Jeffrey LewisItalian Stallions & Plutonium

Oh, you just can’t wait to figure out what the hell that means, can you?

I noticed a fun fact about plutonium pits while reading the AAAS report on the Reliable Replacement Warhead: During the Cold War workers at Rocky Flats worked plutonium (“wrought processing”) to make a spherical pit; today, TA-55 casts them in a furnace (image at right).

The AAAS panel noted the change, explaining that the decision was made in part due to the expense and difficulty of installing the necessary equipment for wrought processing at TA-55. (Haninah also mentioned this in passing in his excellent DefenseTech series on the RRW and included a link to Kautz et al, “The Pit Production Story” from Los Alamos Science Number 28, 2003.)

Anyway, the AAAS panel observed that a sufficient test pedigree exists for cast pits. Someone asked about the difference between wrought and cast pits at an ACA panel. Dick Garwin claimed the two methods result in pits that perform equivalently:

QUESTION: … You stated that if it turns out that there are serious problems with pits we can always remanufacture them to original specifications. At present I believe that’s not exactly true because the original specifications called for wrought plutonium pits and today we only have the capability to build the cast ones. … Do you believe that there is no difference, or are you supporting building a modern pit facility that will allow us to build wrought pits?


GARWIN: This was an early question when people at Los Alamos started manufacturing pits there. They had them made at Rocky Flats, but Rocky Flats was shut down and Los Alamos was made to bite the bullet and build new pits at TA-55 for the W-88 nuclear weapon. I was on the visiting committee that looked at this. We have made certifiable pits. They are made by cast and machine process. The decision was made and people have judged that cast and wrought perform equivalently. One can find minor differences in strength, and the more we know about it, the more they are equivalent within the range that you need.

There is not a significant difference in primary yield and you don’t need to reopen that question. That decision has been made. Remanufacturing to original specification will allow the substitution of cast plutonium for wrought plutonium so long as the dimensions are correct.

[Emphasis Mine]

Turns out, Los Alamos performed a series of subcritical experiments “to investigate differences in material properties between cast and wrought plutonium driven by high explosives (HE).”

Los Alamos named the series … wait for it … Stallion, and the experiments … I am so not making this up … Mario, Rocco, and Armando.

Mario, Rocco and Armando. The Italian Stallians that prove cast Pu performs just as well as wrought.

Of course, casting pits is only the most significant of a number of changes that occurred when production moved to TA_55. Another significant difference is that Los Alamos developed a “dry machining” technique that avoids the use of lubricants that leave a toxic mess (think radioactive Santorum).

Okay, okay, that’s enough. I am sorry. Off to have a beer with John Park.

If you want to know more, just
read Kautz et al.


  1. Alex W. (History)

    For the curious, there are some pictures of the Armando experiment at the DOE/NNSA Nevada Site Office, and a (boring) movie.

    I had seen the pictures before but didn’t really know what they were about! Now they make more sense, even though they are still pretty boring, aesthetically speaking.

    (It looks like there was also another subcritical experiment named Vito. Too bad there wasn’t a Vinnie.)

  2. Haninah (History)

    Great post, Jeffrey, and thanks for the link (and the praise), but I have one question: you mention a “test pedigree… for cast pits.” Is that a reference to the sub-crits, which you discuss later in the piece, or is there a separate pedigree involving pre-moratorium underground testing of cast pits?

  3. Jon (History)

    I laughed out loud when I read the phrase “radioactive Santorum”…

  4. Nukem (History)

    Love the Santorum reference! Dan Savage would be proud.

  5. John Field (History)

    The description of the Armando optical pyrometry experiment is interesting in that it is exactly the experiment that Zel’dovich and Raizer explain does not work in Chapter 11 Section 23. As they point out, the observed temperature corresponds to the temperature of the unloaded material expanding off the surface and not the temperature behind the shock.

    As they are measuring the temperature of the unloaded surface, it appears to me that what they are interested in is the entropy produced by the initial weak shock wave. Presumably, the reason the expected (unloaded surface) temperatures are so high is that the (low-density) gallium stabilized plutonium undergoes a phase transformation behind the shock front which increases the density from something like 16 to maybe 19-21

    Herein lies the reason why they are so concerned about this. This shock induced phase change generates lots of entropy which is bad – limits further compression. Ideally, you would like to initially strike the Pu with the weakest initial shock you can to minimize the phase change shock entropy. Then, further shocks generate little extra entropy. If the initial shock is weak enough, and the method of Pu pit fabrication has changed, you might be worried that the altered phase transition will affect the shock propagation or the amount of entropy produced – which in turn could change the timing or limit the ultimate compression.

    Now, Pu is a very anisotropic material, and I suppose that the grain sizes(as well as anisotropic work hardening) are going to be different between the cast and the wrought process. This could be a source of additional entropy production as the shock propagates through the material. But, I would have made a (very uninformed) guess that the cast pits would be better than the wrought ones.

    None of this would be a problem using Uranium as you don’t need the funny Pu delta-phase stabilized alloy. And, I suppose that for the RRW they will use enough Pu in there to avoid this concern.

    My view is that it is not appropriate public policy to base national defense on extravagant science like this. If, for whatever reason, we deem it necessary to have the capability to murder millions of people, the bombs need to be simple, few, and credible. To this end, our scientific developments have failed to support public policy.

  6. spacemanafrica

    Great point, great post. I still remember when I first read, in grade school, about how a multi-stage missile system with near 97% reliability across its entire flight profile had a total system reliability of ~69% give or take a stage. I guess it was also the first time I was exposed to statistics. Solid numbers like that used to come from live tests. I’m still concerned about nuclear fratricide when it comes to our forces. John’s advice is mostly right but “few” should be replaced by “sufficient in number.” Kinda sad, that.