulation analysis method is used to input the mechanical property parameters of the material, the geometric dimensions of the part and the injection molding process parameters to simulate the injection molding process and obtain the residual stress distribution cloud map inside the shell. The residual stress concentration area is determined by the stress distribution cloud map.

[0020] The dimensional data of the plastic shell structure in the 3D solid model are collected. The elastic coefficient and Poisson's ratio of the material are calibrated, and elastic deformation curve data of the plastic shell are generated based on the calibrated measurement data. Based on the elastic deformation curve data, injection melt temperature, injection pressure, and injection rate data are collected. Load boundary conditions are applied to the sampling points of the injection station, and stress calculation input data is obtained using multiple linear regression. Based on the stress calculation input data, a finite element mesh generation tool is used to mesh the 3D solid model, and the stress distribution dataset of the element nodes is calculated based on Hooke's law. Stress cloud map data is constructed based on the stress distribution dataset, and the principal stress values ​​are calculated using the von Mises stress criterion. Residual stress cloud map data is obtained by calibrating the coordinates of the stress peak points, and the residual stress concentration area is determined through the stress distribution cloud map.

[0021] For example, the structural dimensions of the battery casing are collected from the 3D solid model. The material's elastic modulus and Poisson's ratio are calibrated and measured. Measurement points on the casing are sampled according to material testing standards. Elastic deformation curve data of the casing is generated through least-squares fitting. Based on the elastic deformation curve data, the injection melt temperature, injection pressure, and injection rate data recorded by the process control unit are collected. Load boundary conditions are applied to the sampling points at the injection molding station. Multiple linear regression is used to generate stress calculation input data from the process parameter data. According to the stress calculation input data, the 3D solid model is meshed using a finite element meshing tool in the control unit. The stress at the unit nodes is calculated based on Hooke's law, generating a stress numerical distribution dataset. In the injection molding simulation station, stress cloud map data is constructed according to the stress numerical distribution dataset. The principal stresses are calculated using the von Mises stress criterion. The coordinates of the stress peak points are calibrated, generating residual stress cloud map data. In the 3D solid model acquisition, the structural dimensions of the casing include key features such as wall thickness, corner radius, and reinforcing rib distribution. Measurement points are typically collected at the battery casing connections, the load-bearing structure, and weak edge areas. When calibrating the elastic modulus of the material, the polymer material was measured to have an elastic modulus of 2000 MPa and a Poisson's ratio of 0.35 at 23℃. These basic mechanical parameters constitute the foundational data for subsequent injection molding simulation. The injection molding process involves three main control parameters: melt temperature, mold pressure, and injection rate. At the injection station, the temperature sensor recorded a melt temperature of 220℃, the pressure sensor collected an injection pressure of 15 MPa, and the injection rate was controlled at 25 cm / s. 3 / s. Sampling points are arranged at the cavity inlet, gate, and runner end. Boundary constraints are applied at these locations, and load data is recorded. In the finite element simulation of the injection molding process, hexahedral elements are used for mesh generation, with a mesh size of 2mm. Regions with curvature greater than 60 degrees are locally refined, resulting in a mesh size of 0.5mm. Stress distribution values ​​on the shell wall are obtained by calculating the stress at the mesh nodes. The stress calculation is based on the linear elastic constitutive equation, considering the influence of the temperature field on material properties. During the injection molding process simulation, stress cloud maps are represented by multi-color bands, with red areas indicating high-stress areas and blue areas representing low-stress areas. The stress cloud map data shows stress concentration at the shell corners and at the roots of the reinforcing ribs, with the maximum stress value reaching 35MPa. The stress concentration area covers 5% of the surface area. Based on the von Mises stress criterion, the principal stress direction forms a 45-degree angle with the reinforcing rib direction, and the stress peak points are mainly distributed in the transition area at the shell corners. During injection molding, the material undergoes shrinkage deformation from melt to solidification, with a shrinkage rate of 1.5%, leading to residual stress in the casing. The residual stress distribution cloud map reveals significant stress concentrations in areas of varying casing wall thickness, gate locations, and reinforcing rib connections. By analyzing the stress distribution patterns, high-risk areas exceeding 25 MPa are identified; these areas are prone to cracking or deformation during subsequent use. Injection molding simulation results show that the internal stress distribution of the battery casing is closely related to material flow behavior. During cavity filling, the melt temperature gradient and pressure non-uniformity significantly affect the residual stress distribution. When the injection rate increases from 20 cm⁻¹... 3 / s increased to 30cm 3 At a temperature of 80℃, the peak residual stress increases by 20%, and the area of ​​the stress concentration region expands by 1.5 times. When the mold temperature is controlled at 80℃, the residual stress distribution of the molded part is more uniform, and the peak stress decreases by 15%.

[0022] S102. The microscopic morphology of the surface and interior of the shell in the residual stress concentration area is characterized and analyzed by microscopic observation to obtain the morphological characteristics of microcrack formation. Combined with material mechanical property parameters, including fracture toughness and fracture strain, the mechanical properties of microcrack formation are analyzed.

[0023] The surface morphology of the shell is acquired using a microscopic scanning device. Based on the morphology image, a Sobel edge extractor is used to obtain microcrack boundary contour curve data. A strain sampling array is arranged within the gauge length range for the microcrack boundary contour curve data. The relationship between fracture strain values ​​and microcrack boundary positions is recorded based on the strain sampling array to obtain a strain data field distribution map. Based on the strain data field distribution map, an edge detection operator is used to extract the microcrack morphology features in segments. Crack depth, crack propagation rate, and crack orientation angle are calculated using these microcrack morphology features to obtain a crack morphology feature parameter matrix. Based on the crack morphology feature parameter matrix and the generalized fracture criterion, the crack tip displacement value, stress intensity factor, and fracture toughness are calculated. The normal stress component and shear stress component are calibrated within a micro-region fracture element using the crack tip displacement value to obtain the crack propagation mechanical parameter field.

