A cool collection of vacuum tubes.
So, you know how General Kevin Chilton, STRATCOM Commander, and others have been making to case for the “Reliable Replacement Warhead” by arguing that we must replace the vacuum tubes (more) remaining in the nuclear stockpile?
Kingston Reif and I discovered and explain in the Bulletin of the Atomic Scientists that is a load of crap.
Firstly, vacuum tubes are not used in the physics package of a single nuclear weapon design. Vacuum tubes are used only in the radar-fuse, which tells the firing system when the bomb is at the correct altitude for detonation, in some modifications (mods) of one warhead design, the B61 gravity bomb. In total, the B61 bombs that have vacuum tubes in their radar-fuses account for only about one in ten operationally deployed warheads. (Vacuum tubes are used in the radars of three B61 mods: 3, 4, and 7. Mods 10 and 11 have newer radars that use solid-state electronics.) The fuses in these weapons are old, but perfectly functional. To reiterate, vacuum tubes are not in use in any other warhead design, including the W76 warhead, a portion of which would be replaced by the first RRW warhead, the WR1, if it ever were funded and developed.
Secondly, the Energy Department has routinely replaced radars without nuclear testing or redesigning the physics package. In fact, during the 1990s, Sandia National Laboratories scientists developed the MC4033 common radar, which uses solid-state electronics, for planned refurbishments of the B61 and B83 gravity bombs. All B83 bombs now use the common radar, though similar plans to fit a new radar on all B61s have been repeatedly deferred.
Vacuum tubes are used only in some older mods of one design (the B61) in … wait for it … the radar-fuse! Vacuum tubes are not used for any other purpose or in any other design. They are not used in the physics package.
The implication is that radar fuses can — and are — routinely replaced without yield testing or redesigning the physics package. In other words, vacuum tubes have jack-taco to do with the RRW.
Alan Tomlinson | April 23, 2009
Those military spec vacuum tubes are also pretty damn reliable.
More importantly, at the risk of irritating some, what possible productive purpose is facilitated by using a B-61, or any fission or fusion device?
I know, it’s off-topic.
Cheers,
Alan Tomlinson
yousaf | April 23, 2009
Jeffrey,
as an aside, vacuum tubes (if functional) are generally more resilient to EMP than microelectronics. There may be good reasons to not replace them all with ICs.
This was, in fact, emphasized in the letter response to Chilton’s OpEd.
Hairs | April 23, 2009
Jeffrey, I’ve read the linked article and it seems pretty clear that General Chilton was referring to vacuum tubes in the radar fuse. However, I’m not so convinced about the statement “Vacuum tubes are used only in some older mods of one design (the B61)…”. For example, what about the krytons, klystrons, etc used in firing circuits?
As I understand it, the rise-times for currents in semi-conductor switches was (is?) too slow to permit the exceptional degree of firing synchronisation needed for implosion weapons. In fact, looking at the characteristics of power station semi-conductor switches (e.g. power thyristors) even in the mid 1990s, I’d be willing to bet that many modern weapons still contain “vacuum tubes” – even if they’re not of the type specifically being referred to by the General.
As a second thought, I don’t see why it is so imperative to replace vacuum tubes. They are simple, which means they’re comparatively reliable and easy to test. In comparison, semi-conductors create wonderful charge separation on exposure to ionising radiation, which is why they’re used so widely as radiation detectors (e.g. hyper pure germanium detectors). However, the combination of charge separation and atom dislocation (caused by the occasional spontaneously emitted neutron from the fissile material) leads to crystal defects, and after sufficient time the crystal defects could affect their operating characteristics.
I presume that all of these issues have been well researched by the labs charged with making warheads reliable, and that if semi-conductor switches are used it is in the full knowledge of how they are likely to deteriorate. But in this case, if Gen. Chilton is referring to replacement of all types of vacuum tubes then he would do better to explain why he wants to replace them, and what the alternative is, rather than just wave part of a radar in the air.
Azr@el | April 23, 2009
The neutron initiator is technically a vacuum tube, albeit one on steroids. And the krytron switch is also a vacuum tube.
rkelly | April 23, 2009
Vacuum tubes in the physics package? Now there’s a novel concept. I think I once saw it on some sci-fi movie – plus it had blinking lights and a revolving wheel. It was awesome.
