"technically so sweet"
J. Robert Oppenheimer, 1951

Radiation Implosion

In a paper prepared in February 1950, Teller outlined the Classical Super, and the alternate and more clearly feasible layered design, the Alarm Clock. But the yield of the Alarm Clock was limited to about 1 MT, in the ballpark of what could be achieved with fission weapons, and Teller was focused on greater megatonnage.

In Los Alamos, Ulam and Everett made further calculations. They came up with results of improved quality, which, however, were negative on the feasibility of their approach to H-bombs. There were incomplete and inaccurate mathematical treatments of the "Super", an idea first proposed by Teller and Fermi, and one of the most difficult problems in physics of the postwar period.

Ulam undertook the important task of determining more accurately the amount of T required. His results were spectacular: the amount was calculated to be quite large. More detailed and thorough calculations by other members of the Theoretical Division of Los Alamos confirmed Ulams estimates. These results were entirely opposite to the 1946 assumptions, and made the economic soundness of the H-bomb program highly questionable.

Hansen noted that "By the fall of 1950, LASL estimated that the Super would require three to five kilograms of tritium, at a time when the entire national tritium stockpile was measured in tens of grams and production of just one kilogram of tritium per year would require the construction of many new nuclear reactors. The reactor power required to make one gram of tritium was equivalent to that required to make 70 to 80 grams of plutonium, during a period when a typical moderate-yield atomic bomb required two to three kilograms of plutonium and an equivalent amount of uranium-235. An unproven hydrogen bomb that required three to five kilograms of tritium would in effect "cost" the national weapons stockpile between 70 and 200 proven atomic bombs; however, if a workable hydrogen weapon were developed, it would more than offset this loss of atomic bombs."

In the summer of 1950, Fermi and Ulam showed by an approximate calculation that there can probably be sustained reaction at all in pure D. This conclusion, which was contrary to the assumptions of 1946, was strengthened by the drop in experimental cross sections resulting from the accurate measurements of Tuck and his group in 1951. The calculations of Fermi and Ulam, however, were not definitive, and the final decision about the feasibility of a thermonuclear reaction in liquid deuterium came when a full-scale machine calculation on this problem was carried out which took into account all important physical processes. Such a calculation was prepared, and even though the reaction turned out to be feasible, it remained impractical and uneconomical.

In June 1950 new calculations by Stanislaw Ulam and Cornelius Everett, and by Ulam, Fermi and Garwin, showed that Teller's Classical Super design could not work. Bethe concluded, after additional calculations by Von Neumann and by Ulam, that the classical Super is dead. The main problem is that radiation carried off the needed heat too fast. And the steadily-increasing tritium requirements finally doomed the Classical Super. Up to that time Teller had dismissed the value of compression as a solution to the thermonuclear problem.

In December 1950, Ulam had suggested the use of energy from one fission bomb to compress another fission bomb. This idea was conceived entirely independently of the thermonuclear program, and its aim was to use fissionable materials more economically. He called the arrangement "hydrodynamic lensing." Ulam proposed compression implosion via a shock wave in a neutron fluid (thus "hydrodynamic") from fission primary. This required staging: a primary bomb would set off a second, physically separate bomb. In January 1951, Ulam saw a way to apply his "iterative" scheme to the thermonuclear problem. Ulam's breakthrough was the combination of compression and staging.

In spite of the apparent failure of the program, it was decided in the Fall of 1950 to proceed with the planned thermonuclear experiment at Eniwetok in the Spring of 1951. This experiment [which proved fully successful] was designed primarily to confirm one proposition of 1946, the burning of D-T, about which there had never been serious doubt. In addition the experiment was to try out one of several possible mechanisms which might be used to provide initial ignition if the latter should turn out to be feasible. In this particular mechanism, the energy was conducted by radiation from a fission bomb to the T-D, and the radiation was used not only to heat but also to compress the T-D.

The accidental choice of the radiation scheme, however, proved fortunate because it led to a theoretical consideration of thermonuclear reactions at high densities, as well as of the propagation of radiation. The former line of work demonstrated that high densities lower the "ignition temperature" of a T-D mixture and thus make the reaction more efficient. Since the T-D reaction occurs easily anyway, this was perhaps not very remarkable. However, after several months, it occurred to Teller to make a bold extrapolation.

