Search

A representation of the state of ion driver technology is shown in Table II. As in any such superficial treatment there are some subjective conclusions but the clear generalization is that the light ion machines have the necessary power and energy but lack rep rate and reliability. The heavy ion drivers, on the other hand, are based on developed accelerator technology which is demonstrated to be of high rep rate and reliability, but is still not at the required energy or power levels.

A pulsed electron accelerator has been constructed and is now in operation at the Lawrence Radiation Laboratory, Berkeley. It was designed specifically for and is used as an injector for ERA (electron ring accelerator) studies. Electron bursts of 1200 amperes and 40 nanoseconds duration can be produced at energies of 1.0 MeV to 4.25 MeV. The present repetition rate is approximately one per second. Acceleration is accomplished by means of ferrite loaded accelerating sections, each of which is capable of maintaining a 250 kV voltage pulse across the accelerating gap for a period of 40 ns. The voltage pulse has a rise time of 12 ns and the total timing jitter is of the order of one nanosecond. Five sections stacked together and operated as a unit function as the electron gun. The gun employs a field emission cathode and is operated at approximately 1 MV. Additional acceleration is achieved by means of an iterated system of accelerator sections and solenoidal magnetic lenses.

Several fast-transfer systems using kicker magnets are needed for the proposed 200-GeV Accelerator. A possible system for transfer from the injector synchrotron to the main ring using stationary, full-aperture kicker magnets presents the most stringent requirements. The extraction kicker for the injector synchrotron will be about 4 m in length, have a 4-by-14-cm aperture, and require a field or 300 G. For 100% efficient transfer the rise time of the magnetic field must be shorter than 17 ns, which is the time between beam bunches, and which is nearly an order of magnitude less than the rise times achieved in existing kicker magnets. Because of the required rise time and the limited length, each magnet is divided into 48 modules operated as 24 push-pull pairs. Each module is analyzed as a short-circuited ferrite-and dielectric-loaded transmission line. The short-circuited mode of operation ensures that voltage on the magnet is applied for only about 12 ns, thus reducing HV problems, and at the same time produces a kick twice as great as can be obtained from terminated magnets. Each module is energized by a tapered coaxial pulse line and switched on with a triggered spark gap.

Results of the first study of the conceptual design of a collective-effect proton accelerator of substantial energy (65-100 GeV) are described. Basic theoretical considerations are outlined and components and their arrangement for a conceptual accelerator are discussed. This first study shows that the promise of ERA principle to produce remarkably compact accelerators can be achieved by use of single-pulse surge-excited acceleration cavities. The example single-pass machine provides accelerating gradients of 250 MeV/meter overall where conventional synchrotrons yield 4o-80 MeV/ meter. The relaxed tolerances of the surge-excited cavities and their associated components allow great advantage to be taken of mass-production techniques with ensuing low cost (? M$ 0.24/GeV).

The construction and assembly of a Cs ion injector consisting of a pulsed source and 3 pulsed drift tubes has been complete since April, 1980. The measurement program, underway since then to characterize the beam, has been interspersed with the development of diagnostic equipment. The Cs contact ionization source and each of the 3 drift tubes are driven by 500 kV Marx generators. The injector has been operated reliably at 300 kV/stage at a repetition rate of 1 pulse/4 sec. About 105 pulses have been accumulated. The space charge limited diode and drift tube acceleration system were designed with the aid of the EGUN code of Herrmannsfeldt. Measurements of the beam envelope have been made by means of a movable biased charge collector. Good agreement with the EGUN calculation is found. Measurements of the beam emittance have been made at the exit of the third drift tube. The normalized emittance ? ?N = 2 x 10-6 ? m-rad is of better optical quality than that required for further acceleration and transport in a Heavy Ion Fusion (HIF) Induction Linac Driver.

In an induction linac the accelerating voltage appears along a voltage-graded vacuum insulator column which is a performance limiting and major cost component. Re-X glass ceramic insulators have the long-sought properties of allowing cast-in gradient electrodes, good breakdown characteristics, and compatibility with high vacuum systems. Re-X is a glass ceramic developed by General Electric for use in the manufacture of electrical apparatus, such as vacuum arc interrupters. We have examined vacuum outgassing behavior and voltage breakdown in vacuum and find excellent performance. The housings are in the shape of tubes with type 430 stainless steel terminations. Due to a matched coefficient of thermal expansion between metal and insulator, no vacuum leaks have resulted from any welding operation. The components should be relatively inexpensive to manufacture in large sizes and appear to be a very attractive accelerator column. We are planning to use a standard GE housing in our MBE-4 induction linac.

