Methods, systems, and media for evaluating dielectric stability of resins after long-term thermal-oxidative aging

Abstract

The application provides a method, system and medium for evaluating the dielectric stability of resin after long-term thermal oxygen aging, and belongs to the technical field of insulation material performance testing. The method performs a long-term aging experiment on H-grade resin in a thermal oxygen environment with a temperature of 245 DEG C + / - 2 DEG C to 250 DEG C + / - 2 DEG C and an oxygen concentration of 21% + / - 2%, and collects dielectric constant, dielectric loss factor and volume resistivity data in a frequency range of 50 Hz to 10 MHz. Three core indexes, i.e., dielectric constant fluctuation rate, dielectric loss growth coefficient and resistivity attenuation index, are defined to accurately and quantitatively evaluate the dielectric performance stability of H-grade resin after long-term thermal oxygen aging. Compared with the traditional method, the method solves the defects of short evaluation period, single frequency range and fuzzy indexes of the traditional method, the evaluation result is consistent with the actual service failure with a coincidence degree of greater than or equal to 92%, and is suitable for resin selection and life prediction of high-end equipment.

Description

Technical Field

[0001] This invention belongs to the field of insulation material performance testing technology, specifically involving a quantitative evaluation method, system and medium for the stability of dielectric constant, dielectric loss factor and volume resistivity of H-grade resin (temperature resistance ≥180℃) under long-term thermo-oxidative aging conditions. It is particularly suitable for predicting the service life and reliability verification of H-grade epoxy, silicone, polyimide and other resins in high-temperature electrical equipment. Background Technology

[0002] H-class resin, as a core insulation material in high-end electrical equipment (such as stator insulation for 20MW steam turbine generators and encapsulation for high-speed rail traction converters), needs to maintain excellent dielectric properties under long-term service conditions at 180℃. Excessive fluctuations in dielectric constant can lead to electric field distortion, increased dielectric loss factor can cause localized overheating, and decreased volume resistivity can increase leakage current; all three can potentially induce insulation breakdown failure. According to an IEEE industry report, 45% of failures in H-class resin insulation systems stem from dielectric property degradation caused by long-term thermo-oxidative aging, and this degradation process is characterized by "slow accumulation and sudden failure," making it difficult for traditional evaluation methods to accurately capture.

[0003] (1) Short aging cycle, unable to cover long-term service: The maximum thermo-oxidative aging test of H-grade resin specified by existing standards (such as IEC 60085) is only 1000h, which is far shorter than the actual service requirements (the design life of H-grade resin is usually 35-40 years, and the equivalent aging time is about 10000h), which makes the evaluation results unable to reflect the dielectric degradation trend after long-term aging.

[0004] (2) The dielectric test frequency band is limited and the fit is poor: Traditional methods only test dielectric parameters at 50Hz power frequency (such as GB / T 1409), but the operating frequency of modern electrical equipment (such as high frequency inverters and variable frequency motors) has been extended to kHz~MHz. Single power frequency data cannot reflect the dielectric loss risk under high frequency conditions.

[0005] (3) The evaluation indicators are vague and lack quantitative standards: Existing evaluations mostly use "dielectric loss ≤0.03 after aging" and "resistivity ≥10¹³Ω·cm" as qualitative standards, without considering the rate of change of parameters with aging time. Resins with the same initial performance may have different actual lifespans by 3 to 5 times due to differences in the rate of deterioration, which cannot provide a basis for predicting equipment lifespan.

[0006] Therefore, there is an urgent need to develop a complete technical solution covering "long-term aging - multi-band dielectric testing - quantitative index evaluation" to fill the industry gap in the evaluation of the dielectric performance stability of H-grade resin under long-term thermo-oxidative aging. Summary of the Invention

[0007] To address the shortcomings of existing technologies, the present invention aims to provide a method, system, and medium for evaluating the dielectric stability of resins after long-term thermo-oxidative aging. This method achieves four major functions: long-term aging coverage, multi-band dielectric capture, quantitative index evaluation, and lifetime prediction support, providing a technical basis for the long-term reliable application of H-grade resins in high-end electrical equipment.

[0008] To achieve the above objectives, the technical solution is as follows:

[0009] This invention provides a method for evaluating the dielectric stability of resin after long-term thermo-oxidative aging, comprising the following steps:

[0010] S1. Prepare standard dielectric property test specimens;

[0011] S2. Establish an integrated system for long-term thermo-oxidative aging and dielectric testing;

[0012] S3. Perform long-term thermo-oxidative aging and staged dielectric parameter testing on standard dielectric performance test samples to obtain their dielectric constant, dielectric loss factor and volume resistivity at different frequency bands and different aging times t; the test frequency f is 50Hz~10MHz, and the aging time t is 0~10000h;

[0013] S4. Quantitative evaluation and judgment of dielectric property stability:

[0014] S41. Calculate the core evaluation indicators, which include dielectric constant volatility. Dielectric loss growth coefficient Resistivity decay exponent α and dielectric failure lifetime t f实际 ;

[0015] The dielectric constant volatility The calculation formula is: ,in The dielectric constant at frequency f after 10000h aging. The dielectric constant at frequency f after aging for 0 hours;

[0016] The dielectric loss growth coefficient The calculation formula is: Unit 10 -6 / h; where Since it is a dimensionless parameter, the difference is directly used in the calculation. The dielectric loss factor at frequency f after 10000h aging. The dielectric loss factor at frequency f after aging for 0 hours;

[0017] The formula for calculating the resistivity decay index α is as follows: ,in Volume resistivity;

[0018] The dielectric failure life t f实际 has the following calculation formula: t f实际 = t f × 72;

[0019] S42. Judgment criterion:

[0020] When the resin meets the following conditions simultaneously after 10,000 h of aging, it is determined that the dielectric performance stability of the resin is qualified:

[0021] ① The dielectric constant volatility ≤ 8% for all test frequencies f;

[0022] ② The dielectric loss growth coefficient ≤ 12 × 10 -6 / h for all test frequencies f;

[0023] ③ The resistivity decay index α ≤ 8 × 10 -5 / h;

[0024] ④ The dielectric failure life t f实际 > 10,000 h.

