3-D SEEA charge analyses on surface of insulators in vacuum

With the introduction of an external circuit model and a gas desorption model, the surface flashover on the plane insulator-vacuum interface perpendicular to parallel electrodes is simulated by a Particle-In-Cell method. It can be seen from simulations that when the secondary electron emission avalanche (SEEA) occurs, the current sharply increases because of the influence of the insulator surface charge on the cathode field emission. With the introduction of the gas desorption model, the current keeps on increasing after SEEA, and then the feedback of the external circuit causes the voltage between the two electrodes to decrease. The cathode emission current decreases, while the anode current keeps growing. With the definition that flashover occurs when the diode voltage drops by more than 20%, we obtained the simulated flashover voltage which agrees with the experimental value with the use of the field enhancement factor β = 145 and the gas molecule desorption coefficient γ=0.25. From the simulation results, we can also see that the time delay of flashover decreases exponentially with voltage. In addition, from the gas desorption model, the gas density on the insulator surface is found to be proportional to the square of the gas desorption rate and linear with time.

In bridged vacuum gaps, it is believed that the secondary electron emission avalanche (SEEA) and the resultant insulator charging play a crucial role in the progress of discharge leading to the surface flashover. This paper deals with the electric field just before the flashover that includes the accumulated SEEA charge component (EsF) and applied geometrical field component (EgF) on a cathode and their relation to the flashover voltage. The test gap consists of parallel disc electrodes bridged by an insulating solid cylinder made of Al2O3. The test insulator changes its length from 5 to 50 mm. As it has been clarified in our previous works that surface roughness of the insulator affects SEEA charge accumulation, we have examined the roughness ranging from 0.2 to 6 μm. Aluminum electrodes are mainly examined, and copper and stainless electrodes are used for comparison. A ring type electrostatic probe embedded near the triple junction on the cathode is employed to measure the electric field. As a result, the total field strength (EgF + EsF) for various insulator conditions with aluminum electrodes is 6.7 kVmm-1 as an average. However, the rate (γ) of SEEA charge component that accounts for the total electric field varies according to the insulator length and the roughness. When γ exceeds ca. 15%, the flashover voltage Vb depends on the square root of the insulator length. On the other hand, Vb is a linear function of the insulator length when the SEEA charging is depressed and thus γ is smaller than 15%. On the condition that γ<15%, the material of electrode makes an impact on the flashover voltage.

As is well known, an insulator in vacuum subjected to high-voltage stress may experience rapid electrical breakdown via surface flashover. Although electron multiplication on the insulator surface is generally thought to play an important role in the flashover mechanism, a full understanding of this role is lacking. <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1–6</sup> In this work, the low-impedance discharge is assumed to occur in a layer of plasma generated by the electron avalanche, rather than in a thin layer of adsorbed gas on the insulator surface. <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3,5,7</sup> Heating by the avalanche is considered as a mechanism by which gas is evolved from the insulator surface.

Essentially surface flashover across a solid in vacuum is considered as a kind of complicated surface and interface physical phenomena. Based on the concurrent electrical and optical measurements and microscopic observations, the research works of our group concentrate on the relationship between flashover and surface/interface condition of insulating materials. There are some different optical phenomena detected from the surface of alumina and polymers, and we attribute the preflashover phenomena to the differences between the surfaces of different insulating materials and different electrode contacts. The experimental results reveal that, under low electric field, prior to field electron emission from the cathode triple junction, electroluminescence phenomena occur due to the radiative recombination of electrons and holes injected into the surface states from the electrodes. According to the results, the phenomena mentioned above are related to the surface states in the surface layer of a material. This work is a contribution to the traditional secondary electron emission avalanche (SEEA) model.

The author reviews surface flashover (i.e. voltage breakdown along the surface of insulators), primarily in vacuum. He discusses theories and models relating to surface flashover and pertinent experimental results. Surface flashover of insulators in vacuum generally is initiated by the emission of electrons from the cathode triple junction (the region where the electrode, insulator, and vacuum meet). These electrons usually then multiply as they traverse the insulator surface, either as a surface secondary-electron-emission avalanche or as an electron cascade in a thin surface layer, causing desorption of gas which had been adsorbed on the insulator surface. This desorbed gas is then ionized, which leads to surface flashover of the insulator. The theory and modeling of this phenomena and experimental studies of surface charging, the applied voltage waveform, prestressing, conditioning, discharge delay and speed, insulator geometry AMD material, surface treatment, surface gases, temperature, and pressure are reviewed. Some suggestions are made regarding how to choose the material geometry and processing when selecting an insulator for a particular application.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

C <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">4</sub> F <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">7</sub> N is a potential alternative gas of SF <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">6</sub> due to its higher dielectric strength and lower Global Warming Potential (GWP). In this paper, the charge accumulation phenomenon on insulator surface in C <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">4</sub> F <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">7</sub> N/CO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> mixtures is investigated under needle-plate electrode. The characteristics of charge accumulation are compared with those in air. The influence of AC voltage frequency, gas pressure, and the ratio of C <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">4</sub> F <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">7</sub> N in the mixed gas on charge accumulation are analyzed. The results show that the charge distribution on the insulator surface presents a three-tier concentric circle structure. The distribution range in C <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">4</sub> F <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">7</sub> N/CO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> mixture is smaller than that in air, and it is mainly composed of negative charges. The charge density and distribution range decrease with the growing of the gas pressure or the ratio of C <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">4</sub> F <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">7</sub> N in the mixed gas, while the positive charges on the insulator surface increases with the growing of voltage amplitude or the decrease of AC frequency. It is suggested that the characteristics of charge distribution in the C <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">4</sub> F <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">7</sub> N/CO <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> mixtures are due to the large diameter and strong electronegativity of C <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">4</sub> F <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">7</sub> N gas molecules, which are easy to attach free electrons.