[0024] For example, images of the shell surface morphology are acquired within a preset area using a microscopic scanning device. The microcrack boundaries are depicted using a Sobel edge extractor, and cubic spline interpolation is used to fit the microcrack contour points, generating crack boundary contour curve data. The fracture area of ​​the shell is precisely loaded according to material tensile testing standards. A 128x128 lattice of strain sampling points is arranged within the gauge length, recording the correspondence between fracture strain values ​​and microcrack boundary positions, generating a strain data field distribution map. In the strain data field distribution map, the microcrack morphology features are extracted segmentally using an edge detection operator, from which crack depth, crack propagation rate, and crack orientation angle are calculated to construct a crack morphology feature parameter matrix. Based on the crack morphology feature parameter matrix and combined with the generalized fracture criterion, the crack tip displacement, stress intensity factor, and fracture toughness are calculated. The normal stress components and shear stress components are calibrated within the micro-region fracture element, generating a crack propagation mechanical parameter field. In microscopic scanning, microcrack morphology was measured using a 1000x microscope with a scanning area of ​​10mm × 10mm, a scanning step size of 0.1μm, and an image resolution of 2048 × 2048 pixels. When Sobel edge extraction was used to depict the microcrack boundaries, a grayscale threshold of 128 was set. The detected crack widths ranged from 0.5μm to 5μm, and the lengths varied from 50μm to 500μm. When fitting the crack profile using cubic spline interpolation, points of curvature change were selected as feature points, with an interpolation node spacing of 0.2μm. During tensile loading, the loading rate was controlled at 0.5mm / min, and 128 × 128 lattice strain sampling points were arranged within a 10mm gauge length with a point spacing of 0.078mm. The fracture strain value reached 15% in the tensile fracture region, with strain concentration at the microcrack tip, where the local strain value exceeded 25%. The strain data field distribution map shows that a butterfly-shaped plastic deformation zone forms around the microcrack, with the size of the plastic zone approximately twice the crack length. In the analysis of the microcrack morphology, a 5×5 pixel window was used for the edge detection operator. The crack depth values ​​ranged from 15% to 45% of the material thickness, and the crack propagation rate exhibited an exponential growth characteristic, increasing from an initial 0.1 μm / s to 1.0 μm / s. The crack orientation angle and the principal stress direction were mostly distributed between 85 and 95 degrees, showing a quasi-vertical propagation trend. In the micro-region fracture element analysis, the crack tip opening displacement value showed a parabolic distribution with increasing distance from the crack tip, with a maximum opening value of 2.5 μm. The stress intensity factor was calculated using the displacement extrapolation method, with a critical stress intensity factor value of 2.5 MPa·m^0.5. The fracture toughness measurement value was 15 kJ / m. 3The maximum normal stress component, at 350 MPa, occurs 0.2 mm in front of the crack tip, while the peak shear stress component reaches 175 MPa. Under cyclic loading, the microcrack propagation of the battery casing material exhibits fatigue characteristics, with a power-law relationship between the crack propagation rate and the stress intensity factor amplitude. At a cyclic loading frequency of 10 Hz, the crack propagation rate is 0.01 μm / cycle. The microcrack propagation path is influenced by the material's microstructure, exhibiting deflection at grain boundaries between 15 and 35 degrees. During microcrack formation, the fracture surface morphology displays typical cleavage fracture characteristics, with cleavage step heights ranging from 0.5 μm to 2 μm. The fracture surface roughness Ra is 0.8 μm, and distinct fatigue bands are observed in the crack initiation region, with a band spacing of 0.2 μm. Secondary cracks branch off from the main crack surface at angles between 30 and 60 degrees, forming a network crack morphology.

[0025] S103. By conducting mechanical performance tests on the constructed three-dimensional solid model of the new energy battery shell, the residual stress level is determined based on the yield strength of the material, the influence of the residual stress level on the mechanical performance of the shell is analyzed, and a residual stress control threshold is preset. When the residual stress exceeds the residual stress control threshold, it is determined that the mechanical performance of the shell does not meet the requirements.

[0026] Orthogonal loading points are arranged on the surface of the plastic shell using a three-dimensional solid load loader. Displacement field data of the plastic shell surface is acquired using a digital image correlation optical strain gauge. A uniaxial tensile load is applied to the loaded area to obtain the strain distribution data of the plastic shell. The tensile strain value, compressive strain value, and shear strain value of the measurement points are extracted from the strain distribution data of the plastic shell. The linear segment values ​​of the material are calibrated using a tensile stress-strain curve calculator to obtain the yield point stress-strain curve. For the yield point stress-strain curve, the shear modulus and bulk modulus of the plastic shell material are solved using a Lamé parameter calculator to calculate the residual stress value and stress concentration factor to obtain the residual stress distribution data. Based on the residual stress distribution data, a numerical comparator is used to compare the residual stress value with a preset stress threshold to calibrate the position of the residual stress peak point and the stress level to obtain the mechanical property judgment result.

[0027] For example, eight orthogonal loading points are arranged on the surface of the plastic shell using a three-dimensional solid load loader. Displacement field data of the plastic shell surface is acquired using a digital image correlation optical strain gauge. A uniaxial tensile load is applied to the loading area, generating strain distribution data for the plastic shell. Tensile strain, compressive strain, and shear strain values ​​are extracted from the strain distribution data at each measurement point. A tensile stress-strain curve calculator is used to calibrate the linear segment values ​​of the material, generating a yield point stress-strain curve. Based on the yield point stress-strain curve, the shear modulus and bulk modulus of the plastic shell material are solved using a Lamé parameter calculator to calculate the residual stress value and stress concentration factor, generating residual stress distribution data. Based on the residual stress distribution data, a numerical comparator is used to compare the residual stress value with a preset stress threshold, calibrating the location and stress level of the residual stress peak point, and generating a mechanical performance determination result. In the mechanical performance test of the plastic shell, the load loader uniformly arranges eight orthogonal loading points on the surface of the plastic shell, with a spacing of 20 mm between the loading points, forming a complete strain measurement network. When acquiring the surface displacement field using a digital image correlation optical strain gauge, the resolution reached 0.1 μm, and the measurement area covered a range of 100 mm × 100 mm. Uniaxial tensile loads increased from 0 to 5000 N, with sampling intervals of 100 N, recording the evolution of the strain field. In strain distribution measurements, the tensile strain value increased from 0 to 0.2% within the linear segment, the compressive strain value reached 0.1% in the transverse direction, and the maximum shear strain was 0.15%. The stress-strain relationship in the linear segment of the material exhibited good linearity, with the slope corresponding to an elastic modulus of 2200 MPa. When the strain exceeded 0.2%, the stress-strain curve began to deviate from a straight line, indicating that the material entered the yielding stage. Material parameter calculations showed that the shear modulus of the shell material was 850 MPa, and the bulk modulus was 2350 MPa. The residual stress distribution on the shell surface exhibited non-uniform characteristics, with stress concentrations occurring at corners and the roots of reinforcing ribs, reaching a stress concentration factor of 2.3. The maximum residual stress occurred at the connection between the reinforcing rib and the matrix, reaching 42 MPa. The residual stress distribution pattern shows that significant stress concentrations exist in areas of abrupt changes in shell wall thickness, at stiffener connections, and in corner transition zones. When the preset stress threshold is set to 35 MPa, residual stress exceeds the standard in some areas of the shell. The areas with excessive stress are mainly distributed at the roots of the stiffeners, accounting for 3.5% of the total area. The location of the peak residual stress points corresponds to the geometric discontinuities in the structure. After injection molding, the residual stress level inside the battery shell directly affects its mechanical properties. Process parameters have a significant impact on the residual stress distribution; when the injection pressure increases from 40 MPa to 60 MPa, the average residual stress increases by 25%. Mold temperature also has a significant impact on residual stress; when the mold temperature increases from 60℃ to 80℃, the residual stress decreases by 20%. Mechanical property test results show that when the residual stress exceeds 35% of the yield strength, plastic deformation begins to occur in some areas of the material.Under cyclic loading, high residual stress regions are prone to stress concentration, leading to fatigue crack initiation. When the residual stress reaches 45% of the yield strength, the fatigue life of the material decreases by 50%. Under temperature cycling conditions, the superposition of residual stress and thermal stress exacerbates the damage evolution process of the material. In temperature cycling from -40℃ to 60℃, the residual stress changes with temperature by ±15MPa.

[0028] S104. Conduct electrical performance tests on the dielectric constant and volume resistivity of the constructed three-dimensional solid model of the new energy battery shell to evaluate the electrical insulation performance of the constructed three-dimensional solid model of the new energy battery shell, determine the microcrack density control range, and if the microcrack density exceeds the range, it is judged that the insulation performance of the constructed three-dimensional solid model of the new energy battery shell has decreased.

