Transformer structure response calculation and hierarchical operation and maintenance strategy considering nonlinear characteristics of winding block materials

Abstract

This invention belongs to the field of non-electrical variable prediction technology for power equipment, and relates to a transformer structural response calculation and graded operation and maintenance strategy that considers the nonlinear characteristics of winding pad materials. It simulates the real stress state and eliminates residual moisture interference through pad sample preparation and pretreatment; achieves accurate simulation of environmental humidity through dynamic humidity control and verification; obtains high-precision stress-strain data through nonlinear compression testing; divides the elastic, plastic plateau, and densification key strain stages based on local slope and rate of change, and establishes a piecewise power-law constitutive model to characterize the mechanical properties of each stage; combines 3D modeling and multiphysics simulation, and improves model reliability through vibration measurement and boundary condition adjustment; formulates a graded operation and maintenance strategy based on the calculation of pad plastic deformation under short-circuit conditions; improves the accuracy of transformer structural response prediction, and solves the problem of structural response prediction deviation caused by neglecting pad nonlinearity, humidity influence, and insufficient multiphysics coupling in traditional methods.

Description

Technical Field

[0001] This invention belongs to the field of non-electrical variable prediction technology for power equipment, and relates to a transformer structural response calculation and graded operation and maintenance strategy that considers the nonlinear characteristics of winding pad material. Background Technology

[0002] When a power transformer is short-circuited, the transient electromagnetic force experienced by the windings can reach hundreds of kilonewtons. Failure of the spacer compression and winding displacement may lead to insulation breakdown. Traditional structural response analysis has significant limitations, mainly including:

[0003] (1) Material model defects: The compression of the pad exhibits three nonlinear characteristics: elasticity, plastic plateau and densification. The traditional linear elastic assumption cannot characterize the true stress-strain relationship, resulting in a large displacement prediction error. After multiple compression cycles under impact loads, the pad will suffer permanent thickness loss, leading to a continuous decrease in clamping pressure.

[0004] (2) Ignoring environmental factors: The mechanical properties of the pad change significantly after absorbing moisture, but the existing methods have not achieved dynamic control and quantitative analysis of humidity;

[0005] (3) Disconnect between experiment and simulation: Although the existing technology measures the compression curve of the pad block through static experiment, it does not integrate the nonlinear characteristics into the dynamic structural response model, and lacks a high-precision detection and feedback mechanism for key parameters such as humidity and strain, and cannot achieve accurate coupling of multiple physical fields (electromagnetic field-structural mechanics);

[0006] (4) Insufficient quantitative basis for operation and maintenance: After a transformer short circuit occurs, the ability to judge the transformer status by offline means is very limited, and there is a lack of quantitative analysis methods for transformer status based on high-precision simulation calculation.

[0007] The aforementioned limitations directly lead to the inability to determine whether a transformer can continue to operate after a short circuit, relying solely on offline methods such as short-circuit impedance testing. Therefore, there is an urgent need for a high-precision modeling and simulation method, coupled with reliable experimental and test data verification, to comprehensively consider the impact of material nonlinearity and environmental factors on the transformer's structural response, and to realize a graded operation and maintenance strategy for transformers after a short circuit. Summary of the Invention

[0008] The technical problem to be solved by this invention is to provide a transformer structural response calculation and graded operation and maintenance strategy that takes into account the nonlinear characteristics of winding pad material, and solves the problem of structural response prediction deviation caused by insufficient nonlinear modeling of pad, neglect of humidity sensitivity and multi-physics coupling in traditional methods.

[0009] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is: a transformer structural response calculation and graded operation and maintenance strategy considering the nonlinear characteristics of winding pad material, comprising the following steps:

[0010] S1, Preparation and pretreatment of pad block samples;

[0011] S2, Dynamic Humidity Control and Verification;

[0012] S3, Nonlinear Compression Test;

[0013] S4, Key Response Stages;

[0014] S5, Establishment of a piecewise power-law constitutive model;

[0015] S6, 3D modeling;

[0016] S7, normal operating condition simulation calculation;

[0017] S8, Vibration acceleration measurement under normal operating conditions;

[0018] S9, Boundary condition adjustment;

[0019] S10, Short-circuit condition simulation calculation;

[0020] S11, formulate a tiered operation and maintenance strategy;

[0021] In S4, specifically including:

[0022] S4-1, Calculate the local slope point by point for the processed stress-strain data. and local slope change rate :

[0023]

[0024] in and Indicates the first Strain and stress at a point.

[0025] S4-2, Identify the end point of the elastic phase in the loading process. Find the rate of change of local slope The maximum value, corresponding to the strain value ;

[0026] S4-3, Identify the end point of the plastic plateau region during the loading process. Calculate the rate of change of local slope The average value, the point that reaches the threshold is the end point of the plateau region, and the corresponding strain should be... The threshold is the sum of the mean and the standard deviation.

[0027] In S5, specifically including:

[0028] S5-1, Capturing the Loading Process The experimental data were used to fit the elastic phase coefficients using a first-order polynomial. and initial residual stress The constitutive equation for the elastic stage of the pad is obtained as follows:

[0029]

[0030] in and Representing stress and strain respectively;

[0031] S5-2, Capturing the Loading Process The experimental data were used to optimize the fit using the Levenberg-Marquardt algorithm. ~ The constitutive equation for the plastic plateau stage of the pad block is obtained as follows:

[0032]

[0033] in and Representing stress and strain respectively. ~ These are the proportionality coefficient, exponential coefficient, and constant term for the plastic plateau stage, respectively.

[0034] S5-3, Capturing the Loading Process The experimental data were used to optimize the fit using the Levenberg-Marquardt algorithm. ~ The constitutive equation for the densification stage of the pad is obtained as follows:

[0035]

[0036] in and Representing stress and strain respectively. and The stress and strain at the end of the plastic plateau stage. ~ These are the proportional coefficient and exponential coefficient for the densification stage of the pad block, respectively.

[0037] S5-4, Extracting experimental data from the unloading phase, and using the Levenberg-Marquardt algorithm to optimize the coefficients of the unloading phase to obtain the constitutive equation for the unloading phase:

[0038]

[0039] in and Representing stress and strain respectively. This represents the residual stress after final unloading. Indicates the starting point of unloading strain. and This represents the coefficient for the unloading phase.

