Restricted Data

Today marks the 75th anniversary of the Trinity test. My thoughts on the world’s first nuclear explosion haven’t changed too much over the last five years, so you can read that if you want my “anniversary take.” The only thing that’s really changed in my thinking since then is that I’ve was able to visit the site last summer, and spent some time in the area around it (Socorro, Alamagordo, etc.). It’s beautiful country out there, and compared to my normal (NYC-adjacent) environs, it still feels pretty isolated and sparse. The distances are just large in such a place — it takes you a long time to get anywhere, and driving 50 miles is no big thing. Anyway, I’ll have some opportunity to post a bit more about that trip later in the summer, I believe, so I’ll hold off for now.

The obelisk at the Trinity site, July 2019. Source: Photo by author.

What I’ve been thinking about lately is a question I have been asked a few times in the ramp-up to this year’s anniversary: What is the importance of the Trinity test? I’ve found it surprisingly difficult to answer more than superficially. Of course, I can easily explain the context of what the test was, why it was done, and what followed it. But “importance” implies, to my mind, a counterfactual: that something different would have happened, historically, if it had not occurred, or occurred differently.

I’ve written on counterfactuals before, and I suspect I find them more interesting than many historians. The “official historian response” to counterfactuals is to say, “well, we really can’t know what might have happened, since it is hard enough to know what did happen,” and I can co-sign that. But counterfactual questions can be a way to focus on why we think something was important, and it can be a useful way to think through what we do know about the past. So I often find them to be useful exercises, so long as you don’t put too much stock in their “reality.” (And I was always a fan of Marvel’s What If? series, for a less intellectual justification.)

So my question today is: What if the Trinity test had failed? 

Modes of failure

One of the reasons this feels like a somewhat jarring question is that we slide easily from “this is how history happened” to “this is how history must have happened.” We know the test was a “success,” and that colors everything we think about it and its preparations. But the chances of Trinity failing in one way or another were not all that low. There’s a lot that could have gone wrong with it.

Even after Trinity, J. Robert Oppenheimer estimated relatively high chances of the “combat unit” failing:

The possibilities of a less than optimal performance of the Little Boy are quite small and should be ignored. The possibility that the first combat plutonium Fat Man will give a less than optimal performance is about twelve percent. There is about a six percent chance that the energy release will be under five thousand tons, and about a two percent chance that it will be under one thousand tons. It should not be much less than one thousand tons unless there is an actual malfunctioning of some of the components.

It is probably not desirable to attempt at destination to establish, on a statistical basis, the reliability of Fat Man Components. On the other hand, it is desirable to subject the components scheduled for hot use to inspection and testing with the greatest care. 1

Oppenheimer’s estimates are high-enough given the stakes of the test, but the big question is un-estimated part: the failure of a component. Because the “Gadget,” and its weaponized form, Fat Man, had a lot of components. And they were all capable of failure. The implosion design required a lot of things to work just right, in order to to get the simultaneous detonation (within a tolerance measured in nanoseconds) and correct shaping of the compressive forces that symmetrically shrunk the solid-metal plutonium core to over half its original volume. This is why they were having the Trinity test in the first place: they didn’t know if it could be done at all, and even if it could be done, they didn’t know how well it would work. 2

An official diagram of the “Gadget,” snug in its casing as the Fat Man bomb. This incredible image comes from a manual that John Coster-Mullen received under the Freedom of Information Act, and the overall document describes all of the many activities that have to be done just right to make one of these fire correctly. There’s a lot that could go wrong; these were not “GI-proof.” Source: “Maintenance and Instruction Manual for Mark III Atomic Bomb Fuze,” Project Y, January 1946, courtesy of the intrepid John Coster-Mullen.

