The Integrated Beam Experiment the Next Step for Heavy Ion Fusion
C. M. Celata, for the Heavy Ion Fusion Virtual National Laboratory
Note: For the convenience of those unfamiliar with the heavy ion fusion driver concept, a short description of the driver is provided in Appendix I. A detailed description of the Heavy Ion Fusion Virtual National Laboratory program, program plan, experimental status, and accomplishments can be found in the white paper "Strategic Plan and Research Needs for Heavy-Ion Fusion Energy Development: An Integrated Research Program", by Grant Logan et al.Introduction
Over the next few years the U.S. Heavy Ion Fusion program will be completing a set of proof-of-principle experiments which will have demonstrated the feasibility of all driver beam manipulations which can be investigated at small scale. The results so far project to good driver performance, with power plant cost comparable to other fusion methods. This document presents a description of an intermediate-scale experiment, the Integrated Beam Experiment, which would be the next step beyond these small experiments-- an experiment which, while requiring only a modest increase in program resources, would access important areas of physics, constitute an integrated test of much of the physics of the driver, and produce a base of data and experience which would lay the groundwork for a larger, proof-of-performance experiment, the Integrated Research Experiment.
Scientific Mission of the Integrated Beam Experiment
The scientific goals chosen for the Integrated Beam Experiment are the result of a series of national workshops in which the Heavy Ion Fusion community examined the progress and future direction of the program. In May of 2001 a national Heavy Ion Fusion "Science Workshop" took place, whose purpose was to establish the science goals of the program for the following 3 years. To accomplish this, participants put together a comprehensive list of the science issues for heavy ion fusion, discussed the status of experimental and theoretical understanding of these issues, and prioritized them. A few weeks later the Director of the Heavy Ion Fusion Virtual National Laboratory, Grant Logan, introduced the concept of the Integrated Beam Experiment to the community. Study groups were formed to consider the scientific goals and design of the experiment, in the context of the list of science issues designated by the Science Workshop as most important to the program. Over the 6 weeks that the study groups met, fundamental design choices were studied, and two "strawman" designs were worked out and costed. This provided a preliminary, but realistic, idea of the tradeoff between science parameters and experimental cost. Then in a national "IBX Workshop" in October of 2001, the study groups reported on science opportunities, design tradeoffs, and costs of an Integrated Beam Experiment. Possible experiments were judged according to the importance given by the Science Workshop to the issues they addressed. By the end of the workshop, the participants had reached a nearly unanimous consensus on the scientific goals of an Integrated Beam Experiment, as well as a high level of enthusiasm that significant driver-relevant scientific advances could be made with this intermediate-scale experiment. In this section the physics issues important to Heavy Ion Fusion are briefly described, in order to show the place of the Integrated Beam Experiment in resolving these issues, and to delineate its scientific mission.
The scientific issues for Heavy Ion Fusion derive ultimately from target implosion requirements. In order to accelerate enough ions to deposit the ~5 MJ needed to implode and ignite the target, while minimizing the diameter (and therefore the cost) of the accelerator, the accelerated beams must be very intense-- i.e., their charge per unit length must be as high as possible. The number of ions required is approximately 1,000,000 times the number transported in a conventional high energy or nuclear physics accelerator bunch, and as a result, the physics for heavy ion fusion driver beams is multi-particle physics-- the physics of non-neutral plasmas-- as compared to the single-particle physics of the conventional accelerator. Thus one major task of the heavy ion fusion research program is to determine the stability and behavior of these intense beams.
The large amount of charge which must be transported leads to several parameter and accelerator design choices. Since the charge can not economically be transported in a single beam, the accelerator must transport in parallel a large array of beams (~ 30-100, depending on design). Cost optimization gives a time duration for the beams of tens of microseconds at the beginning of the accelerator. Thus beam-beam interactions, and any limitations to beam pulse length are important areas of interest.