[0040] S1 specifically includes:

[0041] S1-1, a test block with dimensions of 8 mm in length, 8 mm in width and 6 mm in thickness was prepared by directional hot pressing process. The hot pressing parameters were controlled as follows: pressure 10±0.1MPa, temperature 75±2℃ and time 20 minutes, to simulate the actual stress distribution of the transformer pad.

[0042] S1-2, First stage rapid dehydration: Dehydrate in a vacuum drying oven for 12 hours to remove free water;

[0043] S1-3, Second stage of stable drying: High-purity nitrogen gas is introduced, and the temperature is maintained at 60±1℃ and the vacuum degree is <5 kPa for 12 hours to continue drying.

[0044] S1-4, Endpoint determination: Using an online mass spectrometer, the resolution must be ≤0.001 mg / (g·h), the water release rate must be monitored in real time and be <0.01 mg / (g·h), and the initial humidity must be ≤0.5%.

[0045] Specifically, S2 includes:

[0046] S2-1, Prepare a saturated salt solution and place it in a constant temperature container. Pass dry nitrogen gas into the solution to generate a humid gas.

[0047] S2-2, Place the insulating pad into a sealed container, set the target humidity according to the actual operating environment of the transformer, introduce the above-mentioned gas, and dynamically adjust the dry / wet nitrogen ratio by using a high-precision temperature and humidity sensor to ensure that the humidity fluctuation is ≤±0.1%;

[0048] S2-3, Steady-state maintenance: Maintain constant temperature at the target humidity for 24 hours, scan the internal humidity distribution using a near-infrared spectrometer, and verify that the humidity error is less than 0.1% by sampling and titration.

[0049] In S3, specifically including:

[0050] S3-1, using a material mechanics testing device with a range ≥10MPa and an accuracy ≤±0.1% of the full scale FS, load to 10MPa at a strain rate of 0.05% / s, hold for 10 seconds, and then unload at the same rate to 0.3MPa residual stress. Record stress-strain data, with a sampling frequency of 100 Hz. Repeat the experiment 5 times for each set of conditions and take the average value.

[0051] S3-2 uses data processing software to extract discrete data points from experimental curve images, including two columns of data: strain (mm / mm) and stress (MPa);

[0052] S3-3, Data preprocessing: A Savitzky-Golay filter with a window width of 5 and a polynomial order of 2 is used to eliminate high-frequency noise.

[0053] In S6, a 3D model of the transformer was created using SolidWorks, including the main structures of high and low voltage windings, pressure plates, pads, and core. The model was then imported into ANSYS Workbench to generate a hexahedral dominant mesh, which was locally refined to 0.3 mm with a Jacobian ratio > 0.6.

[0054] In S7, specifically including:

[0055] S7-1, set the material parameters for calculating the magnetic field of the winding and core, where the BH curve of the silicon steel sheet uses the factory parameters;

[0056] S7-2, calculate the normal operating conditions, including no-load closing and rated operating current. This provides the excitation for magnetic field calculation;

[0057] S7-3, calculate the Lorentz force on the winding by solving the magnetic field distribution using ANSYS Maxwell.

[0058] Based on Maxwell's equations, the governing equations for the magnetic field are derived. These equations are related to the magnetic vector potential. Related:

[0059] ;

[0060] in, It is a current density vector. It is the permeability. It is electrical conductivity;

[0061] S7-4, Structural Response Analysis under Normal Operating Conditions;

[0062] The magnitude of the current density vector under normal operating conditions is defined as the current. The cross-sectional area of ​​the conductors in the transformer winding The ratio:

[0063]

[0064] in, It is a unit vector indicating the tangential direction at the winding;

[0065] Furthermore, the forces acting on the windings are calculated according to the Lorentz force law:

[0066]

[0067] in, It is the magnetic flux density;

[0068] The differential equation of motion in solid mechanics for calculating the structural response of a transformer is as follows:

[0069]

[0070] in, This indicates that the winding is under stress. , and These represent the mass matrix, damping matrix, and transformer winding deformation, respectively. , and These represent acceleration, velocity, and displacement, respectively.

[0071] In S8, specifically including:

[0072] Accelerometers and stress sensors are arranged on the upper and lower pressure plates of the winding, with two sensors arranged circumferentially to measure stress and acceleration under the same normal and simulated operating conditions.

[0073] In S9, specifically including:

[0074] Compare the results measured in S8 with the simulation results at the corresponding positions in S7. If the simulation and experimental error of the acceleration amplitude is greater than 5%, adjust the key boundary conditions in S7, namely the prestress of the pressure plate, until the simulation and measured results of the acceleration amplitude are less than 5%.

[0075] Specifically, S10 includes:

[0076] S10-1, set the material parameters for calculating the magnetic field of the winding and core, where the BH curve of the silicon steel sheet uses the factory parameters;

[0077] S10-2, Calculate the short-circuit current:

[0078]

[0079] in, It is short-circuit current. It is the effective value of the steady-state current; Indicates time, and These represent the equivalent resistance and equivalent leakage reactance, respectively. It is the initial phase angle. It is angular frequency;

[0080] S10-3, the magnetic field distribution is solved using ANSYS Maxwell, and the Lorentz force density is derived.

[0081] Based on Maxwell's equations, the governing equations for the magnetic field are derived. These equations are related to the magnetic vector potential. A Related:

[0082] ;

[0083] in, It is a current density vector. It is the permeability. It is electrical conductivity;

[0084] Furthermore, the magnitude of the short-circuit current density vector is defined as the ratio of the short-circuit current to the cross-sectional area of ​​a single conductor in the transformer winding:

[0085]

[0086] in, It is a unit vector indicating the tangential direction at the winding;

[0087] Furthermore, the forces acting on the windings are calculated according to the Lorentz force law:

[0088]

[0089] S10-4, Structural response analysis under short-circuit conditions;

[0090] In the finite element simulation software, the material parameters of the winding, core, and pad block for structural response analysis were set. The copper conductor used a linear elastic model with an elastic modulus of 110 GPa; the silicon steel sheet of the core used the factory parameters with an elastic modulus of 200 GPa; the experimental results obtained from S5 were used to set the pad block material to adopt a segmented constitutive model, and the prestress boundary conditions of the pressure plate were set according to the values ​​determined in S9.

[0091] The differential equation of motion in solid mechanics is:

[0092]

[0093] in, This indicates that the winding is under stress. , and These represent the mass matrix, damping matrix, and transformer winding deformation, respectively. , and These represent acceleration, velocity, and displacement, respectively.