[0029] Test electrodes are arranged on the surface of the housing using a three-dimensional solid testing platform. The surface dielectric constant is collected using a capacitance meter, and the volume resistivity is recorded using a high-voltage DC resistance meter to obtain basic electrical performance data. The dielectric constant distribution and volume resistivity distribution are extracted from this basic electrical performance data. The insulation strength of the material is scanned and measured using a voltage-current characteristic curve analyzer to obtain insulation performance evaluation data. For this insulation performance evaluation data, a region growing segmenter is used to segment the microcrack image into blocks. A standard convolutional neural network is used to calculate the number of cracks per unit area and the crack surface area, generating microcrack density feature data. Based on this microcrack density feature data, an association matrix calculator is used to map the microcrack feature values ​​to the insulation performance parameters, determining the calibration results for areas of weakened insulation strength.

[0030] For example, eight test electrodes are arranged on the surface of the housing using a three-dimensional solid testing platform. An AC voltage is applied to the test points at a frequency of 1 kHz using a capacitance meter, and the surface dielectric constant is collected simultaneously. A high-voltage DC resistance meter is used to record the volume resistivity at each point at 1 kV, generating basic electrical performance data. The dielectric constant distribution, dielectric loss angle, and volume resistivity distribution are extracted from this basic electrical performance data. The insulation strength of the material is scanned and measured using a voltage-current characteristic curve analyzer, generating insulation performance evaluation data. Based on the insulation performance evaluation data, a region growing segmenter is used to segment the microcrack image. A standard convolutional neural network is used to calculate the number of cracks per unit area, crack surface area, and crack size spectrum, generating microcrack density feature data. Based on this microcrack density feature data, an association matrix calculator is used to map the microcrack feature values ​​to the insulation performance parameters, calibrating areas of weakened insulation strength and generating insulation strength judgment results. In the electrical performance test, the test electrodes are arranged in an eight-point configuration with a diameter of 10 mm, and the surface is silver-plated to improve conductivity. When a 1kHz AC voltage with an amplitude of 100V is applied, data is collected using a capacitance meter, and the dielectric constant is measured to be distributed between 2.8 and 3.2. During high-voltage DC testing, with a 1kV voltage applied, the volume resistivity reaches 10¹⁴ Ω·cm, indicating good insulation properties. In the dielectric performance evaluation, the dielectric constant distribution exhibits anisotropic characteristics, varying by ±0.2 along the shell wall thickness direction. The dielectric loss angle is measured at 0.02, reflecting low energy loss under alternating electric fields. The volume resistivity distribution map shows that the resistivity decreases to 10¹² Ω·cm in the microcrack region. Insulation strength is measured using the step voltage method, with a breakdown voltage reaching 25kV / mm. In microcrack feature extraction, the region growth segmenter is set with a growth threshold of 8 pixels, and the crack boundary grayscale gradient is greater than 50. The standard convolutional neural network identifies 5 cracks per unit area per mm. 3 The crack surface area accounted for 2.5%. Crack size spectrum analysis showed that the length ranged from 20 μm to 200 μm, and the width ranged from 2 μm to 10 μm. In the insulation performance evaluation, correlation matrix calculation showed that the microcrack density was negatively correlated with the dielectric constant, with a correlation coefficient of -0.85. When the crack density exceeded 3 cracks / mm... 3At this time, the local dielectric constant decreased by 15%. Insulation strength decreased significantly in areas with dense cracks; when the crack surface area exceeded 2%, the breakdown voltage decreased to 75% of its original value. Electrical insulation performance varied significantly with environmental conditions; at 85% relative humidity, the volume resistivity decreased by two orders of magnitude. When the temperature rose to 80℃, the dielectric constant increased by 0.5, and the dielectric loss angle increased to 0.035. Microcracks easily formed conductive channels in high-temperature and high-humidity environments, accelerating insulation degradation. Insulation strength test results showed that crack orientation had a significant impact on the breakdown path. Cracks perpendicular to the electric field direction had less impact on insulation strength, while cracks parallel to the electric field direction were more likely to cause breakdown. The electric field enhancement effect at the crack tip reduced the local breakdown voltage by 30%. Under alternating electric fields, partial discharge at microcracks accelerated insulation degradation, with local temperature increases reaching 15℃. When the microcrack network formed conductive paths, the leakage current density increased significantly. Under standard voltage, the leakage current density in the healthy region was 0.1 μA / cm². 3 The crack density in densely populated areas increased to 1.5 μA / cm. 3 During long-term operation, the corrosion effect caused by partial discharge further expands the crack size, creating a cyclical effect that deteriorates the insulation performance.

[0031] Using a dielectric constant tester and a high-resistivity meter, the dielectric constant and volume resistivity of the shell material are measured. Based on electrical insulation performance standards, it is determined whether the shell material meets the insulation requirements. The number and distribution of internal microcracks are counted, and the correlation between microcrack density and insulation performance is analyzed to obtain the control range and threshold of microcrack density of the shell material.

[0032] AC and DC voltages are applied to the surface of the housing using testing instruments to obtain capacitance and resistance values. Comprehensive electrical characteristic data is obtained by scanning the breakdown voltage. Dielectric constant and dielectric loss angle are extracted from this comprehensive electrical characteristic data and compared with preset insulation standard values ​​using a numerical comparator to obtain insulation performance characteristic data. A region growth algorithm is run on this insulation performance characteristic data to extract the number of cracks per unit area and crack propagation length. An insulation strength feature matrix is ​​constructed using a convolutional neural network. The correlation between crack density parameters and insulation strength parameters is calculated based on the insulation strength feature matrix. If the correlation value exceeds a preset standard threshold, the microcrack control interval data is determined.

[0033] For example, according to standard measurement specifications, test electrodes are arranged in an 8-point array on the surface of the housing. A 1 kHz AC voltage is applied using a dielectric constant tester to collect capacitance values, while a 1 kV DC voltage is simultaneously applied using a high-resistivity meter to record resistance values. A step scan of the breakdown voltage is performed to generate comprehensive electrical characteristic data. Dielectric constant, dielectric loss angle, volume resistivity, and breakdown field strength are extracted from this comprehensive electrical characteristic data. These values ​​are compared with preset insulation standard values ​​using a numerical comparator to collect insulation strength decay curves, generating insulation performance characteristic data. Based on this insulation performance characteristic data, a region grower is used to divide and count microcrack morphology, extracting the number of cracks per unit area, crack propagation length, and crack penetration depth. An insulation strength feature matrix is ​​constructed using a convolutional neural network. Based on this insulation strength feature matrix, a correlation coefficient calculator is used to calculate the correlation between crack density parameters and insulation strength parameters. Based on electrical insulation standards, upper and lower limits for crack density control are calibrated to generate microcrack control interval data.

