….was it a fluke?
I think there are some category errors here:
In the vast majority of mechanical inventions, there have been thousands of trials at a component level, hundreds of partial (e.g. static tests of a rocket in which the engine is run but the rocket is not actually flown) or complete trials of a full system. It usually involves many attempts before a full system such as an atomic bomb actually works. Mechanical inventions that work right the first time are clearly the exception in the history of invention and discovery. Some possible exceptions are Tesla’s alternating current motor (if Tesla is to be believed), the atomic bomb, and the first flight of the Space Shuttle. Inventions that work right the first time do appear to occur, but they are rare, exceptions, outliers, flukes. They probably should not be treated as typical or likely for planning purposes or investment decisions.
…projects that succeed on essentially the first attempt are rare; in this, the Manhattan Project is quite unusual. Yet, this success of the Manhattan Project has greatly helped fund scientific R&D megaprojects that implicitly assume that the full system will work on the first try or with only a few attempts, something that is historically rare. Full scale systems like the ITER tokamak, particle accelerators like the Large Hadron Collider (LHC), and so forth are both extremely expensive and each trial of the full system is likely to cost anywhere from millions to billions of dollars. Thus, one hundred full system trials, perhaps a more realistic planning number, implies vast costs. Not surprisingly, many scientific megaprojects like the NASA Ares/Constellation program recently or the Super Conducting Supercollider (SSC) have foundered in a sea of rising costs.
Ignoring the fact that Constellation wasn’t a “science” project, one of these things is not like the other. Rocketry isn’t really rocket science any more. If you consider Falcon 1 a “training rocket” for SpaceX, consider that Falcon 9 worked almost without a hitch the first time (the only issue was the upper-stage roll), and Dragon worked the first time. If you do enough simulations, it is in fact possible to get it right the first time (though Shuttle had a pretty bad first flight — I’ve learned recently that Young and Crippen actuallywould have considered ejecting due to concern about the body flap damage from overpressure, had they known about it). The problem with vertical takeoff expendable rockets is that they pretty much have to work the first time, or at least in as few a number of tries as possible, because tests are expensive, and they’re not possible to incrementally test. I can’t emphasize enough what a breakthrough the new reusable suborbital vehicles are going to be, in their ability to incrementally test and do gradual envelope expansion. But in the context of incremental development and testing, I’m not sure what “work the first time” even means.
I would also note that Constellation’s problems were cooked in from the beginning, given what an awful design concept, and incompetent management it had. Combine that with the pork aspects, and its failure was inevitable, as many (including me) predicted at the time.
Anyway, despite the mixing of apples and oranges, it’s an interesting, albeit long, read.
The problem is when folks think of the Manhattan Project they forget the delivery system, the B-29, which was plaque with problems.
The reason of course is simplicity. For all its advance physics a nuclear device is rather simple in operation. Like the lighting rod was for Benjamin Franklin. Whereas aircraft and spacecraft are complex with many failure points.
The Missile Defense Agency has had a couple failed tests in the past year on their long range interceptor system. Despite a lot of modeling and simulation work, the interceptors failed to hit their targets, apparently for different reasons. Mod & Sim are great but they only get you so far. Real world tests are very expensive but unavoidable. Even then, problems are often not detected until the system enters daily service (e.g. DC-10 and A-380).
Yep, and the more complex the system, the more problems are in hiding waiting to bite you. Apollo 1 and 13 were both good example of that.
The STS 1 flight was a near-disaster in a number of areas. The shockwave from the solids igniting actually bent or broke several of the struts holding the shuttle to the ET. Several tiles were lost and others severely damaed, resulting in hot gasses entering one of the wheel wells. But I’d never heard that Young and Crippen actually considered bailing out of the shuttle on the way down (ejecting during ascent would have almost certainly killed them both as they would have likely passed through the exhaust trail of the solids and been incinerated). The closest thing I could find online was this comment:
“The same overpressure wave pushed the body flap below the main engines at the rear of the shuttle well past the point where damage to the hydraulic system would be expected, which would have made a safe re-entry impossible. The crew were unaware of this until after the flight, and John Young reportedly said that if they had been aware of the potential damage at the time, they would have flown the shuttle up to a safe altitude and ejected, causing Columbia to have been lost on the first flight.”
So he reportedly said they WOULD have considered ejecting if they had known about the overpressure, but since they weren’t told about it, the went ahead and landed, as it turned out, successfully.
Did your source tell you differently?
No, I probably misremembered what I had been told. That sounds like the story that I heard. I’ll update the post.
It seens kind of unfocussed for a mathematician. I think he’s a little off about the nature of the MED Project. The use of numerical calculations was surely very important, but they also did loads of empirical tests, blowing up charges in various canyons, “tickling the dragon’s tail” with subcritical assemblies, and so forth. The gun design was extremely conservative and almost guaranteed to work (and it should be pointed out its fission efficiency was absurdly low, I recall something like less than 5% of the U-235 fissed). The implosion was another story, but they did test it, at Trinity, and through lots of partial tests at Los Alamos.