The momentum spread required to stabilize longitudinal modes in a proposed high current betatron is computed.

A simple theoretical study and numerical estimate is presented for the transverse amplitude growth of a nonrelativistic heavy ion beam in an induction linac, as envisaged for use in commercial power plants, due to the nonregenerative coherent beam breakup mode. An equivalent electrical circuit has been used to represent the accelerating induction modules. Our calculation shows that for the parameters of interest, the beam breakup amplitude for a heavy ion beam grows extremely slowly in the time scales of interest, to magnitudes insignificant for transport purposes. It is concluded that the coherent beam breakup mode does not pose any serious threat to the stability of a high current (kA) heavy ion beam in an induction linac.

A 500 KeV one-ampere Cs+1 ion beam has been generated by contact ionization with a 30 cm dia. iridium hot plate. Reproducibility of space charge limited ion current wave forms at repetition rates up to 1 Hz has been verified. The beam is characterized to be very bright and suitable as an ion source for the induction linac based heavy ion fusion scheme. The hot anode plate was found to be reliable and self-cleaning during the operation.

A Cs ion-beam transport experiment is in progress to study beam behavior under extreme space charge conditions. A five-lens section matches the beam into a periodic electrostatic quadrupole FODO channel and its behavior is found to agree with predictions. With the available parameters (?200 keV, ?20 mA, ??n ? 10-7 ? rad-m, up to 41 periods) the transverse (betatron) oscillation frequency (?) can be depressed down to one-tenth of its zero current value (?o), where ?2 = ?2o - ?2p/2, and ?p is the beam plasma frequency. The current can be controlled by adjustment of the gun and the emittance can be controlled independently by means of a set of charged grids.

A long-pulse induction acceleration unit has been installed in the high-current Cs+ beam line at LBL and has accelerated heavy ions. A maximum energy gain of 250 keV for 1.5 ?s is possible. The unit comprises 12 independent modules which may be used to synthesize a variety of waveforms by varying the triggering times of the low voltage trigger generators.

The physics of the pellet implosion sets stringent conditions on the accelerator driver. The beam energy should be > 1 MJ, the beam power > 100 TW (implying a pulse length approx. = 10 ns), and the specific energy deposition in the pellet > 20 MJ/g. Thus, considerable current amplification is required, e.g. from some 10 amps at the source to perhaps 10 kiloamps at the pellet. Most of this amplification can be accomplished continuously along the accelerator and the remainder achieved at the end by bunching in the final transport lines to the target chamber. A conceptual schematic of an Induction Linac Fusion Driver is shown, which includes an injector, an accelerator-buncher, and a final transport system. Here only the accelerator portion of the driver is discussed.

A compact monitor has been developed which measures beam current and beam position without intercepting or appreciably affecting the particle beam being measured. It is a broad band device which prevents the development of beam disturbing resonant modes. Although developed specifically for use on a short-pulse electron induction accelerator, it might be applied to other accelerator beams of pulsed or rf-bunched nature. The monitor consists of a resistive band inserted in the beam pipe wall with connections for reading the voltage across the band at each vertical and horizontal axis intercept. For pulsed or rf beams, return current equal in magnitude to the beam current flows on the beam pipe inside wall and through the resistive band. If the beam is centered, equal voltage appears on all four monitor connections. If the beam is off-center, the return currents are unsymmetrical and unequal voltages appear at the monitor connections. Beam current is proportional to the sum of all monitor voltages, while beam position is approximately proportional to the difference in voltage at opposing monitors. Tests and operating experience confirm its operation.

Two beam choppers were constructed which produce an approximately square, 20-ns pulse of 3+-MeV electrons out of the best portion of the 300-ns Astron beam or an 8-?s microwave electron linac beam for injection into the electronring accelerator (ERA) compressor. The choppers are impedance-matched traveling-wave devices which deflect the beam with equal strength electric and magnetic forces, and are powered with either three-electrode spark gaps or thyratrons and a 20-ns pulse line. The electron beam is biased out of the beam-transport system with a dc magnetic field, and switched into it with the chopper.