Yonghong Cheng

Flashover Characteristic of Epoxy Composites Filled With Different Micro-Inorganic Oxide Particles Under Nanosecond Pulse in Vacuum

Research on the vacuum surface flashover mechanisms of insulating material under nanosecond pulses is important for the design and manufacture of high-power pulse generators. Here, the vacuum surface flashover characteristic of epoxy composites filled with different microfillers Al2O3 middot3H2O , Al2O3, BaTiO3, and TiO2 is investigated using a unipolar pulse generator (50-ns rise time and 600-ns full-width at half-maximum). In addition, their thermally stimulated depolarization current and secondary electron emission (SEE) coefficients are measured. The results show that the flashover performance does not depend just on the SEE coefficient, but that the different effects of shallow and deep traps must also to be taken into account.

Modeling and simulation for effective permittivity of two-phase disordered composites

We present a simulation for the effective permittivity of two-phase disordered composites based on a three-dimensional disordered model. In this model, the inclusions can be arranged at random positions in a composite material that is simulated by 10×10×10 cubic cells. Under the condition of the quasistatic approximation, numerical calculations are performed using the finite element method. The dependence of the calculated effective permittivity on the orientation of the disordered model is investigated. In order to get the macroscopical dielectric properties of isotropic composites, an averaging method is adopted because each topological structure of the disordered model is anisotropic. Then, the effect of the permittivity contrast between the inclusions and the matrix phase of composites is studied. The numerical results obtained by the disordered model are also compared with those obtained by the classical mixing rules and the conventional periodic models. The advantage of this presented disordered model is that it can succeed in predicting the effective permittivity in the cases of high permittivity contrast and large volume fractions of inclusions by considering overlapping of inclusions.We present a simulation for the effective permittivity of two-phase disordered composites based on a three-dimensional disordered model. In this model, the inclusions can be arranged at random positions in a composite material that is simulated by 10×10×10 cubic cells. Under the condition of the quasistatic approximation, numerical calculations are performed using the finite element method. The dependence of the calculated effective permittivity on the orientation of the disordered model is investigated. In order to get the macroscopical dielectric properties of isotropic composites, an averaging method is adopted because each topological structure of the disordered model is anisotropic. Then, the effect of the permittivity contrast between the inclusions and the matrix phase of composites is studied. The numerical results obtained by the disordered model are also compared with those obtained by the classical mixing rules and the conventional periodic models. The advantage of this presented disordered mod...

Diagnosis of stator winding insulation of large generator based on partial discharge measurement

The Stator winding insulation is aged by exposure to a combination of thermal, electrical, mechanical, and environmental stresses. Diagnosis of stator winding insulation is an important measure of ensuring the safe operation and extending the remaining life of large generator. In the paper, two new parameters are proposed for the diagnosis of stator winding insulation based on the test results of multistress accelerating aging test of stator bars. The test results show that the skewness of partial discharge phase resolved distribution and the high-frequency crest of partial discharge (PD) impulse change much more remarkably with the aging time than do the conventional parameters. Therefore, the two parameters could be considered as characteristic parameters for the diagnosis of stator winding insulation of the large generator. With these characteristic parameters, authors successfully assessed the insulation condition of actual stator bars removed from several large generators.

Simulation of partial discharge sequences using fluid equations

Simulation methods employing continuity equations to describe particle transportation are used to examine the process of single discharge. However, it is difficult to simulate partial discharge (PD) behaviour accurately by these methods, because it will take too much computation time to simulate consecutive PD sequences. In this report, in order to save computation time, a simple three-dimensional network is used to simulate PD sequences in a small void under ac voltage. The model considers charge generation, charge transport and charge disappearance, and electrical field redistribution in three dimensions. The continuity equations for electrons, positive ions and negative ions, including the effects of ionization, attachment, recombination, electron diffusion and the transport of charge under electrical field are solved simultaneously with Poissons equation. With this model, the phase-resolved PD pattern and the current forms of each PD are obtained. Moreover, considering the propagation process of each PD and its influence on surface charge distribution, the fall part of PD current form and the effect of previous discharge on the subsequent one are also studied.

Total dose response of hafnium oxide based metal-oxide- semiconductor structure under gamma-ray irradiation

The research of the total dose response of high dielectric-constant hafnium oxide under gamma-ray irradiation is important for the anti-irradiation study of the ultra-deep submicron electronic devices in space application. The response of the HfO2-based Metal Oxide Semiconductor (MOS) structure under various total doses of gamma-ray irradiation is investigated in this article. MOS capacitors which are composed of aluminum electrode, HfO2 gate dielectric on p-type silicon (Al/HfO2/pSi) are prepared and tested before and after 60Co gamma-ray irradiation to access the irradiation induced damage in HfO2/Si system. The surface morphology is obtained by the use of atomic force microscope(AFM), the chemical and physical characteristics are obtained using X-ray diffractometer (XRD) and X-ray photoelectron spectrometer(XPS), and the charge trapping characteristic of HfO2 film is calculated from the current-voltage(CV) curve by the semiconductor parameter testing. Crystallized HfO2, SiO2 and HfSiO4 are detected from the XRD spectrum and the crystallinity and grain size are found to decrease with the increase of gamma-ray total dose. Oxygen vacancy is found to increase after irradiation and it dominates the oxide charge trapping in the HfO2/Si system. Schottky emission is found to be the charge transport mechanism from the current-voltage curve at room temperature, and the barrier height of HfO2/Si interface is found to decrease with the increase of irradiation. Irradiation induced defects would lower the HfO2/Si interface barrier height and then give rise to the leakage current, this will consequently lead to the failure of HfO2 film and corresponding MOS structures.

