铀立方开口：海森堡的"只差一点"，究竟差多少

Nuclear Archaeology Revisits Heisenberg's Last Reactor

文化 · 2026-10-19

📖 今日导读

1947 年，海森堡在《自然》杂志上写道：德国最后的 B8 反应堆只差"相对少量的铀"就能实现自持链式反应。八十年后，研究者用两枚幸存的天然铀立方加上计算机模型重新算了一遍——要跨过临界点，需要约两倍的铀和两倍多的重水，远超德国战时的全部库存。

TODAY'S ARTICLE

For seventy years, one sentence shaped how the world remembered a failed wartime project. In 1947, the German physicist Werner Heisenberg wrote that his last reactor experiment had failed only because it lacked "a relatively small amount of uranium." A new study, based on two surviving cubes of natural uranium, suggests that the gap was far wider than he claimed.

七十年来，一句话塑造了世人对一项失败的战时工程的记忆。1947 年，德国物理学家维尔纳·海森堡写道：他最后一次反应堆实验之所以失败，只是因为它缺少"相对少量的铀"。一项基于两枚幸存天然铀立方的新研究则表明，实际差距远比他声称的更大。

The device, the B8 pile, was built in 1945 in a cellar beneath a church in Haigerloch, southern Germany. It held 664 cubes of natural uranium, hung on chains in a tank of heavy water — water whose hydrogen atoms carry an extra neutron. Heavy water slows the neutrons released when uranium nuclei split, so they are more likely to trigger further splitting.

这台被称为 B8 堆的装置于 1945 年建成，位于德国南部海格洛赫一座教堂的地下室里。它把 664 枚天然铀立方挂在链条上，浸入一槽重水——即氢原子多带一个中子的水。铀核裂变时释放中子，重水能使这些中子减速，从而更可能引发下一次裂变。

That repeated splitting, a chain reaction, is what makes a reactor work. When the reaction becomes self-sustaining, physicists call the reactor critical. The B8 pile never reached that point; American forces found and dismantled it in 1945.

这种反复发生的裂变就是链式反应，也是反应堆能够运转的关键。当反应能够自我维持时，物理学家称该反应堆已达到临界。B8 堆从未走到那一步；1945 年，美军发现并拆解了它。

Heisenberg's claim was hard to test: almost all of the project's uranium vanished into the postwar market. Only about 14 of the roughly 1,100 cubes made for the German program are known to survive. Two reached Timothy Koeth, a materials scientist at the University of Maryland; one arrived in a lunch bag in 2013.

海森堡的说法很难检验：该项目的铀几乎全部流入了战后黑市。为德国核计划制造的约 1100 枚铀立方中，已知存世的只有约 14 枚。其中两枚到了马里兰大学材料科学家蒂莫西·科思手中；2013 年，一枚装在三明治纸袋里送到他手上。

Koeth and his colleagues measured the cubes, then reconstructed the pile in a computer model. Their model reproduced what Heisenberg had measured in 1945, which gave the team confidence. It also produced a number: as built, the reactor ran at about 0.94, well short of the 1.0 needed for a chain reaction.

科思与同事测量了这两枚铀立方，再用计算机模型重建了该反应堆。模型复现了海森堡 1945 年实测的数据，这让团队对它有信心。模型同时给出了一个数字：按实际建成的样子，反应堆只运行在约 0.94，距链式反应所需的 1.0 相差甚远。

In reactor physics, that gap is enormous. To calculate how much material would have closed it, the researchers reran the model repeatedly. They estimate that the pile would have needed roughly twice as much uranium and more than twice as much heavy water. Those amounts exceeded the entire stock Germany held — 1.2 times its uranium, 1.8 times its heavy water.

在反应堆物理中，这个差距极其巨大。为了计算填平它需要多少材料，研究者反复运行模型。他们估计，该堆需要约两倍的铀和两倍以上的重水。而这些数量已超出德国全部库存——约为其铀总量的 1.2 倍、重水总量的 1.8 倍。

Could German scientists have substituted graphite for heavy water, as the Americans did? The team examined the graphite available inside German-controlled territory. It contained traces of boron, an impurity that absorbs neutrons. American reactors used graphite made from low-boron petroleum coke; Germany, cut off from that, relied on metallurgical coke. No practical design would have been sufficient.

那么，德国科学家能否像美国人那样，用石墨替代重水？团队考察了德国控制区内可得的石墨。它含有微量硼——一种会吸收中子的杂质。美国反应堆用的是由低硼石油焦制成的石墨；德国失去这一供应，只能用冶金焦。用这种石墨，任何现实的设计都不够用。

The study does not guess what Heisenberg intended, nor whether one more delivery of fuel would have changed the war. It measures what his reactor required, and the conclusion is blunt: Germany lacked the materials for a heavy-water reactor, and for a graphite one too. Eight decades later, two grey cubes have given a quieter answer than the memoir did.

这项研究并不猜测海森堡的意图，也不主张再来一批燃料就能改变战争走向。它衡量的是他的反应堆究竟需要什么，而结论很直白：德国既缺少重水反应堆所需的材料，也缺少石墨反应堆所需的材料。八十年后，两枚灰色立方给出了比回忆录更平实的答案。

生词已用 蓝色加粗 标出，对照下方词汇表食用更佳。

📌 词汇注释 VOCABULARY

survive | /səˈvaɪv/ | v. 幸存；存世 | two surviving cubes of natural uranium / known to survive

claim | /kleɪm/ | v. 声称；n. 主张 | wider than he claimed

self-sustaining | /ˌself səˈsteɪnɪŋ/ | adj. 自我维持的 | the reaction becomes self-sustaining

critical | /ˈkrɪtɪkl/ | adj. 临界的；关键的 | physicists call the reactor critical

reconstruct | /ˌriːkənˈstrʌkt/ | v. 重建；复原 | reconstructed the pile in a computer model

calculate | /ˈkælkjuleɪt/ | v. 计算 | to calculate how much material would have closed it

estimate | /ˈestɪmeɪt/ | v. 估计 | they estimate that the pile would have needed

entire | /ɪnˈtaɪə(r)/ | adj. 全部的；整个的 | exceeded the entire stock Germany held

substitute | /ˈsʌbstɪtjuːt/ | v. 替代 | have substituted graphite for heavy water

impurity | /ɪmˈpjʊərəti/ | n. 杂质 | boron, an impurity that absorbs neutrons

sufficient | /səˈfɪʃnt/ | adj. 足够的 | no practical design would have been sufficient

— 每日一句 DAILY QUOTE —

"A relatively small amount of uranium would in all probability have sufficed."

—— 维尔纳·海森堡，1947 年发表于《自然》（Nature）