[0034] ρ represents the Pearson correlation coefficient between the crack density parameter and the insulation strength parameter, Xi represents the crack density of the i-th observation, and Yi represents the insulation strength of the i-th observation. and The values ​​represent the sample mean values ​​for crack density and insulation strength, respectively. In electrical performance testing, the test electrodes were arranged in an 8-point array with a spacing of 25 mm, covering the main test area of ​​the housing. When a 1kHz AC voltage was applied (voltage amplitude 100V), the capacitance measured by the dielectric constant tester was between 100pF and 120pF. A 1kV DC voltage was applied using a high-resistivity meter, and the measured resistance reached 10¹⁴ Ω. The breakdown voltage scan range increased from 1kV to 30kV. Insulation performance parameters showed that the housing material had a dielectric constant of 2.8, a dielectric loss angle of 0.02, and a volume resistivity of 10¹⁴ Ω·cm. The breakdown field strength test used the step voltage method, with a voltage increment of 1kV / 30s, and a breakdown field strength of 25kV / mm. The insulation strength decay curve exhibited a non-linear characteristic, with the decay rate accelerating in the microcrack region. In microcrack morphology analysis, the region growth algorithm was set with a grayscale threshold of 128 and a growth step size of 2 pixels. The number of cracks per unit area is 5 / mm 3 The crack propagation length ranged from 50 μm to 500 μm, with a penetration depth reaching 25% of the material thickness. The feature matrix extracted by the convolutional neural network contained 16×16 feature points, each recording crack morphology parameters. Correlation analysis between insulation strength and crack density showed a significant negative correlation, with a correlation coefficient of -0.85. When the crack density exceeded 3 cracks / mm... 3 At that time, the local insulation strength decreased by 30%. Based on electrical insulation standards, the upper limit for crack density control is 4 cracks / mm. 3The lower limit is 0.5 strips / mm 3 In practical applications, the electrical properties of the housing material are significantly affected by environmental conditions. When humidity increases from 45% to 85%, the volume resistivity decreases by two orders of magnitude. When the temperature rises to 80℃, the dielectric constant increases by 0.3, and the dielectric loss angle increases to 0.035°. Microcracks are more likely to form conductive paths in high-temperature and high-humidity environments. Insulation breakdown tests show that the presence of microcracks alters the electric field distribution. At the crack tip, the local electric field strength increases by 2.5 times, promoting electrical tree formation. When the crack direction is aligned with the electric field direction, the breakdown voltage decreases by 40%. Under alternating electric fields, partial discharge at the crack accelerates material aging. The microcrack density has a significant impact on leakage current; under standard voltage, the leakage current density in the intact region is 0.1 μA / cm². 3 The crack area increased to 1.2 μA / cm. 3 As operating time increases, material erosion caused by partial discharge further expands the crack size, creating a vicious cycle of insulation degradation. The rate of decay of electrical insulation performance is exponentially related to the crack propagation rate.

[0035] S105. If it is determined that the insulation of the constructed three-dimensional solid model of the new energy battery shell is reduced, the injection molding process parameters, including injection temperature, holding time, and cooling rate, are optimized. The optimal combination of process parameters is obtained through orthogonal experimental design.

[0036] According to the injection molding specifications, a process parameter calculator is used to combine and arrange the injection melt temperature range, holding time range, and cooling rate range. The material viscosity and crystallinity values ​​corresponding to each set of parameters are calibrated using a physical property parameter database to obtain an injection molding process parameter matrix. From this matrix, process configuration parameters are extracted, and the cavity pressure sensing points, temperature sensing points, and flow front sensing points are recorded in real time in the injection molding controller to obtain an injection molding process parameter set. For this parameter set, a gradient boosting tree is used to extract features from the parameter curve data, calculate the insulation contribution value and parameter coupling index of each parameter, and generate a process parameter influence factor set. Based on this influence factor set, a parameter response surface is constructed using a central composite designer. Under the constraint of optimal insulation performance, the parameter space is traversed and searched to determine the optimal values ​​for melt temperature, holding time, and cooling rate.

[0037] For example, injection molding parameter ranges are set in a three-factor, four-level orthogonal table according to the injection molding specifications. A process parameter calculator is used to combine and arrange the injection melt temperature range, holding time range, and cooling rate range. The material viscosity and crystallinity values ​​corresponding to each set of parameters are calibrated using a physical property parameter database to generate an injection molding process parameter matrix. Process configuration parameters are extracted from this matrix. Cavity pressure sensing points, temperature sensing points, and flow front sensing points are recorded in real time in the injection molding controller. Parameter curve data and insulation performance data during the molding process are collected to generate an injection molding process parameter set. Based on this parameter set, a gradient boosting tree is used to extract features from the parameter curve data, calculating the insulation contribution value, parameter coupling index, and physical property parameter equilibrium point for each parameter, generating a process parameter influence factor set. Based on this process parameter influence factor set, a parameter response surface is constructed using a central composite designer. Under the constraint of optimal insulation performance, the parameter space is traversed and searched to calibrate the optimal values ​​for melt temperature, holding time, and cooling rate, generating a collaborative configuration scheme for process parameters. In the optimization of injection molding process parameters, the three-factor, four-level orthogonal design includes three key parameters: melt temperature, holding time, and cooling rate. The melt temperature range is set between 220℃ and 260℃, with a level value every 10℃. The holding time ranges from 5s to 20s, with values ​​taken at 5-second intervals. The cooling rate ranges from 10℃ / s to 40℃ / s, with levels set at 10℃ / s intervals. The material viscosity is 280 Pa·s at 220℃ and decreases to 180 Pa·s when the temperature rises to 260℃. During parameter acquisition, cavity pressure sensors are located at the gate, the end of the runner, and corners, with a sampling frequency of 100Hz. The cavity pressure rises from 0MPa at the initial injection stage to 40MPa during the holding stage, and then drops back to 0MPa at the end of the molding cycle. Temperature sensor records show that during the process of the melt temperature decreasing from 260℃ to 80℃, the local cooling rate difference reaches 15℃ / s. The flow front sensor records a filling time of 2.5s. Analysis of the influence of process parameters shows that melt temperature has the most significant impact on the crystallinity of the material, contributing 45%. The interaction between holding time and cooling rate is significant, with a coupling effect index of 0.72. When the melt temperature is below 230℃, the material viscosity is too high, leading to incomplete filling. When the temperature is above 250℃, thermal degradation of the material intensifies, and the insulation performance decreases by 15%. Response surface methodology optimization results show that under the constraint of optimal insulation performance, the optimal melt temperature is 240℃, at which temperature the material viscosity is 220 Pa·s, and the crystallinity reaches 32%. The optimal holding time is 15 s, and the holding pressure is maintained at 35 MPa. The optimal cooling rate is 25℃ / s, and the total molding cycle time is controlled within 45 s. The insulation performance of the shell is closely related to the molding process, and the material crystallinity is the key bridge connecting process parameters and insulation performance. When the crystallinity is below 25%, amorphous regions form inside the material, and the insulation resistance decreases by two orders of magnitude.When crystallinity exceeds 40%, internal stress concentration occurs, easily leading to microcracks. During injection molding, the holding time directly affects the level of residual stress within the product. Increasing the holding time from 5 seconds to 20 seconds reduces residual stress by 35%. The uniformity of holding pressure plays a crucial role in product deformation; when pressure distribution non-uniformity exceeds 20%, product warpage increases by 0.5 mm. The cooling rate significantly impacts the internal structure of the material. At rates below 15°C / s, crystallization is sufficient but the cycle time is prolonged, reducing production efficiency by 30%. At rates exceeding 35°C / s, uneven cooling between inner and outer layers generates shrinkage stress, doubling the microcrack density. Uniform cooling plays a key role in reducing defects and improving insulation performance.

[0038] S106. Based on the optimal combination of process parameters, produce injection molded housings, anneal the injection molded housings to promote the release of residual stress, apply low-frequency vibration to the housings to induce the redistribution of residual stress, reduce the degree of stress concentration, inhibit the propagation of microcracks, and perform annealing and vibration aging treatments to improve the residual stress state of the housings.

