80 MeV Synchrotron

The synchrotron principle of accelerating relativistic particles, proposed in 1945 by V. I. Veksler¹,² and, somewhat later, by McMillan³, made it possible to raise by an order of magnitude the maximum energy limit attainable for electrons. The principal advantages of the synchrotron in comparison with the betatron are: the insignificance of electromagnetic radiation for the successful operation of the device (up to energies of \(10^9\) eV), and the possibility of using a ring magnet instead of a solid one, since acceleration is produced by a high-frequency electric field, so that the central magnetic flux is not necessary.

As in the cyclotron, acceleration in the synchrotron is produced by means of a high-frequency electric field of frequency \(\omega_0\), concentrated in one or several narrow gaps. To preserve the resonance condition

\[ \omega_0 = \frac{ceH}{mc^2} \tag{1} \]

electric field with the motion of an electron of mass \(m\) in a magnetic field of intensity \(H\), the magnetic field increases with time, compensating the increase of the relativistic mass \(m\) in the acceleration process. The operation of the synchrotron is based on the principle of autophasing \(^{1,2,3}\), which consists in the fact that the particle itself “selects” the moment of passage through the accelerating gap so that its energy increases in accordance with the increment of the magnetic field. A certain gain in energy at the expense of the e.m.f. of induction and the loss of energy to radiation (at high energies) affect only the fact that the electron takes from the electric field a somewhat smaller or somewhat larger energy, so that the total gain in energy corresponds to the increment of the magnetic field. Rigorous calculations carried out by a number of authors \(^{4,5,6}\) confirmed the preliminary calculations in the original papers \(^{1—3}\).

By 1947 several small synchrotrons had been built to test the principle \(^{7,8}\). In the same year a synchrotron of the General Electric Company was put into operation, with a maximum electron energy of 70 MeV (according to the latest data, the maximum energy has been raised to 80 MeV).

A section of the synchrotron magnet is shown in Fig. 1. The magnet is made according to the type of betatron magnets of the General Electric Company \(^{10}\). Its main difference—the absence of the central part of the pole pieces—makes it possible to reduce substantially the weight of the magnet and its reactive power. The pole pieces have an annular shape and are made of thin (0.35 mm) dynamo-steel sheets, from which individual sectors are assembled. The profile of the pole pieces was selected by means of tests on a special model (half full size) so that the magnetic

Fig. 1. Side view of the synchrotron magnet.
1 — yoke, 2 — brass tube, 3 — clamping beam, 4 — coil box, 5 — core, 6 — spiral for water cooling, 7 — electromagnet windings, 8 — chamber, 9 — pole piece, 10 — textolite ring, 11 — adjustable air gap, 12 — opening for cooling, 13 — exciting winding, 14 — clamping bolt, 15 — clamping beam, 16 — cooling tubes.

field between the poles would decrease with radius according to the law \(H \sim r^{-3/8}\), which ensures the stability of the radial and vertical motion of the electrons. The width of the annular poles is 10 cm. In order for the electron orbits during synchrotron acceleration to lie in this narrow ring, the initial energy of the electrons must be sufficiently large. Therefore the initial acceleration to an energy of \(\sim 2\) MeV is carried out in a beta-

...tron mode, with the central accelerating magnetic flux passing through a small central core (see Fig. 1), which, upon transition to the synchrotron mode, is rapidly saturated and has no effect on the further acceleration of the electrons. The total weight of the magnet is about 8 t; the radius of the synchrotron orbit is 29.3 cm (frequency of the electric field 163 MHz); the maximum field on the orbit during acceleration is up to 70 MeV, \(H_m = 8100\) gauss. The magnet employs a system of water cooling for the poles and yokes and air cooling for the central core.

The magnetic field is produced by magnetizing coils (130 turns on each pole) connected into a resonant circuit with a capacitor bank of capacitance 14.4 μF. The frequency of variation of the magnetic field is 60 cps; the power developed (at \(E = 70\) MeV) is 3000 kVA; the voltage is about 23 kV. The circuit is excited by two primary two-turn coils fed from an induction regulator.

The Pyrex vacuum chamber, of toroidal form with wall thickness \(\sim 6\) mm, had two branches for the injector (an electron gun of the usual betatron type) and for the target. The chamber was evacuated through the branch for the target.

A \(57^\circ\) sector was cut out of the chamber and replaced by a sector of special 707-D glass with small dielectric losses, on which the accelerating system of the synchrotron was mounted—a quarter-wave resonant cavity (Fig. 2), excited by a generator through a coaxial line. To obtain a potential difference of 100 V in the gap of the endovibrator, the supplied power should be 250 W.

Fig. 2. Diagram of the endovibrator.