This paper studied the cylindrically bending effects of an fabric-based antenna. The fabric-based antenna consists of a meandered feeding line a patch. Two open stubs are inserted in the meandered feeding line to provide two transmission zeros near the passband. The bandwidth of fabric-based antenna was 2.5-4.6 GHz with the maximal gain of -0.42 - -3.1 dBi. Two bending orientations on hollow and solid cylinders of the fabric-based antenna were compared for simulation of human body structure. For transverse bending, the shape of radiation pattern and gain were almost identical for solid and hollow cylinders. A slightly lower bandwidth was observed due to the dielectric material of solid cylinder cannot neglected. For longitudinal bending, the maximal gain and bandwidth were affected because the resonance path was changed. The significant reduction of maximal gain was observed for longitudinal bending on hollow cylinder because radiation gain was loss to ground. Although the maximal gain at the middle band was almost similar on solid cylinder for longitudinal bending, the directivity at various frequencies was changed on solid cylinder because two dips were found. Results demonstrate the bending behavior of wearable antenna should be carefully monitored for on-body wireless communications.

The magnetic field due to the rotation of a superconductor has been measured for a number of hollow and solid tin cylinders of varying purity. The field generated within a hollow cylinder agrees with the predictions of the London theory:B=−(2mc/e)ω. The value of−2mc/e determined from the experiments was 1.17 × 10−7 G-sec, which agrees with the theoretical value of 1.137 × 10−7 G-sec within the experimental error of ±5%. The reversibility of the production of the field was investigated for both hollow and solid cylinders. In the case of the hollow cylinder, effects analogous to flux trapping were observed: If the specimen is cooled belowTc while in rotation, then the “London field” appeared only when the rotation speed was changed. In the case of the solid cylinder, effects analogous to those associated with “frozen in flux” were observed: the London field appeared reversibly only in those specimens sufficiently pure to exhibit a Meissner effect at fields of the order of 10−5 G. In one run a ferrite cylinder was substituted for the superconductor. A Barnett effect was observed; the magnitude of the field generated by rotation was consistent with the permeability measured at audio frequencies by conventional means. A novel heat-transfer effect connected with the rotation of the sample was also observed during the course of measurements on the hollow cylinders. A discussion of the electrodynamics of the rotating superconductor is given in an appendix.

A three dimensional equivalent charge distribution is estimated from scalp EEGs with the Laplacian weighted minimum norm solution. Computer simulations have been conducted to test the feasibility of the proposed approach. A three-shell concentric head volume conductor model was used in the present simulation study. The algorithm was tested using two kinds of source configurations: 4-dipoles and 4-charges. In both cases, the present equivalent charge distribution can provide a 3D equivalent source image of the assumed sources.

A software for simulation of CCD solid-state imagers has been designed. The software consist of simulators for imager process modeling, two and one dimensional charge and potential distribution modeling and simulators for modeling of main characteristics such as spectral response (SR), spectral sensitivity (SS), full noises (FN), and modulation transfer function (MTF). These characteristics may be so calculated jointly for optical blocks (lens system) and solid-state imager which may be based on charge-coupled devices, photodiodes, and photoresistors. Some methods and algorithms are examined for creating of photoelectric, SR, SS, FN and MTF simulators.© (1996) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

In the operation of HVDC GIL (Gas-insulated Transmission Line), cone-type insulator can separate SF 6 chambers, support the HV conductor and play the role of electric insulation. However, with long-term applied HVDC voltage, charges are prone to accumulate on the insulator surface, which can lead to the decrease of flashover voltage and threaten the stability of power system. Besides, the existence of metallic particles inside the GIL chamber, especially those attached along the insulator surface, can result in a more serious charge accumulation. Therefore, it is necessary to measure the surface charge distribution of clean and metal particle contaminated cone-type insulator in order to analyze its insulation properties under HVDC voltage. In this paper, a 3D 4-axis manipulating system is built to control the position of a capacitive probe to measure the surface potential of a ±200kV cone-type insulator. The probe keeps perpendicular to the concave surface at an identical distance and takes measurement along a pre-set trajectory after the applied voltage has been interpreted. Surface charge distributions of clean and metallic contaminated cone-type insulators are obtained successively. Influences of voltage polarity and application time are equally studied. The results show that metallic contamination has an enormous influence on surface charge polarities and distribution. For a clean surface, negative charge accumulates along the insulation surface under positive HVDC voltage; while under negative HVDC voltage, charges of both polarities are detected in different areas. For an insulator surface with metallic contamination, charge of the same polarity as the applied voltage is found to be significant near the contamination area. Analyses are made from the perspective of electric field distortion and partial discharge caused by the metallic contamination.

