DEVELOPMENT OF COMPACT BETATRONS WITH IMPROVED SPECIFIC CHARACTERISTICS

A. A. Geizer, V. L. Chakhlov, M. M. Shtein

(Presented by the scientific seminar of the PMB sector)

Portable compact betatrons of the PMB-6 type [1], which are currently produced in small series, are used very effectively for radiographic inspection of various materials and products with an equivalent steel thickness from 30 to 250 mm under nonstationary conditions.

An important characteristic for these types of betatrons is the ratio of the exposure-dose power to the weight of the entire installation or to the weight of the radiation unit, which we shall call the specific characteristics of the betatron.

Improving the specific characteristics makes it possible to increase inspection productivity both by reducing exposure time and by reducing the time spent transporting and installing the emitter at the inspected section of the part. In developing portable compact betatrons with specific characteristics improved in comparison with the PMB-6, consideration was given to the possibility of increasing the repetition frequency of radiation pulses, raising the injection voltage, using new materials, and also improving manufacturing technology and creating more optimal electromagnet designs.

Among the known excitation circuits for betatron electromagnets [2, 3] that make it possible to increase the repetition frequency of radiation pulses, the most promising is the pulsed power-supply circuit [4], which ensures the flow of unipolar current pulses through the electromagnet winding with unipolar voltage on the capacitor bank.

As shown in [5], transition from excitation of the electromagnet by alternating current at a frequency of 50 Hz to feeding by unipolar current pulses, with unchanged total losses in the electromagnet and with the same weight, makes it possible to bring the repetition frequency of the current pulses up to 200 per second and, consequently, to increase the dose rate of the bremsstrahlung radiation generated by the betatron by a factor of 4.

To investigate the operating features of a betatron with pulsed excitation of the electromagnet, an experimental betatron installation designed for an energy of 6 MeV was manufactured, consisting of four units: the emitter, the pulse circuits, the capacitors, and the control console.

In the emitter unit of this betatron, an electromagnet from a PMB-6-type betatron [1] was used.

The parameters of the excitation pulses are as follows: current-pulse duration 2500 μsec, pulse repetition duty factor equal to 2, current-pulse amplitude 376 A.

The power-supply circuit of the electromagnet is shown in Fig. 1 and operates as follows. From the moment the triggering pulses are applied to thyristors \(T_1\) and \(T_1'\), the precharged capacitor bank \(C_k\) discharges into the electromagnet winding \(L_k\). At the instant when the voltage across capacitance \(C_k\) is equal to zero, diodes \(D_1\) and \(D_1'\) open, but the saturation choke delays the buildup of current in them for a time sufficient for closing thyristors \(T_1\), \(T_1'\). After the current is transferred to the diodes, the capacitor current changes direction and, as the current in the inductance decays, the capacitance \(C_k\) is charged to a voltage of the initial sign.

Fig. 1.

To compensate losses in the circuit, energy is supplied from the rectifier filter capacitance \(C_f\) through the inductance \(L_p\) and thyristor \(T_2\). If the betatron is powered from a low-voltage source, then the energy-input thyristor \(T_2\) is switched on after the transfer of current from thyristors \(T_1\), \(T_1'\) to diodes \(D_1\), \(D_1'\).

Stabilization of the maximum value of the magnetic field in the working gap of the betatron under unstable supply-source voltage is achieved by regulating the portion of energy introduced into the oscillatory circuit. The amount of energy introduced is regulated by changing the turn-on time of thyristor \(T_2\) relative to the moment of closing of thyristors \(T_1\) and \(T_1'\).

Analysis of the process of energy input, under the assumption that during the input time the current in the \(C_k L_k\) circuit remains constant and the voltage across capacitance \(C_f\) does not change, showed that during the input time

\[ t_v = \frac{2}{\omega}\operatorname{arctg}\frac{U_{cf}-U_1}{\rho I_k}, \]

the following energy is introduced into the circuit,

\[ Q_{\mathrm{v}}=\frac{2U_{\mathrm{cf}}C_k^2\left(U_{\mathrm{cf}}-U_1\right)^3}{3L_p I_k^2} \]

joules, where

\[ \omega=\frac{1}{\sqrt{L_p C_k}} \]

is the resonant frequency of the recharging circuit,

\(U_{\mathrm{cf}}\) is the voltage on the rectifier filter capacitance,

\(U_1\) is the voltage on the circuit capacitor at the moment when energy input begins,

\[ \rho=\sqrt{\frac{L_p}{C_k}} \]