So let’s imagine three possible modes of “failure” for the test. The first is not really a failure at all: that “the Gadget” had gone off with the yield that it had been expected to have, prior to testing. This was around 4-5 kilotons, not the 20 kilotons it turned out to be. So that would have not been considered a failure by the scientists, but it would have made the plutonium bomb considerably smaller than their projections for the uranium bomb. We’ll round that up to 5 kilotons for simplicity’s sake. 

A second possibility could be a result at the low-end of what they thought was plausible: a few hundred tons of TNT equivalent. Let’s say 500 tons, just to pick a number. That means that the Gadget would be seriously underperforming (their goal was at least a kiloton), but still a usable weapon. It would not meet their stated criteria for a usable atomic bomb (they had set that at 1 kiloton), but it would still be something you could drop on an enemy. 3

And an ignominious third possibility would be a total failure, a “fizzle” of zero nuclear yield.  This would be the true component failure that Oppenheimer mentioned: a problem with the detonation system, or a major flaw with the lens system. There would be about 5 tons of TNT equivalent result from the high explosive system, which would destroy the “Gadget” and scatter its plutonium. Again, this is not implausible at all — this was a new weapon, and these components were custom-made, and every technical system has a rate of failure. The scientists knew this was a real possibility; some of them even bet on it!

Prior to the Trinity test, the scientists had considered the (uranium-fueled) Little Boy bomb to be their “big” bomb, because they had always assumed it would be capable of hitting at least 5 kilotons, and probably around 15 kilotons. The (plutonium-fueled) Fat Man was guessed to have a yield that would range from a few hundred tons up to 5 kilotons. The advantage of the Fat Man design was that you could make more of them: their production rates, in late 1945, were about half of a Little Boy bomb’s worth of enriched uranium per month, compared to three Fat Man bombs worth of plutonium per month. So they saw their probable future capabilities as one big atomic bomb every couple months, with a few smaller atomic bombs in between. The actual Trinity test revealed that the Fat Man bombs were in fact more powerful than they had expected the Little Boy bombs to be.

A relevant excerpt from the notes of the second meeting of the Target Committee, from May 1945, which describe the range of expected yields at that time for the two weapons. 4

The technical implications for these different modes of failure, as I see it, are fairly straightforward. If the Trinity test was as powerful as they expected it to be (5 kt), then it would not have changed much about how they saw the situation — 5 kilotons is still nothing to scoff at. If instead it had been on the low end (500 tons), then that would have been disappointing, but still within the realm of possibility, and I don’t think they would have done too much different, technically, other than try and figure out what the issue had been that had resulted in the reduced yield.

But if the test had totally failed to give a nuclear yield — I think they would have had to do another test. That was certainly their original plan in the event of a failure. That would have taken several weeks to prepare, at best. The original Trinity test had taken months to prepare, but a Trinity failure would still have done damage to the tower, and probably contaminated the site with plutonium. A “quick and dirty” Trinity test, where they just set one off somewhere, wouldn’t have given them the data they needed (and in the face of a total failure, I don’t see them thinking “quick and dirty” would be the way to go — especially given how precious their plutonium stockpile was). So I think that would have essentially set back the possibility of using a plutonium bomb on Japan by a month or so at the minimum. 

What if the cause had been something more nefarious, like intentional sabotage? (An idea floated in the sadly-cancelled Manhattan television show that I consulted for some years back.) I think this would have been the “total failure” outcome, plus a lot more security, hand-wringing, and paranoia. Which is to say: something very different, and not something I feel confident at all predicting, but a really interesting question.

Strategic and diplomatic implications

The success of the Trinity test told the US policymakers and military planners that atomic bombs worked, and that they would have a fair number of them over time. Both of these would have been challenged to different degrees by a failure.

