INVESTIGATION OF SWITCHING PROCESSES IN ESGD

A. M. Kuptsov, M. N. Lisetskaya

(Presented by the scientific seminar of the Department of Theoretical Fundamentals of Electrical Engineering and the Laboratory of Rotary ESG, Research Institute of Nuclear Physics)

Electrostatic generators with a dielectric transporter (ESGD) are used in a number of branches of industry [1]. However, because of their low power, the area of their application is very limited. An increase in the power of an ESGD can be achieved both by improving the manufacturing technology (reducing the rotor–stator and rotor–ionizer gaps, using materials with high electrical strength and permeability) and by qualitative changes in the design. The latter, as a rule, becomes possible after a thorough analysis of its operating processes.

Despite the fact that ESGDs have been manufactured industrially for a long time, their operating processes have not been studied sufficiently. In particular, there is no clarity in the switching processes [2, 3].

It is known [2–3] that switching in an ESGD, by ensuring the deposition and removal of charges on the rotor dielectric, creates voltage and current in the corresponding circuit. In this case it is assumed that in the load the current is produced by discharge of the charged transporter (the rotor dielectric) through the load resistance [2].

For generators with conducting transporters, switching questions were considered in [4, 5]. It was shown there that such a simplified approach to the consideration of switching processes is not justified.

In connection with this, in the present article an attempt has been made at an experimental study similar to [4–5], with the aim of clarifying the nature of switching processes in ESGD.

For the study a single-disk electrostatic generator was used, the disk of which was made of epoxy resin. The disk diameter was 140 mm, the thickness 1 mm. The disk rotation speed was 2780 rpm. A schematic representation of the generator with elements of the measuring circuit is shown in Fig. 1. In the diagram the inductors, brush-ionizers made in the form of narrow blades, and measuring resistances are shown schematically.

Fig. 1. Circuit for oscillographing currents:
1, 2—excitation and discharge inductors;
3, 5—discharge and charging brushes; 4—rotor.

The generator operated in the short-circuit mode. To increase the load current and to eliminate the possibility of breakdown from the high-voltage excitation inductor to the brush when the excitation voltage was raised, the inductors were covered with thin insulating plates.

The research procedure, the results of which are presented in this article, is basically similar to that described in [4]. The difference consisted only in the values of the resistances used for oscillography and in the choice of oscilloscope (OK-17 M). Oscillography was carried out by feeding the signals directly to the deflecting plates of the oscilloscope.

Obtained Results and Their Discussion

Figure 2, a, b, v shows typical oscillograms of the current in the circuits of the brush \((i_1)\), the inductor \((i_2)\), and the load \((i_3)\). In appearance, the current in the brush circuit (Fig. 2, a) consists of irregular pulses whose amplitude and duration vary with changes in the applied excitation voltage. The higher the excitation voltage, the greater the pulse amplitude, and vice versa. It has been established that, as the excitation voltage (load current) increases, the repetition frequency of the above-mentioned pulses also increases. To determine the pulse repetition frequency, the oscillogram shows a calibration curve of a sinusoidal current with a frequency of 5 kHz. At moments when there are no pulses on the oscillogram, the current in the circuit is zero; i.e., the dc component of the current in the brush circuit is absent.

Fig. 2. Typical oscillograms of currents

The current in the inductor circuit (Fig. 2, b) consists of pulses similar to the current pulses in the brush circuit, but of opposite polarity. If oscillography of the current in the inductor circuit is performed simultaneously with oscillography of the current in the brush circuit, then it is clearly seen on the oscillogram that these are the same pulses. The current in the inductor circuit has a dc component. Its polarity corresponds to the polarity of the current in the brush circuit. In this case, the ratio between the dc component and the ac component is such that the average value of the current, indicated by the microammeter (M-135), is zero. This ratio remains valid at any excitation voltage on the linear part of the short-circuit characteristic.

The current in the load circuit is a combination of pulsed and dc components. Their polarity is the same and coincides with the polarity of the pulses in the brush circuit. The amplitude of the pulses is substantially smaller than that of the pulses in the brush and inductor circuits. The average value of the current indicated by the instrument is equal to the current in the brush circuit.

During visual observation of the commutation zones it is clearly seen that, between the rotor disk and the brushes, there exists a region

sufficiently bright bluish glow. As the excitation voltage increases, it expands and intensifies its glow. A noticeable displacement of this zone relative to the middle of the inductor toward the direction opposite to the rotation of the rotor disk is observed. The glow is accompanied by the noise characteristic of a corona discharge.