The beams must arrive at the final focusing system with low enough transverse (i.e., transverse to the beam propagation direction) temperature that they can be focused to the small spots required, and low enough longitudinal temperature that aberrations in the final lens system do not enlarge the spot. The beam parameter which measures its phase space volume, and therefore its temperature and focusability, is the "emittance". The basic issues fundamental to feasibility of heavy ion fusion drivers therefore are those of producing low-emittance intense beams, maintaining that good beam quality (i.e., low emittance) while accelerating to a few GeV and performing all beam manipulations necessary (e.g., bending the beam trajectory, or compressing the beam longitudinally), and transporting the focused beam through the target chamber environment without degrading the focus. The important physics questions which must be answered by the program can therefore be summarized as:
Many of these questions are fundamental science questions which are important to beam science, and indeed they appear as design issues for the most modern accelerators, such as the Spallation Neutron Source, in which the space charge forces of the beam are high-- i.e., the beam is intense. As such, these fundamental issues can be explored in small, relatively inexpensive experiments where physics parameters important to particle dynamics are in the same range as in the driver, but the beam is at low energy-- a few hundred keV. In many cases the relevant parameters to the physics are ratios-- e.g., the ratio of the space charge force of the beam to the beam kinetic energy in the accelerator transport-- so that this ratio can be set at values common to driver designs, while both the beam current and its energy are kept low in order to minimize experiment cost. Using this approach of small experiments at low kinetic energy, in the 1980's and '90's the Heavy Ion Fusion program investigated the stability of intense beams; the emittance growth inherent in beam shape and profile changes; the interaction of the intense beam with the walls through image charge effects; and the effect on the beam of certain beam manipulations such as longitudinal compression, transverse combining of beams, and focusing to a spot.
These were experiments with ion density in the driver parameter range, and with sufficient length to allow evolution of beam emittance over several cycles of the frequencies of interest. While extremely productive in terms of confirming stable beam transport and maintenance of beam quality during beam manipulations, they were far from the parameter range of the driver in other key parameters, and therefore could not explore certain issues. The experiments in general used a single beam (the exception was a 4-beam accelerator), the pulse length was 1-2 microseconds, the energy of the beams was £ 1 MeV, and while the ion density was in the driver range, the beam radius was kept small to reduce cost. Thus the beam currents were a few milliamperes, as compared to ~ 0.5 - 1 ampere at the low-energy end of the driver. Furthermore, though the experiments were long enough to explore well the transverse physics of the beam, the growth rates and wave velocities for longitudinal phenomena are much longer, requiring a longer accelerator to explore this physics.
There is a class of important phenomena which can also be explored at low energy, in short-length experiments, but which requires high current-- in other words, a larger radius beam. With driver-scale beam radius comes a high space-charge potential, so that the beam can interact strongly with any electrons produced. Two experiments of the present program are investigating with beams of driver-scale diameter and current (0.1 - 0.5 amperes) the interaction of intense beams with electrons. One experiment, the High Current Experiment, is measuring the quantity and orbits of electrons produced if the beam is allowed to scrape the vacuum wall, and the other (the Neutralized Transport Experiment, or NTX) will look at neutralization-assisted focus of an intense beam to a small spot. These experiments will of course also explore other physics as well-- the effect of final focus lens aberrations on high perveance beams (perveance = ratio of space charge to kinetic energy of the beam), and the limits set by nonlinear forces to the amount of transportable current, for instance.
When the high-current low-energy experiments conclude, the program will have demonstrated production, stable transport, and acceleration of intense space-charge-dominated beams; longitudinal beam compression; transverse beam combining; beam neutralization using plasmas; electron production and control; and the final focus of the beam. With the exception of the "drift compression" of the beam (a longitudinal compression of the beam of a factor of 10-20 before the final focus), this list includes all of the beam manipulations in the driver. For each of these processes, phase space changes have been measured and matched with extensive particle-in-cell computer simulations. But though each process has been individually explored, in the driver they will be sequential, and the errors and changes in the beam distribution function which result from previous manipulations will be fed in as initial perturbations to the next area of transport. The perturbations in the beam shape will cause changes in the space-charge forces which dominate the dynamics. There is therefore a need to not only, for instance, check the focusing of a low-energy beam to a spot, with lens aberrations and all the attendant complexity, in order to understand focusing, but also to take a beam which has been transported and accelerated and drift-compressed, and therefore has a different distribution function giving different space-charge forces, and measure the effect of the focusing system on this beam. This integration will be one of the chief scientific missions of the Integrated Beam Experiment. As a result, the experiment will also benchmark computer codes which will be used to design future experiments with end-to-end computer simulation.