[0094] Specifically, S11 includes:

[0095] Based on the calculated plastic deformation results of the transformer pads, the following graded operation and maintenance strategy is formulated:

[0096] If the maximum plastic deformation of the pad is less than a, the transformer can be put into normal operation and its operating status should be continuously monitored.

[0097] If the maximum plastic deformation of the pad is greater than or equal to a, and less than a c th The transformer is put into operation for 15 minutes and the vibration data is observed. If the vibration data is abnormal, the transformer is shut down and the oil is drained for inspection.

[0098] If the maximum plastic deformation of the pad is greater than or equal to c th If not, the transformer cannot be put into operation; drain the oil and check the transformer windings.

[0099] critical value a and c th Determined based on the maximum oil passage height of the transformer.

[0100] In summary, this invention simulates real-world operating conditions by preparing pad samples and precisely controlling initial humidity, combined with dynamic humidity regulation to achieve realistic environmental simulation, thus solving the problem of traditional methods neglecting the influence of humidity. Through nonlinear compression testing and quantitative division of key strain stages, it provides high-quality data and clear boundaries for constructing a piecewise power-law constitutive model, overcoming the prediction bias of traditional linear models and improving deformation prediction accuracy. Based on 3D modeling and multiphysics simulation, and after verification under normal operating conditions and boundary condition calibration, it achieves precise coupling between electromagnetic fields and structural mechanics, ensuring the reliability of short-circuit simulation. Finally, based on the plastic deformation of the pad samples, a graded operation and maintenance strategy is developed, transforming simulation results into quantitative maintenance standards, solving the problem of lack of basis for traditional decision-making, balancing safety and economy, providing high-precision support for transformer post-short-circuit condition assessment and operation and maintenance, and reducing operation and maintenance costs. Attached Figure Description

[0101] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0102] Figure 1 This is a flowchart of the present invention.

[0103] Figure 2 This is the piecewise power-law constitutive model (stress-strain relationship) of the pad block in this invention.

[0104] Figure 3 This is a three-dimensional model of the 6.3kV transformer in this invention.

[0105] Figure 4 This is the BH curve for silicon steel sheets.

[0106] Figure 5 The results are the calculated results of the magnetic field and structural response under normal operating conditions.

[0107] Figure 6 This is a diagram showing the layout of the accelerometer and stress sensor.

[0108] Figure 7 This is a comparison of the simulated and measured acceleration values ​​under normal operating conditions.

[0109] Figure 8 The results are the magnetic field calculations and structural response calculations under short-circuit conditions. Detailed Implementation

[0110] like Figures 1-8 The present invention relates to a transformer structural response calculation and graded operation and maintenance strategy considering the nonlinear characteristics of winding pad material, comprising the following steps:

[0111] S1, Preparation and pretreatment of pad block samples;

[0112] S2, Dynamic Humidity Control and Verification;

[0113] S3, Nonlinear Compression Test;

[0114] S4, Key Response Stages;

[0115] S5, Establishment of a piecewise power-law constitutive model;

[0116] S6, 3D modeling;

[0117] S7, normal operating condition simulation calculation;

[0118] S8, Vibration acceleration measurement under normal operating conditions;

[0119] S9, Boundary condition adjustment;

[0120] S10, Short-circuit condition simulation calculation;

[0121] S11, formulate a tiered operation and maintenance strategy;

[0122] In S4, specifically including:

[0123] S4-1, Calculate the local slope point by point for the processed stress-strain data. and local slope change rate :

[0124]

[0125] in and Indicates the first Strain and stress at a point.

[0126] S4-2, Identify the end point of the elastic phase in the loading process. Find the rate of change of local slope The maximum value, corresponding to the strain value ;

[0127] S4-3, Identify the end point of the plastic plateau region during the loading process. Calculate the rate of change of local slope The point where the mean reaches a threshold (the sum of the mean and standard deviation) is the end point of the plateau region, and the corresponding strain should be... ;

[0128] The purpose of S4 is to achieve quantitative identification of the "inflection point" of the stress-strain curve by calculating the local slope and the rate of change of slope, thus solving the problem of the strong subjectivity of the boundary in the traditional manual interpretation stage. Using the maximum value and threshold as criteria reduces the error in dividing the elastic stage and the plastic plateau stage, provides accurate boundaries for the establishment of the piecewise constitutive model, and ensures that the model can accurately distinguish the mechanical properties of different stages.

[0129] In S5, specifically including:

[0130] S5-1, Capturing the Loading Process The experimental data were used to fit the elastic phase coefficients using a first-order polynomial. and initial residual stress The constitutive equation for the elastic stage of the pad is obtained as follows:

[0131]

[0132] in and These represent stress and strain, respectively.

[0133] S5-2, Capturing the Loading Process The experimental data were used to optimize the fit using the Levenberg-Marquardt algorithm. ~ The constitutive equation for the plastic plateau stage of the pad block is obtained as follows:

[0134]

[0135] in and Representing stress and strain respectively. ~ These are the proportional coefficient, exponential coefficient, and constant term for the plastic plateau stage, respectively.

[0136] S5-3, Capturing the Loading Process The experimental data were used to optimize the fit using the Levenberg-Marquardt algorithm. ~ The constitutive equation for the densification stage of the pad is obtained as follows:

[0137]

[0138] in and Representing stress and strain respectively. and The stress and strain at the end of the plastic plateau stage. ~ These are the proportional coefficient and exponential coefficient for the densification stage of the pad block, respectively.

[0139] S5-4, Extracting experimental data from the unloading phase, and using the Levenberg-Marquardt algorithm to optimize the coefficients of the unloading phase to obtain the constitutive equation for the unloading phase:

[0140]

[0141] in and Representing stress and strain respectively. This represents the residual stress after final unloading. Indicates the starting point of unloading strain. and These represent the proportional coefficient and exponential coefficient of the unloading phase, respectively.

[0142] The purpose of S5 is to characterize the characteristics of the pad block in the elastic (linear) and plastic plateau (gradual stress growth) and densification (rapid stress growth) stages respectively, significantly reducing the stress prediction error compared to the traditional linear model. By introducing exponential and power functions for fitting, the nonlinear evolution law of the plastic stage is accurately captured, solving the divergence problem of traditional polynomial fitting under large strain. A separate unloading stage equation is established, taking into account the irreversible strain characteristics of the pad block, improving the deformation prediction accuracy under cyclic loading, and better conforming to the short-circuit impact scenario in actual operation.