[0025] Furthermore, in step S1, when the initial dielectric constant deviation of the standard dielectric performance test specimen at 25 °C and 50 Hz is ≤ 3%, it is determined as a qualified standard specimen;

[0026] In step S2, the system includes a programmed temperature-controlled thermal-oxidative aging oven, a broadband dielectric spectrometer, a high-precision high-resistance meter, and a data linkage module;

[0027] In step S3, the test frequencies f are 50 Hz, 1 kHz, 10 kHz, 100 kHz, 1 MHz, 10 MHz, and the aging duration t is 0 h, 1000 h, 2000 h, 4000 h, 6000 h, 8000 h, 10000 h.

[0028] Furthermore, the resin is an epoxy resin, a silicone resin, or a polyimide resin.

[0029] Furthermore, in step S3, the temperature of the aging parameters in the long-term thermal-oxidative aging is 245 °C - 250 °C, and the oxygen concentration is 15% - 25%; in the staged dielectric parameter test, when the set aging duration is reached, the standard dielectric performance test specimen needs to be balanced, and the temperature for balancing is 20 - 30 °C, the humidity is 40% - 60%, and the time is 20 - 30 h.

[0030] Furthermore, in step S42, when t f实际 ≥ 3.6 × 10 5h, divided into the first - level life, that is, equivalent service life ≥ 35 years; when 1.8×10 5 h ≤ t f实际 <3.6×10 5 h, divided into the second - level life, that is, equivalent service life 20 - 35 years; when t f实际 <1.8×10 5 h, divided into the third - level life, that is, equivalent service life < 20 years.

[0031] The present invention also provides a system for testing the dielectric stability of resin after long - term thermal - oxidative aging. The system includes:

[0032] Initial dielectric parameter test module, configured to obtain the initial dielectric constant, initial dielectric loss factor and initial volume resistivity of the resin to be tested at different test frequencies f; the test frequency f is 50 Hz - 10 MHz;

[0033] Long - term thermal - oxidative aging module, configured to perform long - term thermal - oxidative aging operations at different set aging durations t; the aging duration t is 0 - 10000 h;

[0034] Stage - by - stage dielectric parameter test module, configured to obtain the dielectric constant, dielectric loss factor and volume resistivity of the tested resin after long - term thermal - oxidative aging;

[0035] Data algorithm module, configured to联动 with the above - mentioned modules to obtain the collected dielectric parameters and calculate the core evaluation indexes. The core evaluation indexes include dielectric constant volatility , dielectric loss growth coefficient , resistivity decay index α and dielectric failure life t f实际 ;

[0036] Data output module, configured to output an evaluation result according to the core evaluation indexes, and the evaluation criteria are as follows:

[0037] When the resin satisfies the following conditions simultaneously after aging for 10000 h, it is determined that the dielectric performance stability of the resin is qualified:

[0038] ① The dielectric constant volatility ≤8% for all test frequencies f;

[0039] ② The dielectric loss growth coefficient ≤12×10 -6 / hh for all test frequencies f;

[0040] ③ The resistivity decay index α ≤ 8×10 -5 / h;

[0041] ④ The dielectric failure life t f实际 >10000 h.

[0042] Furthermore, the dielectric constant fluctuation rate The calculation formula is: ,in The dielectric constant at frequency f after 10000h aging. The dielectric constant at frequency f after aging for 0 hours;

[0043] The dielectric loss growth coefficient The calculation formula is: Unit 10 -6 / h; where Since it is a dimensionless parameter, the difference is directly used in the calculation. The dielectric loss factor at frequency f after 10000h aging. The dielectric loss factor at frequency f after aging for 0 hours;

[0044] The formula for calculating the resistivity decay index α is as follows: ,in Volume resistivity;

[0045] The dielectric failure lifetime t f实际 The calculation formula is: t f实际 =t f ×72.

[0046] Furthermore, the test frequency f is 50Hz, 1kHz, 10kHz, 100kHz, 1MHz, or 10MHz, and the aging time t is 0h, 1000h, 2000h, 4000h, 6000h, 8000h, or 10000h.

[0047] Furthermore, the aging parameters in the long-term thermo-oxidative aging module are a temperature of 245℃~250℃ and an oxygen concentration of 15%~25%; in the staged dielectric parameter testing module, after the set aging time is reached, the resin to be tested needs to be balanced, and the balancing temperature is 20~30℃, the humidity is 40%~60%, and the time is 20~30h.

[0048] The present invention also provides a storage medium storing a computer program, wherein the computer program is configured to execute the above-described system for testing the dielectric stability of resin after long-term thermo-oxidative aging.

[0049] Compared with the prior art, the present invention has the following significant advantages:

[0050] (1) Comprehensive aging coverage: For the first time, the thermo-oxidative aging time of H-grade resin is extended to 10,000 hours, which is equivalent to an actual service life of more than 35 years. It can capture the dielectric degradation inflection point after long-term aging and avoid the problem of "underestimating the degradation risk" in the traditional 1,000-hour aging.

[0051] (2) The frequency band is in line with reality: the dielectric test covers 50Hz (power frequency) to 10MHz (high frequency inverter operation), which is suitable for the full frequency operation requirements of modern electrical equipment. For example, 1MHz data can be directly used for the performance evaluation of high frequency inverter encapsulation resin.

[0052] (3) Precise quantification of indicators: through , The three major indicators, α, 0, and 1, respectively quantify dielectric constant fluctuation, dielectric loss growth, and resistivity decay, solving the problem that traditional "qualitative judgment" cannot make horizontal comparisons of the stability of different resins.

[0053] (4) Reliable lifetime prediction: based on dielectric failure lifetime t f Establishing a correlation model between laboratory aging and actual service (1 hour of laboratory aging = 72 hours of actual service) can provide a direct basis for formulating equipment maintenance cycles, such as t f For resin with a service life of 10000h, it is recommended to use 3.6 × 10⁻⁶ units per hour. 5 Insulation testing is conducted every h (40 years).

[0054] Obviously, based on the above description of the present invention, and according to common technical knowledge and conventional methods in the field, various other modifications, substitutions or alterations can be made without departing from the basic technical concept of the present invention.

[0055] The following detailed embodiments further illustrate the above-described content of the present invention. However, this should not be construed as limiting the scope of the present invention to the following embodiments. All technologies implemented based on the above-described content of the present invention fall within the scope of the present invention. Detailed Implementation

[0056] The raw materials and equipment used in this invention are all known products, obtained by purchasing commercially available products.

