A Scheme for Increasing the Energy and Its Transfer Rate from an Impact Generator to an Inductive Load

V. V. Ivashin, A. V. Loos, G. A. Sipaylov, K. A. Khor’kov

(Presented by the scientific seminar of the Electromechanical Faculty)

In many experimental installations for obtaining magnetic fields with high energy, pulsed power supply of an inductive load is used; in this case, energy of up to 200 MJ and a transfer rate of up to \(10^6\) MVA may be required [1]. To transfer energy at such a rate, preliminary accumulation of energy over a comparatively long interval of time is used, which reduces the power of the charging devices.

The most economical and effective method is the accumulation of energy in rotating flywheel masses, for example, in the rotor of a synchronous generator, which makes it possible to readily convert the kinetic energy of the rotor into electromagnetic energy taken from the stator winding. Such a generator, which received the name impact generator, was first built and used by P. L. Kapitsa [2].

Energy is stored in the rotor of an impact generator over a prolonged interval of time from a motor of small power. The accumulation of energy is accompanied by small no-load losses of the generator, and the accumulated energy can be preserved for a comparatively long time.

When the impact generator is connected to the load for the duration of one positive half-wave of the emf, the kinetic energy of the rotor is converted into electromagnetic energy. If the generator is connected to the load during several positive half-waves of the emf, with disconnection during the negative half-waves (single-half-period rectification), then the amount of energy taken from the generator can be substantially increased. To preserve the energy when the generator is disconnected, the load must be short-circuited for the duration of, or for part of, the negative half-wave of the emf.

The energy transferred to the load can be expressed through the average power of the machine \(P_{\mathrm{avg}}\) and the energy-transfer time \(t_{\mathrm{p}}\):

\[ Q = P_{\mathrm{avg}} \cdot t_{\mathrm{p}} \]

This paper considers a scheme that uses an increase in the transmitted energy both by increasing \(P_{\mathrm{avg}}\) and by increasing \(t_{\mathrm{p}}\), without increasing the overall power rating of the machine.

An increase in \(P_{\mathrm{av}}\) is achieved by using an auxiliary capacitance connected in parallel with the generator and the load. The operation of the circuit without preliminary charging of the capacitance and operation with preliminary charging of the capacitance are considered.

The advisability of increasing the rate of energy transfer from an impact generator to the load is due to the fact that, first, for a given amount of transmitted energy, a reduction in the transfer time leads to a reduction of energy losses in the transfer process, as a result of which heating of the generator windings is reduced and its utilization is increased; second, for a given transfer time, more energy can be taken from the generator and delivered to the load, which also increases the utilization of the generator.

Principle of operation

The basic circuit for operation of an impact generator into an inductive load with the use of an auxiliary capacitor bank is shown in Fig. 1, where

Fig. 1. Circuit for increasing the energy and the rate of its transfer from an impact generator to an inductive load without preliminary charging of the bank.

UG is a single-phase impulse-power generator;

\(r_{\mathrm{g}}, L_{\mathrm{g}}\) are the active resistance and inductance of the stator winding, taking into account the shielding effect of the field winding and the damping circuits of the rotor [3]. For large impact generators \(L_s \approx 1.15L_s\) (\(L_s\) is the leakage inductance of the stator winding);

\(L_{\mathrm{l}}, r_{\mathrm{l}}\) are the inductance and active resistance of the load;

\(r_{\mathrm{s}}\) is the active resistance of the short-circuit;

\(K\)-1, \(K\)-2, \(K\)-3 are arcless ion-mechanical switching devices [4];

\(C\) is the capacitor bank.

It is known that, by means of a capacitor bank connected in parallel with the load, one can considerably increase the energy delivered by the generator to the load during one positive half-cycle of the emf [5]. The capacitor bank contributes to an increase in the fraction of kinetic energy converted into the energy of the magnetic field of the generator leakage and into the energy of the electric field of the capacitor bank, which are then transferred to the load.

The nature of the variation of currents and voltages in the various elements of the circuit is shown in Fig. 2. To obtain an additional increase in the energy transferred during the first positive half-cycle of the generator emf, the capacitor bank may be preliminarily charged to a certain voltage.

Let us first consider the operation of the circuit without preliminary charging of the bank (Fig. 1). In the initial position the devices \(K\)-1 and \(K\)-2 are open; the generator operates at no load.

At the instant when the generator emf passes through zero, K-1 closes. The capacitor begins to charge, i.e., it consumes the generator energy. When the capacitor voltage reaches its maximum value, the current in it changes direction; the generator and the capacitor together then deliver energy to the load. When the generator current passes through zero at the instant \(t_1\), K-1 opens; however, the current in the load continues to increase, since the capacitor has not had time to discharge. At the instant \(t_2\), when the voltage on the capacitor passes through zero, K-2 closes. The capacitor current is taken over by the short circuit. In the load–short-circuit loop there flows a current decaying exponentially with the time constant of this loop. At the beginning of the next positive half-wave of the generator emf, at \(t_3\), K-1 closes. The generator operates in the sudden short-circuit regime; the generator current rises rapidly, while the current in the short circuit decreases. When the short-circuit current reaches zero at \(t_4\), K-2 opens. The capacitor bank facilitates the operation of K-2, since at the first instant after K-2 opens it is capable of taking over any current. The initial conditions on the capacitor are the same as at the instant \(t=0\), i.e., the capacitor again contributes to an increase in the energy taken from the generator during each positive half-wave of the emf. The cycles are then repeated. The current in the load tends in magnitude toward the sudden short-circuit current of the impact generator.