[0039] A temperature sensor array is arranged on the surface of the shell according to the process control parameters. A temperature acquisition device is used to obtain temperature field distribution and crystallinity change data to obtain an annealing temperature change curve. The heating rate and holding time parameters are extracted from the annealing temperature change curve. A stress relaxation calculator is used to numerically solve the displacement field inside the shell to obtain a stress relaxation field distribution map. For the stress relaxation field distribution map, an acoustic exciter is used to apply a frequency scanning signal to the shell. The frequency scanning signal is processed by a neural network fitter to obtain stress redistribution data. Based on the stress redistribution data, a vibration sensor array is arranged in the stress concentration area of ​​the shell. The vibration sensor array acquisition signal is processed by a phase delay calculator to obtain the microcrack propagation rate and stress intensity factor.

[0040] For example, a 16-point temperature sensor array is arranged on the surface of the shell according to process control parameters. A temperature acquisition device records the temperature field distribution and material crystallinity changes. An infrared thermal imager is used to calibrate the overall temperature field. Crystallization transformation point data of the shell are marked within a standard annealing cycle, generating an annealing temperature change curve. The heating rate, holding time, and crystallinity change values ​​are extracted from the annealing temperature change curve. A stress relaxation calculator is used to numerically solve the internal displacement field of the material. Stress gradient changes are recorded during annealing, generating a stress relaxation field distribution map. Based on the stress relaxation field distribution map, an acoustic exciter applies a frequency scan to the shell, recording displacement response and strain distribution from low to high frequencies. A neural network fitter is used to construct the stress field evolution law, generating stress redistribution data. Based on the stress redistribution data, a vibration sensor array is arranged in the stress concentration area of ​​the shell. A phase delay calculator is used to extract the stress wave propagation law, calibrate the microcrack propagation rate and stress intensity factor, and generate stress regulation optimization data. During annealing, a temperature sensing array was arranged in a 4×4 matrix on the surface of the shell, with a sensor point spacing of 25 mm. The annealing temperature was increased from room temperature to 80℃, with a heating rate controlled at 2℃ / min. The material crystallinity increased from 28% to 35% with increasing temperature, and thermal imaging showed that the temperature field uniformity was within ±2℃. The standard annealing cycle was set to 4 hours, including 1 hour of heating, 2 hours of holding, and 1 hour of cooling. The stress relaxation process showed that the internal displacement field of the material exhibited a significant time dependence during the 80℃ holding stage. Displacement field calculations showed that the residual stress decreased by 45% after 2 hours of holding. The stress gradient value decreased from the initial 5 MPa / mm to 1 MPa / mm, and the stress field distribution tended to be more uniform. The internal structural reorganization of the material further increased the crystallinity to 38%. In acoustic excitation testing, the frequency was scanned from 10 Hz to 200 Hz, and the amplitude was set to 0.5 mm. In the low-frequency range below 50 Hz, the displacement response was in phase with the excitation signal, and the strain distribution was uniform. A resonance peak appears at a frequency of 150Hz, with a displacement amplification factor reaching 2.5. The stress evolution law fitted by the neural network shows that the vibration process redistributes stress, reducing local stress concentration by 35%. A vibration sensor array is arranged in the stress concentration area with a sensor spacing of 10mm. Phase delay calculations show that the stress wave propagation speed in the material is 1200m / s. The microcrack propagation rate decreases by 80% under vibration, and the stress intensity factor decreases from 1.2MPa·m^0.5 to 0.6MPa·m^0.5. The synergistic effect of annealing and vibration aging is manifested in several aspects. Annealing promotes molecular chain movement, reducing the overall stress level. Vibration loading breaks up local stress concentration through stress wave propagation. Increased crystallinity enhances the overall material properties and reduces the tendency for microcrack propagation. The uniformity of the temperature field distribution has a significant impact on stress release. When the temperature field non-uniformity exceeds 5℃, new thermal stress is generated, which in turn exacerbates stress concentration.When the heating rate exceeds 5℃ / min, the temperature gradient between the material surface and interior increases, forming a new stress field. The choice of vibration frequency is closely related to the material's natural frequency. Applying vibration near the natural frequency induces resonance within the material, promoting stress redistribution. Excessive amplitude exceeding 1mm can lead to fatigue damage. Frequency exceeding 180Hz increases material damping, making it difficult for vibrational energy to penetrate the interior. The microcrack suppression effect is directly related to the evolution of the stress field. After stress redistribution, the stress intensity at the crack tip decreases, inhibiting crack propagation. Internal structural reorganization improves fracture toughness and enhances resistance to crack propagation. Annealing and vibration aging treatments significantly extend the material's service life.

[0041] S107. After annealing and vibration aging treatment, the surface of the shell is scanned using infrared thermal imaging to identify local heating areas caused by residual stress. Combined with the stress distribution cloud map in the three-dimensional solid model of the shell, the impact of microcrack propagation in the hot spot area on the insulation performance of the battery tab is predicted, the risk of battery short circuit in the hot spot area is assessed, and short circuit prevention design optimization suggestions are proposed for the shell in high-risk areas.

[0042] Based on the scanning data acquired by the thermal imaging acquisition device, the surface temperature of the shell is calibrated using an infrared detector. The infrared detector uses an area integrator to calculate the thermal zone coverage area to obtain thermal zone distribution characteristic data. Temperature gradient field and thermal stress field are extracted from the thermal zone distribution characteristic data. A region growing algorithm is used to partition and match the stress distribution cloud map to obtain a thermal stress coupling distribution map. For the thermal stress coupling distribution map, a multilayer perceptron is used to identify the microcrack growth path. The distance between the tab region and the crack tip is calculated in a spatial coordinate system to obtain microcrack evolution prediction data. Based on the microcrack evolution prediction data, a resistance threshold detection array is arranged near the tab insulation layer. The short-circuit risk area is calibrated using the insulation resistance decay curve to obtain a structural optimization layout map.

