John Carpenter - Session II

Notice: We are in the process of migrating Oral History Interview metadata to this new version of our website.

During this migration, the following fields associated with interviews may be incomplete: Institutions, Additional Persons, and Subjects. Our Browse Subjects feature is also affected by this migration.

We encourage researchers to utilize the full-text search on this page to navigate our oral histories or to use our catalog to locate oral history interviews by keyword.

Please contact [email protected] with any feedback.

ORAL HISTORIES
Interviewed by
Catherine Westfall
Interview date
Location
Argonne National Laboratory
Usage Information and Disclaimer
Disclaimer text

This transcript may not be quoted, reproduced or redistributed in whole or in part by any means except with the written permission of the American Institute of Physics.

This transcript is based on a tape-recorded interview deposited at the Center for History of Physics of the American Institute of Physics. The AIP's interviews have generally been transcribed from tape, edited by the interviewer for clarity, and then further edited by the interviewee. If this interview is important to you, you should consult earlier versions of the transcript or listen to the original tape. For many interviews, the AIP retains substantial files with further information about the interviewee and the interview itself. Please contact us for information about accessing these materials.

Please bear in mind that: 1) This material is a transcript of the spoken word rather than a literary product; 2) An interview must be read with the awareness that different people's memories about an event will often differ, and that memories can change with time for many reasons including subsequent experiences, interactions with others, and one's feelings about an event. Disclaimer: This transcript was scanned from a typescript, introducing occasional spelling errors. The original typescript is available.

Preferred citation

In footnotes or endnotes please cite AIP interviews like this:

Interview of John Carpenter by Catherine Westfall on 2010 May 3,
Niels Bohr Library & Archives, American Institute of Physics,
College Park, MD USA,
www.aip.org/history-programs/niels-bohr-library/oral-histories/33721-2

For multiple citations, "AIP" is the preferred abbreviation for the location.

 

Abstract

In this interview, John Carpenter discusses topics such as: his graduate school work in nuclear engineering; his early professorship at the University of Michigan; going to the Reactor Testing Station in Idaho to learn about neutron scattering; beginning work at Argonne National Laboratory; developing the first-ever pulsed spallation neutron sources equipped for neutron scattering, ZING-P and ZING-P'; development and implemention of the intense pulsed neutron source (IPNS); becoming an advisor at Oak Ridge National Laboratory; spallation neutron source (SNS); his retirement; slow neutron scattering; Motoharu Kimura and winning the Clifford G. Shull prize.

Transcript

Westfall:

This is the second part of the interview that we began earlier. The last story that you told was how the IPNS got its name.

Carpenter:

Well, starting in mid-1975, we had a small crew of people. I was aboard; Norm Swanson was aboard as the real manager. Among of the people who came to help us was John Ball, who had developed into a right-hand man in the engineering side. Constantine Smyros was supposed to help plan the management side of this project. Our offices were in Solid-State Science Division in the 223 building. We also had enormously useful help from Bob Kleb. David Price, who was then the director of the Solid-State Division, had assigned Bob to us. Bob was the best engineer in the Solid-State Science Division, and was probably the best engineer at Argonne at the time. He had made significant contributions to the 12-foot hydrogen chamber, and so on. Along with Bob came two sidekicks Tom Erickson and Bb Stefiuk, craftsmen who, with Bob, built many of the components of ZING-P, ZING-P’ eventually of IPNS and their instruments. Peggy O’Connor, a young pool secretary worked for us and typed our manuscripts. So we worked as a team to develop our concepts for this and put together a proposal. At the time, what we had to do was submit a Schedule 44, which was a sort of a prospectus for a proposal. We had a half-megawatt model IPNS. We decided that this really needed a prototype at larger scale than Zing P?, first of all to develop the community, the instruments, and the applications, and second to develop the experience needed to build a really big one. We called them IPNS I and IPNS II. It took us several years to develop these proposals. In the meantime, Norm Swanson and I and this team were sending Schedules 44 to Washington and back. We finally documented our proposal in early 1978, the proposal for IPNS, which was a two-step project. By the time we were done, Norm Swanson and I calculated, we added up that we must have submitted 40 different versions of Schedules 44, going back and forth. They would say do this, do that, and we would go do this and do that. In the end they said well, we can’t fund the big one. David Price and I, I pointedly remember standing in front of his office. “Shall we accept a proposal to build only a first stage, or shall we hold out for the bigger one?” He said, “If we accept just to build IPNS I, maybe we won’t ever get to build IPNS II.” We debated this, and decided we would build the IPNS I. That is how it got settled. Then it was a question of how much money, because we had formulated a budget to build IPNS I while IPNS II was being designed and built up. I think we had estimated that in that atmosphere, it would cost about $6.4 million to build the prototype while we were still working on the other parts of the project, and they would support a lot of the overhead. Well, the problem was that then we needed more than the $6.4 million for IPNS I without the large project. So this was a big puzzle, but we accepted this. I can’t remember just now how wise thought we were in doing this, but it was actually a bad time to fix a number on a project, because in those years, 1978 to 1980, were years of bad inflation, so $6.4 million was going to go a lot less far than it was intended to go in the first place, and then we had to eat that. Furthermore, we had no budget then for the instruments. So finally, I forgot how it happened, it was probably David Price’s insistence, we went back and drummed up funding for the instruments. We asked for another $3.2 million dollars, I think, which included a contingency amount, a standard 20% or whatever. They conceded, “Well, you need consistent contingency money.” If you ever make a project, you must have it, because you know that there are things that you can’t anticipate and you will have to make up for. But finally they said (they being the people in what is now the DOE offices) said, “Oh no, we’ll give you the money, but no contingency.” So we accepted this. We went on and built the project, but it was hard. Now I’ll tell you about the interaction with Constantine (Dean) Smyros. He’s a colorful character. He was doing budget management, and he and I hurriedly learned how to do PERT programming, Project Evaluation and Review Technique.