is the characteristic impedance of the recharging circuit. To maintain unchanged energy in the electromagnet circuit, the delay of the turn-on instant of the energy-input thyristor \(T_2\), relative to the current interception in the circuit, obeys the relation

\[ t_3=\frac{1}{\omega_k}\arcsin \frac{\left(4.3\Delta U_{\mathrm{cf}}-\dfrac{\Delta U_{\mathrm{cf}}U_1}{U_{\mathrm{cf}}}\right)} {U_{ckm}e^{-\alpha t/2}}\cdot\frac{\omega_k}{\omega_{0k}}, \tag{3} \]

where

\[ \omega_k=\sqrt{\frac{1}{L_kC_k}-\frac{r_k}{2L_k}} \]

is the natural frequency of the circuit,

\[ \omega_{0k}=\frac{1}{\sqrt{L_kC_k}} \]

is the resonant frequency of the circuit,

\(U_{ckm}\) is the voltage on the circuit capacitor,

\[ \alpha=\frac{r_k}{2L_k} \]

is the damping coefficient of the circuit,

\(\Delta U_{\mathrm{cf}}\) is the change in voltage on the filter capacitance,

\[ T=2\pi\sqrt{L_kC_k-\frac{2L_k}{r_k}} \]

is the period of the natural frequency of the circuit.

Such a dependence is implemented by the delay circuit [6], which includes dinistor \(D_3\), capacitors \(C_3, C_4\), resistors \(R_1\)—\(R_5\), transformer \(Tr_1\), thyristor \(T_3\), and diodes \(D_4, D_5\).

The circuit assumes its initial state after the supply voltage has been applied and the potential on the integrating capacitance \(C_3\) has reached the value at which dinistor \(D_3\) turns on. In this case capacitor \(C_3\), through the internal resistance of the reference-voltage source, discharges into the primary winding of pulse transformer \(Tr_1\). The pulse arising in the secondary winding \(W_2\) brings thyristor \(T_3\) into the on state. The operating mode of the thyristor is selected so that it is held in the on state. This is ensured by the appropriate choice of the value of resistor \(R_3\) from the condition

\[ \frac{E_a\left(R_3+R_5+r_{d5}+r_{d4}\right)} {R_3\left(R_5+r_{d5}\right)} =I_T\ge I_{\mathrm{hold}}, \tag{4} \]

where \(I_T\) is the thyristor current,

\(E_a\) is the voltage of the power-supply source,

\(I_{\mathrm{hold}}\) is the thyristor holding current,

\(r_{d4}, r_{d5}\) are the internal resistances of diodes \(D_4\) and \(D_5\).

The dynistor is switched off after the complete discharge of capacitor \(C_3\). The initial state of the circuit is such that thyristor \(T_3\) is on, while dynistor \(D_3\) is off. Upon arrival of the trigger pulse from winding \(W_3\), thyristor \(T_3\) is turned off. The current flowing in the \(R_5 C_3\) circuit charges the capacitor, and the voltage across its plates increases exponentially:

\[ U_{c3}(t)=U_{c3}(0)+[E_a-U_{c3}(0)](1-e^{-t_{\mathrm{del}}/T_1}), \tag{5} \]

where \(T_1=R_5 C_3\), and \(t_{\mathrm{del}}\) is the pulse delay time.

The dynistor is switched on when

\[ U_{c3}(t)-U_{\mathrm{ref}}=U_{\mathrm{on}}, \tag{6} \]

where \(U_{\mathrm{ref}}\) is the magnitude of the reference voltage.

The pulse arising at transformer winding \(W'_2\) returns the circuit to its initial state, and the pulse from winding \(W_3\) serves to switch on the energy-input thyristor \(T_2\).

From formulas (5) and (6) it is clear that, when the magnitude of the reference voltage is changed, the instant at which the dynistor is switched on changes, i.e., the delay of the turn-on pulse of the energy-input thyristor changes.

Fig. 2.

In testing, the stabilization circuit operated satisfactorily and, when the voltage on the filter capacitor \(C_f\) was varied within the range from 200 to 300 V, it ensured variation of the voltage on the resonant-circuit capacitor \(C_k\) within 1.5 V.

The circuit that sets the turn-on instant of the thyristors of the oscillatory circuit, \(T_1\) and \(T'_1\), and thereby determines the duty factor of the current pulses in the electromagnet winding, is constructed analogously to the delay circuit described above. It is triggered by pulses from the additional winding \(W_y\) of the electromagnet, differentiated by the \(C_7R_2\) chains.