[0043] For example, the surface of the shell is scanned using a thermal imaging acquisition device with a scanning resolution of 0.1 mm. An infrared detector is used to calibrate the surface temperature. If the local temperature exceeds a set multiple of the reference temperature value, the hot zone coverage is calculated using an area integrator, generating hot zone distribution characteristic data. Temperature gradient field, thermal stress field, and hot zone density values ​​are extracted from this hot zone distribution characteristic data. A region growing algorithm is used to partition and match the stress distribution cloud map, marking the abrupt temperature field changes in high residual stress areas, generating a thermal stress coupling distribution map. Based on the thermal stress coupling distribution map, a multilayer perceptron is used to identify the microcrack growth path. The distance between the tab region and the crack tip, the rate of change of insulation layer thickness, and the crack propagation speed are calculated in a spatial coordinate system, generating microcrack evolution prediction data. Based on this microcrack evolution prediction data, a resistance threshold detection array is arranged near the tab insulation layer. The short-circuit risk area is quantitatively calibrated using the insulation resistance decay curve, marking the thickened insulation layer and buffer layer placement positions, generating a structural optimization layout diagram. In thermal imaging inspection, the surface of the housing was scanned with a resolution of 0.1 mm, achieving a temperature measurement accuracy of 0.1℃. A reference temperature of 25℃ was set; any local temperature exceeding the reference temperature by 2℃ was identified as an abnormal hotspot. Hotspot area calculations showed that the coverage area of ​​a single hotspot was approximately 0.5 mm. 3 up to 2mm 3 Between these points, the hotspot density reaches 5 per cm² in high-stress areas. 3 Temperature gradient field analysis shows that the temperature around the hotspot decreases by 0.2℃ for every 0.5mm, forming a significant temperature gradient. Thermal stress field calculations show that a 1℃ increase in temperature generates 0.5MPa of thermal stress. The stress distribution identified by the region growth algorithm overlaps with the hotspot distribution by 85%, and the temperature in the high-value residual stress area exceeds that of the surrounding area by 1.5℃. In microcrack growth path prediction, the multilayer perceptron input layer contains 16 stress field feature points and 16 temperature field feature points. The minimum distance between the crack and the tab region is 2mm, and the insulation layer thickness is reduced by 15% at the crack tip. The crack propagation rate is positively correlated with the temperature field gradient, reaching 0.01mm / hour in the hotspot region. Resistance threshold detection uses an 8×8 dot matrix arrangement with a 1mm spacing between detection points. An early warning is triggered when the insulation resistance drops from the initial value of 10¹⁴Ω to 10¹²Ω. The insulation protective layer is thickened by 0.5mm in high-risk areas, and a buffer layer is placed within 2mm of the tab. After structural optimization, the temperature gradient in the hotspot region was reduced by 30%. The coupling effect of residual stress and temperature field was particularly pronounced in the tab region. When the residual stress exceeded 20 MPa, the local temperature rise reached 3°C. Under thermal stress, microcracks preferentially propagated along the direction of maximum temperature gradient, significantly increasing the risk of forming through-cracks. The insulation performance degradation process has a cumulative effect. Initial microcracks lead to local heating, and thermal stress accelerates crack propagation, forming a positive feedback loop. When the crack density exceeds 3 cracks / mm...3 At that time, the temperature rise rate in the region accelerated, increasing by 0.1℃ per hour. Vibration and thermal cycling in the hot spot area accelerated the damage to the insulation layer. In the short-circuit protection design, the combination of material thickening and a buffer layer was highly effective. The stress in the thickened area decreased by 40%, and the hot spot temperature returned to the reference temperature. The stress dispersion effect of the buffer layer reduced the crack propagation rate by 75%. In temperature cycling tests, the optimized structure maintained stable insulation performance after 500 cycles from -40℃ to 80℃. The key to structural optimization lies in the balance between stress dispersion and heat diffusion. If the thickened area is too large, it will affect the volumetric efficiency of the battery pack; if it is too small, the stress concentration effect will still exist. The thickness of the buffer layer gradually decreases towards the direction away from the tab, forming a gradient structure and avoiding abrupt stress changes.

[0044] S108. Construct a reliability assessment system for injection molding quality of plastic shells. Integrate residual stress, microcrack density, mechanical properties, and electrical properties. Use fuzzy comprehensive evaluation method to quantitatively assess the quality of plastic shells and screen and improve plastic shells with defects.

[0045] The data acquisition device obtains stress distribution, crack density, mechanical strength, and insulation resistance values ​​on the surface of the casing. The device is equipped with stress sampling points, crack density observation points, mechanical performance test points, and insulation performance test points. Based on the stress distribution, crack density, mechanical strength, and insulation resistance values, a range normalizer is used to perform interval mapping on the acquired values ​​to obtain a standardized evaluation parameter matrix. For this standardized evaluation parameter matrix, a principal component analyzer is used to calculate the correlation of characteristic indicators, and a hierarchical structure method is used to quantify the weights of performance parameters to obtain a comprehensive performance evaluation matrix. For this comprehensive performance evaluation matrix, a fuzzy relation matrix is ​​constructed using triangular fuzzy numbers, and an index synthesis operator performs weighted integration of each parameter. A quality evaluation result set and a quality improvement parameter set are then generated by comparing the results against a preset quality level threshold.

[0046] For example, based on the data acquisition device, 16 stress sampling points, 8 crack density observation points, 12 mechanical performance test points, and 10 insulation performance test points are arranged on the surface of the shell. A range normalizer is used to perform interval mapping on the collected values, and a standardized evaluation parameter matrix is ​​generated based on a performance index standard library. From the standardized evaluation parameter matrix, the stress distribution value, crack density value, mechanical strength value, and insulation resistance value of each point are extracted. A principal component analyzer is used to calculate the correlation of the characteristic indicators, and a hierarchical structure method is used to quantify the weights of the performance parameters, generating a comprehensive performance evaluation matrix. Based on the comprehensive performance evaluation matrix, a fuzzy relation matrix is ​​constructed using triangular fuzzy numbers. An index synthesis operator is used to perform weighted integration of each parameter, and the shell is graded and evaluated against a preset quality level threshold, generating a quality evaluation result set. Based on the quality evaluation result set, a defect morphology identifier is used to calibrate improvement parameters in the low-value area of ​​the shell quality rating, including wall thickness optimization values, material strengthening values, and structural reinforcement values. The improvement action points are marked in a spatial coordinate system, generating a quality improvement parameter set. In multi-point performance testing, stress sampling points were uniformly arranged along the circumference of the shell with a spacing of 25 mm. The stress value was 15 MPa in the normal area and reached 35 MPa in the abnormal area. Crack density observation points were arranged in the stress concentration area, with the number of cracks per unit area ranging from 0.5 cracks / mm. 3 Up to 5 strips / mm 3 The results are not consistent. The tensile strength recorded at the mechanical property test points was 45 MPa, and the elongation at break was 150%. The volume resistivity measured at the insulation property test points was 10¹⁴ Ω·cm. Standardization was performed using the range method, mapping all indicators to the 0-1 range. Stress indicators were normalized based on the design allowable stress of 20 MPa. Crack density was set at a safe threshold of 1 crack / mm². 3 The mechanical properties were compared to a standard value of 45 MPa, and the insulation properties were compared to a standard value of 10¹³ Ω·cm. Principal component analysis showed that the correlation coefficient between residual stress and crack density was 0.85, with weights of 0.35 and 0.3 respectively for their impact on quality. The correlation coefficient between mechanical properties and insulation properties was 0.6, with weights of 0.2 and 0.15 respectively. For every 5 MPa increase in residual stress, the crack density increased by 0.5 cracks / mm². 3The fuzzy evaluation uses triangular fuzzy numbers to represent the membership degree of the indicators, setting four levels: excellent, good, qualified, and unqualified. A comprehensive score below 0.6 is considered unqualified. A weighted average operator is used for weight integration, considering the interaction between indicators. The excellent quality range is 0.85 to 1, and the good range is 0.7 to 0.85. Defect improvement parameters include multiple dimensions. The wall thickness optimization value is determined based on stress distribution, increasing the thickness by 0.5 mm in high-stress areas. The material strengthening value is achieved by increasing crystallinity, from 30% to 35%. The structural reinforcement value considers the arrangement of local reinforcing ribs, with a rib height of 0.8 mm. Multi-parameter collaborative optimization shows that there is a coupling effect between the indicators. Increasing wall thickness increases injection molding difficulty and extends the molding cycle by 5%. Material strengthening improves mechanical properties, but excessive crystallization reduces toughness. While structural reinforcement reduces stress, it may introduce new stress concentrations. A balance needs to be struck between improvement parameters during quality improvement. When the wall thickness increase exceeds 0.8 mm, it actually leads to increased internal stress. When the crystallinity exceeds 38%, the material becomes brittle, and its impact resistance decreases. Excessive density of reinforcing ribs can affect material flowability and lead to weld defects. After establishing the evaluation system, quality control benchmarks are formed through batch data accumulation. Excellent quality shells exhibit a stable stress level below 15 MPa and a crack density of less than 0.8 cracks / mm². 3 The tensile strength exceeds 50 MPa, and the insulation resistance remains above 10¹⁴ Ω·cm. These data provide a quantitative basis for process optimization.

