STM-D-1116Paper2008Designed, not yet built

Ignition of a Deuterium Micro-Detonation with a Gigavolt Super Marx Generator

Friedwardt Winterberg

Abstract and summary · read the original at the source · none found

In one page

Two experiments bracket the problem Friedwardt Winterberg is trying to solve. The Centurion-Halite tests showed that a deuterium-tritium micro-explosion can be lit with no more than a few megajoules. The 1952 Mike test showed that pure deuterium can be made to burn — but what lit it was a fission bomb. Winterberg wants the second without the first, because pure deuterium comes out of ordinary water and needs no lithium, and his answer is voltage. An ordinary Marx generator charges a bank of capacitors in parallel and then switches them into series, adding their voltages. Winterberg stacks that trick twice: a hundred conventional Marx banks charge a hundred magnetically insulated coaxial capacitors, which are then switched in series to reach a billion volts. That drives a ten-million-amp proton beam into the end of a thin, compressed deuterium rod, and the beam’s own magnetic field traps the charged fusion products inside the rod, so a detonation wave runs down its length — with the yield set by how long you make the rod.

Why it matters hereChapter 12 asks what it would actually take to make a clean fusion burn go without a fission trigger, and this paper names the machine, the voltage, the current, the rod and the length of tunnel it would occupy. Chapter 9 gets the mechanism that keeps recurring across this site: a current large enough to hold its own reaction products in place with its own magnetic field. The result that would settle the whole line is a single pure-deuterium micro-detonation ignited without fission.

What it claims

01The problem is bracketed by two experiments and by the fuel. The Centurion-Halite experiment demonstrated ignition of a deuterium-tritium micro-explosion with no more than a few megajoules; the Mike test demonstrated a pure deuterium explosion above 10⁶ megajoules — and in both cases a fission bomb supplied the ignition energy. Winterberg wants pure deuterium because deuterium-tritium depends on lithium and puts about 80 percent of its energy into neutrons that activate the reactor, whereas the raw material for deuterium is ordinary water, everywhere abundant, and the reaction releases much less energy into neutrons. The price is a far higher ignition energy.Abstract; Section 1, Introduction; Section 2, Solution in between two extremes, Fig. 1
Settled physics
02The central move replaces a sphere with a rod. For deuterium-tritium the condition for propagating burn in a sphere is a density-radius product of at least 1 gram per square centimetre; for deuterium-deuterium it is at least 10, and with an optimal ignition temperature about ten times higher the ignition energy for a deuterium sphere works out around 10⁴ megajoules — beyond any laser or particle beam. But at a beam current of about 10⁷ amperes the beam’s own magnetic field entraps the charged fusion products inside a thin deuterium cylinder, so the sphere condition is replaced by the same product taken along the length of the rod, and the gain then depends only on how long the rod is.Section 2, equations 1 to 3
Designed, not yet built
03The ignition numbers close. Single GeV protons would pass straight through, but an intense beam is stopped by the proton-deuteron two-stream instability sharpened by a collisionless magnetohydrodynamic shock: in hundredfold-compressed deuterium, with a number density about 5 by 10²⁴ per cubic centimetre, the stopping length is about 1.2 by 10⁻² centimetres. A rod length of about 0.6 centimetres then satisfies the burn condition, and the ignition energy comes out at about 10¹⁶ erg — one gigajoule or less — which is what the 10⁷ ampere gigavolt proton beam delivers in 10⁻⁷ seconds. At thousandfold compression, already reached in laser fusion experiments, the ignition energy is ten times less.Section 2, equations 4 and 5
Designed, not yet built
04The reason the voltage has to be a gigavolt is a pair of settled results. For a capacitor of a given size, the energy stored and the power released both go as the square of the voltage. And a beam is easy to focus only while its current stays below the Alfvén limit, which is 17 kiloamps for electrons but 31 megamps for protons — so a 10⁸ ampere electron beam from a 10 megavolt Marx generator is hopelessly above the limit, while a 10⁷ ampere proton beam accelerated through a billion volts sits below it and carries 10¹⁶ watts, which is enough to ignite deuterium.Section 3, The Importance of High Voltages for Inertial Confinement Fusion, equations 6 to 10
Settled physics
05The machine is specified, not gestured at. About 100 coaxial capacitors, each 15 metres long and built as a magnetically insulated transmission line, are charged by two conventional Marx banks each to plus and minus 5 megavolts, decoupled, and then switched in series by spark gaps to produce 1 gigavolt along a 1.5 kilometre evacuated vessel. Breakdown is held off by magnetic insulation — the condition being that the magnetic field in gauss exceeds 300 times the electric field in volts per centimetre, so an axial field of 2 by 10⁴ gauss insulates up to 9 by 10⁷ volts per centimetre — and the current pulse from the closing spark gaps raises an azimuthal field that insulates further as it runs down the line.Section 4, Super Marx Generator; Figures 3 to 6
Designed, not yet built
06What to watch is whether anyone builds it, and Winterberg is explicit about why he thinks the alternatives will not carry a power plant: the photon burst from a high-gain micro-explosion destroys the laser that lit it, heavy-ion beams have a stopping problem in the target, and deuterium-tritium sends 80 percent of its yield into 14 MeV neutrons. He sets the neutron-free hydrogen-boron reaction aside on the grounds that it cannot be ignited at realistically attainable pressures, which leaves pure deuterium. The stand-off problem is already answered in the design — the stiffness of a GeV proton beam lets the target sit a comparatively large distance from the wall of the confining cavity — so the open item is the machine, and the result that would settle it is one pure-deuterium micro-detonation ignited without fission.Abstract, closing sentence; Conclusion
What to watch