The science issues which could not been investigated in small-scale experiments are those which require high energy, multiple beams, and long length scales. Simply extending the length of the accelerator offers entry into significant new physics without the higher cost of accelerating to near-driver-scale energy or adding a large number of beams. This is the choice that the Integrated Beam Experiment would represent. With one, or a few, beams accelerated to tens of MeV (~1/300 of the energy of a driver beam), which means propagation over ~70 meters, it would be possible include all driver beam manipulation sections (see Fig. 1) and test the integrated source-to-target physics mentioned in the last paragraph. But the longer experiment would also make accessible the following important physics:
The mission of the Integrated Beam Experiment is to demonstrate the integrated source-to-target propagation physics of the driver, and to investigate these long-length-scale issues.
Figure 1 Schematic diagram of a heavy ion fusion driver and target chamber.
Parameters of the Integrated Beam Experiment
After consensus was reached on the scientific mission outlined above, participants at the national IBX Workshop derived the parameters of an experiment which would meet these scientific goals. Guidance was provided by two "strawman designs" prepared by study groups. Both of the designs included beam propagation from the source through the drift compression, final focus, and chamber neutralization, but made slightly different design choices. In order to focus and appropriately model the compression and neutralization of a beam with driver-scale space charge potential, the final energy must be high enough to give beam perveances (ratio of space charge potential energy to kinetic energy) in driver range. Final energy of 5 - 20 MeV was chosen to meet this criterion. The beam charge-per-unit-length of driver designs, namely ~ 0.2 microcoulombs/m, was selected. This will produce the space charge potential necessary for electron dynamics and final transport neutralization studies. The final perveance (measure of the ratio of space charge potential energy to kinetic energy) of the beam after longitudinal compression of a factor of 10 would be ~ 10-3, an aggressive value which would enable exploration of a wide range of final transport options. Short beam pulse length (0.2 - 2 microseconds) was chosen in order to reduce the cost of acceleration materials. Long pulse length effects will be studied on the LLNL 500 kV test stand, and on the High Current Experiment. Finally, there was nearly universal agreement that it would be very valuable to be able to study multiple beam effects, but that this option was likely to be too expensive. Upgradable one-beam systems were discussed, and the issue awaits resolution as more definitive design work proceeds.
The cost of construction of the experiment is estimated to be ~ $45M (2002 dollars). Although this is a preliminary estimate, since design is in its initial stages, it is based for most items on fairly detailed and realistic information: on bottoms-up design and costs of the 2nd arm of the DARHT accelerator, which has been built and is being commissioned; on engineering drawings for previous designs for large experiments; and on costs from heavy ion fusion small-scale experiments, vendor quotes, and prototypes.
Given sufficient funding and approval by the DOE, the Integrated Beam Experiment could be ready for start of (5-year) construction in FY04. This would require increases in the Heavy Ion Fusion Virtual National Laboratory program budget from its present $10M/year to $18.6M in FY04, with slight increases in FY05, 06. Experiments would begin in FY07.
What the Integrated Beam Experiment Cannot Do the Mission of the IRE
The Integrated Beam Experiment (IBX) would constitute a bridging step between the small experiments used to demonstrate feasibility and physics, and the Integrated Research Experiment (IRE), a proof-of-performance experiment which would demonstrate all the physics and technology necessary to a driver. The main differences between the IBX and the IRE are consistent with this difference in scale, capability, and mission. They are: (1) the final kinetic energy of the IRE would be a few 100 MeV, while for the IBX we envision 5-20 MeV, (2) the IBX pulse length would be much shorter than that in the IRE in order to save induction acceleration core cost, and (3) the IRE would be a significant multibeam experiment (~30 - 100 beams), where the IBX is expected to have 1-to-a-few beams. The difference in number of beams and final kinetic energy imply a vast difference in total current in the two experiments.