[0047] The above description is merely a preferred embodiment of one or more embodiments of this specification and is not intended to limit the scope of one or more embodiments of this specification. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of one or more embodiments of this specification should be included within the protection scope of one or more embodiments of this specification.

Claims

1. A method for predicting short-circuit protection in new energy batteries with rubber casings, characterized in that, The method includes: Based on the constructed three-dimensional solid model of the new energy battery shell, the finite element simulation analysis method is used to input the mechanical property parameters of the material, the geometric dimensions of the parts and the injection molding process parameters to simulate the injection molding process and obtain the residual stress distribution cloud map inside the shell. The residual stress concentration area is determined by the stress distribution cloud map. The microscopic morphology of the surface and interior of the shell in the residual stress concentration area was characterized by microscopic observation to obtain the morphological characteristics of microcrack formation. Combined with material mechanical property parameters, including fracture toughness and fracture strain, the mechanical properties of microcrack formation were analyzed. Mechanical performance tests were conducted on the three-dimensional solid model of the constructed new energy battery shell. The residual stress level was determined based on the yield strength of the material. The influence of the residual stress level on the mechanical performance of the shell was analyzed. A residual stress control threshold was preset. When the residual stress exceeds the residual stress control threshold, the mechanical performance of the shell is determined to be unsatisfactory. Electrical performance testing methods for dielectric constant and volume resistivity were used to evaluate the electrical insulation performance of the constructed three-dimensional solid model of the new energy battery shell, and to determine the microcrack density control range. If the microcrack density exceeded the range, it was determined that the insulation performance of the constructed three-dimensional solid model of the new energy battery shell was reduced. If it is determined that the insulation of the constructed three-dimensional solid model of the new energy battery shell is reduced, the injection molding process parameters, including injection temperature, holding time, and cooling rate, are optimized, and the optimal combination of process parameters is obtained through orthogonal experimental design. Based on the optimal combination of process parameters, injection molded shells are produced. Annealing is performed on the injection molded shells to promote the release of residual stress. Low-frequency vibration is applied to the shells to induce the redistribution of residual stress, reduce the degree of stress concentration, and inhibit the propagation of microcracks. Annealing and vibration aging treatments are performed to improve the residual stress state of the shells. After annealing and vibration aging treatment, infrared thermal imaging is used to scan the surface of the shell to identify local heating areas caused by residual stress. Combined with the stress distribution cloud map in the three-dimensional solid model of the shell, the impact of microcrack propagation in the hot spot area on the insulation performance of the battery tab is predicted, the risk of battery short circuit in the hot spot area is assessed, and short circuit prevention design optimization suggestions are proposed for the shell in high-risk areas. A reliability assessment system for injection molding quality of plastic shells is constructed. By comprehensively considering residual stress, microcrack density, mechanical properties, and electrical properties, a fuzzy comprehensive evaluation method is adopted to quantitatively assess the quality of the plastic shells and screen and improve those with defects.

2. The method according to claim 1, characterized in that, The three-dimensional solid model of the constructed new energy battery shell is used to simulate the injection molding process using the finite element method. The mechanical property parameters of the material, the geometric dimensions of the parts, and the injection molding process parameters are input. This simulation yields a residual stress distribution cloud map inside the shell, which is used to determine the areas of residual stress concentration, including: Collect the dimensional data of the plastic shell structure in the three-dimensional solid model, and generate the elastic deformation curve data of the plastic shell based on the calibration measurement of the material elastic coefficient and Poisson's ratio. Based on the elastic deformation curve data of the plastic shell, the injection melt temperature, injection pressure and injection rate data are collected. Load boundary conditions are applied to the sampling points of the injection station, and stress calculation input data are obtained by multiple linear regression. The finite element mesh generation tool is called based on the stress calculation input data to perform meshing on the three-dimensional solid model, and the numerical distribution dataset of element node stress is calculated based on Hooke's law. Stress cloud map data is constructed based on the stress numerical distribution dataset. The principal stress values ​​are calculated using the von Mises stress criterion. The residual stress cloud map data is obtained by calibrating the coordinates of the stress peak points. The residual stress concentration area is determined by the stress distribution cloud map.

3. The method according to claim 1, characterized in that, The microscopic observation method is used to characterize and analyze the surface and internal micromorphology of the shell in the residual stress concentration area, obtain the morphological characteristics of microcrack formation, and combine them with material mechanical property parameters, including fracture toughness and fracture strain, to analyze the mechanical properties of microcrack formation, including: The surface morphology image of the shell is acquired by a micro-scanning device, and the microcrack boundary contour curve data is obtained by using a Sobel edge extractor based on the morphology image. Strain sampling points are arranged within the gauge length for the microcrack boundary profile curve data. The strain data field distribution map is obtained by recording the correspondence between the fracture strain value and the microcrack boundary position based on the strain sampling points. Based on the strain data field distribution map, the microcrack morphology features are extracted in segments using the edge detection operator. The crack depth, crack propagation rate, and crack orientation angle are calculated using the microcrack morphology features to obtain the crack morphology feature parameter matrix. Based on the crack morphology characteristic parameter matrix and the generalized fracture criterion, the crack tip opening displacement value, stress intensity factor and fracture toughness are calculated. The normal stress component and shear stress component are calibrated in the micro-region fracture element by the opening displacement value, and the crack propagation mechanical parameter field is obtained.

4. The method according to claim 1, characterized in that, The process involves conducting mechanical performance tests on a three-dimensional solid model of the constructed new energy battery casing, determining the residual stress level based on the material's yield strength, analyzing the influence of the residual stress level on the casing's mechanical properties, and setting a preset residual stress control threshold. When the residual stress exceeds this threshold, the casing's mechanical properties are deemed unsatisfactory. This includes: Based on the three-dimensional solid load loader, orthogonal loading points are arranged on the surface of the shell. The displacement field data of the shell surface is collected by the digital image correlation optical strain gauge. A uniaxial tensile load is applied to the loaded area to obtain the strain distribution data of the shell. The tensile strain, compressive strain and shear strain values ​​at the measurement points are extracted from the strain distribution data of the plastic shell. The linear segment values ​​of the material are calibrated using a tensile stress-strain curve calculator to obtain the yield point stress-strain curve. For the yield point stress-strain curve, the shear modulus and bulk modulus of the shell material are solved using the Lamé parameter calculator, and the residual stress value and stress concentration factor are calculated to obtain the residual stress distribution data. Based on the residual stress distribution data, a numerical comparator is used to compare the residual stress value with a preset stress threshold, and the location of the residual stress peak point and the stress level are calibrated to obtain the mechanical performance determination result.

