It's premature to endorse direct drive laser fusion Daniel L. Jassby
Retired from Princeton Plasma Physics Lab., dljenterp @aol.com
To accomplish nuclear fusion by “indirect drive,” a suitable energy source is first converted to X-rays in a hohlraum, and those X-rays implode a fusion fuel capsule by ablation of the capsule surface. Radiative implosion of the fusion fuel is the only terrestrial technique that to date has proved capable of igniting thermonuclear burn, whether the energy source is a multi-terajoule fission explosion [1] or megajoule laser beams [2]. By contrast, in the process of “direct drive” the laser beams themselves are focused upon a spherical fusion fuel capsule to implode it [3].
In the January 2024 Newsletter of this Forum, an article by Bodner [4] declared that indirect drive laser fusion is irrelevant to planned fusion power reactors and should be abandoned in favor of direct drive, while an article by Manheimer [5] urged that the US DOE transform the mission of a magnetic confinement laboratory to research and perfection of direct drive laser fusion, including development of the optimal laser. The authors identified the alleged fatal flaws that will eventually doom the application of indirect drive to IFE (Inertial Fusion Energy) as the inefficiency of converting laser energy to X-rays of which only a small fraction is absorbed by the fuel capsule, and the extra cost of fuel targets that employ a hohlraum.
In favoring direct drive, both Bodner and Manheimer propose programs to develop a reactor-grade excimer laser (ArF or KrF) followed by a research program to achieve thermonuclear ignition. But their proposals employ coupled speculations, as Bodner has shown elsewhere [6] that only the ArF laser has the potential to effect ignition with direct drive, principally because of its short wavelength (193 nm), broad bandwidth (10 THz), and ability to zoom. If an ArF laser of the required energy (e.g., 30 kJ per module) and pulse length (~ 3 ns) cannot be developed, then ignition with direct drive is likely impossible.
NIF Experiments With Direct Drive
The NIF (National Ignition Facility) near Livermore, California is the world’s most powerful laser-driven fusion facility, featuring a frequency-tripled Nd-glass laser that produces up to 2.2 MJ at 351 nm. While the NIF fusion program is concerned mainly with indirect drive, the project uses direct drive when it wants a “quick and dirty” neutron source for some application, because it’s cheaper and faster than setting up indirect drive.
However, the NIF laser beam configuration is far from ideal for optimizing the performance of direct drive. For indirect drive the NIF laser beam array is oriented so that beams enter the two opposite ends of a hohlraum, called the “polar” configuration. In the absence of a hohlraum there is asymmetric laser illumination of the spherical D-T fuel capsule, so the beams must be defocused to increase uniformity on the target. With direct drive NIF has produced up to 35 kJ of fusion yield with 1.2 MJ of laser energy, giving target gain Q = 0.03 [7]. This value is about 1/60th of the highest target Q reached with indirect drive on the NIF.
OMEGA Experiments with Direct Drive
The OMEGA facility at the Univ. of Rochester (URLLE) houses the second largest laser for fusion research in the US, which is also a frequency-tripled Nd-glass laser that delivers up to 35 kJ at 351 nm.
Very recently the URLLE published an extensive account of the best direct drive results with their OMEGA laser imploding a spherical D-T fuel capsule [8]. The facility produced 900 J of fusion energy with a 28 kJ laser pulse, giving Q = 0.03, which by coincidence is the same as the highest Q obtained by direct drive on the NIF with 40 times as much laser energy.
The URLLE “hydro-equivalently” scaled OMEGA results to the NIF with 2.1 MJ laser pulses and with the model using spherically symmetric laser illumination of the fuel capsule (unlike NIF’s actual polar direct drive). This extrapolation gave a fusion yield up to 1.6 MJ but did not quite ignite [9]. Nevertheless, this yield scaled from the OMEGA results was 45 times larger than the amount achieved experimentally with a 1.2 MJ laser pulse and polar direct drive.
By contrast, with indirect drive the NIF has actually ignited fuel capsules numerous times with just 250 to 300 kJ of X-rays absorbed by the capsule [10]. The fact that ignition can be achieved by X-rays with just 1/7 of the laser energy is a testimony to the greater penetrating power of the short wavelength X-rays and resistance to instability growth offered by their incoherent nature.
Laser Development
For the next decade, laser development for IFE is likely to be pursued mainly by private companies in the US with modest support from the DOE and by government labs abroad. The half-dozen private companies in this field claim that they will develop a reactor- grade laser system based on either the DPSSL (diode-pumped solid-state laser), or KrF, or ArF. This R&D may result in a superior MJ-level, nanosecond-pulse, high-rep-rate laser for direct drive. If not, then the only route to IFE is indirect drive, which can tolerate relaxed driver specifications. If the laser development is successful, an experimental test of the ability of direct drive to attain thermonuclear ignition is still 10 to 12 years away and the outcome is uncertain.