[0143] In the preferred embodiment, S1 specifically includes:

[0144] S1-1, a test block with dimensions of 8 mm in length, 8 mm in width and 6 mm in thickness was prepared by directional hot pressing process. The hot pressing parameters were controlled as follows: pressure 10±0.1MPa, temperature 75±2℃ and time 20 minutes, to simulate the actual stress distribution of the transformer pad.

[0145] S1-2, First stage rapid dehydration: Dehydrate in a vacuum drying oven (95±1℃) for 12 hours to remove free water;

[0146] S1-3, Second stage of stable drying: High-purity nitrogen gas is introduced, and the temperature is maintained at 60±1℃ and the vacuum degree is <5 kPa for 12 hours to continue drying.

[0147] S1-4, Endpoint determination: Using an online mass spectrometer, the resolution must be ≤0.001 mg / (g·h), the water release rate must be monitored in real time and be <0.01 mg / (g·h), and the initial humidity must be ≤0.5%.

[0148] The purpose of the above scheme is to simulate the stress distribution state of the transformer pad during actual operation through directional hot pressing process, so that the mechanical properties of the test block are closer to the real working conditions and the deviation from reality is reduced. Two-stage drying and high-precision moisture monitoring ensure that the initial humidity of the test block is ≤0.5%, eliminating the interference of residual moisture on the mechanical properties of the material and providing a reference state for humidity control experiments.

[0149] Specifically, S2 includes:

[0150] S2-1, Prepare a saturated salt solution and place it in a constant temperature container. Pass dry nitrogen gas into the solution to generate a humid gas.

[0151] S2-2, Place the insulating pad into a sealed container, set the target humidity according to the actual operating environment of the transformer, introduce the above-mentioned gas, and dynamically adjust the dry / wet nitrogen ratio by using a high-precision temperature and humidity sensor (accuracy ±0.05%RH) to ensure that the humidity fluctuation is ≤±0.1%;

[0152] S2-3, steady-state maintenance: maintain constant temperature (25±0.1℃) at the target humidity for 24 hours, scan the internal humidity distribution using a near-infrared spectrometer, and verify that the humidity error is less than 0.1% by sampling and titration.

[0153] The purpose of the above scheme is to improve the realism of humidity environment simulation. During the steady-state maintenance stage, near-infrared spectroscopy scanning and titration verification are used to ensure that the humidity distribution inside the pad is uniform. This provides a reliable humidity parameter basis for accurately studying the impact of humidity on the mechanical properties of the pad, and avoids the problems of environmental simulation distortion or excessive parameter error in traditional humidity control.

[0154] In S3, specifically including:

[0155] S3-1, using a materials mechanics testing device with a range ≥10MPa and an accuracy ≤±0.1% of full scale FS, load to 10MPa at a strain rate of 0.05% / s, hold for 10 seconds, then unload at the same rate to 0.3MPa residual stress, and record the stress. -strain Data was sampled at a frequency of 100 Hz, and each set of conditions was repeated 5 times and the average value was taken.

[0156] S3-2 uses data processing software to extract discrete data points from experimental curve images, including two columns of data: strain (mm / mm) and stress (MPa);

[0157] S3-3, Data preprocessing: A Savitzky-Golay filter with a window width of 5 and a polynomial order of 2 is used to eliminate high-frequency noise.

[0158] The purpose of the above scheme is to simulate the dynamic compression process of the pad block under short circuit with a higher strain rate, overcoming the problem of large deviation in the traditional static test rate, and ensuring that the stress-strain data can reflect the transient response characteristics. The high-precision test equipment captures the subtle changes in the stress-strain curve, avoiding the loss of feature points caused by low-precision equipment. The average value of 5 repeated experiments and Savitzky-Golay filtering effectively reduce random errors and high-frequency noise, improve the signal-to-noise ratio of the data, and provide high-quality raw data for subsequent stage division.

[0159] In S6, a 3D model of the transformer was created using SolidWorks, including the main structures of high and low voltage windings, pressure plates, pads, and core. The model was then imported into ANSYS Workbench to generate a hexahedral dominant mesh, which was locally refined to 0.3 mm with a Jacobian ratio > 0.6.

[0160] In S7, specifically including:

[0161] S7-1, set the material parameters for calculating the magnetic field of the winding and core, where the BH curve of the silicon steel sheet uses the factory parameters;

[0162] S7-2, calculate the normal operating conditions, including no-load closing and rated operating current. This provides the excitation for magnetic field calculation;

[0163] S7-3, calculate the Lorentz force on the winding by solving the magnetic field distribution using ANSYS Maxwell.

[0164] Based on Maxwell's equations, the governing equations for the magnetic field are derived. These equations are related to the magnetic vector potential. Related:

[0165] ;

[0166] in, It is a current density vector. It is the permeability. It is electrical conductivity;

[0167] S7-4, Structural Response Analysis under Normal Operating Conditions;

[0168] The magnitude of the current density vector under normal operating conditions is defined as the current. The cross-sectional area of ​​the conductors in the transformer winding The ratio:

[0169]

[0170] in, It is a unit vector indicating the tangential direction at the winding;

[0171] Furthermore, the forces acting on the windings are calculated according to the Lorentz force law:

[0172]

[0173] in, It is the magnetic flux density;

[0174] The differential equation of motion in solid mechanics for calculating the structural response of a transformer is as follows:

[0175]

[0176] in, This indicates that the winding is under stress. , and These represent the mass matrix, damping matrix, and transformer winding deformation, respectively. , and These represent acceleration, velocity, and displacement, respectively.

[0177] The purpose of the above scheme is to use normal operating condition simulation as a benchmark to verify the effectiveness of the model under normal operating conditions and provide credibility support for short-circuit condition simulation.

[0178] In S8, specifically including:

[0179] Accelerometers and stress sensors are arranged on the upper and lower pressure plates of the winding, with two sensors arranged circumferentially to measure stress and acceleration under the same normal and simulated operating conditions.

[0180] Its purpose is to directly compare the measured data with the simulation results, providing quantitative indicators for subsequent boundary condition adjustments.