The scientific mission of the IRE covers the beam physics which can only be done at high energy or with the high total current implied by high energy and large numbers of beams. Most of the multiple beam physics of the driver would be studied first on the IRE, especially inductive effects that occur only at high velocity, and the interaction of multiple beams and the electrons neutralizing them in the target chamber. The effect on longitudinal stability of the interaction of the beams with the induction cores is another important topic. This "beam loading" will be negligible in the IBX because of low beam velocity and the small number of beams. The study of electron accumulation in the beam requires a long pulse length, and therefore though much electron physics could be explored on the IBX, a significant part of the problem remains for the IRE. The IRE would also give the first transport for lengths of the order of the driver length, providing a definitive test of the effect of long transport on beam quality. For this reason much of the physics first explored on the IBX would be tested again over long length scales in the IRE. Finally, the IRE is expected to be capable of exploring some heavy-ion target-interaction physics, while the IBX will not have the kinetic energy for this mission. The IRE would be capable of upgrade to an Engineering Test Facility (ETF) which could test target chamber design.
Value of the IBX to the IRE
The value of the Integrated Beam Experiment to future experiments must be emphasized, now that its value to driver-relevant physics has been discussed. The Integrated Beam Experiment would give physics data which is needed to optimize design quantities such as the accelerator aperture, pulse length, and final focus strategy of an IRE. This inevitably would lower the IRE cost and raise its performance capability. The Integrated Beam Experiment would also prototype diagnostics and technology necessary to an IRE. And, extremely important, it would benchmark the computer codes and physics models which would be used to design the IRE. With the IBX data and integrated models increasing confidence in calculating the IRE focal spots, we can then design specific target physics experiments using multiple beams from the IRE, such as beam intensity distributions at the target with multiple overlapping beams, and, when using two-stage focusing, for other experiments to test symmetry changes due to range-shortening in scaled hohlraums. Thus the IBX would enable the IRE to perform more and higher precision ion-driven target experiments.
Summary
In conclusion, the Heavy Ion Fusion program has done small scaled experiments demonstrating the validity and promise of the heavy ion concept for fusion energy production. Present experiments are investigating electron dynamics, optimization of accelerator parameters involving the highly nonlinear physics near the beam aperture, and final focus and neutralization with driver-scale. All of these experiments have been done at low energy (£ 1 MeV) in order to keep costs low. Both because of the physics which cannot be explored in the short length of such small experiments, and because of the necessity to demonstrate that all of the beam manipulations can be successively performed in an integrated experiment which produces optimal physics at every stage, the program must take the next step to an integrated source-to-target experiment with higher final kinetic energy. The Integrated Beam Experiment that we have described in this document makes this step by going to a 1-to-a-few beam accelerator/drift compression/final focus system which is significantly longer than present experiments and is a factor of 5-20 higher in final beam kinetic energy. This is a key step for heavy ion fusion, which would integrate the physics of previous experiments and introduce new capability for exploration of physics of high importance to the program. It is a necessary step in order for heavy ion fusion to proceed, and this expansion of experimental scope and complexity, with its attendant increase in budget, has been recommended by several national review committees. It can be incorporated in the program given a modest increase in funding.
Appendix 1. The Heavy Ion Driver Concept
A heavy ion fusion driver must deliver 5-8 MJ (depending on target design) of beam energy to the target in ~ 10 ns, with ion range ~ 0.03 g/cm2. The challenge to the accelerator designer is to transport, without degradation of beam quality (focusability), the large number of ions necessary to convey this much energy namely, ~1016 ions, a number which is about 6 orders of magnitude larger than that transported in, for instance, high energy physics colliders. This large amount of charge results in two major design choices which shape the driver implementation for the Heavy Ion Fusion program. First, it strongly favors the use of an induction linear accelerator (induction linac), since induction acceleration is extremely efficient (~ 48%) at high currents. And second, efficiency optimization for such high currents also means that the charge should be transported in multiple parallel beams, which can then simultaneously utilize the accelerating field produced by the induction elements.