As for Manheimer’s proposal [5] to change the mission of a US DOE lab., one notes that reaching ignition or very high Q in the tokamak or stellarator is indeed as speculative as ignition with direct drive. However, the DOE and the magnetic confinement labs are convinced that the tokamak is a sure thing and in any case would not give up one speculation for another.
Increasing the Efficiency of Indirect Drive
The drawback with discarding or ignoring indirect drive is that after 75 years of controlled fusion R&D it is the only technique of a hundred fusion concepts that has demonstrated “scientific feasibility,” ie., the ability to reach thermonuclear ignition, or for magnetic confinement systems to demonstrate fusion energy gains of 10 or more.
A different approach to overcoming the objections to indirect drive is to attempt to mitigate its relative inefficiency and the extra cost associated with hohlraum targets.
There are 5 components in the NIF chain of efficiencies:
- the laser itself (0.5% electrical efficiency)
- wavelength conversion (50%)
- conversion to X-rays, including loss of laser light from the hohlraum (80%)
- fraction of X-rays lost from the hohlraum (20%)
- fraction of X-rays absorbed by the fuel capsule (15%)
Laser inefficiency is the “low-hanging fruit” and the electrical efficiency can be increased to at least 10% with either the DPSSL or a gaseous excimer laser. Developing the ArF laser system to generate repetitive MJ-level, nanosecond pulses, as proposed by Bodner and Manheimer, would be invaluable for indirect drive as well as direct drive. For indirect drive it obviates the need for frequency tripling (item #2), and will produce an electrical efficiency 40 times greater than the NIF laser system.
In present-day hohlraums, only about 15% of the laser energy is absorbed by the fuel capsule. However, the efficiency advantage of direct drive is no more than a factor of four because of reduced ablation effectiveness relative to that of X-rays, loss of laser light by backscatter exacerbated by cross-beam energy transfer, and deleterious laser- plasma instabilities [11]. Furthermore, the fraction of X-rays absorbed by the fuel capsule can be increased significantly with newly optimized hohlraum designs (item #5). For example, both rugby-shaped [12] and frustrum-shaped [13] hohlraums have been shown experimentally to increase the X-ray absorption efficiency from 15% to 30%, which reduces the advantage of direct drive to a factor of two. The indicated failure to reach ignition when the OMEGA direct drive results are “hydro-equivalently scaled” to NIF’s 2.1 MJ raises the possibility that indirect drive may eventually have the efficiency advantage if only longer wavelength lasers are available. Excimer laser drivers would presumably give better results with direct drive than the glass lasers currently used at OMEGA and NIF [11].
Particle Beam Drivers
The residual factor of two advantage of direct drive indicated above will be erased if particle beams are the energy source for indirect drive because the electrical efficiency of those beams is tens of percent, and because the hohlraum will be closed, eliminating X-ray loss (item #4). Ion beams can penetrate the thin outer surface of the hohlraum into the material that converts their energy to X-rays.
Recent advances in accelerator development suggest that affordable compact fusion drivers will become available in the foreseeable future [14]. In the case of heavy-ion beams, the first experiments using existing ion-beam facilities would not need any fuel capsule, but must show that an adequate X-ray field can be established inside a hohlraum by an ion beam impinging on the exterior. While a working hohlraum will be closed, an experimental hohlraum must have holes for diagnostic access. The intensity and spectrum of X-ray production can be monitored by time-resolved X-ray spectrometers such as deployed on the NIF.
There have not yet been experiments with particle beams and fusion-relevant targets, and no private company is proposing to pursue this route. If the DOE (other than NNSA) is to fund anything substantial in inertial fusion research, given the success of indirect drive it ought to fund particle beam experiments with hohlraum targets using existing ion accelerators or those nearing completion.
Multi-megajoule pulses are inherent to heavy-ion drivers as well as some schemes using laser drivers [15]. The extra cost of a hohlraum target can be accommodated when driver pulses are tens of MJ instead of the more commonly cited 1 or 2 MJ, so that each pulse produces at least ten times as much fusion energy. For a given averaged power output, “oversized” pulses also allow a much reduced repetition rate and correspondingly longer time to clear target debris from the reaction chamber and for the FLIBE or molten metal first wall to recover from the impulse loading.
Summary
Radiatively driven implosion via soft X-rays is uniquely successful in the fusion game and should be exploited in all its potentially useful variations. There exist practical avenues for decisively increasing its overall electrical efficiency and for reducing the relative cost of target hohlraums. Its competitor, direct drive laser fusion, shows great promise but attaining ignition is far from assured. Indirect drive should not be set aside as an energy source until it has been shown to compare unfavorably with a direct drive system that actually produces ignition in the laboratory.
References
[1] First demonstrated in the Ivy Mike shot of November 1952. See Wikipedia entry on “Ivy Mike”.
