Consider the impact of a single nuclear explosion – temperatures higher than the surface of the Sun vaporize everything within tens to hundreds of meters; pressure and shock waves rippling outward at the speed of sound shred buildings; and fissioning uranium or plutonium atoms produce huge amounts of radioactivity. There have been over 2000 nuclear weapons tests, ranging from sub-kiloton to over 50 megatons in yield, since 1945 – more than 1000 of these were American and most of the rest by the Soviet Union.  I’ve spoken to people who witnessed nuclear weapons tests – both atmospheric and below-ground – and they all describe something that sounds more like a force of nature than a man-made event. There is general agreement that banning nuclear testing is a good thing – but it might be interesting to look at why that is, and to speculate about what purpose such testing might serve.

At this point I should state explicitly that I am agnostic on the subject of nuclear weapons testing – I can see good points in favor of testing, just as I can see the attraction of a ban – of saying that there will never again be a nuclear explosion on the face of the Earth (or in its depths). As you read this, please keep this in mind so that, if you feel I am making an argument either for or against nuclear weapons testing, you can attribute this to my writing rather than to my preferences! That being said, on with this posting….

The US never bothered to test the nuclear weapon design that was dropped on Hiroshima – the gun-type uranium weapon was considered so fool-proof that there simply wasn’t a need to waste precious U-235 by running a test. But the implosion design used a few days later over Nagasaki was much more complex – that design was tested before it was used in combat to assure ourselves that it would work as designed and, as we all know, the test worked perfectly. In the early days of nuclear weapons we ran tests for two main reasons – to test new weapon designs and to learn about their effects. The first sort of testing is how we learned whether or not the ever-more advanced designs would work – nuclear weapons testing helped us design miniaturized nuclear weapons that can be packed a dozen to a missile, high-yield tritium-boosted and thermonuclear devices, low-yield tactical weapons, and more. Without this testing we would not have had the confidence to trust our national security to a lump of uranium or plutonium. And at the same time, this testing told the Soviet Union that our nuclear weapons worked – that we could indeed rain destruction on them as promised. Without this testing, for all we (and our foes) knew, our nuclear umbrella might just not hold water.

The other reason for the testing – especially in the early years – was simply to see what these weapons would do. How big was the fireball, and what happened to everything inside? How intense was the pressure wave and how did it affect structures? How much radiation was emitted and how far away was it fatal? How did different types of structures hold up and what did this mean for our population in the event of a nuclear war? There were hundreds of questions that needed answered so that we could build our war plans, as well as our civil defense plans, and the only way to answer them was to set off nuclear weapons to see what happened.

Over time, the reasons for nuclear testing changed. The physics of a nuclear weapon design doesn’t change over a half-century – 20 kg of plutonium acts the same now as it did in 1962, and high explosives compress a sphere of plutonium the same as it did then as well. But even if the physics stayed the same the weapons were changing – what happens (for example) to a plutonium warhead after the plutonium has gone through a few decades of radioactive decay? Will the built-up decay products alter the weapon’s characteristics and, if so, will this affect the way we can use them? What about radiation damage to the components of the weapons – will a few decades of alpha, x-ray, and neutron radiation affect the working of the high explosives or electronics? It’s easy to make a guess, but it’s nice to confirm these guesses with an actual test – one of my professors used to say that a beautiful theory could be ruined by a single ugly observation – no matter how good our mathematical models might be, running an actual test is the ground truth.

Having said that, our weapons scientists have created some fantastically sophisticated software to run on the fastest computers in the world. With this software they can run amazingly detailed simulations to try to understand how our nuclear weapons will behave under a variety of conditions. We can calculate to a high degree of precision what should happen under any of a number of conditions – the question is whether or not we feel comfortable with the results of these calculations. And please note – I am not saying that I trust these calculations implicitly and neither am I saying that I don’t trust them. Truth be told, I don’t know enough to have an informed opinion on the matter so anything I can say on this particular matter is opinion, but not necessarily any better an opinion than yours.

So – say we are trying to develop a new and improved nuclear weapon. We can run sophisticated computer models detailing the behavior of the proposed weapon in minute detail, the characteristics of the explosion unfolding nano-second by nano-second in computer memory. Similarly, we can build detailed models of how a plutonium core behaves with a decade’s worth of decay products built up, or with two decades’ worth for that matter. But even the best of our models contain simplifying assumptions – they are an approximation of reality, but only an approximation. No matter how good our models are, there’s nothing like a real-live test to prove (or disprove) our assertions. This is both the single best reason for – and against – nuclear testing.