Read it · abstract

Abstract

The Centurion–Halite experiment demonstrated the feasibility of igniting a deuterium–tritium micro-explosion with an energy of not more than a few megajoule, and the Mike test the feasibility of a pure deuterium explosion with an energy of more than 10⁶ megajoule. In both cases the ignition energy was supplied by a fission bomb explosive. While an energy of a few megajoule, to be released in the time required of less than 10⁻⁹ sec, can be supplied by lasers and intense particle beams, this is not enough to ignite a pure deuterium explosion. Because the deuterium-tritium reaction depends on the availability of lithium, the non-fission ignition of a pure deuterium fusion reaction would be highly desirable. It is shown that this goal can conceivably be reached with a “super Marx generator”, where a large number of “ordinary” Marx generators charge (magnetically insulated) fast high voltage capacitors of a second stage Marx generator, called a “super Marx generator”, ultimately reaching gigavolt potentials with an energy output in excess of 100 megajoule. An intense 10⁷ Ampere-GeV proton beam drawn from a “super Marx generator” can ignite a deuterium thermonuclear detonation wave in a compressed deuterium cylinder, where the strong magnetic field of the proton beam entraps the charged fusion reaction products inside the cylinder. In solving the stand-off problem, the stiffness of a GeV proton beam permits to place the deuterium target at a comparatively large distance from the wall of a cavity confining the deuterium micro-explosion.

Friedwardt Winterberg, Ignition of a Deuterium Micro-Detonation with a Gigavolt Super Marx Generator, Journal of Fusion Energy 28, pages 290 to 295, published online 18 December 2008 and in print in September 2009. The published article is at doi.org/10.1007/s10894-008-9189-3; the author’s preprint of the same paper, from the University of Nevada, Reno, is at arxiv.org/abs/0812.0394.

(Abstract only — the abstract above is the author’s own, from his preprint; no other part of the paper is reproduced here. See the rights note.)

On this site, the 1968 Physical Review paper in which Winterberg first proposed reaching inertial confinement by high-voltage pulse power rather than by lasers — reference 1 of this paper — is at /library/stm-b7211e4006, his 1975 treatment of dense thermonuclear plasmas made by intense ion beams is at /library/stm-ed9e1e7165, and the metastable super-explosives he proposed the same year as this paper are at /library/stm-31be696254. The Defense Intelligence Agency reference document that carries the same architecture into a spacecraft — magnetic insulation, magnetic entrapment of the fusion products, and a Super Marx generator as the ground test — is at /library/stm-c75846c5a6.

The way in

https://doi.org/10.1007/s10894-008-9189-3THE YEAR. The article appeared online on 18 December 2008 and in print in the September 2009 issue — Journal of Fusion Energy volume 28, issue 3, pages 290 to 295. The year field carries 2008, the online date, because that is the date the work entered the record; the print issue is 2009. THE PUBLISHED VERSION WAS NOT REACHED. Springer holds it closed, Crossref carries only Springer’s text-and-data-mining terms, which are not a reuse licence, and on 2026-09-08 Unpaywall and OpenAlex both reported open access status closed with no repository copy. WHAT WAS READ. The author’s own preprint of the same paper, arXiv:0812.0394v1, submitted 1 December 2008 and dated October 2008 on its title page, University of Nevada, Reno — 14 pages, 7 figures — was fetched and read in full on 2026-09-08. The abstract reproduced below is that preprint’s abstract, which is the author’s own, and every claim is located to a numbered section or equation of the preprint. Nothing beyond the abstract is reproduced, because arXiv’s default deposit licence is not a Creative Commons licence and does not permit republication of the body. TEXT FIDELITY. The preprint’s abstract as extracted from the PDF flattened its exponents, writing 10 6 megajoule, 10-9 sec and 107 Ampere; those three are restored below as superscripts, the spaced en dashes in Centurion – Halite and deuterium – tritium are closed up, and nothing else is altered. Readers should note one wording difference between the two deposits: the arXiv listing page carries non-fusion ignition where the PDF and the sense of the argument have non-fission ignition; the PDF reading is used.

How to cite it

Friedwardt Winterberg (2008) Ignition of a Deuterium Micro-Detonation with a Gigavolt Super Marx Generator. doi:10.1007/s10894-008-9189-3

Where it sits in the curriculum

Lattice confinement fusion Plasmoids, charge clusters and the orbs