[0057] The technical solution of the present invention will be described in detail below with reference to the embodiments and comparative examples. The specifications of the equipment and raw materials used in the embodiments and comparative examples of the present invention are uniform as follows:

[0058] 1. Grade H resin to be evaluated:

[0059] a. Resin A: Modified bisphenol F type H grade epoxy (NPEF-170, Shanghai Kaiyin Chemical Co., Ltd., with naphthalene ring antioxidant groups and methylsiloxane flexible segments introduced into the molecular chain, epoxy value 0.52eq / 100g, curing process 120℃ / 30min+180℃ / 2h).

[0060] b. Resin B: Commercially available bisphenol A type H grade epoxy (model EP-01441, Henan Aviation Materials Technology Co., Ltd., no antioxidant modification, epoxy value 0.48eq / 100g, curing process 120℃ / 30min+180℃ / 2h).

[0061] c. Resin C: Organosilicon modified H-grade epoxy (model RSN-0805, Dow Corning Incorporated, USA, organosilicon content 20wt%, epoxy value 0.45eq / 100g, curing process 130℃ / 60min+180℃ / 1h).

[0062] d. Resin D: H-grade polyimide resin (model D006, Xuchang Yingtai Insulation Materials Co., Ltd., glass transition temperature 220℃, curing process 200℃ / 120min).

[0063] e. Resin E: Halogen-free flame-retardant H-grade epoxy (model FR-H180, Dongying Hebang Chemical Co., Ltd., flame retardant is phosphazene compound, addition amount 15wt%, epoxy value 0.46eq / 100g, curing process 140℃ / 30min+180℃ / 2h).

[0064] f. Resin F: Phenolic H-grade epoxy (model F-51-H, Dongying Hebang Chemical Co., Ltd., phenolic group content 35wt%, epoxy value 0.50eq / 100g, curing process 150℃ / 30min+180℃ / 2h).

[0065] 2. Testing equipment:

[0066] a. Programmable temperature controlled thermo-oxygen aging chamber (model TH-800, Suzhou Taist, temperature control range 100℃~250℃, temperature control accuracy ±2℃, temperature uniformity ≤±3℃, oxygen concentration adjustment range 10%~30%, accuracy ±2%, equipped with an oxygen circulation system with a flow rate of 50mL / min, maximum continuous operation time ≥6000h).

[0067] b. Wideband dielectric spectrometer (model Novocontrol Alpha-A, German Novocontrol, test frequency band 50Hz~10MHz, frequency accuracy ±0.1%, dielectric constant ( Test accuracy ±0.01, dielectric loss factor ( Test accuracy ±0.0001, test voltage 1V ±0.1V).

[0068] c. High-precision high-resistance meter (model Agilent 4339B, Agilent Technologies, USA, testing range) Test voltage adjustable from 100V to 1000V, accuracy ±5%).

[0069] d. Laser flatness tester (model LS-900, Shanghai Lingliang, testing accuracy 0.01mm / m).

[0070] e. Vacuum evaporation coating machine (model ZK-300, Beijing Zhongke Keyi, vacuum degree -0.095~-0.098MPa, evaporation temperature 150℃±10℃).

[0071] f. Data linkage module (automatically records "aging time - temperature - oxygen concentration - ...)" - - (Data, sampling interval adjustable from 1 hour to 100 hours)

[0072] 3. Preparation process of standard dielectric property test specimens:

[0073] a. Prepare sheet samples of the H-grade resin to be evaluated according to its standard curing process. The sample size is 50mm×50mm×1.0mm±0.1mm, and the surface flatness is ≤0.05mm / m.

[0074] b. Grind both sides of the sample with 1000-grit sandpaper, ultrasonically clean it in anhydrous ethanol for 15 minutes, and dry it at 80°C for 30 minutes.

[0075] c. Circular silver electrodes are prepared on both sides of the sample using a vacuum evaporation process. The electrode diameter is 20 mm ± 0.5 mm, the edge distance from the sample edge is ≥ 10 mm, and the contact resistance between the electrode and the sample is ≤ 5 Ω.

[0076] d. Randomly select 3 samples and test the initial dielectric constant at 25℃ and 50Hz. When 3 samples If the deviation is ≤3%, it is determined to be a qualified standard sample;

[0077] 4. Long-term thermo-oxidative aging and dielectric property testing procedures:

[0078] a. Initial dielectric parameter test: The qualified standard sample was placed in an environment of 25℃ and 50%±5% humidity for 24 hours to equilibrate. The initial dielectric constant of the sample was then measured at 50Hz, 1kHz, 10kHz, 100kHz, 1MHz, and 10MHz using a wideband dielectric spectrometer (test voltage 1V±0.1V). ) and initial dielectric loss factor ( The initial volume resistivity was measured using a high-precision high-resistivity meter (test voltage 500V, data was read after holding the voltage for 1 minute). ), where f is the test frequency;

[0079] b. Long-term thermo-oxidative aging: After testing the initial parameters, the sample is placed in a programmable temperature-controlled thermo-oxidative aging chamber. The aging parameters are set as follows: temperature 245~250℃±2℃, oxygen concentration 21%±2%, and aging time is set as 0h, 1000h, 2000h, 4000h, 6000h, 8000h, and 10000h.

[0080] c. Staged dielectric parameter testing: After each set aging time is reached, the sample is removed and equilibrated for 24 hours in an environment of 25℃ and 50%±5% humidity. The dielectric constant of the corresponding frequency band is then tested according to method S31. ), dielectric loss factor ( ) and volume resistivity ( ), where t is the aging time; if the sample breaks down at a certain stage, < Or dielectric loss factor in the 1MHz band >0.05, stop the test and record the aging time at this point as the dielectric failure lifetime (t). f );

[0081] 5. Quantitative evaluation and judgment of dielectric property stability:

[0082] a. Calculate the core evaluation indicators:

[0083] ① Dielectric constant fluctuation ,in The dielectric constant at frequency f after 10000h aging; if the sample fails prematurely, based on 0~t f Time period The linear fitting equation (fitting correlation coefficient R² ≥ 0.90) is extrapolated to 10000h of data;

[0084] ② Dielectric loss growth coefficient Unit 10 -6 / h (where (where the parameter is dimensionless and the difference is directly used in the calculation), The dielectric loss factor at frequency f after 10000h aging; if the sample fails prematurely, based on 0~t f Time period The linear fitting equation (fitting correlation coefficient R² ≥ 0.90) is extrapolated to 10000h of data;

[0085] ③ Resistivity decay exponent α: based on The correlation coefficient R² was ≥ 0.95 when the log-log fit was calculated.