Trapped charges namely, the electrons and ions would be present in the bulk and surface of solid insulation, due to electrical stressing or other reasons. These charges modify the main field (MF), thereby affecting the initiation and growth of electrical trees. Whenever the total stress due to the MF and the Local Field (LF) exceeds the critical stress at any point, the physical and molecular structure is broken and a channel (tubule) is formed. Initiation of breakdown here is defined as the appearance of the breakdown channel (tubule) followed by the growth of further breakdown channels, either in the form of branch type trees with well defined branches or bushes consisting of an aggregate of large number of fine tubules. The field condition at any point is dependent on the MF and the charges on the tubules and their orientation. The effect of the tubules on the field distribution in the needle-plane gap may be estimated, by defining charges located suitably and calculating the fields thus simulated. In this work, a PMMA sample aged at 50Hz A.C. that had developed electrical trees was selected. The electrical field resulting from the superposition of main field (needle-plane) and the tubules (implying tree channels) is calculated and the estimation of future growth direction is presented.

Permanently grounded transformers must be given the insulation lest by inducing the necessary voltage across the windings. The A. I. E. E. Standards specify that the time of induced voltage tests be the same as for high potential test. In certain cases where the transformers are of very large capacity the induced voltage must be made at a frequency several times higher than normal. Since the dielectric strength of most insulating materials decreases with an increase in the frequency, an investigation has been made to determine the proper and fair length of time to make induced voltage test when the frequency is higher than normal. Following are the main points brought out in the investigation. Above a certain voltage, time of voltage application as well as voltage causes failure of insulation. The dielectric strength can be expressed as a function of both time and voltage by an equation of the form <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$Kv. = A(a + {1-a \over \root {1} \of T})$</tex> in which A is the kilovolts necessary to cause failure in one minute, “a” is a constant representing the ratio of strength for infinite time to the one minute strength and T is time in minutes. The value of “a” varies for different materials apparently depending mostly on the dielectric loss. The breakdown by creepage over solid insulation with the electrodes either on the same side or on opposite sides of the sample (arranged in such a manner that the solid insulation is under considerable stress) is not affected by time nearly so much as is the puncture voltage of solid insulation. The behavior of oil without barriers is so erratic that no very definite relation can be obtained between time and dielectric strength. In general time decreases the strength quite rapidly for the first few seconds after which the effect decreases and probably disappears entirely after two or three minutes. The momentary strength ranges from 25 to 30 per cent higher than the one minute strength. The effect of time on the strength of solid insulation and oil in series is about the same as for solid insulation alone until the oil distance exceeds the solid insulation thickness after which it begins to be the same as for oil without barriers. The strength-time curves for solid insulation are of approximately the same shape for all frequencies from 60 to 420 cycles, although the strength decreases with an increase in frequency F approximately in accordance with the formula kv. = 1.75/ F <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0·137</sup> Failure by creepage over the surface of solid insulation which is under no stress (i. e., with the electrodes on the same side of the barrier) takes place at approximately the same voltage for all frequencies from 60 to 420 cycles; but if the electrodes are so arranged (on opposite sides) that the insulation is under considerable stress the failure voltage decreases with an increase in frequency in about the same order as the puncture voltage of solid insulation does. The rupture voltage of oil is the same for both 60 and 420 cycles. The effect of frequency on the puncture voltage of solid insulation and oil in series is the same as for solid insulation until the oil distance exceeds the thickness of solid insulation, after which the effect decreases and as the oil distance increases the effect approaches that for oil without barriers. By considering the effects of both time and frequency on dielectric strength it is shown how to determine the proper length of time to make the voltage strain at higher frequencies equal to the strain at 60 cycles for one minute. 8. ACKNOWLEDGEMENTS It is desired to acknowledge the valuable assistance rendered in the Testing Department by Messrs. N. M. Albert, H. L. Garver, C. F. Green, L. D. Martin and W. F. Weikel in carrying on the long and tedious tests, covering a period of three to four months.

We have studied the composition of plasma formed during nanosecond high-voltage discharge at the surface of various insulators in vacuum. The flows of ions with velocities on the order of 102 km/s have been revealed. The energy distribution of ions in the plasma extends from several hundred electronvolts to a few kilo-electronvolts.

A computer simulation of the insulator surface charging in vacuum under the application of a high voltage, is presented for alumina (A1203) insulator and for alumina insulators coated with cuprous oxide (Cu2O) and chromium oxide (Cr2O3). It is shown that the mode of surface charge distribution is dependent on the secondary emission yield of the insulator surface. Mleasurements of insulator- electrode junction fields, using an electronic microforce balance, show the effect of surface coatings on the cathode and anode junction fields. High speed streak photography showing the time development of an arc in the case of coated insulators is presented. The experimental results support the insulator surface charging model which is simulated on a digital computer.