[0181] In S9, specifically including:

[0182] Compare the results measured in S8 with the simulation results at the corresponding positions in S7. If the simulation and experimental error of the acceleration amplitude is greater than 5%, adjust the key boundary conditions in S7, namely the prestress of the pressure plate, until the simulation and measured results of the acceleration amplitude are less than 5%.

[0183] Its purpose is to use acceleration amplitude as a calibration index to achieve precise adjustment of boundary conditions, solving the problem of setting parameters based on experience in the traditional method; it controls the error within 5%, ensuring the reliability of the model in subsequent short-circuit condition simulations, and reducing the basic error of short-circuit displacement prediction to an acceptable range.

[0184] Specifically, S10 includes:

[0185] S10-1, set the material parameters for calculating the magnetic field of the winding and core, where the BH curve of the silicon steel sheet uses the factory parameters;

[0186] S10-2, Calculate the short-circuit current:

[0187]

[0188] in, It is short-circuit current. It is the effective value of the steady-state current; Indicates time, and These represent the equivalent resistance and equivalent leakage reactance, respectively. It is the initial phase angle. It is angular frequency;

[0189] S10-3, the magnetic field distribution is solved using ANSYS Maxwell, and the Lorentz force density is derived.

[0190] Based on Maxwell's equations, the governing equations for the magnetic field are derived. These equations are related to the magnetic vector potential. A Related:

[0191] ;

[0192] in, It is a current density vector. It is the permeability. It is electrical conductivity;

[0193] Furthermore, the magnitude of the short-circuit current density vector is defined as the ratio of the short-circuit current to the cross-sectional area of ​​a single conductor in the transformer winding:

[0194]

[0195] in, It is a unit vector indicating the tangential direction at the winding;

[0196] Furthermore, the forces acting on the windings are calculated according to the Lorentz force law:

[0197]

[0198] S10-4, Structural response analysis under short-circuit conditions;

[0199] In the finite element simulation software, the material parameters of the winding, core, and pad block for structural response analysis were set. The copper conductor used a linear elastic model with an elastic modulus of 110 GPa; the silicon steel sheet of the core used the factory parameters with an elastic modulus of 200 GPa; the experimental results obtained from S5 were used to set the pad block material to adopt a segmented constitutive model, and the prestress boundary conditions of the pressure plate were set according to the values ​​determined in S9.

[0200] The differential equation of motion in solid mechanics is:

[0201]

[0202] in, This indicates that the winding is under stress. , and These represent the mass matrix, damping matrix, and transformer winding deformation, respectively. , and These represent acceleration, velocity, and displacement, respectively.

[0203] Specifically, S11 includes:

[0204] Based on the calculated plastic deformation results of the transformer pads, the following graded operation and maintenance strategy is formulated:

[0205] If the maximum plastic deformation of the pad is less than a, the transformer can be put into normal operation and its operating status should be continuously monitored.

[0206] If the maximum plastic deformation of the pad is greater than or equal to a, and less than a c th The transformer is put into operation for 15 minutes and the vibration data is observed. If the vibration data is abnormal, the transformer is shut down and the oil is drained for inspection.

[0207] If the maximum plastic deformation of the pad is greater than or equal to c th If not, the transformer cannot be put into operation; drain the oil and check the transformer windings.

[0208] critical value a and c th Determined based on the maximum oil passage height of the transformer.

[0209] Its purpose is to transform simulation results into actionable maintenance thresholds, such as a and b, which are determined based on the oil passage height, thus solving the problem of the lack of quantitative standards in traditional maintenance decisions. The tiered strategy balances safety and economy, avoids over-maintenance, such as shutting down or missing inspections due to slight deformation, and neglecting to address severe deformation, thereby reducing operation and maintenance costs.

[0210] Example

[0211] Taking a 6.3kV transformer as the object, structural response simulation calculations and graded operation and maintenance strategies after a short circuit are performed. The winding state of the transformer after a short circuit impact is determined according to the following steps:

[0212] A transformer structural response calculation and graded operation and maintenance strategy considering the nonlinear characteristics of winding pad materials, the specific steps of which are as follows:

[0213] 1. Preparation and pretreatment of the pad sample;

[0214] 1.1 A test block with dimensions of 8 mm in length, 8 mm in width, and 6 mm in thickness was prepared using a directional hot pressing process. The hot pressing parameters were controlled as follows: pressure 10 ± 0.1 MPa, temperature 75 ± 2 °C, and time 20 minutes, to simulate the actual stress distribution of the transformer pad.

[0215] 1.2, First stage rapid dehydration: Dehydrate in a vacuum drying oven (95±1℃) for 12 hours to remove free water;

[0216] 1.3 Second stage of stable drying: High-purity nitrogen gas is introduced, and the temperature is maintained at 60±1℃ and the vacuum degree is <5 kPa for 12 hours to continue drying;

[0217] 1.4 Endpoint determination: Using an online mass spectrometer, the resolution should be ≤0.001 mg / (g·h), and the water release rate should be monitored in real time to be <0.01 mg / (g·h), ensuring that the initial humidity is ≤0.5%.

[0218] 2. Dynamic humidity control and verification;

[0219] 2.1 Prepare a saturated salt solution and place it in a constant temperature container. Pass dry nitrogen gas into the solution to generate a humid gas.

[0220] 2.2 Place the insulating pad into a sealed container, set the target humidity according to the actual operating environment of the transformer, introduce the above-mentioned gas, and dynamically adjust the dry / wet nitrogen ratio by using a high-precision temperature and humidity sensor (accuracy ±0.05% RH) to ensure that the humidity fluctuation is ≤±0.1%.

[0221] 2.3 Steady-state maintenance: The humidity was maintained at a constant temperature (25±0.1℃) for 24 hours at the target humidity. The internal humidity distribution was scanned using a near-infrared spectrometer, and samples were taken for titration to verify that the humidity error was less than 0.1%.

[0222] 3. Nonlinear compression test;

[0223] 3.1 Using a materials mechanics testing device (range ≥ 10 MPa, accuracy ≤ ± 0.1% of full scale (FS)), load to 10 MPa at a strain rate of 0.05% / s, hold for 10 seconds, and then unload to 0.3 MPa residual stress at the same rate. Record stress-strain data (sampling frequency 100 Hz). Repeat the experiment 5 times for each set of conditions and take the average value.

[0224] 3.2 Use data processing software to extract discrete data points from the experimental curve images, including two columns of data: strain (mm / mm) and stress (MPa);

[0225] 3.3 Data preprocessing: Savitzky-Golay filter (window width 5, polynomial order 2) was used to eliminate high-frequency noise.