1 — tuning strip, 2 — coupling strip, 3 — gaps in the silver coating (to reduce Foucault currents), 4 — insulated pad for the measuring instrument, 5 — connection of adjacent strips, 6 — accelerating gap.

Of interest are reports of experiments carried out on the synchrotron to study the dependence of its efficiency on various parameters. As a measure of intensity, the intensity was taken of the bremsstrahlung X-radiation arising when electrons strike the internal or external target after the high-frequency electric field is switched off (the injector served as the external target).

The tests showed that the exact position of the betatron orbit has little effect on the operation of the synchrotron. A large fraction of the electrons is transferred from the betatron mode to the synchrotron mode, and changing the time at which the amplitude of the voltage in the gap is established (from 2 to 20 μsec) and increasing the amplitude of this voltage above a certain value have no appreciable effect on the intensity; apparently, practically all electrons are then captured. The measured curves of the dependence of the output on the moment of switching on the high frequency, on the moment of injection, and on the injection energy show a strong dependence on these factors: the optimum values of the moment of switching on the high frequency and of the moment of injection are clearly expressed; with increasing injection energy the intensity increases continuously (in the investigated energy interval from 15 to 37 MeV). Study of the dependence of the intensity on the position of the injector showed that the maximum intensity is obtained with the injector located at a radius slightly exceeding the radius corresponding to \(n = 1/(1 - r^{-n})\). In a supplementary paper \(^{11}\) measurements are reported that were carried out to study...

of the influence of the chamber dimensions. The effective dimensions of the chamber were set by the position of special movable wire frames that limited the region of the electrons’ excited oscillations. The experiments showed a sharp dependence of the intensity on the height of the chamber (the intensity rapidly decreases and falls to zero at a height of \(\sim 1\) cm) and on its radial dimensions. (At first the intensity falls because the region of free betatron oscillations is reduced, and then—as the wall approaches the synchronous orbit—because the electrons hit the wall upon transition to the synchronous regime.) The experiments showed a sharp dependence of the intensity on the “clearance” of the injector—the distance between the filament and the inner edge of the injector: adding a small shield of 5 mm on the inner side reduced the intensity by a factor of 20.

On the General Electric synchrotron, the electromagnetic radiation predicted by Ivanenko and Pomeranchuk from an electron moving in a magnetic field was discovered and studied.\(^{12,13}\) At an energy of \(\sim 70\) MeV the maximum intensity falls in the region of visible radiation. Therefore this radiation is directly visible to the eye. In the last paper\(^{13}\) the spectral composition of the radiation, theoretically calculated by Artsimovich and Pomeranchuk,\(^{14}\) was determined. The light emitted by the electrons passed out of the chamber through a quartz window and was directed into a spectrometer with a Buda grating containing 15,000 lines per inch. The intensity of various portions of the spectrum was measured with an IP22 photomultiplier at maximum energies of 42.5, 60, 70, and 80 MeV. The measurements showed good agreement between the experimental results and the theoretically expected ones. The magnitude of the total radiation, determined by means of a thermocouple, makes it possible to determine the number of electrons circulating in the orbit. It proved to be equal to \(10^8\).

E. L. Burshtein

CITED LITERATURE

1. V. I. Veksler, DAN SSSR, 44, 393 (1944).
2. V. I. Veksler, Journ. of Phys. 9, 153 (1945).
3. E. M. McMillan, Phys. Rev. 68, 143 (1946).
4. M. S. Rabinovich, Journ. of Phys. 10, 523 (1946).
5. N. H. Frank, Phys. Rev. 70, 177 (1946).
6. D. Bohm and L. Foldy, Phys. Rev. 70, 249 (1946).
7. R. D. Hill, Nature 159, 774 (1947).
8. F. K. Goward and D. E. Barnes, Nature 158, 413 (1948).
9. F. R. Elder, A. M. Gurewitsch, R. V. Langmuir, and H. C. Pollock, Journ. Appl. Phys. 18, 810–818 (1948).
10. V. F. Vestendorf and E. E. Charlton, UFN 30, 119 (1946).
11. F. R. Elder, R. V. Langmuir and H. C. Pollock, Rev. Sci. Instr. 19, 121–122 (1948).
12. F. R. Elder, A. M. Gurewitsch, R. V. Langmuir and H. C. Pollock, Phys. Rev. 71, 829 (1947). See the abstract by M. S. Rabinovich, UFN 33, 277 (1947).
13. F. R. Elder, R. V. Langmuir, and H. C. Pollock, Phys. Rev. 74, 52–56 (1948).
14. L. A. Artsimovich and I. Ya. Pomeranchuk, ZhETF 16, 379 (1946).