Westfall:

So this was PERT, and this was an early project management tool?

Carpenter:

Yes, right. We learned how to do this in its crude way, well sort of learned. This is one of the things that Smyros was supposed to do. There was another person working at the same time on shielding questions. These days, in the present time, he have elaborate MCMP and FLUKA and other codes that help calculate radiation dose rates and do shielding calculations for you. In those days it was done mostly on the basis of analytic formulas and estimates and correlations, and back-of-the-envelope type of calculations. We brought to work with us the very well known, respected shielding analyst, Marcel Barbier who had worked with us on shielding questions for Zing P?, and now he was going to do this for IPNS. He designed the shielding, and he got it right — there was never a question of this. So he would be working in one office with several others of us along the hallway in 223. It was the summertime, and it was really hot weather. Marcel was a very conservative dresser. He arrived at work in the heat of the summer in a three-piece dark suit. He wore French-cuff shirts, but used paperclips to hold the cuffs together. He was a very straight-laced kind of guy. A very effective scientist in what he did. Now, back to this Smyros interaction. Smyros, it seemed after a couple of months, like three months at this, he was not very interested anymore in what he was doing because he wasn’t famous yet — three months into a major project, he wasn’t famous yet — so he didn’t show up for a while. And it was the heat of the summer. One day, he did show up, and he had short pants on. He was Greek, he has dark skin and a hairy body, and he walks in in his shorts and sleeveless T-shirt. Marcel Barbier, who dressed in exactly the opposite attire, a three-piece black suit with a shirt with paperclip cuff links — he is Swiss so he speaks in a thick French accent — looked up and said, “Smyros, where have you been? Have you been mowing your lawn?” [Laughs] I think we never saw anything more of “Dean” Smyros. I don’t remember what he might have answered, but I love that little story. Barbier stayed on to carry out our shielding calculations. We formed up a construction project, and we moved into building 372 for that purpose, a little building in the area close by where IPNS was to be built. . There were Swanson and myself, Tom Worlton, on the computing side, with John Ball the assistant for the engineering manager. Swanson brought budget man Al Stark with him. We had another Argonne engineer, Alvin Knox, who worked with us, who was very pleasant and effective. And three secretaries, Midge Thompson, Peggy O’Connor, and Dianne Hoffman. Some are still here at Argonne. This was a convenient place to work. We used to have project meetings there. Now I want to tell another story about a couple of these project meetings. Swanson was the manager. I was the project director. They would go along doing things. Swanson, of course, was the one who wanted to get things done, and I was the one who wanted to get things right. It took more time to get things right than it did to just get them done, so there were a lot of compromises that were made in this process. We always had to appeal also to the higher management. Betsy Anker-Johnson was the associate lab director. So it turned out that she was much more interested in what Swanson said and didn’t trust what I said. I never mentioned that we had a crew of people working on the accelerator side with us. There were Jim Simpson, Yanglai Cho, and Bob Kustom and Charlie Potts. These were some of the principal people. I didn’t work so much directly with them — they had their own accelerator to build, and just delivering protons was the idea. So I did not have the ear of Betsy Anker-Johnson. Betsy liked to talk to Bob Kustom, for good reasons I think, and she liked to listen to Swanson. So I was really not very high on that totem pole.