Steven Dolley | April 23, 2009
I could be completely wrong, but I think I recall reading somewhere that vacuum tubes are less susceptible than solid-state technology to EMP, which is certainly a consideration in warhead design.
Smith | April 23, 2009
General Chilton’s comments interest me. Is he familiar with the engineering and components that comprise warhead design? If not, I might be inclined to give him a pass; if so… shameful.
Someone please give me a high-level explanation of why this RRW proselytizing continues. Given what little I’ve read and my extremely limited knowledge on the subject as a whole, I cannot contrive a rational reason to allocate billions of dollars on this particular program when the stockpile has been shown to be reliable for years. Is this a pissing contest for government contracts? Someone’s pet project?
rkelly | April 23, 2009
“Kingston Reif and I discovered that is a load of crap.”
Aw come on Jeffery. Do you really expect a Pulitzer Prize for discovering a “load of crap” that people in the business know to be a “load of crap”?
It’s perfectly acceptable to not be in favor of the RRW but at least be honest as to your reasons why and not post a “load of crap” trying to justify your position.
I know, I know, the guys a general & you saw it on the internet & therefore…
Pakistan is purportedly, in a position to potentially lose control of its nuclear weapons.
Lord be it a modern IHE system designed by America’s best.
George William Herbert | April 24, 2009
There is still considerable debate among open sources as to the level of demonstrated stockpile reliability.
The DOE establishment opinion has solidified on a consensus that the stockpile is very reliable, but at the current level of programmatic secrecy in that establishment, it’s impossible for outsiders to validate that what they’re saying is compatible with good engineering judgement and materials science.
Regarding the RRW, the opposite is true in a sense as well – it’s impossible to validate the claim that RRW-1’s design could be reliable enough without actually test firing at least one. The RRW-1 project lead hinted that the primary was an already tested design, for example, but also hinted that there were some primary modifications and materials changes.
I personally find both positions disturbing, given the level of birth defects in weapons up through at least the W80 (and its LX-17 corner turning at cold temperatures problem), much less going further back at historical problems with earlier warheads. The W80 problem was a common mode failure that apparently / alledgedly affected the whole B61 series where they use insensitive HE (all the ones still in service), and had been latent in the B-61 primaries using IHE for some years prior to the W80 cold soak proof test failure and reengineering effort that eliminated that risk across the B61 family.
That particular problem is not likely to reappear. But it indicates that latent flaws can last in systems for some time before being discovered, and that relatively modern designs with decades of experience can still have fatal flaws.
The whole government and our weapons program took a leap of faith – that the unknown risks here (that the whole fleet of weapons, or a large subsegment of them, are flawed and will misfire if used in anger due either to design defects or aging problems) are outweighed by the proliferation limiting benefits of the test ban treaty and the lowered risks of someone having to fire them off.
From a geopolitical perspective that’s a well argued and commonly held belief, and I don’t disagree with the conclusion.
However, from an engineering perspective, I find the hubris of engineers or physicists highly disturbing when they insist that it’s ok to rationalize not testing and those built-in failure risks away by asserting that the new designs and non-test-firing testing of old designs are good enough to have militarily high confidence going forwards.
Past history provides reliability estimation ground truth – the experts screw up regularly and predictably, on a time and design basis. This is true across engineering disciplines. That’s why aircraft get extensive test programs, why rockets generally fail half the time in their first few launches, and why despite NASA’s best ongoing programmatic efforts they have lost a pair of very expensive shuttles and 14 astronauts.
DOE and NNSA aren’t any less vulnerable to fooling themselves about the level of reliability of these systems. In the absence of testing, the level of uncertainty is high.
Deciding to live with that uncertainty, for the larger geopolitical benefits, is ok. Asserting that there’s no uncertainty is a terrible error.
Alex W. | April 24, 2009
As an aside, hi-fi enthusiasts pay big bucks to get formerly-military-grade vacuum tubes from Russia for their amps (or whatever). Apparently good tubes are very hard to find these days outside of some sort of old military connection. (At least, so my audiophile friend tells me.)
MarkoB | April 24, 2009
Another RRW “talking points memo” that bites the dust. These have been accumulating steadily since the work on the longevity of Pu pits. If the stated rationales of RRW don’t hold water, the really big most important question is…what’s it for?