[0086] ④ Actual dielectric failure lifetime t f实际 =t f ×72, where 1 hour of laboratory aging is equivalent to 72 hours of actual service time;

[0087] b. Judgment criteria: It is judged as qualified if the following conditions are met after 10,000 h of aging:

[0088] ① For all frequencies in the 50 Hz - 10 MHz band ≤8%;

[0089] ② For all frequencies in the 50 Hz - 10 MHz band ≤12×10 -6 / h;

[0090] ③ Resistivity attenuation index α ≤ 8×10 -5 / h;

[0091] ④ Dielectric failure life t f > 10,000 h.

[0092] Example 1. Evaluation of the dielectric property stability of Resin A

[0093] To achieve the above object, this example provides a method for evaluating the dielectric stability of resin after long-term thermal oxygen aging. The specific steps are as follows:

[0094] S1. Prepare standard dielectric property test specimens:

[0095] S11. Make the H-class resin to be evaluated into sheet specimens according to its standard curing process. The specimen size is 50 mm × 50 mm × 1.0 mm ± 0.1 mm, and the surface flatness ≤ 0.05 mm / m;

[0096] S12. Polish both sides of the specimen with 1000-mesh sandpaper, place it in absolute ethanol and ultrasonically clean for 15 min, and dry at 80 °C for 30 min;

[0097] S13. Use the vacuum evaporation process to prepare circular silver electrodes on both sides of the specimen. The electrode diameter is 20 mm ± 0.5 mm, the distance from the edge to the specimen edge ≥ 10 mm, and the contact resistance between the electrode and the specimen ≤ 5 Ω;

[0098] S14. Randomly select 3 specimens and measure the initial dielectric constant ( ) at 25 °C and 50 Hz. When the deviation of the 3 specimens ≤ 3%, it is determined as a qualified standard specimen;

[0099] S2. Build an integrated system for long-term thermal oxygen aging - dielectric testing:

[0100] The system includes a programmable temperature controlled thermo-oxidative aging chamber, a wideband dielectric spectrometer, a high-precision high-resistivity meter, and a data linkage module. The programmable temperature controlled thermo-oxidative aging chamber has a temperature control range of 100℃~250℃, a temperature control accuracy of ±2℃, and a temperature uniformity of ≤±3℃. The oxygen concentration adjustment range is 10%~30%, with an accuracy of ±2%. It is equipped with an oxygen circulation system with a flow rate of 50mL / min and a maximum continuous operating time of ≥12000h. The wideband dielectric spectrometer has a test frequency range of 50Hz~10MHz, a frequency accuracy of ±0.1%, and can simultaneously test the dielectric constant (…). ) and dielectric loss factor ( The test accuracies are respectively ±0.01、 ±0.0001, the test voltage is 1V±0.1V; the high-precision high-resistance meter has a test range of ±0.0001. The test voltage is 100V~1000V, with an accuracy of ±5%; the data linkage module is linked with the above-mentioned equipment to automatically record "aging time - temperature - oxygen concentration - - -Volume resistivity ( The sampling interval for the data is adjustable from 1 hour to 100 hours.

[0101] S3. Perform long-term thermo-oxidative aging and dielectric performance testing:

[0102] S31. Initial Dielectric Parameter Test: The qualified standard sample is placed in an environment of 25℃ and 50%±5% humidity for 24 hours to equilibrate. The initial dielectric constant of the sample is then measured at 50Hz, 1kHz, 10kHz, 100kHz, 1MHz, and 10MHz using a wide-band dielectric spectrometer (test voltage 1V±0.1V). ) and initial dielectric loss factor ( The initial volume resistivity was measured using a high-precision high-resistivity meter (test voltage 500V, data was read after holding the voltage for 1 minute). ), where f is the test frequency;

[0103] S32. Long-term thermo-oxidative aging: After testing the initial parameters, the sample is placed in a programmable temperature-controlled thermo-oxidative aging chamber. The aging parameters are set as follows: temperature 245℃±2℃, oxygen concentration 21%±2%, and aging time is set as 0h, 1000h, 2000h, 4000h, 6000h, 8000h, and 10000h.

[0104] S33. Staged dielectric parameter testing: After each set aging time is reached, the sample is removed and equilibrated for 24 hours in an environment of 25℃ and 50%±5% humidity. The dielectric constant of the corresponding frequency band is then tested according to method S31. ), dielectric loss factor ( ) and volume resistivity ( ), where t is the aging duration; if breakdown occurs in a certain stage of the specimen, < or the dielectric loss factor in the 1 MHz frequency band >0.05, stop the test and record the aging duration at this time as the dielectric failure life (t f );

[0105] S4. Quantitative evaluation and determination of dielectric property stability:

[0106] S41. Calculate the core evaluation indicators:

[0107] ① Dielectric constant volatility , where is the dielectric constant at frequency f after 10000 h of aging; if the specimen fails prematurely, extrapolate the 10000 h data based on the linear fitting equation (fitting correlation coefficient R²≥0.90) for the 0~t f period ;

[0108] ② Dielectric loss growth coefficient , unit 10 -6 / h (where is a dimensionless parameter, and the difference is directly involved in the calculation), where is the dielectric loss factor at frequency f after 10000 h of aging; if the specimen fails prematurely, extrapolate the 10000 h data based on the linear fitting equation (fitting correlation coefficient R²≥0.90) for the 0~t f period ;

[0109] ③ Resistivity decay index α: Based on double logarithmic fitting calculation, fitting correlation coefficient R²≥0.95;

[0110] ④ Actual dielectric failure life t f实际 =t f ×72, where 1 h of laboratory aging is equivalent to 72 h of actual service time;

[0111] S42. Judgment criteria: If the following conditions are met after 10000 h of aging, it is judged as qualified:

[0112] ① For all frequencies in the 50 Hz~10 MHz frequency band, ≤8%;

[0113] ② For all frequencies in the 50 Hz~10 MHz frequency band, ≤12×10 -6 / h;

[0114] ③ Resistivity decay index α≤8×10 -5 / h;

[0115] ④ Dielectric failure lifetime t f >10000h.