Westfall:

Let me say that in the meantime I’ve talked with Bob Kustom, did an interview with him about the APS, particularly the advanced photon source. One of the things that I definitely get the sense of when I talk to him was that at the laboratory at this time there’s kind of a pecking order, if you will, that you have this group of accelerator specialists who had worked together on the Zero Gradiant Synchrotron. They were a team, and they were the lab’s accelerator-big-guy experts. So what you’re saying is this Betsy Anker-Johnson is listening to the established experts. Jack Carpenter at this time — subsequently you became a big deal, too — but at this time you’re a newcomer, or an up and coming guy.

Carpenter:

I’m a newcomer, and didn’t have that kind of a track record, so that’s true. Swanson had a track record from construction work that he had done at Argonne West. He was really a high-powered guy.

Westfall:

I see, so he had worked at Argonne West, which meant that he had worked with the reactors.

Carpenter:

Yes, that’s correct. So he’s a really high-powered guy. The main reason he came to Argonne East was the prospect of a half-megawatt proton source project, the big one, the IPNS II. That’s the main reason he came. To his great credit in my mind, even though he was hard to work with on my part, he stayed with it through IPNS-I, and saw it through until it was done. That’s guts. That’s commitment. And he gets my enduring praise for having done this. Just some of our project meetings would go haywire. Now you know the pecking order question. We were building things, and we’d have weekly meetings with the project team, like half a dozen people. I remember several times when things were not going as Swanson wished, he would say, “I’m not throwing stones, but…”, and then he would criticize something that I had held up or done not according to what he thought it should be. Whenever he said, “I’m not throwing stones, but…,” I knew I should get my hard hat.

Westfall:

[Laughs] To protect yourself from the stones!

Carpenter:

He was throwing them at me. Yes, so that was interesting. But it was very hard. In another instance of this, we needed to have some assistance building the cryogenic system for the cold moderators in IPNS, so we brought in a consulting firm. I don’t remember the name of them; there were two principal people. They were working for Swanson and John Ball to design the cryogenic moderator system that I had specified. So we would leave it with these people, and I always thought it sounds like something is wrong. When we asked these guys questions, the answers didn’t come back in the same direction as I thought. Week after week we had these discussions, and finally I understood that the instructions from John Ball and the engineering people were that they were supposed to cool these moderators with liquid methane. My specification is that these moderators were to be containers filled with liquid methane, not that you just cooled them with liquid methane. It took a long time to sort it out. The message didn’t get across that they were supposed to fill these moderators with liquid methane and that would circulate. They didn’t give any attention to whether they would be full or not. Well, finally it got straightened out. A lot of things got left behind. Like we knew the atmosphere around the target would be corrosive because the proton beam went through the air and ionized products that would corrode the iron shielding. Hence, I think I had a promise from Norm Swanson that there would be nickel plaiting that is resistant to this kind of corrosion, nickel plaiting on the internal components of the iron shield. I didn’t find out for a long time, until after, that that was never installed. We had a lot of rust, and we still have a lot of nasty radioactive rust inside that container. However, everything worked up to a point. We finished the project in August — Well, pardon me; we finished up to a point in May of 1981. We had two target systems and two target arrangements. One of them for radiation damage work. That was an initiative of Tom Blewitt, who is a much-beloved fast-neutron damage expert at Argonne at the time. Bruce Brown had worked for him in his project at the CP-5 reactor, so they were well acquainted. We had a target system with two locations, one for fast neutron damage work and the other for neutron scattering. We were going to divide the time something like two to one, one part for radiation damage, two parts for neutron scattering. So the first time we really radiated a target was the first one that was ready and that was the fast neutron damage radiation effects target. That was in May, 1981. We turned the target on, it worked, but people remember curious things that happen when you do things for the first time. We had thermocouples in the target, and we were waiting for them to bring proton pulses in that we could detect from the thermocouples. However, instead of the temperature going up it went down when they turned the proton beam on. It was a slight embarrassment because we had the leads reversed. The thermocouples put out the opposite of the voltage that we had expected, so we just had to turn the leads around. Then when we turned the proton beam on the temperature went up, which was a good thing [laughs] — it’s what we expected. So that was the first operation. Later in August of 1981 they turned the proton beam onto the neutron scattering target. A good deal of effort went into making targets, because we had decided that we would use zirconium-clad depleted uranium for targets, because it provides more neutrons. So a lot of effort went into building targets. They were made of uranium, cast uranium discs carefully machined. The disks were placed in a cup carefully machined out of zircaloy that fit just so around it, and all the air pumped out. Then you put an electron beam weld around the circumference to make the seal. We made a lot of those. Henry Thresh Argonne’s super uranium metallurgist, led that effort, but it was marvelous. Al Knox was a part of this and Al Hins was one of the chiefs of that operation. It was a lot of work to find out how you bond the uranium to the cladding around. The bonding question was very significant from the point of view of heat transfer, to get the heat out of the uranium and into the outer surrounding zircaloy discs and to the cooling water. So we had a lot of examinations, physical examinations and metallurgical tests to refine the hot isostatic pressure (HIP) bonding process to make sure that there was metal-to-metal contact in the joint all around and between the uranium and the zircaloy. Then the disks needed to be annealed in a beta-quenching process to produce the desired metallurgical alpha phase uranium. So we made a lot of discs. Out of maybe 40 discs, we only needed 8 good ones. We produced a surplus of good ones, but we had a lot of inferior ones that by ultrasonic methods you could determine where there was a good bond and where there was not a good bond. We had people like Dave Kupperman who carry out ultrasonic examinations of these discs. We finally had a good process and made a lot of discs. Those were the basis for the target. I won’t describe the details of the target. It was a flowing water system that passed water between a stack of discs, and that was it. So when the accelerator turned on, I think it was delivering 4 microamperes of current at 500 million volts. That’s far less than we thought we would be able to get out of the Rapid Cycling Synchrotron (RCS). We expected to be able to get to 25 microamperes, but we were getting 4. Well, it was early time, so we started out at 4. As time wore on, we took pride of delivering 5 and 6 microamperes. Then the Brinkman committee met and said well, this is far below what our expectations are, far below our expectations, and furthermore, we expect you’ll never get more than 8 microamps out of this machine — that’s like a third of what we promised to deliver. So this is somewhat of an embarrassment, but it was early days. Going through another different aspect of this same thing, because it looked like we were only going to get a third as much of the IPNS proton current as we had promised. I had designed moderators that would be most effective if they were operated in the full current. Betsy Anker-Johnson was pretty disappointed in this. She said, “Your neutron fluxes are lower than you expected, so do something.” Well, reasoning to myself is a factor of three is really disappointing. Maybe we’ll be able to do better in the future. I don’t want to change the moderators in such a way as to give away instrument resolution, which is the quality of the experiment, for neutron flux, which is the tradeoff that you’d have to make. I didn’t want to make that tradeoff, and she let it be known that I would have to do that. So I did. And the resolution was not as good as it was supposed to be. There’s more to this story. I just basically made the compromise and got on with it. I think it was under threat that I lose my job at that time. So I did make the change, but I nevertheless didn’t make myself popular with Betsy Anker-Johnson.

Westfall:

So this was during the period of time…

Carpenter:

IPNS was just starting up.

Westfall:

So it was at this perilous time when it’s not clear that the IPNS is going to be able to continue, and under the threat that it might be cancelled without…

Carpenter:

Yeah, without it hardly running at all, yes. This was at that time. This was 1981. Mike Nevitt was very supportive. He was my great supporter. He heard the news, and said, “Well, you’ll have to find yourself a better position.” I thought, you know, this whole enterprise won’t work if I’m not working on it. David Price, who was the Solid-State Science Division Director, agreed that if I were not working on it, IPNS just would not go. He supported the idea that I should stay, and I redefined my position as Technical Director. No power; I was just Technical Director. I had only advisory power. I could do my work and I accepted this, and I remained the Technical Director for all of IPNS existence. Bruce Brown, who later became IPNS Director, posted a New Yorker cartoon on his bulletin board in which an executive introduces a clownish figure,”This is my crazy idea man.” Guess I was.