From comparison of the presented oscillograms of the current in the circuits of the discharge system and the results of visual observation of the commutation zones, it follows that the pulses in the brush circuits are caused precisely by the corona current existing in the ionizer--rotor gap. This is confirmed by the fact that an increase in the current through the gas gap is accompanied by an increase in the repetition frequency of the current pulses, which is characteristic of corona current [6].

Since, in the inductor circuit, on the basis of the above oscillograms, the irregular pulses have exactly the same shape and frequency as in the brush circuit, it is obvious that these are one and the same pulses. Hence it follows that the corona current of the gas gap (brush--dielectric) flows in the brush--inductor circuit. But in the inductor circuit the mean value of the current, measured by a microammeter, is equal to zero. This corresponds to the fact that another current, of the opposite direction, also flows in the same circuit. The presence of a constant component is noticeable on the oscillogram. A constant component of current of the same magnitude (according to Kirchhoff's first law for a node) must be present in the load circuit.

Thus, we conclude that the transporter discharge current bypasses the load circuit and flows through the inductor circuit. This fully coincides with the phenomenon of discharge of transporters in an ESGD.

The presence of pulses in the load circuit can be explained by the fact that part of the charge of the dielectric does not participate in inducing charge on the inductor, but is directly connected with the earth. The more completely the inductor shields the commutating sections of the dielectric, the smaller the pulses in the load circuit.

Based on the foregoing, it can be shown that at any moment of time the current in the load circuit is a practically constant quantity. For this purpose let us consider the circuit shown in Fig. 3, depicting the discharge commutation zone of the ESGD. To simplify the calculation, it is assumed that the generator rotor is a cylinder. At any moment of time the current in the circuit is determined by the rate of change of the inductor charge

Fig. 3

\[ i_{\mathrm{n}}=\frac{dq_{\mathrm{in}}}{dt}, \tag{1} \]

where \(q_{\mathrm{in}}\) is the inductor charge, determined by the difference between the charges induced by the rotor charges and the magnitude of the compensated charge

\[ q_{\mathrm{in}}=q-q_{\mathrm{sk}}. \tag{2} \]

Since the geometrical position of the rotor and the commutation zone relative to the inductors does not change for the given operating mode,

and, consequently, the arrangement of charges on the rotor, then the quantity \(q\) is constant \((q=\mathrm{const})\).

\[ q_{\mathrm{sk}} = clvt = 2\pi lnRt\sigma, \tag{3} \]

where

\[ v = 2\pi Rn, \]

\(\sigma\) is the density of dielectric charges per cm of length;
\(l\) is the length of the brush;
\(n\) is the rotational speed;
\(R\) is the radius of the transporter;
\(t\) is time.

Finally,

\[ i_{\mathrm{n}}=\frac{dq_{\mathrm{in}}}{dt}=-\frac{dq_{\mathrm{sk}}}{dt}=-2\pi\sigma lRn. \tag{4} \]

Since all the quantities in expression (4) do not depend on time, the current in the load circuit is a constant quantity, which is fully confirmed by the experimental results.

Conclusions

An analysis of the experimental data on the investigation of commutation in ESGD showed that the commutation processes in them are similar to the processes in ESG with a large number of conductor-transporters per pole pitch.

In both cases, commutation in the discharge system is represented as follows. The charges deposited on the conducting transporter in ESGP or on an element of the dielectric surface in ESGD induce charges of the opposite sign on the inductor as they approach it. When charges are induced, an almost constant current flows in the load and inductor circuit. As the transporter or the charged dielectric surface approaches the brush, compensation of the charges of the transporter and the inductor takes place through a spark discharge in ESGP or through corona in ESGD. The compensation of charges is accompanied by the appearance of a pulse component of the current in the circuit of the brush and the inductor.

References

1. Edited by V. I. Levitov, A. P. Lyapin. Application of the Electric Field in Industry and Agriculture. VNIIEM, Moscow, 1964.

2. A. F. Ioffe. Electrostatic generator. ZhTF, vol. IX, 1939.

3. V. I. Levitov, A. G. Lyapin. Electrostatic Generators with a Rigid Rotor, Part II. VNIIEM, Moscow, 1965.

4. A. M. Kuptsov, V. V. Patsevich. Analysis of commutation in electrostatic generators. Elektrichestvo, No. 7, 1968.

5. A. M. Kuptsov. Physical Processes in Electrostatic Generators. Dissertation, Tomsk, 1968.

6. J. Meek, J. Craggs. Electrical Breakdown in Gases. IL, Moscow, 1961.