[0116] The above formula is an empirical formula obtained based on a large number of basic experiments and long-term exploration. It reflects the dielectric stability (dielectric constant fluctuation rate) of the resin after long-term thermo-oxidative aging. Dielectric loss growth coefficient In the process of calculating the resistivity decay exponent α, etc., the core logic is always data-driven and mechanism-guided, relying on the long-term accumulated basic research results to build a complete derivation system. The specific process is as follows:

[0117] First, a multi-dimensional, highly reliable basic experimental database is constructed. Taking the actual service scenarios of H-grade resin in high-end electrical equipment (such as steam turbine generator stators) as a reference, the core influencing factors of dielectric performance stability are identified. Test parameters are strictly defined based on the above content, covering the core dielectric performance indicators (initial dielectric constant). Initial dielectric loss factor Initial volume resistivity The test focused on six key parameters: thermo-oxidative aging parameters (aging temperature: 245~250℃, oxygen concentration: 15%~25%, aging time t: 0~10000h) and dielectric test conditions (test frequency f: 50Hz, 1kHz, 10kHz, 100kHz, 1MHz, 10MHz). Over 350 sets of cross-validation experiments were designed using standardized testing methods. Grubbs criterion was used to remove outliers, and Shapiro-Wilk test was employed to verify data normality. This resulted in a standardized database containing 912 sets of valid data, ensuring that the data foundation for subsequent formula derivations fully matches the testing system.

[0118] Secondly, a genetic algorithm is introduced to mine multi-parameter nonlinear correlations. This takes into account the stability of the resin's dielectric properties (dielectric constant fluctuation rate). Dielectric loss growth coefficient The resistivity decay exponent α is the result of the coupling effect of the core parameters mentioned above (e.g., prolonged aging time will simultaneously exacerbate dielectric constant fluctuations and volume resistivity decay, and the two are synergistically deteriorating). Traditional linear regression is difficult to accurately capture the implicit interaction between parameters. Genetic algorithms, however, possess strong global search capabilities and multi-objective optimization advantages. They can simulate the selection, crossover, and mutation mechanisms of biological evolution, efficiently screening key correlation features within the parameter space defined by the above content: using initial dielectric parameters, aging parameters, and test frequency as algorithm input variables, and combining the predicted value from the formula with the experimentally measured change in dielectric properties (…). , The objective function is to minimize the deviation of α. By iteratively updating the population fitness (setting crossover probability = 0.8 and mutation probability = 0.05), the weight coefficients of each parameter and the function combination form are gradually locked, and finally a semi-empirical formula framework containing 4 core evaluation indicators is output.

[0119] Finally, the accuracy of the formula was optimized through multiple rounds of experimental calibration and physical mechanism verification. Based on the initial formula output by the genetic algorithm, a three-step optimization strategy was adopted: least squares coefficient fine-tuning, K-fold cross-validation (K=10) error analysis, and parameter boundary verification.

[0120] The first step was to calibrate the initial coefficients using the least squares method to obtain the first version of the formula. The root mean square error (RMSE) was calculated to be 0.32. It was found that the fitting deviation between the test frequency and the dielectric constant fluctuation rate was relatively large.

[0121] The second step involves performing piecewise fitting at the six test frequencies (50Hz~10MHz) specified above, combined with controlled variable experiments (fixed aging temperature 248℃, oxygen concentration 21%, and individual changes in test frequency) to verify the effect of frequency on [the desired effect]. The actual impact was that the coefficients for the high-frequency bands (1MHz, 10MHz) were optimized from 0.021 to 0.025, reducing the RMSE to 0.18;

[0122] The third step is to verify the physical meaning and boundaries of all coefficients. After each round of optimization, 25 sets of experimental data for different types of H-grade resins are added for verification until the consistency between the predicted value and the experimental value is ≥93%, and all calculation results can match the index definitions and judgment criteria mentioned above, ensuring that the final formula is consistent with both scientific validity and patent protection.

[0123] The calculated results were verified with the test results of the resin experiment, and the results were in high agreement, thus confirming the dielectric constant fluctuation rate in the patent. Dielectric loss growth coefficient Formulas for calculating resistivity decay index α, etc.

[0124] The entire derivation process ensures the authenticity of the data through systematic experiments, achieves accurate mining of multivariate relationships by relying on ant colony algorithms, and avoids overfitting problems of purely data-driven models through physical mechanism verification, forming a closed loop of experiment-algorithm-verification to ensure that the formula has both statistical significance and engineering application value.

[0125] Furthermore, the standard curing process described in S11 is determined according to the type of H-grade resin: for epoxy resins, it is 120℃ / 30min+180℃ / 2h; for silicone resins, it is 130℃ / 60min+180℃ / 1h; and for polyimide resins, it is 200℃ / 120min.

[0126] Furthermore, the vacuum degree of the vacuum evaporation process described in S13 is -0.095~-0.098MPa, the evaporation temperature is 150℃±10℃, and the evaporation time is 15~20min.

[0127] Furthermore, the ambient humidity of the wideband dielectric spectrometer described in S31 is ≤50% to avoid the influence of ambient moisture on dielectric parameters.

[0128] Furthermore, the temperature and oxygen concentration of the programmable temperature-controlled thermo-oxygen aging chamber described in S32 are calibrated every 100 hours during the aging process to ensure that the parameter deviations are within the set range.

[0129] Furthermore, in S33, three parallel samples are tested at each aging time node, and the average value of the dielectric parameters is taken as the test result of that node, with the deviation of parallel samples ≤5%.

[0130] Furthermore, S42 also includes a lifespan classification: t f实际 ≥3.6×10 5 h represents the first-level service life (equivalent to ≥35 years of service), 1.8×10 5 h≤t f实际 <3.6×10 5 h represents the second-level service life (equivalent to 20-35 years of service), t f实际 <1.8×10 5 h represents Level 3 service life (equivalent to <20 years of service life, not recommended for long-life equipment).

[0131] Furthermore, if there are 3 samples in S14 If the result is greater than 3%, repeat the sample preparation steps S11 to S13 until a qualified standard sample is obtained.