Westfall:

So then Price became the…?

Carpenter:

When I left it, he became the Director of the IPNS project, and he remained for a few years. And then there was the Brinkman I event in which Bill Brinkman chaired a committee at the behest of the Office of Science, or whatever it was called at the time. That committee recommended, after a grace period of some time, that IPNS be shut down. So my own attitude was about that was, I don’t believe it. I believe in IPNS — we’d stick to it and we’ll make this thing work. The accelerator folks didn’t like being told that their limit is 8 microamperes, so eventually they got it up to 16 or 17 microamperes. So the people who said 8 microamperes were just, I don’t know, they had some basis for this but it wasn’t a correct basis. Eventually Frank Brumwell and Bob Kustom and Charlie Potts and these guys made it work in the end, running at 15, 16, 17 microamps. It was probably the best they could do. Furthermore, the 25 microampere estimate that had been my belief — I believed in this — people took it wrong. When people designed and evaluated accelerator concepts, they did this on the basis of an idealized proton beam distribution in the accelerator. You never can get that, experienced people know, but I didn’t know that. So the 25 microamperes came from a calculation based on this idealized proton beam distribution in the accelerator, which is a limiting case. Although there are things you can do to get closer than we did, 16 or 17 was darn good. We made it to there with the efforts of the accelerator crew. We were always working on methods to get more current in the machine. I think they were still working on that when IPNS was shut down. At the same time, they accomplished enviably high reliability, RCS operated in the high 90’s as a percentage of the time that it was supposed to be running. It turns out that high reliability is more important to users of a facility like IPNS that supports a large number of short experiments than the last 30% of intensity. Now, I want to go back to the late 1970’s, roughly speaking, and recall that the neutron scientists who had been working at the CP-5 reactor were still working at the CP-5 reactor, and we needed them to work on the instruments for IPNS. So David Price decided to shut down the CP-5 reactor in 1979 to chase the people over to IPNS. Some of them had already been using Zing P?, of course. At the same time, the laboratory decided to give up on the ZGS, and that was shut down in 1979 as well. So we knew that was coming. We didn’t know that the shutdown of CP-5 was coming, but the laboratory knew that the ZGS would be shut down. That happened, and IPNS had been able to pursue this and undertook the maintenance and ownership of the low energy part of the accelerator system, the 50 MeV linac and the 500 MeV Rapid Cycling Synchrotron then belonged to IPNS. We built the IPNS into the 375 building because that had been abandoned with the ending of the ZGS operation. We got a lot of shielding materials from the ZGS people. We got people from the ZGS operation and they brought with them the culture of ZGS, which is to help the users, to get their experiments done. There were some outstanding people from that crew. They were the technicians on the side. Don Bohringer and David Leach were leaders of this group. Others came over to help with the accelerator operations and for the technical aspects of the neutron source and the instruments. We gathered people from CP-5, Tom Worlton, Denis Wozniak and Gus Schulke, and Tom Blewitt and Terry Scott (who did cryogenics) from MSD. At the same time we formed this crew, the people were operating the neutron instruments, building instruments, operating the accelerator, and all working together. It was a culture that grew in service of the users. One of the hallmarks of IPNS is that we had such dedicated service to the users and such cooperation among the technical people on all sides. So that culture was very important to the final success of IPNS, and it endured for the life of the facility. And that culture made it fun to work. Maybe these statements about the culture of IPNS are the most significant observations that I will make in these interviews. We started out with four instruments on IPNS, with provisions for twelve. One easy one was a single crystal diffraction (SCD) instrument that we had operated as a prototype at Zing P?. SCD had a detector on it at that time that we built at Argonne Lab. Raul Brenner was one of the principals. Mike Strauss was the principal who undertook this design. It was an adaptation of the x-ray camera called an Anger camera because it was invented by a person named Harold Anger. The idea that Mike Strauss brought to this was, to make this into a neutron detector, and use a neutron scintillator instead of the gamma ray scintillator for x-rays. And this worked. That detector worked at the ZING P?, and we put that on to just basically move the ZING P? instrument into IPNS, so we had a single crystal diffraction instrument. We had two powder diffraction instruments, both of very similar design, worked out with Jim Jorgensen. We brought a chopper spectrometer, formerly called TNTOFS, from the CP-5 reactor, which I had worked on ten years before when I came to Argonne on sabbatical leave. So we adapted that to the pulsed source and called it LRMECS. Chuck Pellizzari, by then an Argonne postdoc and a Michigan former doctoral