[0226] 4. Division of key response stages;

[0227] 4.1 For the processed stress-strain data, calculate the local slope point by point. and local slope change rate :

[0228]

[0229] in and Indicates the first Strain and stress at a point.

[0230] 4.2 Identify the end point of the elastic phase in the loading process Find the rate of change of local slope The maximum value, corresponding to the strain value The strain corresponding to the embodiment It is 0.0604.

[0231] 4.3 Identifying the end point of the plastic plateau region during the loading process Calculate the rate of change of local slope The average value, the point that reaches the threshold is the end point of the plateau region, and the corresponding strain should be... The threshold is the sum of the mean and the standard deviation; the corresponding strain in the example It is 0.1299.

[0232] 5. Establishment of a piecewise power-law constitutive model;

[0233] 5.1 Capturing the loading process The experimental data were used to fit the elastic phase coefficients using a first-order polynomial. and initial residual stress The constitutive equation for the elastic stage of the pad is obtained as follows:

[0234]

[0235] in and These represent stress and strain, respectively. (Example) =3.2119, =0.0013.

[0236] 5.2 Capturing the loading process The experimental data were used to optimize the fit using the Levenberg-Marquardt algorithm. ~ The constitutive equation for the plastic plateau stage of the pad block is obtained as follows:

[0237]

[0238] in and Representing stress and strain respectively. ~ These are the proportionality coefficient, exponential coefficient, and constant term for the plastic plateau stage, respectively. In the example... =6.2505, = 11.2622, =-6.0552.

[0239] 5.3 Intercepting the loading process The experimental data were used to optimize the fit using the Levenberg-Marquardt algorithm. ~ The constitutive equation for the densification stage of the pad is obtained as follows:

[0240]

[0241] in and Representing stress and strain respectively. and The stress and strain at the end of the plastic plateau stage. ~ These are the proportional coefficient and exponential coefficient for the densification stage of the pad, respectively. In the example... =7.6875, =223.8142, =1.0545.

[0242] 5.4 Extracting experimental data from the unloading phase, the Levenberg-Marquardt algorithm is used to optimize the coefficients of the unloading phase, yielding the constitutive equation for the unloading phase:

[0243]

[0244] in and Representing stress and strain respectively. This represents the residual stress after final unloading. Indicates the starting point of unloading strain. and These represent the proportional coefficient and exponential coefficient for the unloading phase, respectively. (Example) =10.7393, =0.148626, =108.0217, =0.8435.

[0245] Piecewise power-law constitutive model of pad blocks, such as Figure 2 As shown.

[0246] 6. 3D modeling;

[0247] A 3D model of the transformer was created using SolidWorks, including the main structures such as high and low voltage windings, pressure plates, pads, and the core. This model was then imported into ANSYS Workbench to generate a hexahedral dominant mesh, locally refined to 0.3 mm (Jacob ratio > 0.6). Figure 3 As shown;

[0248] 7. Simulation calculation under normal operating conditions;

[0249] 7.1 Set the material parameters for the winding and core magnetic field calculation, where the BH curve of the silicon steel sheet uses the factory parameters, such as... Figure 4 As shown;

[0250] 7.2 Calculate the current under normal operating conditions (no-load closing, rated operation, etc.) This provides the excitation for magnetic field calculation;

[0251] 7.3 Solve the magnetic field distribution using ANSYS Maxwell and derive the Lorentz force on the winding.

[0252] Based on Maxwell's equations, the governing equations for the magnetic field are derived. These equations are related to the magnetic vector potential. Related:

[0253] ;

[0254] in, It is a current density vector. It is the permeability. It is electrical conductivity;

[0255] Furthermore, the magnitude of the current density vector under normal operating conditions is defined as the ratio of the current to the cross-sectional area of ​​the conductors in the transformer windings:

[0256]

[0257] in, It is a unit vector indicating the tangential direction at the winding;

[0258] Furthermore, the forces acting on the windings are calculated according to the Lorentz force law:

[0259]

[0260] in, It is the magnetic flux density.

[0261] 7.4 Structural response analysis under normal operating conditions;

[0262] In the finite element simulation software, material parameters for the winding, core, and pads in the structural response analysis were set. The copper conductor used a linear elastic model with an elastic modulus of 110 GPa; the silicon steel sheet of the core used factory parameters with an elastic modulus of 200 GPa; and the pad material adopted a piecewise constitutive model. Using the experimental results obtained in step 5, boundary conditions such as pressure plate prestress were set, given according to the factory's empirical value of 3 MPa.

[0263] The differential equation of motion in solid mechanics for calculating the structural response of a transformer is as follows:

[0264]

[0265] in, This indicates that the winding is under stress. , and These represent the mass matrix, damping matrix, and transformer winding deformation, respectively. , and These represent acceleration, velocity, and displacement, respectively.

[0266] The calculation results of the magnetic field under normal operating conditions and the structural response contour plot are as follows: Figure 5 As shown.

[0267] 8. Vibration acceleration measurement under normal operating conditions;

[0268] Accelerometers and stress sensors are arranged on the upper and lower pressure plates of the winding, with two sensors arranged circumferentially on each plate. Figure 6 As shown, the stress and acceleration are measured under normal operating conditions (the same as the simulation conditions).

[0269] 9. Boundary condition adjustment;

[0270] Compare the results obtained from measurements in step 8 with the simulation results at the corresponding positions in step 7, such as... Figure 7 As shown, if the simulation and experimental error of the acceleration amplitude is greater than 5%, adjust the prestress of the pressure plate in step 7 until the simulation and measured results of the acceleration amplitude are less than 5%.

[0271] 10. Short-circuit condition simulation calculation;

[0272] 10.1 Set the material parameters for the magnetic field calculation of the winding and core, where the BH curve of the silicon steel sheet uses the factory parameters;

[0273] 10.2 Calculate the short-circuit current:

[0274]

[0275] in, It is short-circuit current. It is the effective value of the steady-state current; Indicates time, and These represent the equivalent resistance and equivalent leakage reactance, respectively. It is the initial phase angle. It is angular frequency.

[0276] 10.3 Solve the magnetic field distribution using ANSYS Maxwell and derive the Lorentz force density.

