Выбрать главу

In the report, the subject of “Compression” is itself compressed to just five pages. This is because so much of the work on compressing the thermonuclear fuel was done in Los Alamos. Our focus at Matterhorn was burning (occupying 73 pages in the report). As I described earlier, my efforts on the SEAC dealt with the so-called steady-state burning in the late stages of the explosion. These calculations gained the code name Swordtail (I no longer remember the reason). At the same time, Larry Wilets was pursuing calculations on Philadelphia’s UNIVAC concerned with earlier stages of thermonuclear burning. His calculations, code-named Chief (after the Santa Fe Railroad train, if I remember correctly) sought to answer the questions: Does the thermonuclear fuel ignite? Does a flame start to propagate? Chief could tell whether an explosion was likely. Only Swordtail could predict how much energy would be released (the “yield”).

There were some 25 scientists and mathematicians (all male) who contributed for times short and long to Matterhorn. In his final report, Wheeler put the average number at any one time as about thirteen (see, for instance, the 1952 picture on page 141). Exclusive of the support staff, the project’s backbone—those who worked for fifteen months or more—consisted of just eight physicists: Wheeler, Toll, and I, plus Walter Aron, David Carter, Pierre Noyes, Ralph Pennington, and Larry Wilets. Like soldiers in a platoon, we formed a close-knit group, and worked cooperatively toward a common goal, with little if any thought of competition or personal eminence. And, like soldiers, once our service ended, we mustered out and went on to other careers. As it turned out, those careers were almost all academic. Only a few of the twenty-five scientific staff members did later weapons-related work (mainly at the new Livermore Lab in California). Nevertheless, Wheeler dubbed us the “U.S. Nuclear Physics Reserve,” able, because of our experience and clearability, to contribute further in case of national emergency.{11}

Matterhorn was, above all, focused. Focused on thermonuclear burning. Apart from one small but encouraging test at Greenhouse George in May 1951, thermonuclear burning existed on Earth only in the minds of theoretical physicists and in their notebooks and computer codes. Its big test was scheduled for Ivy Mike late in 1952. To be sure, we theorized about other configurations and about radiation flow and compression, but mostly it was all burning all the time.

The author at Project Matterhorn in 1952, preferring then, as now, to use a fountain pen.

One topic we left entirely to our colleagues at Los Alamos. That was the design of Mike’s fission-bomb trigger, the device that would provide the flood of radiation leading to the implosion of the cylinder of deuterium. According to various reports, including that of the independent analyst Carey Sublette,{12} the choice for this fission bomb was the TX-5,[82] a spherical implosion bomb similar to, but smaller than, the Hiroshima bomb (smaller in size, not necessarily in yield). It had been used successfully in the earlier Greenhouse test series (and, according to public reports, became the basis for deployed fission weapons for many years thereafter).{14} The main requirement for the fission trigger was that it not surprise. It had to predictably do what was expected of it. There were enough uncertainties in the rest of the Mike device.

Yet, surprisingly, the design did surprise Marshall Rosenbluth. In a 2003 interview,{15} he says that because he had been focusing on the thermonuclear part of Mike, he had not studied the fission primary. But this bright and energetic twenty-four-year-old wanted to understand all facets of the design.

And so once before the test I went down and talked to the primary design group [at the Los Alamos Lab] and I found to my horror that in order to make sure it worked they had packed in so much plutonium it was sure to pre-detonate. So in fact a separate primary was flown out to the Marshall Islands a week or two before the shot.[83]

Rosenbluth’s well-founded fear was that the TX-5 core, champing at the bit to go critical, would do just that before it reached full compression and would then explode anemically (by nuclear-weapon standards), bathing the sausage in much less radiation than was needed. Harold Agnew, a Los Alamos scientist who was later to become the lab’s director, went as far as to say, “Marshall may have saved the Mike shot.”{13} (To embellish this assertion, Agnew tells a story about Rosenbluth eating too much shrimp at a sumptuous dinner on Enewetak, not being able to sleep, and worrying about the fission core in his insomnia. A nice story, but Rosenbluth was not at Enewetak.{15})

Even after the design of Mike was frozen in mid-1952, we at Matterhorn could continue to refine our calculations of its performance and its yield. At the same time, the small army of experimental physicists and engineers who intended to take full advantage of the microseconds available to them during the blast could continue to work away, revising the deployment of their wires and mirrors and electronic counters.

When October 1952 arrived, John Wheeler left for the Pacific, where he would observe the Mike test from a ship some twenty-five miles from the blast. He promised to report back to us. Edward Teller stayed in Berkeley but had his own plan for “observing” the shot.

Chapter 16

More than a Boy

There’s another number that got stuck in my head—7 megatons. The sequence of events in Mike was quick, extremely quick, but it was a sequence, one thing after another. First the explosion of the fission trigger, then the flow of radiation, then the implosion of the “sausage,” and finally the thermonuclear burning and the secondary fission. The last part of this sequence is what I calculated for many weeks of overnights on the SEAC. From all of this flowed a single number, 7 megatons, our predicted energy release, or “yield,” of Mike. As it turned out, Mike flexed its muscles and did better than that.

Enewetak, like other atolls in the Pacific, is a chain of islands surrounding a large lagoon.{1} Located in the Marshall Islands about 2,700 miles west-southwest of Hawaii, Enewetak consists of forty islands (formerly forty-one) lined up in an oval some fifty miles around. Mike was only one (the largest one) of 43 nuclear explosions detonated there between 1948 and 1958.[84] Had Elugelab, one of Enewetak’s islands, had a voice in its selection as the place where Mike would be built, it might have opted out. After Mike’s explosion, a crater more than a mile across and 164 feet deep occupied the spot where Elugelab had been.{3}

The northern part of the Enewetak atoll, before and after Mike.
Satellite imagery courtesy of Los Alamos National Laboratory Archives.

To witness Mike, John Wheeler joined Task Force 132, a joint military-civilian operation responsible for preparing, detonating, and measuring the two “shots” that made up Operation Ivy.{4} The first Ivy test was Mike, on which we had worked so hard, fired just after dawn on November 1.[85] The second was King, a record-setting fission bomb fired two weeks later. Preparatory work at Enewetak had begun more than six months earlier, in March 1952. The task force, established well before that (on July 8, 1951), moved its headquarters to Parry Island in Enewetak on September 17, 1952, just six weeks before Mike, and closed down that base of operations on November 21, 1952, a little more than a month after King.{4}

вернуться

82

Also called TX-V, according to Richard Rhodes.{13}

вернуться

83

According to Rhodes’s even more dramatic account, the new core arrive at Enewetak “a few days before the target date.”{13}

вернуться

84

From 1946 to 1958, there were an additional two dozen tests on Bikini atoll (two hundred miles east of Enewetak and also part of the Marshall Islands). A long history of relocation and resettlement of native people on the islands is a dismal chapter in American history.{2}

вернуться

85

To be exact, it was at 7:14:59 local time—a hundredth of a second before 7:15 a.m.—on November 1, 1952, or 19:14:59 on October 31 expressed in what was then called Greenwich Mean Time and is now called Universal Time, or Zulu Time.{3}