One oft stated rationale is that it’s a make work scheme for the weapons complex. Perhaps, that is partly true. Why is Chilton so gun-ho on RRW? If Strategic Command wants RRW just as much as the DoE, even making pretty amazing public interventions (like this one, coming not long after Obama’s speech in Europe; think about the politics of that) to support RRW, suggests to me that it also involves some strategic rationale.
This is especially the case if arguments due to reliability and safety and so on don’t hold water. You have caught Chilton red-handed in one of the most significant pieces of research on RRW, in my opinion. Chilton is not really worried about the reliability of the stockpile…you have caught him out.
According to the official RRW story (and your own view!) the RRW project carries no intended implication for new or enhanced military capabilities.
That being the case why the amazing Chilton public advocacy? Your own vacuum tube analysis can be interpreted as undermining the “no intended military implications” position on RRW. Chilton and STRATCOM want RRW because it will enhance their ability to do what they are charged with doing…striking against a desired set of strategic and military targets…this at the very least now becomes a working hypothesis.
George William Herbert | April 24, 2009
Short answer? Because anyone who studies the history knows there have been enough design failures (tritium cross section, W80/B61 cold temp aka LX-17 cold corner turning failure) and component aging failures in the system that we do not have high confidence that “the stockpile has been shown to be reliable for years” is an accurate statement.
The DOE / NNSA are good. But in other technology areas, nobody believes anything but real tests, because analysis and partial tests regularly equal failures in the field.
If geopolitics says its better not to test, that’s ok. But downplaying technical risk in an effort to feel better about not testing, or about being told not to test, is not technically defensible. There is latent fleet risk in the whole program, and buying into that was part and parcel of the test ban, component or ongoing subcritical testing notwithstanding.
Jeffrey Lewis | April 24, 2009
RKelly.
Even if you know it is a load of crap, a lot of people in senior positions do not.
As for “at least be honest as to your reasons why” I opposed the late RRW, I think I was pretty straightforward on that matter: I did not see evidence that replacement was more cost-effective or technically appropriate (ie minimum risk of needing to test) than life extension.
I thought I spelled that out pretty clearly — and forthrightly — in Arms Control Today:
Jeffrey Lewis | April 24, 2009
Also, I should say, we did remain silent about components that, strictly speaking, are vacuum tubes but not replaceable by solid-state devices — I think of that as the “light bulb exception” to the debate.
In any event, those applications — neutron generators and AFF — are also external and replaceable without yield testing or redesign.
Smith | April 24, 2009
Thank you all – your responses are very informative. I suppose part of my problem is that it’s extremely difficult to find information when you don’t know what questions to ask (and when you know no one with an interest in such topics).
I don’t know enough about stockpile reliability to make a truly informed decision (yet). One of you should compile a reading list for people like me.
Carey Sublette | April 24, 2009
BTW: Krytrons aren’t used in current weapons, they were replaced in the 60s by a related tube device called the sprytron. These may also have been replaced by solid state devices though.
But more to the point: it is a fallacy to equate “tube” with “obsolete”.
If the tube device fulfills a particular specialized role (extremely fast high power switching, or generating a microwave pulse, or – as another poster here pointed out, a neutron pulse) and meets operational requirements (ruggedness, reliability, etc.) in what sense is it a problem?
This is simply an attempt to harness a technological prejudice.
Carey Sublette | April 24, 2009
George William Herbert and I take a very similar view about nuclear weapon engineering: that it is just like every other area of engineering, not imbued with special (aka “magical”) properties.
But I think we take somewhat different lessons from this.
George emphasizes the necessity of testing for ensuring reliability. With this I agree, but I strongly question whether this necessitates nuclear testing.
In the past, when it was the weapon lab’s position that nuclear testing was essential and that maintaining our arsenal would be impossible without it, it was argued that the behavior of a device in a nuclear explosion was actually poorly understood and that virtually any change in the physics package might lead to outright failure. Only nuclear testing could address this.
Some years down the road, once it became clear that the future of the labs was in stockpile stewardship, and after some turn-over in staff I imagine (some older scientists who were indoctrinated into testing culture having retired), it is now felt that testing is not required after all, though a more robust and conservatively designed device might be an asset in addressing a problem we really don’t have(!).