A 5 kt Trinity probably wouldn’t have changed that much. Again, this was the expectation. It might have changed how the bombs were deployed, though. A 5 kiloton explosion would do roughly 40% as much damage as 15 kiloton one. 5 To put that in terms of raw effect, if you detonated a 5 kiloton bomb over Hiroshima today at the ideal blast height, you’d kill maybe 47,000 people (according to NUKEMAP), as compared to over 80,000 people for 15 kilotons. That’s still a pretty powerful weapon. But it would be conspicuously less powerful than the Little Boy bomb. So it’s possible they might have imagined using it for purposes other than destroying entire cities, such as targeting specific military bases. But overall I think this is probably “close enough” that their existing assumptions would still have been likely maintained.

NUKEMAP screenshots of the effects of three blasts over Hiroshima, each set at the ideal height of burst for maximizing the range of the 5 psi blast range for their yield: 15,000 tons of TNT (15 kt), 5,000 tons of TNT (5 kt), and 500 tons of TNT (0.5 kt). From largest to smallest (roughly, as the effects scale differently over this range), the rings show: 1 psi blast pressure (very light gray), the maximum distance for 3rd degree burns from the thermal radiation (orange), the area of 5 psi blast damage (gray), the area of 500 rem ionizing radiation exposure in free air (green), and size of the fireball (yellow). Source: NUKEMAP / Map data © OpenStreetMap Contributors, CC-BY-SA, Imagery © Mapbox.

A 500 ton Trinity, on the other hand, is much less powerful than they had wanted the atomic bomb to be. It would still kill 15,000 people if dropped on Hiroshima today. But much of the city would still be intact — the psychological effect would be far diminished. And compared to the 15 kiloton bomb, it would have looked relatively paltry. Again, I think they might still have considered using it, but I don’t see them “wasting” any their precious-few “reserved” targets on it — they’d be saving those for the big bombs, and using the smaller atomic bombs for other purposes. So I think this would really shake up how they saw their arsenal and their use of it.

It’s also possible that depending on how well they thought they understood the failure, that it might impact their sensibilities on the Little Boy bomb as well. The scientists had high confidence that the gun-type design would work, and it was easier to confirm the principles behind it without a full-scale test. Would their confidence have been shaken? If their diagnostics of the Trinity test told them that the detonator system had worked as planned, then they might have worried that their deeper understanding of a fission bomb was incorrect. But if they thought it was just an assembly problem — something unique to the implosion design — then they’d probably have still been confident about the gun-type arrangement. 

But the policymakers and military brass would probably have been a lot less confident. Outside of Groves, none of the other military leaders had a deep understanding of the bomb, and several expressed extreme pessimism about its prospects prior to Trinity. A Trinity failure would have reinforced these perspectives. It’s possible they might have judged the entire thing not ready for prime time, and scuttled any use plans until they were confident that it wouldn’t be an embarrassment.

And a failed Trinity would, as noted, probably mean that they would have extreme delays in their plutonium bomb capabilities. I think they’d still want to use the uranium bomb as soon as possible. But they’d know that they would not be able to follow it up with more attacks for some time. Maybe they’d try to bluff about that, or maybe they’d just downplay how much destruction they’d be delivering that way, I don’t know. But I think they’d consider it a pretty different situation.

Stalin, Truman, and Churchill at the Potsdam Conference on the day after Trinity (July 17, 1945). Source: Harry S. Truman Presidential Library.

What political implications would this have? Truman received the news of the successful Trinity test with great excitement, and it bolstered his confidence greatly with regards to the end of the war. I suspect that wouldn’t have changed with a 5 kiloton result, but with a 500 ton result, that would probably have been diminished. With a total failure, I think the opposite would have occurred: he might have gotten even more depressed about dealing with the Soviet Union, and with the prospects of an unconditional surrender by the Japanese.

So what might he done differently if that was the case? The two places I can see him modifying his approach would be in his dealings with Stalin, and the question of unconditional surrender and the Potsdam Declaration.