[0132] Furthermore, when testing volume resistivity in S31, the test voltage is applied and held for 1 minute before data is read to ensure current stability and reduce test errors.

[0133] Furthermore, in S41, if the sample reaches its dielectric failure lifetime t before aging for 10,000 hours... f Based on existing aging data (0~t) f Linear extrapolation calculation of 10,000 hours of aging time period) , And α, extrapolation error ≤10%.

[0134] The tests were performed according to the above method, with aging parameters set at 245℃ and oxygen concentration at 21%. The dielectric parameter test results for each aging stage are shown in Table 1. No breakdown was observed in the samples after 10,000 hours of aging. , =0.0183 < 0.05.

[0135] Table 1

[0136]

[0137] Calculation of core evaluation indicators:

[0138] (1) The percentages are as follows: 50Hz: 5.9% (|4.05-3.82| / 3.82×100%), 1kHz: 4.5%, and 1MHz: 4.7%, all ≤8%.

[0139] (2) 50Hz is 5.7×10 -7 / h ((0.0142-0.0085) / 10000), 1kHz is 5.9×10 -7 / h, 1MHz is 5.8×10 -7 All are ≤12×10 -6 / h;

[0140] (3) α: α = 5.65 × 10⁻⁶ obtained by double logarithmic fitting. -5 h -1 ≤8×10 -5 h -1 ;

[0141] (4)t f >10000h, t f实际 =10000×72=7.2×10 5 h.

[0142] Judgment result: All qualification conditions are met, dielectric performance stability is qualified, and lifespan level is Grade 1.

[0143] Example 2: Evaluation of the dielectric stability of resin B

[0144] The test was performed according to the method in Example 1, with aging parameters of 247°C and 21% oxygen concentration, and the sample was aged for 7000 hours. This triggers dielectric failure, stopping subsequent tests. Data from previous stages are shown in Table 2; linear fitting based on 0~7000h data ( Fitting equation: =1.628×10 -4 t+3.75, R²=0.98; Fitting equation: =3.286×10 -6 Extrapolate 7000h data (t+0.0090, R²=0.97).

[0145] Table 2

[0146]

[0147] Calculation of core evaluation indicators:

[0148] (1) Extrapolate 10000 h 50Hz is 43.4% (|5.378-3.75| / 3.75×100%, 5.378=1.628×10). -4 (×10000+3.75), 44.0% at 1kHz and 45.5% at 1MHz, both >8%;

[0149] (2) Extrapolate 10000h 50Hz is 3.286×10 -6 / h((0.04186-0.0090) / 10000,0.04186=3.286×10 -6 ×10000+0.0090), 1kHz is 3.5×10 -6 / h, 1MHz is 4.8×10 -6 / h, all ≤12×10 -6 / h;

[0150] (3) α: obtained by fitting data from 0 to 6000 h =-1.2×10 -4 t+15.2 (based on double logarithmic fitting of volume resistivity data from 0 to 6000 h, R²=0.96), α=1.2×10 -4 h -1 >8×10 -5 h -1 ;

[0151] (4)t f 7000h,t f实际 =7000×72=5.04×10 5 h.

[0152] Judgment result: >8% and α>8×10 -5 h -1 The dielectric stability was deemed unqualified, and the lifespan rating was determined to be Level 2.

[0153] Example 3: Evaluation of the dielectric stability of resin C

[0154] The test was performed according to the method in Example 1, with aging parameters of 248°C and 21% oxygen concentration. After 10,000 hours of aging, the sample showed no breakdown. , =0.0172 < 0.05, and the data for each stage are shown in Table 3.

[0155] Table 3

[0156]

[0157] Calculation of core evaluation indicators:

[0158] (1) The percentages are as follows: 3.8% at 50Hz, 3.9% at 1kHz, and 3.4% at 1MHz, all ≤8%.

[0159] (2) 50Hz is 5.0×10 -7 / h, 1kHz is 5.0×10 -7 / h, 1MHz is 6.0×10 -7 / h, all ≤12×10 -6 / h;

[0160] (3) α: The fitted result is =-6.5×10 -5 t+15.5, α=6.5×10 -5 h -1 ≤8×10 -5 h -1 ;

[0161] (4)t f >10000h, t f实际 =7.2×10 5 h.

[0162] Judgment result: All qualification conditions are met, dielectric performance stability is qualified, and lifespan level is Grade 1.

[0163] Example 4: Evaluation of the dielectric stability of resin D

[0164] The test was performed according to the method in Example 1, with aging parameters of 250°C and 21% oxygen concentration. After 10,000 hours of aging, the sample showed no breakdown. , =0.0138 < 0.05, and the data for each stage are shown in Table 4.

[0165] Table 4

[0166]

[0167] Calculation of core evaluation indicators:

[0168] (1) The percentages are as follows: 4.5% at 50Hz (|3.68-3.52| / 3.52×100%), 4.6% at 1kHz, and 4.8% at 1MHz, all ≤8%.

[0169] (2) 50Hz is 3.0×10-7 / h ((0.0095-0.0065) / 10000), 1kHz is 3.0×10 -7 / h, 1MHz is 4.0×10 -7 / h, all ≤12×10 -6 / h;

[0170] (3) α: obtained by double logarithmic fitting =-4.2×10 -5 t+15.8, α=4.2×10 -5 h -1 ≤8×10 -5 h -1 ;

[0171] (4)t f >10000h, t f实际 =10000×72=7.2×10 5 h.

[0172] Judgment result: All qualification conditions are met, dielectric performance stability is qualified, and lifespan level is Grade 1 (polyimide resin molecular chain has strong rigidity, optimal resistance to thermal and oxygen aging, and minimum decay index).

[0173] Example 5: Evaluation of the dielectric stability of resin E

[0174] The test was performed according to the method in Example 1, with aging parameters of 248°C and 21% oxygen concentration. After 10,000 hours of aging, the sample showed no breakdown. , =0.0205<0.05, and the data for each stage are shown in Table 5.

[0175] Table 5

[0176]

[0177] Calculation of core evaluation indicators:

[0178] (1) The percentages for 50Hz are 7.6% (|4.25-3.95| / 3.95×100%), 1kHz is 7.2%, and 1MHz is 7.1%, all ≤8%.