My impression is that the use of massive mathematical calculation to do fancy nuclear designs start to finish didn’t come in until the more intensive development efforts after the war, and particularly for the Super, and (1) they did plenty of tests, and (2) some of those tests were fizzles.
I think he’s 100% right that breakthroughs always depend on serendipitous discovery, of enabling technology, unexplained and surprising observattions, and so on. That’s even true of The Bomb. Until Hahn discovered (much to everyone’s astonishment) traces of barium when he bombarded uranium with neutrons in 1938, no one knew nuclear fission existed. So far as anyone knew, the only source of nuclear power was fusion — and the technological difficulties of duplicating the Sun’s core were an obvious stumbling block. Probably they thought fusion power lay far in the future, e.g. 1960 or so. So the breakthrough — the enabling observation — occured long before the MED Project got started. My impression is that there were only minor technological breakthroughs in the Project itself, e.g. the idea of using merging jets instead of a smooth shock wave to compress the pit. Indeed, it was expected that a bomb would require only modest engineering innovations, as otherwise it would not have been ready by the end of the war, and the money could have been put to better use. (This was, after all, what the Germans decided, in part because they were confused about the critical mass of U-235 and thought a breakthrough was needed to reduce it, or produce it. They also didn’t know about Pu-239 of course.)
Indeed, one can readily argue that a breakthrough is by definition unforseeable — if you could map out a plan to get there, it wouldn’t be a breakthrough, right? So that means it’s inherently unpredictable, a real Black Sawn.
But I don’t fully agree that means there is no correlation between breakthroughs and the amount of men, money and time you throw at a problem. Where the needle lies in the haystack, or even whether it’s there at all, may well be unpredictable. But without doubt the more people who are looking, the sooner it will be found. Even if all you’re doing is a random walk through idea space, you’ll find something interesting sooner the more walkers you have.
they’re not possible to incrementally test
Not quite true, you could start with just a first stage, maybe even a battleship first stage.
Interesting point that at least one of the H-bomb tests was the exact opposite of a fizzle. The second one, at Bikini, was much more powerful than anticipated.
“I would also note that Constellation’s problems were cooked in from the beginning, given what an awful design concept, and incompetent management it had. Combine that with the pork aspects, and its failure was inevitable, as many (including me) predicted at the time.”
Of course, if one substitutes “SLS” for “Constellation” in the above statement, as Yogi Berra put it, “It’s deja vu all over again.” Especially if one considers certain politicians in the role of “incompetent management” (as in politically enforced design specs) while also enabling “the pork aspect”. We don’t have Griffin to blame this time for the management problem.
“The second one, at Bikini, was much more powerful than anticipated.”
If I remember my readings correctly, the blast itself was not anymore powerful than the first one (the bombs themselves were identical). It was the huge difference in the extent of damage produced by underwater detonation that was unexpected.
Quite right, Fletch. Castle Bravo was something like twice design yield, IIRC. I believe all the accidental overexposure in the Pacific — the Japanese fishing boat, the native Pacific Islanders — was from this shot.
There’s only one possible conclusion: The success of nuclear fission is an authentic miracle. God wants us to use nukes.
The only similar miracle I can think of is the Eiffel Tower, built on schedule and within budget.
I note that there are shelves of alt-history/fantasy books to be written about different outcomes of Rudolf II’s alchemical (etc) R&D program …
I take it his point is that Big Science is not a productive as widely believed. I agree with this but calling the Manhattan Project a fluke misses the point.
The Manhattan Project was successful for at least 3 reasons that are not found in modern big science efforts:
1) They had well defined and bounded goal. It was stated to every scientist who arrived:
They knew what the goal was. The goal wasn’t some vague mission statement about shortening the war or using physics to help out. The goal was clear. Compare this with the modern big science goals of “developing renewable energy sources” or even “curing cancer.”
2) As Carl says, they knew a lot about how to get from where they were to where they wanted to go. The people working on the project knew what needed to be done to achieve the well defined goal. Much of big science these days is just a vague urge followed with the hope that by developing component technologies and software models something awesome will happen.
3) The people working on the Manhattan project knew they were working on something important and urgent. This means they would be more willing to make compromises and suborn their egos to the leadership for the bigger good. I can’t imagine anyone thinking that of, say, an ARPA-E project.
The Manhattan Project was successful not because it was a fluke but because it had the elements that can make big science successful. Modern big science is frequently unsuccessful – in my opinion – because it is frequently missing two or more often all three of the above elements.
@TL, I think you may be talking about the fission bomb tests, Able and Baker http://en.wikipedia.org/wiki/Operation_Crossroads#Nicknames, whilst Fletcher Christian is talking about the fusion tests. I saw a program the other night “The biggest bomb in the world” which talked about the Castle Bravo http://en.wikipedia.org/wiki/Castle_Bravo test, which was three times expected yield, due to lithium-6 and lithium-7 being present. The neutron cross-section in one of the isotopes was much larger than the other, and this caused the much larger yield.