5. The method according to claim 1, characterized in that, The method for testing the dielectric constant and volume resistivity of the constructed three-dimensional solid model of the new energy battery shell is used to evaluate the electrical insulation performance of the constructed three-dimensional solid model of the new energy battery shell, determine the microcrack density control range, and if the microcrack density exceeds the range, it is judged that the insulation performance of the constructed three-dimensional solid model of the new energy battery shell has decreased, including: Test electrodes are arranged on the surface of the shell according to the three-dimensional solid test platform. The surface dielectric constant value is collected by a capacitance meter and the volume resistivity value is recorded by a high voltage DC resistance meter to obtain the basic data of electrical performance. The dielectric constant distribution value and volume resistivity distribution value are extracted from the electrical performance basic data, and the insulation strength of the material is scanned and measured by a voltage-current characteristic curve instrument to obtain insulation performance evaluation data. For the insulation performance evaluation data, a region growing segmenter is used to segment the microcrack image into blocks, and a standard convolutional neural network is used to calculate the number of cracks per unit area and the crack surface area to generate microcrack density feature data. Based on the microcrack density characteristic data, the microcrack characteristic values ​​and insulation performance parameters are mapped using an association matrix calculator to determine the calibration results of the insulation strength weakening area. It also includes: using a dielectric constant tester and a high-resistivity meter to measure the dielectric constant and volume resistivity of the shell material; judging whether the shell material meets the insulation requirements based on electrical insulation performance standards; counting the number and distribution of internal microcracks; analyzing the correlation between microcrack density and insulation performance; and obtaining the control range and threshold of microcrack density of the shell material.

6. The method according to claim 5, characterized in that, The method utilizes a dielectric constant tester and a high-resistivity meter to measure the dielectric constant and volume resistivity of the housing material. Based on electrical insulation performance standards, it determines whether the housing material meets insulation requirements. The method also statistically analyzes the number and distribution of internal microcracks, examines the correlation between microcrack density and insulation performance, and obtains the control range and threshold for the microcrack density of the housing material, including: AC and DC voltages are applied to the surface of the housing using testing instruments to obtain capacitance and resistance values. Comprehensive electrical characteristic data are obtained based on breakdown voltage scanning. Based on the comprehensive electrical characteristic data, the dielectric constant and dielectric loss angle are extracted, and the insulation performance characteristic data are obtained by comparing them with the preset insulation standard value through a numerical comparator. The region growth algorithm is run on the insulation performance feature data to extract the number of cracks per unit area and the crack extension length, and an insulation strength feature matrix is ​​constructed by a convolutional neural network. The correlation value between the crack density parameter and the insulation strength parameter is calculated based on the insulation strength feature matrix. If the correlation value exceeds the preset standard threshold, the microcrack control interval data is determined.

7. The method according to claim 1, characterized in that, If it is determined that the insulation of the constructed three-dimensional solid model of the new energy battery shell is reduced, the injection molding process parameters, including injection temperature, holding time, and cooling rate, are optimized. The optimal combination of process parameters is obtained through orthogonal experimental design, including: According to the injection molding specifications, the process parameter calculator is used to combine and arrange the injection melt temperature range, holding time range, and cooling rate range. The material viscosity value and crystallinity value corresponding to each group of parameters are calibrated through the physical property parameter database to obtain the injection molding process parameter matrix. Extract each set of process configuration parameters from the injection molding process parameter matrix, and record the cavity pressure sensing point, temperature sensing point and flow front sensing point in real time in the injection molding controller to obtain the injection molding process parameter set; For the parameter set of the injection molding process, a gradient boosting tree is used to extract features from the parameter curve data, calculate the insulation contribution value and parameter coupling index of a single parameter, and generate a set of process parameter influence factors; Based on the set of process parameter influencing factors, a parameter response surface is constructed through the central composite designer. Under the constraint of insulation performance optimization, the parameter space is traversed and searched to calibrate the optimal values ​​of melt temperature, holding time, and cooling rate.

8. The method according to claim 1, characterized in that, The process involves producing injection-molded housings based on optimal process parameter combinations, annealing the injection-molded housings to promote residual stress release, applying low-frequency vibration to the housings to induce residual stress redistribution, reducing stress concentration, and inhibiting microcrack propagation. Annealing and vibration aging treatments are then performed to improve the residual stress state of the housings, including: According to the process control parameters, a temperature sensor array is arranged on the surface of the shell, and a temperature acquisition device is used to obtain the temperature field distribution and crystallinity change data to obtain the annealing temperature change curve. The heating rate and holding time parameters are extracted from the annealing temperature change curve. The stress relaxation calculator is used to numerically solve the displacement field inside the shell and obtain the stress relaxation field distribution map. For the stress relaxation field distribution map, an acoustic exciter is used to apply a frequency scanning signal to the shell, and the frequency scanning signal is processed by a neural network fitter to obtain stress redistribution data; Based on the stress redistribution data, a vibration sensor array is arranged in the stress concentration area of ​​the shell. The signals acquired by the vibration sensor array are processed by a phase delay calculator to obtain the microcrack propagation rate and stress intensity factor.

9. The method according to claim 1, characterized in that, After annealing and vibration aging treatment, infrared thermal imaging is used to scan the surface of the casing to identify localized heating areas caused by residual stress. Combined with stress distribution cloud maps in the three-dimensional solid model of the casing, the impact of microcrack propagation in hot spots on the insulation performance of the battery tabs is predicted, the risk of battery short circuits in hot spots is assessed, and short-circuit protection design optimization suggestions are proposed for high-risk areas, including: Based on the scanning data acquired by the thermal imaging acquisition device, the surface temperature of the shell is calibrated by an infrared detector. The infrared detector uses an area integrator to calculate the thermal zone coverage area to obtain thermal zone distribution characteristic data. Temperature gradient field and thermal stress field are extracted from the thermal zone distribution feature data, and thermal stress coupling distribution map is obtained by partitioning and matching the stress distribution cloud map using a region growing algorithm. Based on the aforementioned thermal stress coupling distribution map, a multilayer perceptron is used to identify the characteristics of the microcrack growth path, and the distance between the tab region and the crack tip is calculated in a spatial coordinate system to obtain microcrack evolution prediction data. Based on the microcrack evolution prediction data, a resistance threshold detection array is arranged near the tab insulation layer, and the short-circuit risk area is calibrated by the insulation resistance decay curve to obtain the optimized structural layout diagram.

10. The method according to claim 1, characterized in that, The constructed injection molding quality reliability assessment system for plastic shells comprehensively considers residual stress, microcrack density, mechanical properties, and electrical properties. It employs a fuzzy comprehensive evaluation method to quantitatively assess the quality of the plastic shells, and screens and improves defective shells, including: The detection data acquisition device obtains stress distribution value, crack density value, mechanical strength value and insulation resistance value on the surface of the shell. The detection data acquisition device is equipped with stress sampling points, crack density observation points, mechanical performance test points and insulation performance test points. Based on the stress distribution value, crack density value, mechanical strength value and insulation resistance value, the range normalizer is used to perform interval mapping on the collected values ​​to obtain a standardized evaluation parameter matrix. For the standardized evaluation parameter matrix, the correlation of the feature indicators is calculated by principal component analysis, and the weights of the performance parameters are quantified by the hierarchical structure method to obtain the comprehensive performance evaluation matrix. For the aforementioned performance comprehensive evaluation matrix, a fuzzy relation matrix is ​​constructed using triangular fuzzy numbers. The parameters are then weighted and integrated using an index synthesis operator. A quality evaluation result set and a quality improvement parameter set are generated by comparing the results with a preset quality level threshold.

Images