Electrical characteristics of nanometer gaps in vacuum under direct voltage

The electrical characteristics of nanometer gaps in vacuum were studied with the tungsten electrodes under dc and pulsed voltage. A novel experimental technique to study the electrical characteristics of nanometer gaps was presented in the paper. In the experimental setup, the tungsten was fabricated and shaped to be a perfect sphere through the electrochemical etch and Joule melting method, and the nanogap was controlled precisely by the scanning electron microscope (SEM) and nanometer manipulator. The effects of electrode geometry, gap separation and injected voltage waveform were investigated. The current-voltage curves and Fowler-Nordheim plots showed the difference of the field emission process before breakdown between the sphere-sphere electrodes and needle-sphere electrodes. The gap separation dependence of dielectric strength demonstrated the similar trend to the previous work but better performance. The breakdown voltage for pulsed voltage was 4-5 times higher than that for the dc voltage. The analysis of the physical damage indicated that the current, duration time and electrode geometry played important roles in electrode modification. In addition, a possible mechanism of nanoscale vacuum breakdown was also proposed in the paper.

Effect of UV illumination on surface flashover characteristics of epoxy resin under pulse

Surface flashover in vacuum is a primary limitation to pulsed-power devices. Recently, there are more studies on the pulsed vacuum surface flashover focused on the flashover mechanism. In this paper, the UV illumination is used to study the mechanism of vacuum surface flashover. The effects of different illumination positions, different illumination durations and different illumination intensities are investigated. The results show that the pulsed vacuum surface flashover characteristic is affected by the UV illumination significantly. When the UV light illuminates the cathode triple-point junction, the flashover voltage is reduced; this reduction is almost unaffected by the illumination duration, but significantly increased with the enhancement of illumination intensity. When the UV light illuminates the sample surface, the flashover voltage reduction is increased both by the increases of illumination duration and illumination intensity. Moreover, UV illumination can also change the surface potential and thus influence the surface flashover characteristic.

Variation of surface charge distribution in PD sequences

Partial discharge (PD) occurring in gas void introduced during the manufacturing process of insulating constructions forms a serious threat to the life of insulation, hence its necessary to clarify its mechanism. In our paper, the surface charge measurement system based on Pockels effect was constructed to detect the residual surface charge distribution after PD occurred in a cylindrical void enveloped by PE. The real discharge current (ns level) of the PD sequences was obtained simultaneously. Depending on the combined measurement system, the current wave of a PD and corresponding discharge area were clearly observed. During continuous discharge sequences, PDs tended to take place at the sites having heterocharges at dielectric surface after voltage polarity reversed. Sometimes, the pulse current waveforms of more than two discharges showed overlapped. This phenomenon could be easily observed at initial discharges after voltage was applied. This may be due to the discharge time lag and photoionization effect. Besides, there were no discharges observed at the decreasing part of applied voltage though some area without surface charges existing.

Partial discharge on-line monitoring system based on FPGA

According to the demand of partial discharge on-line monitoring system, this paper puts forward a new scheme employing field programmable gate array (FPGA) instead of the general-purpose high-speed date acquisition (DAQ). Then, the reconfigurable hardware platform is designed and partial discharge algorithms based on IEC270 are provided. The form of reconfigurable structure and method for realizing the remote reconfiguration of the programmable FPGA are investigated. Using the dataflash chip, DSP can manage and store data up to 512 Mbits by serial peripheral interface (SPI) bus. At the same time, this paper discusses the results of software simulation, including data acquisition, data storage, partial discharge algorithm, and communication with DSP. At last, this paper analyzes the possibilities of applying the reconfigurable platform, FPGA-DSP, to the study of new methods for partial discharge on-line monitoring system.

Dielectric spectroscopy analysis of the h-BN ceramic

In this paper, we report the dielectric spectroscopy of hexagonal boron nitride (h-BN) ceramic which has been widely used due to a very low dielectric coefficient. In order to obtain the relaxation time distributions of h-BN, the Havriliak–Negami empirical equation is selected to analyse the measured dielectric data due to the Cole–Cole plots of h-BN samples. In the processes of solving the empirical equation, a nonlinear least-square technique is used and the boundary condition for unknown parameters in the equation is defined. The relaxation time and its distribution parameters are obtained by this method. The polarization mechanisms are also discussed based on the obtained relaxation time distributions. This research may be helpful for applications of h-BN ceramic. (Some figures in this article are in colour only in the electronic version)