To investigate the gas density fluctuations caused by the XeCl laser discharge, the deflection of the beam was monitored by passing He-Ne laser beam through the discharge region. Three kinds of the gas density fluctuations were observed as follows: (1) propagating to the transverse directions of the discharge, (2) making round trips between the main electrodes, (3) being generated by the preionization pin electrode sparks.
We also measured the propagation velocities of the gas density fluctuations, shifting the beam parallel to the excimer laser optical axis and compared them with the sonic velocity at 20 degree centigrade. The fluctuation (1) propagates 1.6 times faster than the sonic velocity at the start, and soon slows down to the sonic velocity. The fluctuation (2) propagates slightly faster than the sonic velocity because the discharge region is in high temperature. The fluctuation (3) travels at the same as the sonic velocity.

View full abstract

On an UV-preionized discharge-pumped KrF excimer laser with a charge transfer system, the electrical characteristics are discussed under the conditions of ambient pressure of 1.5 to 3.0 atm and discharge gap length of 14 to 21 mm for 20 to 34 kV of charging voltage of a primary capacitor bank at an optimum gas mixture of Kr/F₂/He/Ne=2.0/0.25/14.45/83.3% for a maximum laser output energy. A similarity relationship of gas breakdown voltage V_b and discharge current I_d, which are decisive factors for the laser characteristic, are found as a function of the product of gap length d and ambient pressure p. The gap breakdown voltage can be predicted by using a constant value of 20 as the product of the effective ionization coefficient and the gap length. Adapting the above similarity relationships of I_d=f(pd) and V_b=g(pd) to a normalized voltage-current characteristic studied previously, an actual voltage-current characteristic under any electrical condition can be estimated.

View full abstract

In many high voltage switchgears used at substations, it is very important to determine the optimal shape of an electrode. A number of techniques have been used for the design of electrodes by combining the finite element or the boundary element method (BEM) with the direct search method or the conjugate gradient method. In these techniques, only one optimal shape is determined by searching the shape corresponding to the minimal objective function. It is not always the optimal design when the objective function has considerably different gradients in the vicinity of the minimal point. In addition, the predicted shape is not always the optimal one when there are multi-minima, because only one of them is predicted by the usual optimal design.
In the present paper, a new method which is proposed by the authors, is used. It is the method for determination of the "Optimal Region" satisfying the objective function after introduction of shape fluctuation criterion in work process. To determine the optimal region, a new computer code is developed by combining the usual BEM with the Powell method. Two dimensional problems for the optimal design of electrode shapes are analyzed by means of the proposed method. The optimal region is determined for arbitary shapes within the limits of the shape fluctuation criterion. It can be seen that the proposed method is useful, in particular, to guarantee the quality of high voltage switchgear, etc.

View full abstract

A combined radiation and convection furnace has heaters radiating directly to a charge and, in addition, a fan to circulate the air inside it. The circulating air transfers by convection a part of the heat that the heaters would otherwise radiate to the charge. Although the heating rate of the charge is undoubtedly improved, the air flow is so indefinite that the calculation of the heat transfer is still not possible.
For the purpose of visualizing and measuring the air flow inside the furnace, author makes a simple model analogous to the furnace for the air flow inside it and measures the air flow inside the model by a hot-wire smoke stream method.
From the measurements, he visualizes the air flow inside the furnace and obtains the velocity distribution of the air inside it. He obtains the heat transfer coefficients between the air and each part (that is, the heaters, inner furnace-wall and charge) inside the furnace.
He ascertains a degree of accuracy of the heat transfer coefficients according to the measurements made for an experimental furnace and the calculated results of its heat transfer.

View full abstract

The thermally stimulated current (TSC) curves obtained by experiments are usually complicated. It is very important to separate such curves into single TSC curves. The autoseparation method proposed by us enables one to separate the complicated curves into individual TSC curves by using the asymptotic estimation method exactly. In these calculation processes, however, several times of calculation of the theoretical TSC curves are necessary.
In this paper, the detailed discussion to establish the autoseparation method is reported. In order to utilize the self-convergence characteristic of the autoseparation method availably, it is reported that the best interval h of Simpson's rule should be determined in two cases devided by the critical temperature, and in each case the smaller h than the best interval can give the ideal results without any errors and the adopted common h leads one to decrease the processing time largely.
The accuracy of the calculated TSC curves with this common h has been confirmed by evaluating the energy depth E_t by the asymptotic estimation method, and the result has led to establishing the autoseparation method.

View full abstract

The asymptotic estimation ν method that enables one to evaluate the escape frequency factor ν value from the fast order thermally stimulated current (TSC) curve has been proposed for the first time. This method enables one to get the exact ν value by using three coordinates determined upon the TSC curve freely. In other words, this enables one to evaluate exact ν value directly from any parts of the TSC curve without visible peak coordinates. Generally, the ν value cannot be evaluated in such cases as the thermal cleaning method and the partial heating method. However, the ν method is simultaneously possible to estimate the energy depth E_t of the carrier trap exactly. Two concepts determining the key coordinates, the fixed coordinates method and the fixed interval method shifted upon the curve, have been discussed in order to avoid unnecessary fluctuation involved in the results evaluated by the proposed method, too.
This method has been also applied to the partial parts of the cleaned curves simulated by computer. And it has been discussed that some ambiguities might happen in the processes evaluated only by the initial rise method from the curves obtained experimentally.

View full abstract

A new estimation method which enables one to evaluate the energy depth E_t of carrier trap from the 1st order thermally stimulated current (TSC) curve was proposed. In this new method, the peak temperature T_p and the slope of tangential line on any coordinates of Arrhenius plot are utilized availably as the Fundamental Elements. The new method was named as the Slope Estimation Method. This method can be dealed with by computer. This fact enables one to evaluate the E_t values even in various observed temperature ranges and to shorten the evaluation time. By utilizing a series of E_t values evaluated through the slope of any coordinates, the T_p can be accurately detected and the E_t value can be evaluated as exactly as possible in the following step. Additionally, it has been discussed for the first time that the cletection of background noises and their removal were made possible by using the clear standard. These original characteristics involved in the proposal are very useful because there exist some unavoidable errors in any data obtained experimentally.

View full abstract

This paper describes the influence of space charge on the growth of an electrical tree which is an important factor of the dielectric breakdown in polyethylene-insulated power cables. Polyethylene and ethylene vinyl-acetate copolymer with a semi-conducting needle were used as the treeing specimens. The statistical analysis showed that the growth rate and the unbalance of positive and negative partial discharges increased as the vinyl-acetate content increased, and that the growth per cycle decreased with the increase in frequency. A model of the tree growth accompanied by the space charge behavior was proposed considering the above experimental results. The tree growth is considered to be caused by the dielectric breakdown due to the field enhancement at the tree tip during the positive half cycle of the applied voltage, the field enhancement being due to residual negative space charge injected during the previous negative half cycle. The negative space charge in polyethylene would be mainly supplied through the partial discharge inside the tree.

View full abstract