[2] A. L. Kritcher et al., “Design of the first fusion experiment to reach target energy gain G > 1,” Phys. Rev. E 109, 20521 (Feb. 2024).
[3] E. M. Campbell, et al., “Direct-drive laser fusion: status, plans and future,” Phil. Trans. R. Soc. A, 379: 2021 Jan. 25.
[4] Stephen E. Bodner, “The Irrelevance of the NIF for Fusion Power,” APS Forum on Physics and Society, Vol. 52, No. 2 (January 2024).
[5] Wallace Manheimer, “A Necessary New Approach for the American Fusion Effort,” APS Forum on Physics and Society, Vol. 52, No. 2 (January 2024).
[6] Stephen E. Bodner, “Comparison of an ArF gas laser to a solid-state laser for application to laser fusion energy,” J. Fusion Energy 42, 33 (2023). https://doi.org/10.1007/%20s10894-023-00372-w
[7] C. B. Yeamans et al, “High yield polar direct drive fusion neutron sources at the NIF,” Nucl. Fusion Vol. 61 (2021) 046031. https://doi.org/10.1088/1741-4326/abe4e6
[8] C. A. Williams, et al. “Demonstration of hot-spot fuel gain exceeding unity in direct-drive ICF implosions,” Nature Physics 05 Feb. 2024. https://doi.org/10.1038/s41567-023-02363-2
[9] V. Gopalaswamy, et al,”Demonstration of a hydrodynamically equivalent burning plasma in direct drive ICF “ Nature Physics 05 Feb. 2024. https://doi.org/10.1038/s41567-023-02361-4
[10] Jeff Tollefson, “US nuclear fusion lab enters new era,” Nature News 15 Dec. 2023. https://doi.org/10.1038/d41586-023-04045-8
[11] Andrew Schmitt and Stephen Obenschain, “The importance of laser wavelength for driving ICF targets, II, Target Design,” Physics of Plasmas 30, 012702 (2023) https://doi.org/10.1063/5.0118093
[12] Yuan Ping, et al., “Reaching 30% energy coupling efficiency in a rugby hohlraum on NIF,” LLNL report no. LLNL-CONF-826193, presented at 28th IAEA Fusion Energy Conf., May 2021.
[13] K. L. Baker, et al, “First large capsule implosions in a frustum-shaped hohlraum,” Physics of Plasmas, Vol. 30, Sept. 2023. https://doi.org/10.1063/5.0163396
[14] Thomas Schenkel et al, “Ion Beams and Inertial Fusion Energy,” presented at the IFE Strategic Planning Workshop, Feb. 23, 2022. https://lasers.llnl.gov/nif-workshops/ife-workshop-2022/white-papers
[15] Conner Galloway, et al., “ASPEN Laser and a New IFE Power Plant Concept,” presented at the IFE Strategic Planning Workshop, Feb.23, 2022. https://lasers.llnl.gov/nif-workshops/ife-workshop-2022/white-papers
Top
A Memorandum Concerning Uranium Power and Bombs Cameron Reed
Department of Physics (Emeritus) Alma College, Alma, MI, reed@alma.edu
In the Fall of 2023, I had the great pleasure of being asked by a local seniors organization to offer a short course on the Manhattan Project in response to interest generated by the Oppenheimer movie. My class was one of several into which seniors could enroll at very modest cost; no prior knowledge was assumed. The menu of courses varied from biblical history to current world events.
I had a very enthusiastic group of about 20 participants. Courses ran over six weeks, with one two-hour session per week. A course like this is always a bit of a balancing act: One wants to delve into enough of the scientific background to give a credible sense of what the scientists and engineers of the Manhattan Project were up against, but at the same time not overwhelm the audience. My group contained many retirees who had been professionals such as engineers and teachers; they were not going to let me get away with loose or shallow explanations. I followed a semi-quantitative approach supported with numerous slides, graphs, and photos.
The first couple weeks of the class offered a whirlwind tour of the scientific background: Atomic structure, radioactivity, isotopes, and on up to the discovery of fission. This brought us to the well-known story of how the Szilard-Einstein letter warning of the potentialities of nuclear energy made its way to President Roosevelt in the Fall of 1939. In discussing this, I saw an opportunity to prepare a convenient summary of the preceding background material: I imagined myself being asked to prepare a two-page, single-spaced memorandum giving a qualitative summary of the possibility of nuclear power and weapons. I imagined my reader to be a busy government official who did not have time for a seminar on nuclear physics, but I wanted to touch on all of the relevant essentials: The difference between the 235 and 238 isotopes of uranium, how they respond to bombardment by fast and slow neutrons, how plutonium might be bred and isolated, the possibility of using reactors to give submarines and aircraft carriers essentially unlimited range, and how, if large-scale enrichment techniques could be developed, immensely powerful fast-neutron bombs were a distinct possibility.
