
Natural ester insulating oils are considered promising alternatives to mineral oils due to their renewability and environmental friendliness; however, their relatively high viscosity limits heat dissipation and cold-start performance. To address this issue, a viscosity prediction model based on a motif–aware dual-channel deep learning architecture (MA-VP) is proposed, which integrates global molecular features with local structural motifs to improve prediction accuracy while maintaining interpretability. The proposed MA-VP model achieves high predictive performance on the viscosity prediction task, with an R² of 0.9908 and an RMSE of 0.8024 on the test set. Integrated gradients analysis reveals that viscosity is primarily influenced by carbon-related molecular frameworks, whereas the effects of oxygen-containing groups depend on their specific types; additionally, unsaturated bonds and aromatic structures tend to reduce molecular viscosity. This study provides a mechanism-driven computational framework for the rational design and performance optimization of natural ester insulating oils, supporting the development of environmentally friendly insulating materials.
SF6 gas mixture is widely used as an insulating medium in various gas insulation equipment. When SF6 is mixed with buffer gases such as CO2 and N2, it will show a certain synergistic effect, thus greatly reducing the use of SF6, but its insulation strength is still lower than that of pure SF6 gas, which is difficult to apply in electrical equipment with ultra-high voltage and above voltage levels. In order to seek an insulating gas with better synergistic effect after mixing with SF6 and further enhance the insulation strength of the mixed insulating gas, a new insulating gas HFO1234ze(E) was mixed with SF2, and the power frequency breakdown voltage of 0% - 100% SF6/HFO mixed insulating gas was measured at a slightly uneven electric field at a pressure of 0.05 - 0.30 MPa, and the breakdown characteristics and synergistic effect of the gas mixture were analyzed. The experimental results show that the insulation strength of SF6/HFO gas mixture is higher than that of SF6 pure gas in the range of 18% - 100% mixing ratio, showing a positive synergistic effect, and the synergistic effect of gas mixture weakens with the increase of SF6 proportion. In order to reveal the positive synergistic mechanism of SF6/HFO gas mixture, the microscopic parameters of three dimer models formed by the interaction between SF6 and HFO molecules were calculated based on density functional theory. The results show that the intermolecular interaction between SF6 and HFO leads to a significant increase in the positive surface area of the electrostatic potential of the whole molecule, and the electronegativity of SF6 molecule is further enhanced due to the change of atomic charge distribution, which may be associated with the positive synergistic effect of SF6/HFO mixture gas macroscopically.
Cellulose insulation paper in transformers is prone to mechanical damage during operation, leading to microcracks that degrade performance. Although extensive research has focused on improving the physical and electrical properties of cellulose insulation, its self-healing mechanism remains insufficiently understood. This study employed molecular dynamics simulation to investigate the diffusion behavior and interface interactions of flaxseed oil repair agents in microcracks. The results indicated that the repair process originated from the mutual diffusion between the repair agent and cellulose, mainly driven by van der Waals forces and hydrogen bonds. The repair efficiency was evaluated based on the elastic properties and adhesion energy before and after the repair, with the highest reaching 68.3% . Temperature had a significant impact on the repair performance, and increasing the temperature (from 298 K to 380 K) could significantly accelerate the diffusion rate and improve the repair efficiency. This research revealed the intelligent repair behavior of cellulose composite materials at the molecular level, providing a theoretical basis for the design of self-repairing insulating materials.
High-voltage generators are key equipment in electric power transmission systems, and accurate assessment of the insulation aging condition of their stator bars is of great significance. Existing dielectric-response-based evaluation methods mostly rely on macroscopic characteristic parameters, making it difficult to deeply reveal the evolution of microscopic defects within insulation materials. To address this issue, this paper proposes a trap-current-based insulation aging assessment method. The trap current that characterizes the charge de-trapping process is extracted from polarization and depolarization currents. A double-exponential decay model is used to fit the trap current, yielding parameters associated with deep- and shallow-trap currents. Based on the trap-current inversion theory, the relationship curve between trap current Itrapt and trap depth ΔE is obtained. On this basis, key dielectric characteristic parameters are extracted, and a feature-parameter database is established using fitting techniques. The aging condition of stator-bar insulation in generators is then evaluated using this database. The results show that the proposed method enables quantitative analysis of the microscopic insulation degradation mechanism from the perspective of trapped-charge release, thereby providing a more accurate assessment of stator insulation aging.
To suppress residual stress and enhance electromechanical reliability, a process parameter optimization approach is proposed by jointly considering the curing temperature profile, cooling rate, and Al2O3 filler content. Material parameters of epoxy/Al2O3 composites are experimentally measured to support a coupled thermo–electro–mechanical FEM model, and a two-layer stacking ensemble surrogate model integrated with MOPSO is developed for process optimization. The results demonstrate that, after optimization, interfacial residual stress is reduced by approximately 24.1%, while flashover voltage, mechanical reliability, and breakdown probability are improved by 0.7%, 18.8%, and 45.4%, respectively, accompanied by a 21.1% increase in hydraulic withstand pressure. Ultrasonic measurements further confirm the reduction in interfacial residual stress, and the proposed process parameters exhibit favorable economic performance.