Fig. 2. Shape of the current and voltage curves in various elements of the circuit:
\(i_{\text{н}}\)—load current, \(i_{\text{г}}\)—generator current, \(i_{\text{з}}\)—short-circuit current, \(U_c\)—battery voltage, \(E_{\text{г}}\)—generator emf.

The operation of the circuit with preliminary charging of the capacitor differs somewhat from the operation of the circuit without preliminary charging (Fig. 3). In the initial position K-1, K-2, K-3 are open; the capacitor bank \(C\) is charged to the voltage \(U_{c0}\). The polarity of the bank charge is shown in Fig. 3. At the instant when the generator emf passes through zero, K-1 closes. The capacitor begins to discharge, and when the voltage on the capacitor passes through zero, K-3 closes. The circuit then operates like the circuit without preliminary charging of the capacitor bank. The increase in the rate of energy transfer occurs owing to the increase of the current in the first cycle. However, this increase is substantial and is determined not only by the addition of the energy of the preliminary charge of the bank, but also, to a considerable extent, by the influence

preliminary charging of the capacitor bank on the rate of energy transfer from the impact generator to the inductive load.

Fig. 3. Circuit for increasing the energy and the rate of its transfer to the load from an impact generator with preliminary charging of the capacitor bank.

The transient processes from the moment the circuit begins operating up to \(t_1\) are described, in operator form, by equations (1) and (2):

\[ pL_g i_g(0)+U_c(0)+e(p)=i_g(p)\left(pL_g+r_g+\frac{1}{pC}\right), \tag{1} \]

\[ pL_n i_n(0)+U_c(0)=i_n(p)(pL_n+r_n)+i_c(p)\frac{1}{pC}. \tag{2} \]

The transient process in the time interval from \(t_1\) to \(t_2\) is described by equation (2), where the initial values of the current \(i_n(0)\) and voltage \(U_c(0)\) are determined from the results of the solution at the preceding stage.

The operation of the circuit from \(t_2\) to \(t_3\) is described by equations (2) and (3)

\[ pL_n i_n(0)=i_n(p)(pL_n+r_n)+i_3(p)r_3. \tag{3} \]

The operation of the circuit from the beginning of the next positive half-wave of emf, \(t_3\), to \(t_4\) is described by the system of equations (1), (2), (3). After the shorting switch is opened at \(t_4\), the stages of operation of the circuit are repeated.

The complexity of the analytical investigation of this circuit is explained not only by the difficulty of solving the equations obtained, but also by the difficulty of determining the initial conditions for each of the stages.

As a result of solving equations (1), (2), (3), cumbersome expressions are obtained for the currents and voltages in the individual elements of the circuit. Therefore, in order to save space, the results of the solution are given in the form of curves (Fig. 2).

The load current, which in this case is of greatest interest, may, in a first approximation without allowance for active resistances, be determined from (4)

\[ \begin{aligned} i_n={}&-\frac{1}{2}i_g(0)+I_m-\frac{\nu^2}{\nu^2-1}I_m\sin \omega t -\frac{\nu^2}{\nu^2-1}I_m\cos \omega t+{}\\ &+\frac{2}{\nu}\frac{U_c(0)}{E_m}I_m\sin \nu\omega t -\frac{1}{2}\frac{I_g(0)}{I_m}I_m\cos \nu\omega t+i_n(0), \end{aligned} \tag{4} \]

where

\[ i_g(0),\ U_c(0),\ i_n(0) \]

are the initial values of the currents and voltage, determined from the results of solving the equations at the preceding stage;

\[ \nu\omega=\sqrt{\frac{L_n+L_g}{L_nL_gC}} \]

is the natural oscillation frequency of the \(L_g-C-L_n\) circuit.

\[ I_m=\frac{E_m}{\omega(L_g+L_n)}. \]

The considered scheme for accelerated transfer of energy from an impact generator to an inductive load broadens the possibilities of the impact generator as an energy-storage device and source.

REFERENCES

1. R. Carruthers, Energy storage fo thermonaclear research, Proc. Inst. Electr. Engs., A—106, No. 2, 1959.

2. V. R. Karasik. Physics and Technology of Strong Magnetic Fields. Nauka Publishing House, 1964.

3. M. P. Kostenko. Electrical Machines. Special Part, GEI, 1949.

4. V. V. Ivashin, G. A. Sipailov. Arcless interruption of large currents, Elektrotekhnika, No. 9, 1964.

5. G. A. Sipailov, K. A. Khor’kov. On the joint operation of an impact generator and a capacitor bank on an inductive load. Izvestiya TPI, Vol. 132, 1964.