student of mine, worked with me on LRMECS and its ZING-P’ prototype. We had then the two powder machines, GPPD and SEPD. We had the single crystal machine. Pretty soon we built a small angle diffraction instrument (SAD) that we had operated at ZING P?. So we were moving things from the prototypes and borrowing instruments that had been abandoned at the reactor. The two powder diffractometers were basically built for our purpose, and we were working on building another chopper spectrometer. Eventually we filled up the neutron beamlines. In about 1985, Gian Felcher built a reflectometer for studying magnetic structures in thin-layered samples, layered magnetic structures. This was a unique instrument because it used polarized neutrons. POSEIDON (polarized observation of surfaces using extremely innovative diffraction of neutrons), POSY for short. It had apparently made such a mark that pretty soon everybody had these in the reactors in their house ware. Then it was also true that one of the physicists who was doing the small angle studies, Tom Russell (he is now at the University of Massachusetts) thought well, if it’s so good for magnetism, why don’t we make another one? He got some money to help do it and built one that was used as just as unpolarized beam of neutrons for studying thin layers of nonmagnetic materials. POSY-II. That turned out to be a big business. So we were filling up these things. Eventually we built another small angle scattering instrument, SAND, so we had two. It’s a very big business. Mel Mueller from the Materials Science Division was the principal in the early small angle scattering instrument with Ernest Epperson. Charles Borso from the biology division was also a party to this effort. And David Mildner, my former doctoral student, pitched in as a frequent visitor. They had been working on SAD in the prototype and brought some of the technology into the IPNS instrument. We knew we needed a certain kind of collimation to do this, but the usual small angle scattering instrument just had a series of pinholes, just two pinholes: a larger one and a smaller one. That was not a particularly efficient way of using a neutron source, so we wanted a multiple aperture collimation system. John Faber of the Materials Science Division and I worked out what we wanted to do, but we didn’t quite know how to do it. I think in the end it was Kent Crawford who suggested that we not do it in one go, do it in two pieces: one a horizontal piece, and the other a vertical piece. We can buy these things, and we did. That was effective. Although the collimation that you can get as measured by the fineness of the angle is not quite as good as you might ultimately wish, still it was good enough to do a lot of high-quality work in small-angle scattering and a pulsed source. Many people didn’t believe you could do small-angle scattering on a pulsed source, but we found out how to do it with the instrument called SAD — a small-angle diffractometer, not because it was unhappy, but that was what we called it. It turned out to be a very happy instrument. A lot of people loved it. It was such a success that we built another better one on the neighboring team, so we had two of those. Then there were two reflectometers. There was always an extra beam unused, so it went around from one use to another from time to time. Paul Sokol from the University of Illinois was the student of the Chairman of the Physics Department at the University of Illinois (the name won’t pop right now), but he sponsored this work somehow and Paul Sokol built another chopper spectrometer special for the purpose of measuring excitations in superfluid helium. Between that instrument and the second chopper spectrometer (at that time we had three), between those instruments they sort of cleaned up all of the things that the theorists were concerned about at the time in the field of superfluid helium and its mixtures with Helium 3 and the effects of temperature and pressure on the superfluid fraction, All of these things that the theoretical physicists had thought to ask, they pretty much cleaned that up. We weren’t the only people working on it, but the instruments that we brought to bear were very effective. That was a big success in the midyears of IPNS. Finally they cleaned it up and there was nobody asking more questions. I’m just imagining this to be the situation. Maybe people were no longer asking questions for the experimentalists to answer, so I think that the interest simply waned. There’s another aspect to this that I want to bring it up. We were hiring new people at the same time, and a young man appeared out of Iowa State University, who came to work with David Price, who put him to work on IPNS. It was Chun Loong. One of the first things he did was come work with me on the adaptation of the chopper spectrometer, LRMECS, which we brought over from CP-5 and for which I had assumed lead responsibility. He then took off on his own, and we’ve been collaborating colleagues ever since. He more or less on his own — I can’t tell a person of his stature what to do. He just made his own way and made enormous marks in the science using of neutron instruments. He marched from one instrument, taking responsibility for a year and he’d go over and do another kind of science on another one. We’ve worked together for all of these years. Now, we’re writing a book together.

Westfall:

He went off someplace, right?