A schematic of the resulting heavy ion fusion power plant is shown in Fig. 1 (above). The (~100-200) beams are produced by multiple ion sources, and their acceleration begins in an injector consisting of a diode and initial electrostatic acceleration. The beams then enter the accelerator. To prevent the transverse blowup of the beams due to space charge, each beam is periodically focused by superconducting quadrupole magnets. The accelerator thus consists of focusing element arrays for the multiple beam array, alternating with accelerating induction "cores" of magnetic material which encircle the beam array. The cores provide both the accelerating electric field for all beams and, by shaping of the time dependence of the field, the longitudinal focusing to keep the beam from lengthening.
At the end of the accelerator, the beams are separated into two arrays or bundles, which will be sent one to each end of the target. The beams have attained their final kinetic energy, but the beams' pulse length is ~100 ns, so that they must be compressed longitudinally by about an order of magnitude. This is accomplished by accelerating with a time dependence which produces a longitudinal velocity profile, accelerating the beam tail to higher velocities than the head. In the "drift compression" section the beam then compresses longitudinally to 10 ns pulse length. At the final focusing system the beams' space charge stagnates the longitudinal compression, and the final quadrupole lenses (note that these are clear-bore magnets-- at no time do the beams pass through material until they enter the target chamber) focus them for their transport through the ~5m of the target chamber to the target [1-3]. The target chamber [4] contains jets of hot liquid FLiBe (a molten salt of fluorine, lithium, and beryllium which is neutron-thick) to protect the walls from neutrons, target debris, and the target explosion pressure wave. The beams, propagating through channels left between the FLiBe jets, are stripped by interactions with atoms from the ambient FLiBe vapor pressure, and neutralized by electrons from this charge-exchange and also, later, by electrons from FLiBe ions photoionized by x-rays from the target. 3D time-dependent simulations show that the increase in space-charge caused by the stripping is compensated by the neutralization, so that the beam focuses to a spot size of radius equal to a few millimeters. Finally, the beams strike the converters at the ends of the target, heat them and produce x-rays which then implode the target.
The physics of the beam transport, between source and target, is dominated by the beam space charge. This is quite different from high energy and nuclear physics accelerators where the physics, except at the very low energy end, consists of single-particle dynamics. Though new collider designs and the U.S. Spallation Neutron Source design are beginning to encounter significant space charge effects, heavy ion fusion has been virtually unique in investigating the regime of space-charge-dominated ion beams. Heavy ion fusion beams act like nonneutral plasmas, with the external focusing fields acting as the "neutralizing species". These beams therefore have their own rich class of normal modes and stability questions, which are studied theoretically using the Vlasov-Maxwell equations and other plasma methods, as well as a variety of computer simulation codes, the most commonly utilized being the particle-in-cell codes.
References
[1] D. A. Callahan-Miller and M. Tabak, "A Distributed Radiator, Heavy-Ion Target Driven by Gaussian Beams in a Multibeam Illumination Geometry," Nuclear Fusion, 39, 883 (1999).
[2] D. A. Callahan-Miller and M. Tabak, "Progress in Target Physics and Design for Heavy Ion Fusion," Phys. of Plasmas, 7, 2083 (2000).
[3] M. C. Herrmann, M. Tabak, J. D. Lindl, "Ignition scaling laws and their application to capsule design," Phys. of Plasmas, 8, 2296 (2001).
[4] P.F. Peterson, "Design Methods for Thick-Liquid Protection of Inertial Fusion Chambers," Fusion Technology, 39, No. 2, pp.702-710, 2001.
**********************
The work of the Heavy Ion Fusion Virtual National Laboratory is performed under the auspices of the U.S. Department of Energy by the University of California, Lawrence Berkeley and Lawrence Livermore National Laboratories under Contract Numbers DE-AC03-76SF00098 and W-7405-Eng-48, and by the Princeton Plasma Physics Laboratory under Contract Number DE-AC02-76CH03073.