Residual stress generated during curing is a critical factor related to the explosive electrical breakdown of GIL/GIS epoxy Insulators. To further investigate the influence of residual stress on the breakdown process, this study establishes a residual stress model of the basin-type insulator as well as a phase-field breakdown simulation model incorporating electro–mechanical coupling effects. The results indicate that during the curing process, sedimentation of Al2O3 particles leads to a non-uniform density distribution within the basin-type insulator, which in turn results in an uneven distribution of residual stress. The residual stress is mainly concentrated at the interface between the basin-type insulator and the metal insert, and this concentration is more pronounced in low-density regions. Moreover, an excessively high cooling rate further intensifies stress concentration in basin-type insulator. Further breakdown simulation results demonstrate that the first principal stress induced by residual stress during curing promotes the breakdown development of the basin-type insulator. However, as the temperature gradient ΔT increases, the first principal stress is gradually released, thereby suppressing the breakdown process to a certain extent. In addition, the proposed simulation model can accurately reproduce a recent practical failure case, verifying the reliability of the model.
Epoxy resins are used widely as insulating materials in power equipment because of their favorable processability and strong adhesion. Nevertheless, their intrinsic properties often fall short in advanced electrical insulation applications. This study compares neat epoxy and epoxy composites filled with micro-sized SiO2 and h-BN, focusing on their thermal, electrical, and mechanical properties. The thermal performance is evaluated by measuring the thermal conductivity, a key indicator of the heat transfer capability. The electrical performance is assessed by examining the dielectric properties and measuring the AC breakdown strength. The mechanical performance of the epoxy composites is evaluated through tensile and shear strength tests, reflecting their structural suitability. The findings reveal distinct performance variations across filler types, attributable to the inherent properties of SiO2 and h-BN. A comprehensive evaluation of micro-filler-reinforced epoxy composites is performed by integrating thermal, electrical, and mechanical assessments. These results provide valuable reference data for selecting epoxy insulation materials in power equipment, particularly for use at joints and interfaces with metallic conductors and structural components.
Spacecraft are made of many insulating materials, and electrostatic discharge occurs due to the irradiation of the charged particles. This phenomenon affects the stable operation of spacecraft. Therefore, it is important that we understand the charging behavior of insulating materials to operate the spacecraft stably. In this study, we investigated the charge accumulation behavior of spacecraft insulating materials during proton irradiation. Specifically, we examined the space charge distribution and the external circuit current of two fluorinated and two polyimide insulating materials. Furthermore, we modeled the accumulated positive charge during proton irradiation, focusing on its correlation with the conductivity enhancement. As a result, positive charges accumulated around the calculated penetration depth, and the electric field increased. The increase in the electric field indicates a higher risk of discharge on the spacecraft. Furthermore, the conductivity of the proton irradiated samples increased. It means a decrease in insulating properties. These results suggest that the change in conductivity contributes to charge accumulation during proton irradiation. In modeling the amount of accumulated positive charge, it was possible to reproduce the amount of charge of the polyimide insulating material.
In this paper, a polypropylene (PP) based nanocomposite dielectric doped with metal organic framework (MOF) material, MIL-100(Fe), is proposed. Iron atom metal groups are introduced as deep traps to capture carriers excited by electromagnetic fields, and benzene rings are used to construct organic frameworks to enhance the bonding properties between the MIL and the PP substrate. It is found that the addition of MIL significantly improves the mechanical properties of the films. The elastic modulus of the composites increases by 279% and the fracture strength increases by 11.24%. By enhancing polarization and trapping charges, MIL can simultaneously realize the increase in dielectric constant and breakdown strength, both in conventional 0 T and 12 T strong magnetic field environments. The composite film exhibits an energy storage density of 6.55 J/cm3 and a charge/discharge efficiency of >95% (measured at 0 T and 600 kV/mm) and exhibits optimized performance in the 12 T strong magnetic field (6.54 J/cm3, calculated with the dielectric constant and breakdown field strength at 12 T). The findings in this research provide a new design paradigm for achieving high energy storage density and processable PP dielectrics, as well as improving performance under strong magnetic fields.