A classic example of learning more from doing the experiment I guess…
Oddly myopic.
What the heck qualifies something as a “breakthrough”? I find it moderately disturbing that a physicist dismisses modern particle physics, as defined by experiments at large accelerators, as a failure. In 1974, physicists at Brookhaven and Stanford discovered the charmed quark, which completely revolutionized our understanding of high-energy physics. (In fact we refer to it as the “November Revolution”.) It is not an exaggeration to say that our understanding of the foundation of physics, as exemplified in the Standard Model, is due to the events of 1974. Maybe this hasn’t filtered down to the “quark cellphone”, but from an intellectual standpoint, it is a landmark that will be remembered for ages.
And of course, one has to wonder: why no mention of the moon landing? Yes, there were a lot of steps along the way, and yes, some people died in the Apollo 1 fire, but Apollo 11 really was our first attempt to “land a man on the moon and return him safely to Earth” — and it worked the first time (though not the third!). Anyone who thinks mathematics played little role in that endeavor is advised to browse through the reports at NTRS.
One of my pet peeves is the grouping together of mathematics and computational methods. These are different disciplines. Mathematics is always “true”, in a Platonic sense: if you do the math correctly, by definition you will get the right answer to the problem (although it may not be the problem you need to solve!) Simulation, on the other hand, is fraught with hidden traps, and I say that as a computational physicist who has fought his way through some pretty thick forests. Not only does simulation bear all the burden of mathematics in showing that the right equations are being solved, but also the coder must demonstrate that the algorithms being used to solve the equations work correctly, and that the algorithms are being implemented correctly, and that systematic errors are controlled and minimized. So, sure, I am always skeptical of the utility of simulations, but try designing a launch vehicle, or a nuclear power plant, or even a cell-phone antenna, without mathematics and I think you won’t get very far.
I think he has a point about the diminishing returns of big science — although I think that Hubble and some of the other big astronomy projects have profoundly affected our understanding of the universe. But the only big science projects I can think of that have the potential to Change Everything are LHC and the Human Genome Project. I think it’s too early to tell if HGP will provide big paybacks but I think optimism is justified. And the LHC has not yet solved the riddle where mass comes from in the Standard Model (the Supercollider would have been a better tool) — but I think we cannot imagine what the world will be like in 100 years if we can learn to manipulate mass at a subatomic level.
But there are also “this changes everything” innovations that happen at much smaller scale. Do you remember the first time you used the world-wide web, or a graphical web browser? (In my case those events were separated by about a year!) (Whatever happened to Marc Andreessen?) I remember the first time I noticed a banner in the end zone of a football game reading “www.nfl.com” — and thinking, hmm, this internet thing seems to be catching on….
Speaking of nuclear weapons: Happy Threadsay, everyone.
Thought-provoking article. The comments here are thoughtful, as well. They did a lot of testing in the Manhattan Project that gets little air time, such as the Ra-La Tests which helped solve the extremely difficult problems associated with implosion assembly. I still think Hiroshima was the gutsiest move in military history. No amount of testing that stops short of a fission explosion would really prove the bomb would work. It was a gun assembly device that used U-235. The only bomb that had ever worked was an implosion device using Pu-239. The two had nothing in common, except the radars, firing units, and maybe (it’s still probably classified) the trigger.
The real breakthough, though, was the discovery of the fissile proprties of uranium and plutonium. What followed is engineering. Same with Apollo. Once the rocket equation was known, and chemical rockets found to have sufficient exhaust velocity, the rest was engineering. It helped that both had well-defined goals, too.
MfK: The real breakthough, though, was the discovery of the fissile proprties of uranium and plutonium. What followed is engineering. Same with Apollo.
I agree that the breakthrough for the atom bomb was the discovery of fission, but there was a lot of basic physics that had to be worked out in the few years between the discovery of fission and the dropping of the bomb. Neutron cross-sections as a function of energy; multiplication factors; energy transport phenomena during the chain reaction; the physics and chemistry of the enrichment process; and many other pieces of the puzzle. The line between physics and engineering is sometimes a big gray area, but none of these pieces could be found in the engineering handbooks of the time, right? The team of physicists and engineers assembled in Los Alamos and Oak Ridge had to work out a lot of these details through a synthesis of theoretical and experimental efforts that certainly resembles what we call science.
And I know Rand protests the promiscuous use of “rocket scientist” in the context of modern space exploration. I think there were more actual rocket scientists in the 1960’s than today. But in my own small corner I consider a good deal of what I do as “science” (and the rest is engineering, of course).
There’s only one possible conclusion: The success of nuclear fission is an authentic miracle. God wants us to use nukes.
That’s two conclusions.
Of course, the possibility exists that He could have given the clues and sat back to see what we, through our flawed political minds, did with it — whether or not Truman knew ahead of time that Japan clearly wanted to surrender and the deaths of tens of thousands of Japanese was to send a message to Moscow is pretty well documented.
It’s hard to ignore the number of technological advances that were lab accidents.