After getting the successful results from Trinity, Truman took a very hard line with Stalin. He believed that the bomb gave him leverage for both the end of World War II and the peace that would follow. Though he did not try to argue that the Soviets should not declare war on Japan or stop their invasion plans, he was less convinced he would need the Soviet entry into the war, and did not encourage them. Without the confidence from Trinity, would he have pushed so hard? I’m not sure he would have; he might have felt the Soviet invasion too necessary for the end of the war to risk alienation. And if he had taken a more compromising approach, what would the impact of that had been on the later Cold War to follow? The Cold War was a complex thing, not the result of a single interaction, but there are scholars who have attributed some of its formation and angst to Truman’s post-Trinity bravado, so it’s not outside the realm of contemplation.

On the Potsdam Declaration, there were members of Truman’s cabinet and the military staff who were trying to put a sentence into the Allied dictum towards Japan that would clarify the position of the Emperor. They knew, from intercepted Japanese communications, that this was a sticking point for even those members of the Japanese high command who were interested in pursuing an end to the war. But Truman, with the pushing of his Secretary of State, James Byrnes, pushed back hard on this, and deliberately did not make things “easier” for the Japanese in this respect. One can argue whether that was the right thing to do or not (that is a separate question), but would he have taken such a hard-line with Japan if he didn’t have positive news from Trinity? Again, I doubt it — he would have been less sure of his own position, and may have listened to those who cautioned him towards moderation.

Would they have still used the Little Boy bomb on the original schedule, to be dropped just after the Postdam Conference ended? I think this depends on what they thought the nature of the failure was for the Trinity test — if it was something particular to the implosion device specifically, then I think they would have continued as planned. If it was something that caused them more fundamental doubts, then I they might have waited until those doubts could be resolved. I don’t think their messaging on the atomic bomb would be significantly affected, however; Truman’s declaration at Hiroshima is basically a “one bomb” announcement (and I am not sure he realized that more bombs would be coming soon afterwards). 

War outcomes

Would a failure at Trinity have changed the outcome of World War II in any significant way? Ultimately this question relies on what you think caused the Japanese to offer up a conditional surrender on August 10th, and then an unconditional surrender on August 14th. In particular, it depends on how important you think the Nagasaki bombing was for the decision-making of the Japanese high command. 

There were a number of overlapping factors that contributed to Japanese surrender, and it is not always clear how much weight to assign each of them. The bombing of Hiroshima, the Soviet declaration of war and invasion of Manchuria, the bombing of Nagasaki, the US rejection of the conditional surrender offer, an abortive coup by junior Japanese officers, and an intensification of conventional bombing all took place over the space of less than a week. If we imagine the only thing missing from that list was Nagasaki, does it matter? Or if Nagasaki had used a much less damaging atomic bomb, would it matter?

On the left, the wispy remains of the mushroom cloud rise over the Trinity test site in New Mexico (and one of my favorite, unusual photographs of the test — one of the few that emphasize Trinity’s cloud). At right, a photograph I took just outside the White Sands Missile Range in summer 2019 — not exactly the same vantage point (I suspect the original was taken on a high ridge or from a plane), but pretty close. Source (left): Los Alamos National Laboratory, TR-239.

While it is hard to disentangle these things, the bombing of Nagasaki clearly left less of a distinct impression on the Japanese command than the other factors. It may have reinforced the feeling of hopelessness that lead to the surrender offers, but it’s not clear it was necessary for them. The Japanese do not seem to have seriously doubted that the US had only one atomic bomb, and in any event, the other factors — such as the Soviet invasion — clearly weighed very largely on their minds. 

So it’s not implausible to suggest that the war would have ended around when it did, even without Trinity being successful. That being said, the Japanese decision to surrender was not over-determined in any way. If the Japanese don’t offer up a surrender of some sort by August 10th, what then? Do they offer it up before a second Trinity can be tested, or a second gun-type bomb is ready? Here we end up in the weeds of speculation, beyond what we can know with much confidence. Or to put it another way, here is where whatever preconceptions you have about the end of the war will dominate: if you’re a “bombs did it” sort of person, then you’ll favor that kind of interpretation, and if you’re a “Soviets did it” sort of person, you’ll favor that one. I’m on the fence, though I lean towards the interpretations that show Nagasaki as not being that important, so I see the war ending around the same time it did anyway. This is a separate question from, “what if no atomic bombs were dropped on Japan,” which is even more contentious.