With the approval and construction of multiple railway lines in plateau and mountainous regions of China, railways are extending into high-altitude areas, where train roof insulators are subjected to the combined challenges of icing and low air pressure during operation. Investigating their flashover characteristics under such conditions is of great significance for revealing the flashover mechanism and providing guidance for engineering design. However, the insulation failure mechanism of roof insulators under the combined effects of low air pressure and icing remains unclear. In this paper, the train roof insulator is selected as the object of study. Based on a multifunctional climate chamber for icing flashover tests and finite element simulation of the electric field, the influences of icing parameters under low-air-pressure high-altitude conditions on the flashover voltage and discharge path are systematically investigated. The results show that the flashover voltage of the roof insulator decreases with increasing icing thickness and icing-water conductivity, while decreasing air pressure causes a more pronounced reduction within the tested range. Icing-induced local electric-field concentration promotes discharge initiation between adjacent sheds, and low air pressure further enhances discharge initiation and arc maintenance. As a result, the arc is more likely to develop into final flashover and exhibit limited deviation and swinging during its development.
Constrained by the limited charge-characterization techniques available under applied voltage, surface charge behaviors during DC corona discharge and voltage removal remain insufficiently understood. We employed a non-invasive surface-potential measurement in a needle–plane electrode configuration to characterize the evolution of surface charge throughout the entire process from corona discharge to voltage removal. The results show that the accumulated surface charges are predominantly homopolar with respect to the applied voltage, while their spatial distribution evolves during corona discharge. Under negative corona, the pattern transitions from isolated localized charge regions to stripe-like charge patterns, whereas under positive corona, it spreads radially from the needle toward the grounded electrode. After voltage removal, heteropolar charges accumulate near the needle electrode, and their magnitude positively correlates with the amount of charge accumulated during corona discharge. These heteropolar charges originate from the reverse discharge triggered by residual surface charges and may adversely affect subsequent voltage re-application after maintenance. Overall, this study helps clarify surface charge evolution during DC corona discharge and voltage removal, providing evidence for understanding charge accumulation behaviors of DC insulators during operation and maintenance.
The cross-linked polyethylene (XLPE)/ethylene propylene diene monomer (EPDM) interface is one of the weakest elements in the cable insulation system, where failure is frequently due to space charge accumulation. Even if the interface charge characteristics have been studied in the past, the space charge characteristics at coated interfaces under temperature gradient (TG) and polarity reversal (PR) remain unclear, and the underlying mechanisms require further understanding. This study investigates the interfacial charge characteristics of EPDM/XLPE systems coated with non-polar (PDMS) and polar (PMTFS) silicone oils using the pulsed electro-acoustic (PEA) method. By integrating carrier mobility, trap distributions, and interfacial potential barriers, the underlying mechanisms were elucidated. The results show that in the EPDM/PDMS/XLPE interface structure, where the interfacial barrier is relatively low, deep traps in the EPDM play a significant role in the hysteresis of interfacial charge polarity during PR. The difference between electron and hole traps in the EPDM leads to different interfacial charge accumulation before and after PR. In the EPDM/PMTFS/XLPE interface structure, due to the high hole potential barrier at the EPDM/PMTFS interface, the interfacial charge polarity does not follow the applied voltage polarity on the XLPE side under negative voltages and low TG. Under high TG and negative voltages, negative interface charges decrease. In addition, the XLPE/PMTFS and EPDM/PMTFS electronic potential barriers cause electrons to readily be trapped within the PMTFS layer, which accelerates the transition of interfacial charge polarity toward negative polarity and promotes the accumulation of negative charges. This may explain why, at low TG, more negative charges accumulate at the interface than positive charges when the applied voltage switches from positive to negative.
Pollution-induced surface arcs on lightning protection composite insulators are governed not only by electric-field distortion but also by transient thermal processes that affect arc morphology and flashover development. This paper investigates the temperature evolution of local surface arcs and the role of thermal buoyancy in promoting arc-column uplift under contaminated conditions. A silicone-rubber plate with thin sheet electrodes was established to emulate the local external-insulation region of a lightning protection composite insulator, and artificial contamination with different severities was applied. A calibrated schlieren system was used to capture the transient refractive-index disturbance around the arc channel. The arc temperature field was reconstructed through Abel inversion combined with the Gladstone–Dale relation, while the arc-column uplift increment was extracted from sequential schlieren images. The results show that increasing contamination severity enhances the continuity, brightness, and spatial expansion of the surface arc. The maximum reconstructed arc temperature increases from approximately 4360 K under low contamination to 5024 K and 5690 K under medium and heavy contamination, respectively. Meanwhile, the high-temperature region changes from a localized distribution near the arc root to a continuous distribution along the arc column. A positive correlation is observed between arc temperature and arc-column uplift increment. The temperature rise reduces the gas density around the arc channel, strengthens the thermal buoyancy, and thereby promotes the uplift, bending, and spatial expansion of the arc column away from the silicone-rubber surface. These findings clarify the coupling relationship among contamination severity, arc thermal state, and spatial morphological evolution, providing insight into the thermally assisted development mechanism of polluted surface arcs on lightning protection composite insulators.