The synchronization, injection, and displacement circuits (Fig. 2), implemented entirely with semiconductor devices, in addition to their simplicity, ...

differ from the known circuits in that, to delay the injection instant, a single-cycle magnetic amplifier, MA, is used. The use of such an amplifier with a bias winding, the current of which is proportional to the injection voltage, makes it possible automatically to maintain the optimal injection instant when the voltage on the electromagnet changes within small limits. The bias winding \(W_{\mathrm{p}}\) serves for manual adjustment of the injection instant. The ampere-turns of the windings \(W_{\mathrm{p1}}\) and \(W_{\mathrm{p2}}\) are set approximately equal.

The voltage from the turns \(W_1\), which enclose the central magnetic flux of the betatron, is fed through diode \(D_1\) to the operating winding of the magnetic amplifier. After the thyristors of the oscillatory circuit are switched on, after a certain time determined by the parameters of the MA and by the ampere-turns \(W_{\mathrm{p1}}\), \(W_{\mathrm{p2}}\), the core of the MA is remagnetized. A voltage drop appears across resistor \(R_4\), which, through the stabilitron \(C_{\mathrm{T}}\) and the differentiating circuit \(C_1R_6\), is applied to the control electrode of thyristor \(T_1\).

Thyristors \(T_1\) and \(T_2\) form the circuit of a triggered multivibrator, the pulse duration of which is regulated by resistance \(R_9\) and determines the instant at which the electrons are displaced onto the target.

The displacement and injection circuits consist of charging valves \(D_3\) and \(D_4\), charging resistors \(R_{14}\) and \(R_{15}\), and storage capacitors \(C_4\) and \(C_5\). Thyristors \(T_3\) and \(T_4\), with an average current of 25 A and a voltage of 600 V, are used as switches.

Transformer \(Tr_4\) serves to obtain an injection pulse voltage with an amplitude of about 30 kV and a duration of 10 ms. A high-voltage filament transformer \(Tr_3\) is structurally combined into one unit with the pulse transformer.

Since, in the transition to a pulsed power-supply mode, the distribution of losses among the elements of the electromagnet depends on the pulse parameters, and a theoretical investigation of this question involves certain difficulties, the temperature regime of the betatron electromagnet and the distribution of losses in its parts were studied in detail on an experimental model of a 6-MeV pulsed betatron, along with other questions.

The losses in the magnetizing windings, found experimentally, are 746 W, and in the magnetic circuit, 551 W.

A calculation of Joule losses, without taking into account the additional losses caused by the presence of leakage fields, shows that in the magnetizing windings, under this power-supply mode, they should amount to only 164.2 W. Thus, in calculating and designing the magnetizing windings it is necessary to take into account the additional losses from leakage fields, which, as is evident, considerably exceed the Joule losses.

A study of the magnetic characteristics of the working gap of the betatron electromagnet, powered by unipolar current pulses, and their comparison with analogous characteristics of the same electromagnet excited by sinusoidal current from a 50-Hz mains supply, showed that the transition to pulsed power supply did not lead to any substantial change in them.

Adjustment of the betatron for radiation showed that, with one and the same accelerating chamber and with the same injection-pulse voltage, the exposure dose rate of the radiation increased in proportion to the frequency of the supply current and amounted to 72 R/h at a distance of 1 m from the target.

References

1. L. M. Anan’ev [et al.]. Small-size betatrons and their application in flaw detection. Defektoskopiya, 1968, No. 6, pp. 60–61.

2. L. M. Anan’ev [et al.]. Induction Electron Accelerator—Betatron. Gosatomizdat, 1961, pp. 136–163.

3. Electron Accelerators. Proceedings of the Fourth Interuniversity Conference on Electron Accelerators. Vysshaya Shkola, 1964, pp. 243–247, 280–285.

4. V. V. Ivashin. An economical thyristor circuit for exciting an accelerator electromagnet with unipolar current pulses. Electron Accelerators, issue 3. Atomizdat, 1970, pp. 27–29.

5. L. M. Anan’ev, Yu. A. Otrubyannikov. Some questions concerning pulsed power supply of electromagnets of induction accelerators. Pribory i tekhnika eksperimenta, No. 3, 1967, pp. 40–41.

6. A. A. Geizer [et al.]. Delay device. Author’s Certificate No. 302824. Byulleten’ izobretenii, No. 15, Moscow, 1971.