My view is that the engineering of nuclear weapons is not exceptional. The last generation of weapons, which pushed the limits the most severely, were validated with very few nuclear tests being required. This indicates that the devices generally performed as designed for reasons that were well understood, and mysterious factors, which would have required a large number of trial-and-error shots to overcome, did not intervene.
The examples that George mentions, where problems were uncovered in nuclear shots (e.g. the cold soak test), did not actually require nuclear tests. At the time, nuclear tests were simply standard practice, and so that is when the problems were discovered. The kind of meticulous non-nuclear component testing and examination conducted under current stockpile surveillance practice were not used. A non-nuclear shot of the B-61 IHE implosion system would have revealed the performance problem, if it had been done.
So is RRW necessary? The belief that RRW is a good idea for stockpile reliability (that performance can be predicted from design and non-nuclear testing) appears to include the assumptions that render it unnecessary.
yousaf | April 24, 2009
Smith,
the real issues are not terribly difficult to grapple with. There is a lot of unnecessary obfuscation in the debate — intentional or otherwise.
Couple of main points re. the RRW debate:
1. actual warhead reliability has little relation to its psychological deterrent value — even if this reliability figure is known by one’s enemy. Your enemy (assuming he is rational and deterrable, to begin with) does not care if your warheads are 99.9%, 95%, 80% or — if they have any sense — even 25% reliable, they will be scared of them.
We are talking about massively destructive nuclear weapons, after all.
2. The second point is that any new warheads will (supposedly), according to DoD and DoE, be certifiable without testing. This may well be, but it does not speak to the psychological impact of having untested weapons in one’s arsenal as perceived by one’s enemy.
i.e. Why would your enemies believe untested new weapons work better than thoroughly tested legacy weapons? If anything, they may believe the reverse.
More prosaically, would you rather fly on an airliner that had never had a test flight, even though its aerodynamics may be well understood, or on a 747?
3. The third point is that there are systems external to the physics package that, in sum, are comparably, or even less, reliable than the warhead, especially when you fold in everything from the button on the console to the ICBM itself.
RRWs would be detrimental to US security or — in the ‘best’ case — no better than what we have now, in terms of their deterrent value.
You can find links to some relevant papers via the URLs highlighted in my 2 articles on the subject
Stephen Young | April 24, 2009
Adding to a point made by Corey, and countering somewhat the points made by George, the well-known fact is that very few of the 1000+ explosive nuclear tests conducted by the United States were for reliability. They were largely to develop or confirm new designs. The weapons designers I talk to have confidence in the current designs and believe that confidence will grow over time, not diminish.
Second, parsing of the language used to promote RRW shows little concern about existing warheads or the Life Extension Programs (LEPs) to maintain them, but about repeated LEPs that 1) drift from the original design and 2) are expensive and difficult because of the exotic materials and manufacturing techniques. So, the RRW was an attempt to answer a problem that is decades away, when we might be doing LEPs of LEPs. If we have a problem at that point, we can just go back to Hiroshima-style bombs that will never require testing.
Finally, Jeffrey, sources? Is this just a tip from an insider, or do you have any documentation you can provide?
A. Weaponeer | April 25, 2009
“… vacuum tubes are not used in the physics package of a single nuclear weapon design.”
I’m impressed – clearly the sign of someone who thinks he knows something about warhead design.
“… Mods 10 and 11 have newer radars …”
You’re absolutely certain about that, are you?
“Vacuum tubes are not used for any other purpose or in any other design.”
What certitude!
———————————-
While Chilton was certainly dumbing down to make a point for the WSJ, it’s clear that “Dr.” Jeffery and his pal engaged in an exercise that stretched their grasp of weapons engineering to counter it.
George William Herbert | April 25, 2009
Addressing some of Carey’s points and some other ones –
Yes, non-nuclear testing would uncover the W80 cold soak / TATB cold corner turning failure mode. However, it had not been detected in the prior engineering work and development work, and the W80 wasn’t the first IHE nuclear weapon system.
The actual failure there was a failure to understand that there was a factor requiring test.
That’s the sort of failure that indicates that we (as humans, and particularly as engineering organizations) aren’t as smart as we think we are.