This study investigates radiation induced electrical degradation in superconducting and instrumentation and control (I&C) wire insulations for particle accelerators through dielectric spectroscopy, DC insulation resistance, breakdown voltage, and FTIR analysis. Two different electrical insulation systems, namely polyvinyl alcohol (PVA) and polyimide (PI) and their degradation with radio-chemical aging are investigated. PVA shows clear chemical modification, confirmed by the growth of carbonyl, hydroxyl, and methyl FTIR bands. This resulted into enhanced dipolar relaxation, dielectric losses, DC conductivity, and moderate reduction of breakdown voltage. In contrast, PI undergoes negligible chemical change; its dielectric evolution is instead linked to physical aging and defect formation within the multilayered architecture of the wire. This leads to increased interfacial polarization and reduced dielectric voltage. The identified structure-property relationships demonstrate the effectiveness of electrical tests as nondestructive tools for aging evaluation in radiation harsh environments.
The Conduction behavior in oil-impregnated pressboard insulation has been extensively investigated under various conditions. However, few studies have addressed the potential influence of the electrode’s metallic nature, and the fibrous structure of the pressboard material, on the conduction process at the metal/pressboard interface. In this work, conduction currents were recorded using the PDC technique, and a comparative analysis was performed on three Metal/Pressboard/Metal (M/P/M) systems under different DC voltages. The results show an effect of the metal electrode on the transient currents feature. The conduction current is relatively high and shows slight variations depending on the type of metal electrode, following the same trend as the work function. This high current could be attributed to the moisture content of the pressboard, which may introduce ionic species (such as H⁺ and OH⁻ generated by water dissociation within the cellulose structure of the pressboard) and induce charge accumulation at interfaces under an applied electric field. It was also highlighted that an increase in oil content in the M/P/M configuration with steel electrode leads to space charge effects. Moreover, experimental data align with the thermionic emission mechanism at the M/P interface and the bulk limited ionic conduction. This helped in estimating the height of the energy barrier at the M/P interface, which was found to be around 1 electron-Volt. In addition, the impact of both metal electrode and oil content (in the impregnated pressboard) on the energy barrier height at the M/P contact is also observed.
Zinc oxide (ZnO) varistor blocks, commonly integrated into cable sheath protectors, generate substantial active power loss under harmonic voltage stress to accelerate the thermal aging. To investigate the aging behavior of ZnO varistors of protectors subjected to high-order harmonic voltages, this study implements a high-temperature accelerated aging test to simulate long-term thermal stress, while the key electrical degradation indicators are obtained through the leakage current measurements. The extrapolated aging time derived from the experiments is modeled by using a three-parameter Weibull distribution. In combination with a likelihood ratio testing method, the life distribution characteristics and reliability variations of ZnO varistors under different harmonic conditions are quantitatively analyzed. The results show that the life evaluation based on the three-parameter Weibull distribution reveals a 34% reduction in average service life under the 51st harmonic voltage compared with that under the 3rd harmonic voltage for aged ZnO varistors. These findings provide critical insights for the reliability design and maintenance decision-making of ZnO varistors under harmonic stress scenarios, ultimately contributing to the enhancement of the safety and longevity of power systems.
As a critical insulating material for key power equipment, insulating oil is prone to space charge accumulation under high electric fields or temperature fluctuations, making its space charge characteristics academically and industrially significant. In this work, the apparent carrier mobility of insulating oil is measured via the pulsed electro-acoustic method. Results show that carrier transport in insulating oil is incompatible with the quasi-“Poole-Frankel” relationship; instead, mobility correlates positively linearly with the hyperbolic sine of electric field, and the fitting slope’s absolute value rises with temperature, indicating that increasing temperature enhances the dependence of carrier mobility on the electric field. The mobility-temperature relationship follows the segmented Arrhenius model with a turning point T0 (4.9–7.7 °C). Below T0, insulating oil has mild molecular thermal motion and high viscosity, with the dominant transport process being slow hopping between shallow traps of large molecular ions, leading to a relatively low apparent activation energy (0.43~0.58eV). Above T0, the dominant transport mechanism transitions to fast hopping between shallow traps of small impurity ions, raising activation energy to 1.06–1.23 eV.
Frequent overvoltage, thermal cycling, and vibration accelerate arrester aging and threaten power system’s reliability. This paper reviews “electro‑thermal‑mechanical” aging mechanisms of ZnO arresters, along with experimental methods, modeling approaches, and performance enhancing strategies. Degradation characteristics and interactions among electrical, thermal, and mechanical aging are summarized; testing indices, standards, and condition monitoring techniques are compared; macro-/micro-scale equivalent-circuit and physical models are reviewed. Material and structural optimization strategies are analyzed, and future research directions are proposed.