Early HIF Experiemnts

Early HIF Experiments

The US heavy-ion fusion program is developing the accelerator physics and technology data base for high-current induction linacs with multiple ion beams and current amplification. Before the formation of the Virtual National Laboratory for Heavy Ion Fusion in 1998, many important ion-accelerator experiments were carried out at Lawrence Berkeley National Laboratory, among with the following are particularly noteworthy.

High-current heavy-ion sources were developed in the 1970s

A suitable injector for a driver should deliver a beam pulse as large as several amperes at a few MeV. One of the first issues was the availability of suitable sources.

Initially, we imagined a driver to use just one (big) beam starting with a few MeV at a few amperes, In the late 70's we built a 1-ampere, 1-MV cesium injector and determined that its quality was 100 times better than needed in a driver. The surface-ionization source is 30 cm in diameter and is mounted on the high-voltage terminal of a Marx generator. The beam was further accelerated through three pulsed drift tubes to 2 MeV.

The Single-Beam Transport Experiment (1980-1986) established linac transport limits

Of key importance to driver design is understanding the beam space-charge limit in a strong-focusing linac. In a synchrotron, the Laslett tune-shift criterion allows space-charge defocusing to depress the betatron frequency by less than 10% of its low-current value. The Single-Beam Transport Experiment (SBTE) established in the laboratory that, in a long AG transport system (87 electrostatic quadrupoles), space-charge tune depressions of factors of at least 10 can be tolerated with little loss of beam quality. Down to this value, both emittance and current are unaffected provided the betatron phase advance per lattice period is kept below approximately 85 degrees. Simulation shows that below this value image effects may be damaging if the beam is not centered properly.

These studies, which allowed us to consider drivers with ion charge-states greater than one, opened up the field of "space-charge dominated beams" in which the Debye length is a small fraction of the beam diameter.

The Multiple-Beam Experiment (MBE-4) amplified current in four beams

Current amplification in an ion induction linac was first demonstrated in the four-beam MBE-4 experiment, operated at LBNL from 1985 to 1991. The experiment showed that properly shaped acceleration waveforms could control the length of pulses with high space-charge by countering the beam electrostatic forces that normally cause such pulses to lengthen during transport. As the oscillograms below illustrate, the MBE-4 beams were successfully compressed by a factor of three, giving a threefold amplification of the current. By taking care to match, center, and align the MBE-4 beams, LBNL scientists showed that the four beams could be accelerated at constant normalized emittance at least over the 14-m length of the experiment. Beam space charge were also found to reduce significantly the effects of longitudinal acceleration errors.

A driver-scale 2 MeV, 0.8 A potassium injector has been developed
Another important HIF issue is the availability of suitable injectors that form ions from the sources into beams that can be accelerated by the accelerator. Injector development has been an important part of the heavy-ion fusion program at Berkeley for more than fifteen years.

LBNL built and successfully tested an injector that provides potassium ions at the energy, current, and emittance of a full-scale driver. The beam is produced in a conventional 0.75 MV diode. It is further accelerated to 2 MV by a electric-quadrupole-focused high-voltage column. During 1994, the new injector successfully met the design goals for beam emittance (normalized edge emittance of less than 1 pi mm-mr), energy (>2 MV), and current (800 mA of K+). Detailed measurements of transverse phase space over a broad range of parameters (in energy and current) have shown excellent agreement with computer simulations done with the three-dimensional particle-in-cell code WARP3D.

By fine-tuning the Marx generator and extraction pulser waveforms, we have improved the head-to-tail energy flatness of the ion beam, reducing the energy variation to less than +/-0.15% over the duration of the beam pulse, which is 1 microsecond or slightly greater. A picture of the injector column is shown here.
We built and successfully tested an injector that provides potassium ions at the energy, current, and emittance of a full-scale driver. The beam is produced in a conventional 0.75 MV diode. It is further accelerated to 2 MV by a electric-quadrupole-focused high voltage column. During 1994, the new injector successfully met the design goals for beam emittance (normalized edge emittance of less than 1 pi mm-mr), energy (>2 MV), and current (800 mA of K+). Detailed measurements of transverse phase space over a broad range of parameters (in energy and current) have shown excellent agreement with computer simulations done with the three-dimensional particle-in-cell code WARP3D.

By fine-tuning the MARX generator and extraction pulser waveforms, we have improved the head-to-tail energy flatness of the ion beam, reducing the energy variation to less than +/-0.15% over the duration of the beam pulse, which is 1 microsecond or slightly greater. A picture of the injector column is shown here.

Photograph of the high-voltage accelerating column

This apparatus is still in use. Recent simulation and engineering work has reduced the beam-hollowing seen during the early testing, and the column will serve as the injector for the new High Current Experiment (HCX).

The Heavy-Ion Fusion Systems Assessment (HIFSA) gave encouraging results
DOE laboratories and industry collaborated in a broad parametric study of a power plant with an induction linac driver (Fusion Technology, February 1988 issue). A surprisingly broad latitude was found in the choice of accelerator parameters, the projected cost of electricity is favorable (5.5 cents/kWh), and plant sizes down to 500 MWe look reasonable. Driver efficiencies typically lie in the 20-40% range. The sketch below shows the induction accelerator driver studied in the assessment. The ion charge state of three enabled 10 GeV of acceleration to be achieved with 3.3 GV of accelerator.
A cost-effective driver architecture (from HIFSA)

The Induction Linac Systems Experiments (ILSE) would have tested many driver features
After 1986, much of the LBNL program effort was directed toward developing a next step toward a driver called ILSE, the Induction Linac Systems Experiments. ILSE would have addressed many of the remaining beam-control and beam-manipulation issues--many of them (including charge per unit length or "line charge density" and beam diameter) at full driver scale, others, such as ion kinetic energy, in a scaled fashion.

Sketch of ILSE in the LBNL Bevatron Building

In the ILSE conceptual design, four beams would have been accelerated and electrostatically focused, then combined into one. The beams were to have been further accelerated in a subsequent section, this time with magnetic focusing. In both the electrostatic-focus and the magnetic-focus sections, the beam would have been compressed, as in a driver. Then the beam would then have been used for experiments in drift compression and final focus. In the table below, some parameters of the ILSE design are shown and compared to those of previous experiments, as well as to typical driver parameters.

Comparison of ILSE parameters with other HIF accelerators

	SBTE	MBE-4	ILSE	Driver
Ion Species	Cs+	Cs+	K+	Hg+
Number of beams	1	4	4 -> 1	~100
Final Energy (MeV)	0.2	1	10	4000
Initial current per beam (A)	0.02	0.01	0.8	0.4
Pulse width (µsec)	20	2.4 -> 0.2	1->0.4	24->0.1

As planned, the ILSE project would have been fabricated in 3 to 4 years at a project cost of approximately $50M. Although the project mission (KD-0) was recognized in March1992 and a decision to proceed (KD-1) with the first section of ILSE called Elise (~$25 M) in FY96 was made by the Department of Energy in December 1994, the project was not started. Because of the restructuring of the US fusion program that occurred in 1996, Elise was postponed indefinitely, effectively killing the project.

For comments or questions contact WMSharp@lbl.gov or DPGrote@lbl.gov. Work described here was supported by the Office of Fusion Energy at the US Department of Energy under contracts DE-AC03-76SF00098 and W-7405-ENG-48. This document was last revised June, 2002