Say, for example, there was a flaw in the implosion-type nuclear weapon dropped on Nagasaki and we hadn’t tested the design. Instead of convincing the Japanese that we could bomb them at will – a city per bomb – we’d have simply contaminated their city with plutonium. Not exactly the impact we would have hoped for. And during the Cold War – when our nation’s security was legitimately at stake – it made sense to test our nuclear weapons to assure ourselves (and our foes) that our nuclear deterrent was functional.

But what about today? Do we still need to test our nukes?

Well…maybe not, but it also depends. First, we are no longer in the Cold War and we can’t really pretend that a credible nuclear deterrent is all that stands between us and our enemy’s global domination. Sure – Russia still has enough nuclear weapons to do more than put a dint in our lifestyle, but I remember the fears we had during the Cold War and we are nowhere close to that level today. And outside of Russia there is really no other nation that poses a nuclear threat to our nation’s existence. So the threat we once faced – nuclear annihilation at the hands of an implacable foe – no longer exists. When our existence is at stake then we had better be able to know that we can trust our weapons to work, but do we still have this imperative when our existence is not at stake?

Consider – what would really happen if one of our nuclear warheads isn’t quite able to work as designed? Well – one question to ask is who we expect to use nuclear weapons against and how important it is to prove to ourselves (and to them) that they’ll work as advertised. For the first part – who DO we expect to use our nuclear weapons against? Iran? North Korea? China? Pakistan? Russia? Honestly, I can’t think of too many plausible scenarios that would call for us to actually use a nuclear weapon – like a house with a “Beware of Dog” sign, our nuclear weapons seem intended more as a threat than anything we intend to use. But if we aren’t expecting to actually use our nuclear weapons, do we need to test them, or is the threat that they pose enough? For example – say that the risk of nuclear retaliation is the only thing keeping North Korea from selling a working nuclear weapon to a terrorist group. Would North Korea take the risk that our nuclear weapons don’t work and sell something to an al Qaeda affiliate? Would they risk their nation’s existence on the supposition that our weapons are too old to work properly? Or would they assume that, even without testing, we are still capable of destroying Pyongyang? Any rational nation would almost certainly assume that we know what we’re doing because the downside of a wrong guess would likely be fatal. Even without testing it’s likely that our enemies will assume our weapons work because the consequences of a mistake are simply too dire.

On the other hand, our nuclear weapons are part of our arsenal – whether we expect to use them or not, shouldn’t we make sure we can count on them if we do need to use them? After all, what’s the sense in having something in our arsenal if we’re not sure that it’ll work if we need it? The threat of nuclear retaliation is pretty sobering – but it would be sort of embarrassing if we were to drop a bomb that fizzled. And the longer we go without testing our stockpile, the more uncertainty creeps in as to whether or not they’ll work properly.

At the same time, much of the world has signed on to a treaty banning nuclear weapons testing and there is a general feeling that banning nuclear tests is a good thing. Even if we want to prove that our nuclear deterrent works, do we want to risk international opprobrium by joining North Korea in testing nuclear weapons?

This is one where I don’t have an answer – there are great arguments for and against testing nuclear weapons. I’d hate to argue that we should defy the rest of the world (except for North Korea) and have a test, but I’d also hate to argue that we should just assume – without proof – that an important component of our national defense should be assumed to work without proof. No matter what we do, there’ll be grounds for argument.

Tags: ,

5 Responses to “Testing?”

  1. George Chandler October 15, 2012 at 12:10 PM #

    Sort of like assembling a race car and rolling it onto the starting line without ever having driven it even around the block. You ran the engine on a test bed and you balanced the wheels and you greased the suspension. You know all the parts work and it ought to work … .

  2. Roger Williams October 15, 2012 at 4:51 PM #

    Without nuclear testing we would not have the foresight or confidence to design propulsion systems and we all should know what happens when humanity loses it sense for exploration.

    • Mark L. October 16, 2012 at 11:30 AM #

      Are you referring to the Orion drive?

  3. Casey Thormahlen October 16, 2012 at 5:31 PM #

    There is no defense-based reason for continued testing of nuclear weapons. Once Teller developed the hydrogen bomb, nuclear weapons became a deterrent device. The weapon is psychological and the reality of its power is irrelevant so long as people believe it works.

    The specifics of the test ban are the issue. There is real value in considering peaceful uses of the explosive power of nuclear weapons. Programs like PACER, ORION, or Operation PLOWSHARE.


  1. North Korea’s latest test | ScienceWonk - February 14, 2013

    [...] why in the world would they risk more condemnation and sanctions by testing a third one? There are good reasons for nuclear weapons testing – here are a few that might apply in this [...]

Leave a Reply