[0179] (2) 50Hz is 6.5×10 -7 / h ((0.0157-0.0092) / 10000), 1kHz is 6.5×10 -7 / h, 1MHz is 7.0×10 -7 / h, all ≤12×10 -6 / h;

[0180] (3) α: obtained by double logarithmic fitting =-7.8×10 -5 t+15.3, α=7.8×10 -5 h -1 ≤8×10 -5 h -1 (Fit correlation coefficient R² = 0.95, error ± 0.1 × 10⁻⁶) -5 h, because flame retardants slightly accelerate resin degradation).

[0181] (4)t f >10000h, t f实际 =7.2×10 5 h.

[0182] Judgment result: All qualification conditions are met, dielectric performance stability is qualified, and lifespan level is Grade 1.

[0183] Example 6: Evaluation of the dielectric stability of resin F

[0184] The test was performed according to the method in Example 1, with aging parameters of 250°C and 21% oxygen concentration, and the sample was aged for 9000 hours. =0.052>0.05, triggering dielectric failure and stopping subsequent tests. Data from previous stages are shown in Table 6; linear fitting based on 0~9000h data ( Fitting equation: =1.378×10 -4 t+4.02, R²=0.96; tanδ fitting equation: =8.444×10 -6 Extrapolate 10000h of data (t+0.0140, R²=0.98).

[0185] Table 6

[0186]

[0187] Calculation of core evaluation indicators:

[0188] (1) Extrapolate 10000h 50Hz is 34.3% ((5.398-4.02) / 4.02×100%, 5.398=1.378×10 -4 ×10000+4.02), 34.8% at 1kHz and 35.8% at 1MHz, both >8%;

[0189] (2) Extrapolate 10000h 50Hz is 1.356×10 -6 / h ((0.02306-0.0095) / 10000), 1kHz is 2.4×10 -6 / h (tanδ=1.42×10 -6 t+0.0103, R²=0.97, extrapolating 10000h tanδ=0.0245, K=(0.0245-0.0103) / 10000=1.42×10 -6 / h), 1MHz is 8.444×10 -6 / h, all ≤12×10 -6 / h;

[0190] (3) α: obtained by fitting data from 0 to 8000 h =-9.5×10 -5 t+15.2, α=9.5×10 -5 h -1 >8×10 -5 h -1 ;

[0191] (4)t f 9000h,t f实际 =9000×72=6.48×10 5 h.

[0192] Judgment result: >8% and α>8×10 -5 h -1 The dielectric performance stability was deemed unqualified, and the lifetime rating was Level 2 (phenolic groups are prone to chain breakage in a hot and oxygen environment, resulting in faster fluctuations in dielectric constant and decay in resistivity).

[0193] The following is a comparative example.

[0194] Comparative Example 1: Traditional Short-Cycle Aging Evaluation (Resin A)

[0195] According to the existing IEC 60085 standard method, only resin A was subjected to 1000h thermo-oxidative aging (245℃, oxygen concentration 21%), and only the dielectric parameters of a single 50Hz frequency band were tested. The results are shown in Table 7:

[0196] Table 7

[0197]

[0198] Traditional methods for determination:

[0199] After 1000 hours of aging, tanδ = 0.0097 ≤ 0.03. , the determination is qualified. Compared with the present invention: The traditional method fails to capture the dielectric degradation trend of Resin A after 10,000 h, and the single-band data cannot reflect the dielectric loss risk at 1 MHz high frequency (10,000 h = 0.0183), which is likely to overestimate the long-term stability of the resin.

[0200] Comparative Example 2, Evaluation without Band Coverage (Resin D)

[0201] Perform 10,000 h aging according to the method of the present invention, but only test the dielectric parameters of the single band of 50 Hz. The results are shown in Table 8:

[0202] Table 8

[0203]

[0204] Single-band determination: 50 Hz = 4.5% ≤ 8%, = 3.0×10 -7 / h ≤ 12×10 -6 / h, the determination is qualified. Compared with the present invention: The single-band evaluation misses the dielectric data at 1 MHz high frequency (10,000 h = 0.0138), which cannot meet the application requirements of high-frequency variable-frequency equipment, while the multi-band data of the present invention can comprehensively cover the dielectric stability of the resin under different working conditions.

[0205] In summary, the present invention provides a method, system and medium for evaluating the dielectric stability of resin after long-term thermal-oxidative aging. This method constructs a technical system of "stepwise thermal-oxidative aging - synchronous acquisition of multi-band dielectric spectroscopy - analysis of quantitative indicators". In a standard thermal-oxidative environment of 245°C ± 2°C to 250°C ± 2°C and an oxygen concentration of 21% ± 2%, perform 0 to 10,000 h long-term aging on H-class resin specimens, and synchronously collect the dielectric constant ( ), dielectric loss factor ( ), and volume resistivity ( ) data in the frequency band of 50 Hz to 10 MHz at different aging stages, and define three core evaluation indicators of "dielectric constant volatility ( )", "dielectric loss growth coefficient ( )", and "resistivity decay index (α)", so as to achieve precise quantitative evaluation of the dielectric property stability of H-class resin after long-term thermal-oxidative aging. This method solves the defects of traditional evaluation such as "short aging cycle, single frequency band, and fuzzy indicators". The aging coverage duration is extended to 10,000 h (equivalent to the actual service life of 40 years of H-class resin). The dielectric test frequency band covers the full working condition frequency range of the motor, and the coincidence degree between the evaluation result and the actual service failure is ≥ 92%. It can be used for the selection and life prediction of H-class resin in high-end equipment such as large generators and high-speed rail traction converters.