By the end of 1950, Teller had the novel and positive answer. Liquid deuterium is much more readily compressed, and such compression would make thermonuclear reactions possible after the energy transfers had reached equilibrium. Compression was far from original, and had been suggested by various people on innumerable occasions in the past. Because of the wartime work, it was known how to strongly compress the thermonuclear fuel, and, in the compressed fuel, radiation would be less important and would not inhibit the reaction.

According to Teller, the Los Alamos administration discouraged new approaches, so for the time being Teller restricted himself to a few private discussions. One of these occurred between Ulam and Teller in February 1951. Ulam suggested compression, for which Teller was fully prepared. Teller knew how to accomplish it -- via radiation, rather than neutrons -- and Teller knew how it would help. Teller and Ulam put all this down in a joint report LAMS-1225 "On heterocatalytic detonations 1: hydrodynamic lenses and radiation mirrors" , dated 09 March 1951.

In this paper Teller added the critical concept of using X-Rays as the source of radiation compression of the thermonuclear secondary, rather than neutron fluid. This radiation implosion compression made high yield with reasonable size device possible. Teller pushed Ulam aside afterwards, and refused to deal with him. Ulam failed to support this concept in subsequent discussions.

The "three concepts" of modern thermonuclear design were thus

1. Separate stages -- a physically separate fission explosive (stage) and a capsule (stage) of thermonuclear fuel, centered at separate points.
2. Radiation coupling -- channeling (ducting) of thermoradiation from the first stage to ignite the second stage.
3. Compression -- implosion of the thermonuclear fuel capsule prior to ignition to achieve maximum yield.

On 04 April 1951 Teller placed his signature on LAMS-1230 report, which presented additional calculations related to the feasibility study of a new superbomb performed by Frederic de Hoffmann. This report also added a second fission component, at the core of the thermonuclear material. This "sparkplug" consisted of subcritical U235 that would be compressed to criticality. The outward push from this second explosion and the initial implosion reach equilibrium, sustaining a hot compressed layer burning deuterium efficiently. Thus the term "an equilibrium hydrogen bomb." Tritium was generated in situ and no external supply was required.

In the radiation implosion concept, the fusion fuel is first compressed. This makes possible the ignition and effective burn of fusion fuel. Radiation is channeled from the exploding primary fission bomb to the secondary fusion device, causing it to implode and compress its fusion fuel.

Ulam’s and Teller’s radically new approach involving staged radiation implosion and the success of the George shot in the Greenhouse series in demonstrating a thermonuclear reaction now made a workable weapon likely. What was “a tortured thing that you could well argue did not make a great deal of sense,” as J. Robert Oppenheimer observed, became by mid-1951 “technically so sweet that you could not argue about that.”

Only Teller's persistent belief in the practicality of thermonuclear reactions led to this completely novel concept. It would be a most remarkable coincidence if the Russian project had taken a similar course, and they had not. It was immediately clear to all the scientists concerned that Teller's new suggestion provided for the first time a firm basis for a thermonuclear program. Without hesitation, Los Alamos adopted the new program.

The fantastic requirement on calculation imposed by an attempt to explore the question of the classical Super as envisaged in 1946 did not, of course, apply to the same extent with respect to the thermonuclear devices in the form considered since early 1951; but even those requirementa still far exceeded the ones which had to be met for the successful design of fission weapons.

By the time (1951) when the thermonuclear program began to emerge, the log-jam in computing resources was rapidly breaking. There was a period in 1952 when the Los Alamos MANIAC, a model of the UNIVAC in Philadelphia, and the SEAC in Washington were all engaged essentially full-time on Loa Alamos (and Matterhorn) calculations for the new thermonuclear program.

The General Advisory Committee [GAC] a nine-man committee, established in 1947 and chaired by Oppenheimer until 1952, held a meeting on this subject in Princeton in the middle of June 1951. This meeting was also attended by the members and the manager of the AEC and by a considerable number of Staff Members and consultants of the Los Alamos Laboratory. Teller took the opportunity of an SAB meeting (spring 1951) to present the plan for "an equilibrium hydrogen bomb," in which compressed fuel would be used. Teller gained the unanimous support of the SAB.

The meeting was unanimously in favor of active and rapid pursuit of work on the device with a test to be prepared as soon as it was clear what exactly was to be tested. At that time, progress at Los Alamos had been assured, and Teller felt that it would be better to start work at an additional laboratory. This possibility materialized (with the essential help of Ernest Lawrence) in Livermore.