[0277] Based on Maxwell's equations, the governing equations for the magnetic field are derived. These equations are related to the magnetic vector potential. Related:

[0278] ;

[0279] in, It is a current density vector. It is the permeability. It is electrical conductivity;

[0280] Furthermore, the magnitude of the short-circuit current density vector is defined as the ratio of the short-circuit current to the cross-sectional area of ​​a single conductor in the transformer winding:

[0281]

[0282] in, It is a unit vector indicating the tangential direction at the winding;

[0283] Furthermore, the forces acting on the windings are calculated according to the Lorentz force law:

[0284]

[0285] 10.4 Structural response analysis under short-circuit conditions;

[0286] In the finite element simulation software, set the material parameters for the winding, core, and pad for structural response analysis. The copper conductor uses a linear elastic model with an elastic modulus of 110 GPa; the silicon steel sheet for the core uses the factory parameters with an elastic modulus of 200 GPa. Using the experimental results obtained from S5, set the pad material to adopt a segmented constitutive model, set boundary conditions such as pressure plate prestress, and determine them according to step 9.

[0287] The differential equation of motion in solid mechanics is:

[0288]

[0289] in, This indicates that the winding is under stress. , and These represent the mass matrix, damping matrix, and transformer winding deformation, respectively. , and These represent acceleration, velocity, and displacement, respectively.

[0290] The calculation results of the magnetic field under short-circuit conditions and the structural response contour plot are as follows: Figure 8 As shown.

[0291] 11. Develop a tiered operation and maintenance strategy.

[0292] Based on the plastic deformation results of the transformer pad obtained in the previous step, the following graded operation and maintenance strategy is formulated:

[0293] 1) If the maximum plastic deformation of the pad is less than a=0.02mm, the transformer can be put into normal operation and the operating status should be continuously observed;

[0294] 2) If the maximum plastic deformation of the pad is greater than or equal to a = 0.02 mm, and less than... c th =0.05mm, the transformer is put into operation for 15 minutes and the vibration data is observed. If the vibration data is abnormal, the transformer is shut down and the oil is drained for inspection.

[0295] 3) If the maximum plastic deformation of the pad is greater than or equal to c th If the diameter is less than 0.05mm, the transformer cannot be put into operation. Drain the oil and check the transformer windings.

[0296] In this embodiment, the maximum plastic deformation is less than 0.02 mm, therefore it can be put into normal operation.

[0297] The above embodiments are merely preferred technical solutions of the present invention and should not be considered as limitations on the present invention. The embodiments and features described in these embodiments can be arbitrarily combined without conflict. The scope of protection of the present invention should be limited to the technical solutions described in the claims, including equivalent substitutions of the technical features described in the claims. That is, equivalent substitutions and improvements within this scope are also within the scope of protection of the present invention.

Claims

1. A transformer structural response calculation and graded operation and maintenance strategy considering the nonlinear characteristics of winding pad material, characterized in that, Includes the following steps: S1, Preparation and pretreatment of pad block samples; S2, Dynamic Humidity Control and Verification; S3, Nonlinear Compression Test; S4, Key Response Stages; S5, Establishment of a piecewise power-law constitutive model; S6, 3D modeling; S7, normal operating condition simulation calculation; S8, Vibration acceleration measurement under normal operating conditions; S9, Boundary condition adjustment; S10, Short-circuit condition simulation calculation; S11, formulate a tiered operation and maintenance strategy; In S4, specifically including: S4-1, Calculate the local slope point by point for the processed stress-strain data. and local slope change rate : in and Indicates the first Strain and stress at a point; S4-2, Identify the end point of the elastic phase in the loading process. Find the rate of change of local slope The maximum value, corresponding to the strain value ; S4-3, Identify the end point of the plastic plateau region during the loading process. Calculate the rate of change of local slope The average value, the point that reaches the threshold is the end point of the plateau region, and the corresponding strain should be... The threshold is the sum of the mean and the standard deviation. Specifically, S2 includes: S2-1, Prepare a saturated salt solution and place it in a constant temperature container. Pass dry nitrogen gas into the solution to generate a humid gas. S2-2, Place the insulating pad into a sealed container, set the target humidity according to the actual operating environment of the transformer, introduce the above gas, and dynamically adjust the dry / wet nitrogen ratio by real-time feedback data through a high-precision temperature and humidity sensor, with humidity fluctuation within ±0.1%. S2-3, Steady-state maintenance: Maintain constant temperature at the target humidity for 24 hours, scan the internal humidity distribution using a near-infrared spectrometer, and verify that the humidity error is less than 0.1% by sampling and titration. In S5, specifically including: S5-1, Capturing the Loading Process The experimental data were used to fit the elastic phase coefficients using a first-order polynomial. and initial residual stress The constitutive equation for the elastic stage of the pad is obtained as follows: in and Representing stress and strain respectively; S5-2, Capturing the Loading Process The experimental data were used to optimize the fit using the Levenberg-Marquardt algorithm. ~ The constitutive equation for the plastic plateau stage of the pad block is obtained as follows: in and Representing stress and strain respectively. ~ These are the proportionality coefficient, exponential coefficient, and constant term for the plastic plateau stage, respectively. S5-3, Capturing the Loading Process The experimental data were used to optimize the fit using the Levenberg-Marquardt algorithm. ~ The constitutive equation for the densification stage of the pad is obtained as follows: in and Representing stress and strain respectively. and The stress and strain at the end of the plastic plateau stage. ~ These are the proportional coefficient and exponential coefficient for the densification stage of the pad block, respectively. S5-4, Extracting experimental data from the unloading phase, and using the Levenberg-Marquardt algorithm to optimize the coefficients of the unloading phase to obtain the constitutive equation for the unloading phase: in and Representing stress and strain respectively. This represents the residual stress after final unloading. Indicates the starting point of unloading strain. and This represents the coefficient for the unloading phase.

2. The transformer structural response calculation and graded operation and maintenance strategy considering the nonlinear characteristics of winding pad material as described in claim 1, characterized in that: S1 specifically includes: S1-1, a test block with dimensions of 8 mm in length, 8 mm in width and 6 mm in thickness was prepared by directional hot pressing process. The hot pressing parameters were controlled as follows: pressure 10±0.1MPa, temperature 75±2℃ and time 20 minutes, to simulate the actual stress distribution of the transformer pad. S1-2, First stage rapid dehydration: Dehydrate in a vacuum drying oven for 12 hours to remove free water; S1-3, Second stage of stable drying: High-purity nitrogen gas is introduced, and the temperature is maintained at 60±1℃ and the vacuum degree is <5 kPa for 12 hours to continue drying. S1-4, Endpoint determination: Using an online mass spectrometer, the resolution must be ≤0.001 mg / (g·h), the water release rate must be monitored in real time and be <0.01 mg / (g·h), and the initial humidity must be ≤0.5%.