It’s a common failure mode in aerospace. “Oh, we hadn’t thought of that” (or hadn’t talked enough about that, effectively, with enough people) killed Columbia and Challenger and at least one Soyuz crew and umpteen dozens of unmanned launches and spacecraft in flight.
Certainly, lessons from that led into the current stockpile monitoring and preventive testing and refresh cycles. We’re not likely to make that specific mistake again.
What’s not clear from the open discussions is how extensively the stockpile stewardship teams seek out new potential failure modes that might be gaps in prior conceptualization or understanding of the things we need to test.
Failures to succeed at that process well enough abound in the engineering reliability literature.
Part of the problem with intelligently debating RRW is that not enough of the technical parameter space in play – in terms of design concepts, materials, manufacturing processes, design margins in components, etc. – is available outside the NNSA and associated military functions.
Design of a RRW in such a way that it was inherently much higher margin than the last generations of production weapons, and inherently more testable than the last generations of production weapons, could produce an ultimately much more reliable system in the long term.
But one can’t argue that point coherently either way without sigma level clearance.
The information available to me does not lead me to have confidence that NNSA and the military have higher long term organizational engineering perfection than other organizations such as engineering companies, NASA, etc.
They may have it, but it’s not clearly evident from what we know from the outside. They do seem to have a more robust program of peer review (cleared outsiders, associated with the other lab or not currently associated with either) than many other engineering organizations.
George William Herbert | April 25, 2009
Earlier in the comments thread, Jeffrey wrote:
In any event, those applications — neutron generators and AFF — are also external and replaceable without yield testing or redesign.
Minor point – most neutron generators appear to be placed within the radiation case, but external to the primary. Gross changes to such components could affect the radiation transfer and secondary behavior, though it would have to be a pretty significant gross change in order to have any end functional result on the secondary due to the relatively low mass and surface area, and high energies involved.
rkelly | April 25, 2009
…”I opposed the late RRW, I think I was pretty straightforward on that matter: I did not see evidence that replacement was more cost-effective or technically appropriate (ie minimum risk of needing to test) than life extension.”
Fine, that’s an honorable position that I can respect. However, imo, honorable positions/opinions based on less than the full facts can be wrong because of false facts & arguments by both sides of the issue (to wit: vacuum tubes or the supreme “Reliability” debate).
I recognize the debate for the RRW, as you have noted, (“… the late RRW”) is a moot issue since President Obama’s administration does not support development of the RRW,& therefore it’s a non-issue for the next 4-8 yrs. Therefore, there’s not a lot of motive to keep this issue on the front burners.
Nevertheless, I have no doubt that I represent a minority opinion on the RRW debate. However, I argue that this debate about Reliability is the lesser of the reasons for a RRW. As many here know, Nuclear Weapons Surety is a function of safety, security, reliability, and control of nuclear weapons.
The quantum leaps in safety, security & control of new (modern) nuclear weapons are more than enough (imo) to justify the RRW. Come on, Congress spills more than this in a week.
As for Reliability of an untested warhead, I’ll trust that to our country’s nuclear design labs – not to a bunch of politicians or the JASON Group that is a bunch of old farts that have their own agenda (yes, their opinion’s valuable but should not be the final say as though they are His Holiness). I trust that our adversaries have a much higher opinion of our weapons labs and the reliability of our “untested” nuclear weapons than America’s RRW opponents.
The Protection of America’s nuclear weapons is based on a protection strategy of “Denial” vs Containment for SNM. The Americans who put their butt on the line every day trying to meet & maintain these protection criteria well know the advantages of a modern IHE NE versus those based on HMX. A full FOF exercise can be a real eye opener.
Argument s today for maintaining an HMX based NE make even less sense now than it did when the Navy was defending its use during the Phase I & II studies for the W76 (throw-weight my A—).
As I noted earlier, (imo) should Pakistan lose control of its nuclear weapons to the Taliban, I would be very, very concerned but not nearly so if those weapons were designed as a modern IHE system by America.
No, they would not be as Gen. James Cartwright of STRATCOM noted, “as useless “as a paperweight” if it falls into terrorist hands”. But there are some major improvements and I do believe they are much, much safer & secure than those that would be replaced.