If that’s the case, then the Trinity test was not so important for that particular end, though those diplomatic decisions might have had long-term consequences. One thing is clear, though: if Trinity had failed, we’d talk about the Manhattan Project in a different light. It wouldn’t seem so “inevitable” that something “Manhattan Project-like” would succeed, and perhaps we wouldn’t be so quick to deploy allusions to it when talking about big science projects. 

The irony of this whole discussion is that the Trinity test is almost always framed as an important and impressive achievement. But what if it only meant that Nagasaki was avoided, and the war still ended? In that light, maybe it would have been a better thing if it hadn’t had worked so well. To say that feels heretical, and I’m not sure I believe it. But it’s a provocative idea.

Either way, it’s easy to conclude, I think, that the Trinity test “mattered,” at least by the counterfactual criteria set out at the beginning: if it had failed, we’d have ended up in a somewhat different world. The interesting question to ask is whether that different world would be better or worse than the one live in now — and that is surprisingly unclear.

1. J. Robert Oppenheimer to Thomas Farrell and W.S. Parsons (23 July 1945), Nuclear Testing Archive, NV0103571.[↩]
2. Update, September 2020: I happened to stumble across a document in the Digital National Security Archive in which Oppenheimer was expressing his reservations about the Operation Crossroads tests to President Truman. This line seemed relevant: “Even if all components work correctly, the bombs which are scheduled for use in these tests have a chance of about one in fifteen of giving an ineffective explosion, that one might call a dud. Bombs have been designed which do not have this weakness, but it is not planned to use them.” The Crossroads bombs were of the same model as the Nagasaki and Trinity designs. J. Robert Oppenheimer to Harry S. Truman (3 May 1946), Digital National Security Archive, George Washington University, document NP00024.[↩]
3. By May 1944, the directive for the Manhattan Project was a weapon of a minimum 1,000 tons of TNT equivalent. William S. Parsons to Leslie R. Groves (19 May 1944), Correspondence (“Top Secret”) of the Manhattan Engineer District, 1942-1946, microfilm publication M1109 (Washington, D.C.: National Archives and Records Administration, 1980), Roll 1, Target 6, Folder 5, “Events Preceding and Following the Dropping of the First Atomic Bombs at Hiroshima and Nagasaki,” Subfile 5F, “Memorandums from (Capt. W. S.) Parsons.”[↩]
4. J.A. Derry and N.F. Ramsey to L.R. Groves, “Summary of Target Committee Meetings on 10 and 11 May 1945,” in Correspondence (“Top Secret”) of the Manhattan Engineer District, 1942-1946, microfilm publication M1109 (Washington, D.C.: National Archives and Records Administration, 1980), Roll 1, Target 6, Folder 5D, “Selection of Targets.”[↩]
5. Blast damage scales as a cubic root (as a power of W^1/3, where W is the yield), not linearly. So to double the blast effect, you need to increase the yield by a factor of about eight. Thus a weapon of 15 kilotons is not three times more damaging than a 5 kiloton weapon, even though it releases three times as much energy.[↩]

Tags: 1940s, Bomb design, Counterfactuals, Harry S. Truman, J. Robert Oppenheimer, Manhattan Project, Trinity, Truman

This entry was posted on Thursday, July 16th, 2020 at 9:07 am and is filed under Meditations. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.

Citation: Alex Wellerstein, "What if the Trinity test had failed?," Restricted Data: The Nuclear Secrecy Blog, July 16, 2020, accessed April 24, 2024, https://blog.nuclearsecrecy.com/2020/07/16/what-if-the-trinity-test-had-failed/.