3. The transformer structural response calculation and graded operation and maintenance strategy considering the nonlinear characteristics of winding pad material as described in claim 1, characterized in that: In S3, specifically including: S3-1, using a material mechanics testing device with a range ≥10MPa and an accuracy ≤±0.1% of the full scale FS, load to 10MPa at a strain rate of 0.05% / s, hold for 10 seconds, and then unload at the same rate to 0.3MPa residual stress. Record stress-strain data, with a sampling frequency of 100 Hz. Repeat the experiment 5 times for each set of conditions and take the average value. S3-2 uses data processing software to extract discrete data points from experimental curve images, including two columns of data: strain (mm / mm) and stress (MPa). S3-3, Data preprocessing: A Savitzky-Golay filter with a window width of 5 and a polynomial order of 2 is used to eliminate high-frequency noise.

4. The transformer structural response calculation and graded operation and maintenance strategy considering the nonlinear characteristics of winding pad material as described in claim 1, characterized in that: In S6, a 3D model of the transformer was created using SolidWorks, including the main structures of high and low voltage windings, pressure plates, pads, and core. The model was then imported into ANSYS Workbench to generate a hexahedral dominant mesh, which was locally refined to 0.3 mm with an Jacobian ratio > 0.

6.

5. The transformer structural response calculation and graded operation and maintenance strategy considering the nonlinear characteristics of winding pad material as described in claim 1, characterized in that: In S7, specifically including: S7-1, set the material parameters for calculating the magnetic field of the winding and core, where the BH curve of the silicon steel sheet uses the factory parameters; S7-2, calculate the normal operating conditions, including no-load closing and rated operating current. This provides the excitation for magnetic field calculation; S7-3, calculate the Lorentz force on the winding by solving the magnetic field distribution using ANSYS Maxwell; Based on Maxwell's equations, the governing equations for the magnetic field are derived. These equations are related to the magnetic vector potential. Related: ； in, It is a current density vector. It is the permeability. It is electrical conductivity; S7-4, Structural Response Analysis under Normal Operating Conditions; The magnitude of the current density vector under normal operating conditions is defined as the current. The cross-sectional area of ​​the conductors in the transformer winding The ratio: in, It is a unit vector indicating the tangential direction at the winding; Furthermore, the forces acting on the windings are calculated according to the Lorentz force law: in, It is the magnetic flux density; The differential equation of motion in solid mechanics for calculating the structural response of a transformer is as follows: in, This indicates that the winding is under stress. , and These represent the mass matrix, damping matrix, and transformer winding deformation, respectively. , and These represent acceleration, velocity, and displacement, respectively.

6. The transformer structural response calculation and graded operation and maintenance strategy considering the nonlinear characteristics of winding pad material as described in claim 1, characterized in that: In S8, specifically including: Accelerometers and stress sensors are arranged on the upper and lower pressure plates of the winding, with two sensors arranged circumferentially to measure stress and acceleration under the same normal and simulated operating conditions.

7. The transformer structural response calculation and graded operation and maintenance strategy considering the nonlinear characteristics of winding pad material as described in claim 1, characterized in that: In S9, specifically including: Compare the results measured in S8 with the simulation results at the corresponding positions in S7. If the simulation and experimental error of the acceleration amplitude is greater than 5%, adjust the key boundary conditions in S7, namely the prestress of the pressure plate, until the simulation and measured results of the acceleration amplitude are less than 5%.

8. The transformer structural response calculation and graded operation and maintenance strategy considering the nonlinear characteristics of winding pad material as described in claim 1, characterized in that: S10 specifically includes: S10-1, set the material parameters for calculating the magnetic field of the winding and core, where the BH curve of the silicon steel sheet uses the factory parameters; S10-2, Calculate the short-circuit current: in, It is short-circuit current. It is the effective value of the steady-state current; Indicates time, and These represent the equivalent resistance and equivalent leakage reactance, respectively. It is the initial phase angle. It is angular frequency; S10-3, the magnetic field distribution was solved using ANSYS Maxwell, and the Lorentz force density was derived; Based on Maxwell's equations, the governing equations for the magnetic field are derived. These equations are related to the magnetic vector potential. Related: ； in, It is a current density vector. It is the permeability. It is electrical conductivity; Furthermore, the magnitude of the short-circuit current density vector is defined as the ratio of the short-circuit current to the cross-sectional area of ​​a single conductor in the transformer winding: in, It is a unit vector indicating the tangential direction at the winding; Furthermore, the forces acting on the windings are calculated according to the Lorentz force law: S10-4, Structural response analysis under short-circuit conditions; In the finite element simulation software, the material parameters of the winding, core, and pad block for structural response analysis were set. The copper conductor used a linear elastic model with an elastic modulus of 110 GPa; the silicon steel sheet of the core used the factory parameters with an elastic modulus of 200 GPa; the experimental results obtained from S5 were used to set the pad block material to adopt a segmented constitutive model, and the prestress boundary conditions of the pressure plate were set according to the values ​​determined in S9. The differential equation of motion in solid mechanics is: in, This indicates that the winding is under stress. , and These represent the mass matrix, damping matrix, and transformer winding deformation, respectively. , and These represent acceleration, velocity, and displacement, respectively.

9. The transformer structural response calculation and graded operation and maintenance strategy considering the nonlinear characteristics of winding pad material as described in claim 1, characterized in that: S11 specifically includes: Based on the calculated plastic deformation results of the transformer pads, the following graded operation and maintenance strategy is formulated: If the maximum plastic deformation of the pad is less than a, the transformer can be put into normal operation and its operating status should be continuously monitored. If the maximum plastic deformation of the pad is greater than or equal to a, and less than a c th The transformer is put into operation for 15 minutes and the vibration data is observed. If the vibration data is abnormal, the transformer is shut down and the oil is drained for inspection. If the maximum plastic deformation of the pad is greater than or equal to c th If not, the transformer cannot be put into operation; drain the oil and check the transformer windings. critical value a and c th Determined based on the maximum oil passage height of the transformer.

Images