Claims

1. A method for evaluating the dielectric stability of resin after long-term thermo-oxidative aging, characterized in that, It includes the following steps: S1. Prepare standard dielectric property test specimens; S2. Set up an integrated system for long-term thermal-oxidative aging-dielectric testing; S3. Conduct long-term thermal-oxidative aging and stage-by-stage dielectric parameter testing on the standard dielectric property test specimens to obtain their dielectric constants, dielectric loss factors, and volume resistivities at different frequency bands and different aging durations t; the test frequencies f for different frequency bands are 50 Hz to 10 MHz, and the aging duration t is 0 to 10000 h; S4. Quantitative evaluation and determination of dielectric property stability: S41. Calculate the core evaluation indicators, which include dielectric constant volatility. Dielectric loss growth coefficient Resistivity decay exponent α and dielectric failure lifetime t f实际 ; The dielectric constant volatility The calculation formula is: ,in The dielectric constant at frequency f after 10000h aging. The dielectric constant at frequency f after aging for 0 hours; The dielectric loss growth coefficient The calculation formula is: Unit 10 -6 / h; where Since it is a dimensionless parameter, the difference is directly used in the calculation. The dielectric loss factor at frequency f after 10000h aging. The dielectric loss factor at frequency f after aging for 0 hours; The formula for calculating the resistivity decay index α is as follows: ,in Volume resistivity; The dielectric failure lifetime t f实际 The determination is made as follows: during long-term thermo-oxidative aging, when the sample breaks down... < Or dielectric loss factor in the 1MHz band When the value is greater than 0.05, stop the test and record the aging time at this point as the dielectric failure lifetime t. f , t f实际 The calculation formula is: t f实际 =t f ×72; S42. Determination criteria: When the resin meets the following conditions simultaneously after 10000 h of aging, it is determined that the dielectric property stability of the resin is qualified: ① Dielectric constant fluctuation at all test frequencies f ≤8%; ② Dielectric loss growth factor at all test frequencies f ≤12×10 -6 / h; ③ Resistivity decay exponent α≤8×10 -5 / h; ④ Dielectric failure lifetime t f实际 >10000h.

2. The method according to claim 1, characterized in that, In step S1, the initial dielectric constant of the standard dielectric performance test sample under conditions of 25°C and 50Hz is... If the deviation is ≤3%, it is determined to be a qualified standard sample; In step S2, the system includes a programmable temperature-controlled thermal-oxidative aging oven, a broadband dielectric spectrometer, a high-precision high-resistance meter, and a data linkage module; In step S3, the test frequencies f are 50 Hz, 1 kHz, 10 kHz, 100 kHz, 1 MHz, 10 MHz, and the aging durations t are 0 h, 1000 h, 2000 h, 4000 h, 6000 h, 8000 h, 10000 h.

3. The method according to claim 1, characterized in that, The resin is an epoxy resin, a silicone resin, or a polyimide resin.

4. The method according to claim 1, characterized in that, In step S3, the temperature of the aging parameters in the long-term thermal-oxidative aging is 245 to 250 °C, and the oxygen concentration is 15% to 25%; in the stage-by-stage dielectric parameter testing, when the set aging duration is reached, the standard dielectric property test specimens need to be balanced, and the temperature for balancing is 20 to 30 °C, the humidity is 40% to 60%, and the time is 20 to 30 h.

5. The method according to claim 1, characterized in that, In step S42, when t f实际 ≥3.6×10 5 h is classified as Level 1 service life, meaning an equivalent service life of ≥35 years; when 1.8×10 5 h≤t f实际 <3.6×10 5 h is classified into two service life levels, equivalent to 20-35 years of service; when t f实际 <1.8×10 5 h is divided into three service life levels, which is equivalent to less than 20 years of service.

6. A system for testing the dielectric stability of resin after long-term thermo-oxidative aging, characterized in that, The system includes: An initial dielectric parameter test module configured to obtain the initial dielectric constant, initial dielectric loss factor, and initial volume resistivity of the resin to be tested at different test frequencies f; the test frequency f is 50 Hz to 10 MHz; A long-term thermal-oxidative aging module configured to perform long-term thermal-oxidative aging operations at different set aging durations t; the aging duration t is 0 to 10000 h; A stage-by-stage dielectric parameter test module configured to obtain the dielectric constant, dielectric loss factor, and volume resistivity of the tested resin after long-term thermal-oxidative aging; The data algorithm module is configured to work in conjunction with the above modules to acquire the collected dielectric parameters and calculate core evaluation indicators, including dielectric constant fluctuation. Dielectric loss growth coefficient Resistivity decay exponent α and dielectric failure lifetime t f实际 Wherein, the dielectric failure lifetime t f实际 The determination is made as follows: during long-term thermo-oxidative aging, when the sample breaks down... < Or dielectric loss factor in the 1MHz band When the value is greater than 0.05, stop the test and record the aging time at this point as the dielectric failure lifetime t. f , t f实际 The calculation formula is: t f实际 =t f ×72; A data output module configured to output an evaluation result according to the core evaluation index, and the evaluation criteria are as follows: When the resin meets the following conditions simultaneously after 10000 h of aging, it is determined that the dielectric property stability of the resin is qualified: ① Dielectric constant fluctuation at all test frequencies f ≤8%; ② Dielectric loss growth factor at all test frequencies f ≤12×10 -6 / hh; ③ Resistivity decay exponent α≤8×10 -5 / h; ④ Dielectric failure lifetime t f实际 >10000h.

7. The system according to claim 6, characterized in that, The dielectric constant volatility The calculation formula is: ,in The dielectric constant at frequency f after 10000h aging. The dielectric constant at frequency f after aging for 0 hours; The dielectric loss growth coefficient The calculation formula is: Unit 10 -6 / h; where Since it is a dimensionless parameter, the difference is directly used in the calculation. The dielectric loss factor at frequency f after 10000h aging. The dielectric loss factor at frequency f after aging for 0 hours; The formula for calculating the resistivity decay index α is as follows: ,in It represents the volume resistivity.

8. The system according to claim 6, characterized in that, The test frequencies f are 50 Hz, 1 kHz, 10 kHz, 100 kHz, 1 MHz, 10 MHz, and the aging durations t are 0 h, 1000 h, 2000 h, 4000 h, 6000 h, 8000 h, 10000 h.

9. The system according to claim 6, characterized in that, The temperature of the aging parameters in the long-term thermal-oxidative aging module is 245 °C to 250 °C, and the oxygen concentration is 15% to 25%; in the stage-by-stage dielectric parameter test module, when the set aging duration is reached, the resin to be tested needs to be balanced, and the temperature for balancing is 20 to 30 °C, the humidity is 40% to 60%, and the time is 20 to 30 h.

10. A storage medium, characterized in that, The storage medium stores a computer program, wherein the computer program is configured to execute the system for testing the dielectric stability of resin after long-term thermo-oxidative aging as described in any one of claims 6 to 9 when running.

Images