I prepared my memo as a handout for my class, and it was well-received. Ultimately, it did (barely) fit on a single double-sided page, albeit with some judicious spacing.
The text of the memorandum is reproduced below in the hope that it might be of use to instructors seeking a compact overall description of the nuclear situation as it was understood in late 1939/early 1940. I would be grateful for any feedback from interested readers.
Memorandum Concerning Uranium Power and Bombs
Executive Summary
Recent research indicates that the element uranium might serve as a source of energy or as the active explosive in extremely powerful bombs. A uranium “reactor” engine could be used to power large naval vessels and submarines, while uranium bombs could release millions of times as much energy as conventional bombs on a pound-for pound basis. We recommend that all uranium ores and uranium-bearing products be prohibited from export, that the Administration appoint an official to maintain contact with researchers, and that funding be authorized to support continued research. The following points summarize the situation.
(1) Neutron-induced fission
“Nuclear fission” is a process whereby the nucleus of a uranium atom breaks apart and releases “atomic energy” upon being struck by an incoming neutron. The probability that a nucleus will fission depends on the speed of the striking neutron, but, in essence, a “fissile” nucleus will potentially suffer fission if struck by any neutron, no matter how little energy it has; a “fissionable” nucleus needs to be struck by a neutron of a certain minimum energy to undergo fission.
Uranium ore comes in two forms: U-235 and U-238. U-238 comprises about 99.3% of natural uranium, with the remaining 0.7% being U-235. U-235 nuclei are fissile, while those of U-238 are fissionable. U-238 does, however, capture incoming neutrons and subsequently decay to a new element, plutonium. This is of potentially great value in the context of bombs as described in (4) below.
(2) Chain reactions via neutrons released in fissions
Each fission of a uranium nucleus releases 2 or 3 very energetic “fast” neutrons which can go on to precipitate a chain reaction in other U-235 nuclei. But since U-238 captures neutrons and constitutes the vast majority of natural uranium, it is impossible to create a fast-neutron chain reaction with natural-abundance uranium. This rules out a high-energy uranium bomb with natural uranium. But see point (3) below regarding the possibility of powerful U-235 bombs.
(3) Possibility of “fast-neutron” bombs
In contrast to point (2), it is predicted that an immensely destructive “fast-neutron bomb” could be possible if tens of kilograms of U-235 (or plutonium) could be isolated. The difficulty is that since nuclei of U-235 and U-238 are forms of the same element, no ordinary chemical process can be used to separate them. To achieve any separation it will be necessary to utilize a technique that takes advantage of the slight difference in the masses between the two types of nuclei. Three such techniques are currently in the nanogram-level research stage, but tens of kilograms will be required to make a bomb.
(4) Possibility of low-energy neutron power sources; plutonium “breeding”
A subtlety in the physics of fission is that the probability that a bombarding neutron will induce a fission in U-235 rises dramatically as the speed of the neutron decreases; in effect, slow-moving, low-energy neutrons have more time to cause a fission. For slow neutrons, the U-235 fission
probability is over 200 times greater than the capture probability for U-238; this is just enough to compensate for the small natural abundance of U-235 to the extent that a controllable, energy- producing chain-reaction becomes possible. In actuality, both neutron-induced fission of U-235 and neutron capture by U-238 would proceed in parallel in a slow-neutron environment; this is an important point.
To slow neutrons after they are emitted in a fission but before they strike another uranium nucleus, it will be necessary to disperse the uranium as small lumps throughout a medium which slows neutrons without capturing them. The entire assemblage is called a pile or reactor. Graphite is the most practical slowing medium now known. The neutrons’ energies from slowing and energy liberated in the fission reactions will be released as heat, which can be used to drive an engine. Such a power plant would require no fuel or oxygen, and would be ideal for use in a naval vessel such as an aircraft carrier or submarine; the range of such a vessel would be effectively unlimited.
After nuclei of U-238 capture incoming neutrons, they decay to a new element called plutonium. Laboratory tests have verified that plutonium behaves like U-235 and is in fact more fissile than U-235. An operating reactor would both provide power and “breed” a new fissile material which can be chemically extracted from the parent uranium fuel and used to power a bomb as in (3) above. It is estimated that a large-scale land-based reactor would supply enough plutonium for about a dozen bombs per year while also supplying power.
(5) Conclusions
The prospect of realizing nuclear power and weapons seems more probable than not. Competent authorities estimate that separation techniques and reactors could be developed within two to three years, but will require industrial-scale infrastructures costing hundreds of millions of dollars.
The ideas outlined here have also likely occurred to scientists in adversary countries. Any country which possesses such a source of power would have a crushing advantage over one that does not. In view of this, we recommend that both nuclear pile and fissile materials research and development be given the highest possible priority.
Top