Like it or not, nuclear weapons are going to be around for several more decades at least. I don’t believe we should let our nuclear weapons labs decay into mediocrity just to appease those who think this as a grand strategy for un-inventing nuclear weapons.
Azr@el | April 25, 2009
I think discussing the placement of the neutron initiator or the degree to which the hohlraum is shielded from external radiation would be in the same category as discussing replacing aerogel beryllium-borates and beryllium-silcates with a glucina composite; i.e. speculative, undebatable and pyrrhic.
bruno | April 25, 2009
Yousaf argues in the Bulleting of Atomic Scientists pieces that are referred to in his post of 24 April that France introduced untested warheads in its arsenal during Mitterrand’s unilateral moratorium. He uses this as an argument in the RRW debate. To the best of my knowledge, this assertion is incorrect. No untested warheads were ever introduced in the French arsenal. (All new types were subject to hot tests.) The resumption of testing by Chirac in 1995 was justified by two reasons: (1) There was a need for one final “qualifying” test for the new TN75, which was introduced in 1996; (2) The government wanted to test “RRW-like” warheads in order to prepare for the complete cessation of testing. The “French RRW” (called “armes robustes” in French) are to be introduced in the stockpile in 2009 and 2015.
Yossi | April 27, 2009
The shallow attitude of US generals towards technology is not new. It probably stems from a very active industrial complex always pushing new products to gullible customers. Vacuum tubes are an excellent example.
Let’s examine, for example, Maj. Gen. Patrick O’Reilly (deputy director of the Missile Defense Agency) visit to one of the most secret Russian early-warning radars located in Azerbaijan. His experience was documented in NYT.
O’Reilly came away with three “significant impressions”:
* The radar is huge, almost twice the size of a similar American system
* The radar relies on outdated vacuum-tube technology
* The system is extremely able as an early-warning radar scanning the skies over the Middle East
The first two points are negative judgments based on irrelevant data, the third is sincere surprise. The ambivalence is clear in:
Despite the system’s older technology, “I was impressed by what I saw,” General O’Reilly said. “It would be a false impression to dismiss the capabilities they have. They just chose another way of achieving it. It is an excellent radar for the case of early warning.”
O’Reilly seems to be aware that he is prejudiced or tries to avoid insulting his friendly hosts:
While American military technology adopted solid-state semiconductor electronics decades ago, the Russian radar still operates using vacuum tubes, General O’Reilly said. He noted that Russia had improved tube technology far beyond where the United States had taken it before moving to electrical semiconductors.
Are vacuum tubes really an older/outdated radar technology? Some UK experts think the opposite is true.
The journal “Aircraft Engineering and Aerospace Technology” published in 2008 a short piece titled “Vacuum tubes still rule the roost in high-performance radar. It says vacuum tubes are unsurpassed in reliable and efficient microwave generation. They are twice as efficient as solid state devices in the S band and nearly three times in the X band. The high efficiency contributes to lower temperature and thus to higher reliability. Moreover, these advantages are widely recognized “[Traveling Wave Tubes] populate the world’s satellite communications networks, and they are at the heart of practically every state-of-the-art military radar, electronic warfare and complex weapon system in service today.”
Several radar books confirm this UK view. Vacuum tubes continue to be the main-stay of radar technology.
This may remind one of the problems with the US Sea based X band radar. Maybe this constantly under repairs “white elephant” was more useful if it employed more vacuum tubes?
The US pays a high price for its shallow attitude towards technology. A few decades ago it discarded gas centrifuge technology and concentrated efforts on “super advanced” laser enrichment. It’s now still stuck with the extraordinarily inefficient gas diffusion while the Russians and URENCO energy bills are about 50 times less. Even Iran does it better than the US.
kg | April 29, 2009
The pathetic dying breaths of the nuke age in America. Thank goodness.
Yossi | May 15, 2009
Azr@el, how do you know that Fogbank is made of aerogel beryllium-borate and beryllium-silicate? Is beryllium used because of its neutron reflecting quality to assist the implosion of the secondary and guide the primary neutrons to it?
Suitable tunneling of the Fogbank could guide fission neutrons to split the Lithium-6 and make a complementary action to what the Hohlraum does for the X-rays. The Teller-Ulam design may be very sophisticated if it baffled the British scientists.