Carpenter:

He retired and went to China.

Westfall:

That’s what I thought, okay, but you’re still working on the book together.

Carpenter:

Yes. Chun is one of the outstanding people we had with us. I’m going to miss mentioning a lot of people for reasons that I can’t really put these things in proper order. But Chun was one of the great ones, and still is. Now, there are some big events that took place in science that just happened at a time when IPNS was ready for it, or Argonne was ready for it. It was 1986, so instruments were pretty well tuned up with the people, and the teams were working well. The discovery of superconductivity at high temperatures in ceramic material was revealed when Bednorz and Mülle announced it in the late 1986. The composition of the compound is yttrium barium copper oxide, YBCO. I forgot the stoichiometry just now, but it’s always known as YBCO. The stoichiometry was known as to how many grams of this and how many grams of that, but the crystal structure was not at all known. It was an astonishing increase in superconductivity. People were working on getting superconductors above 23 K, and maybe get 23.1 thinking that’s a great advance, and all of the sudden here comes one that’s good at 77K. the temperature of liquid nitrogen. There was then not much known about it, but enormous attention was devoted to this, and Bednorz and Müller got the Nobel Prize for this announcement, and people went to work. Well, it’s a fairly complicated crystal structure. The preparation of the material was rather special. Dave Hinks and his group in the Solid-State Science Division knew what to do and prepared a sample of YBCO. They brought it to Jim Jorgenson, who put it into SEPD, the special environment powder diffractometer. Now we’re in December of 1986. It wasn’t Jim who published the results, but rather Mark Beno and his colleagues. It was Jim’s initiative and his instrument, but they were the first with the structure determination for YBCO, indeed on a pulsed source diffractometer because it has the resolution and the range of wave vectors required for the job. Müller needed actually to get the structure of such a complicated compound. They figured it out. They reported one day at the end of December in 1986, and the next couple of days the Japanese did the same thing using their powder diffractometer at KENS. as did the scientists at ISIS. I’m not sure the order of things; all I know is that Jim, that team, and SEPD were first.

Westfall:

KENS?

Carpenter:

Yes, that’s right. It means Ko Enerugi for high energy in Japanese, and is the name of the host laboratory, and Neutron Source. Anyway, so they had to put their powder machine on it. They had some material, and they got the same structure. The ISIS people did the same thing. This was all within a few days of the same time that Jorgenson was there first with his instrument. It’s Mark Beno’s name that’s first on the paper. That was a great success. The method applied to all of the variants of the YBCO structure, as people have been playing with the stoichiometry and different preparation methods until now. That class of superconductors is now operating at something like 125 K. Everything is known about the structures. In fact the magnetic structure in the material as well has been determined by neutron powder diffraction at the pulsed sources. So they’re perfectly adapted to this; and it has been a great business. People did the studies also with the reactor sources. I think that the range and the effectiveness of the pulsed source diffractometers have contributed most to the evolution of that field. About 1985, we were running liquid methane moderators but I knew from my Michigan experience that the solid would be more effective than liquid methane. So we installed a solid methane moderator to feed the beams for the small angle scattering instruments and the reflectometers. It really was effective, as I had predicted. They were really good. You could run them at temperatures of 20 K. We got to a lot of very cold long wavelength neutrons from them. That was the game: make longer wavelength neutrons and preserve resolution. But every once in a while, the moderator would spontaneously warm up, and then go back down again to low temperature. This was inconvenient because we wanted it to run a nice steady temperature. So it was doing this all by itself. Nobody knew quite why. It depends on whom you talk to and what the trouble was. Some people said you’re storing energy in the aluminum container. We know that if you radiate aluminum at low temperatures, energy is stored and the crystal lattice has defects. It’s unstable and it dumps out, every once in a while. This is one explanation. Other people said well no, air is leaking and it condenses on the cold surfaces and the oxygen portion under radiation becomes ozone, and then when the ozone and nitrogen concentration gets too big, then it explodes and gives off this energy. That’s the explanation, they thought. I suppose there were more explanations, like every once in a while you’ve got a blast of air in there and you get warm. There could be leaks, and all kinds of explanations like this. Or there was a chemical reaction between methane and the aluminum foam that we used to conduct heat from the methane. The temperatures would cycle up and down, and then the moderator container would begin to leak because of the heating up and cooling off. So frequently, we’d every once in a while have to replace the container, which we could do very easily, but it was awkward. So we thought well let’s get it colder. Let’s see if it helps. So I told Kerry Scott, who was our principal cryogenics man on the job (I should have mentioned him before), but he had worked with Tom Blewitt in the cryogenic irradiation facility at CP-5, and he was our cryogenics man at the time. He had his own theories about why this was happening, but I asked him well, just open all the valves and get it as cold as you can, then we’ll see what happens. He did, and managed to cool off the methane to 7 K, which is mighty cold in the radiation-heated environment! And it ran there for two whole weeks without these interruptions. We called them burps, we didn’t know what else it was or the cause of it. So it ran for two weeks, a whole operating cycle, without a burp. Previously they’d occur every day or two. It got hot, and then we’d cut it. This is what we have to do, run it colder, we thought. So at the end of the operating cycle, we let things warm up. As this system was warming up, it went bang, big time, in a huge mechanical explosion. It destroyed the box, destroyed the container for the methane. Fortunately there was not much radioactive material in there because we had been down for an hour or two, and that’s enough time for the Carbon-11 to decay. We would have otherwise had a shot of Carbon-11 to go up the radioactive gas exhaust stack, but it was pretty much gone. Anyway, this blew up the container, after which I could then say there’s only one theory left. No, there’s not enough energy in the aluminum to do this. No leaks and no ozone, no water vapor, none of that. It’s intrinsic to the methane, and I thought I will work this out. I had an idea, a theory of this. I worked day and night, in 1985 or ’86 it was. It was dedicated to figuring this out. I had the theory, kind of made up equations, which are an adaptation of equations that people used to describe the Wigner effect in reactors, although the origin of that is entirely different, and also an adaptation of Kaminetski-Frank behavior, that was a theory of spontaneous combustion. We put these things together. I had a compact theory that I thought might describe this. We had a great deal of data from records of the burping. We knew how much the temperature went up and how frequently it took place, and so on. I needed help at the next stage of things, as the equations are highly nonlinear differential equations. I needed somebody who could program a way to solve these equations. The method that applies to such highly nonlinear systems is called Runga-Kutta integration. But I needed somebody to help program that solution. There was a post-doc working in the Solid-State Division, with whom I had struck up a good relationship, Ulrich Walter, a German post-doc. I borrowed him for a while to make code. We put in some numbers that were guesses at the time. We started to run this code, and he had it rigged up so that you could follow a screen plot of the temperature history as you went. So he put in one set of numbers. Those just didn’t do anything much, so he put in some other numbers. At one point we found that the temperature would be steady, we’d kind of work along, and then pow! I was watching this thing on the computer screen as it was plotting the calculations live. I rose up right out of my chair when this thing went pow! There it is! When you look more at the theory, you could see well really that there are a lot of phenomena that we could assume and get some numbers for the phenomenon. We decided that it was due to the fast neutron damage in the methane molecular solid, it would knock the molecules apart. In the CH4, that’s methane, a fast neutron comes in and knocks out a proton, which goes careening through the rest of the material, knocks a lot more protons off. So you have a bunch of loose protons stuck in the methane lattice, and a lot of CH3 ions left behind. The protons, you can reason, recombined into H2 right away, give off their energy. But the CH3’s don’t. So I could calculate how many they should be, how much energy is given off when they recombine, and that’s it — then the numbers work. So I published a paper in Nature magazine about thermochemical instability of irradiated cold solid materials. On the basis of the theory and the numbers that we had, we could then manage the process. So we knew when we had to release this energy before it got out of hand. So we knew how to do that, and eventually we had a management system that was based on this theory. It turned out that we weren’t the only ones who observed this kind of behavior. At NIST reactor they had a solid D2O cold moderator, and it behaved in the same way. So that’s the explanation for it. It’s the defects of some kind that carry energy but don’t recombine until the temperature rises or the density of stored energy rises and it becomes unstable. So I felt good about that. I worked myself to exhaustion for six weeks in getting it all together, but that was it! There’s more to the story. It’s still playing out at ISIS. Anyway, that’s just very recent stuff that ISIS had trouble. They know about the low temperature stuff.

Westfall:

This is ISIS, and this is the pulsed source that’s in England some place?

Carpenter:

That’s right, at Rutherford-Appleton Laboratories, which is really a high-power pulsed source. It is what IPNS II might have been. So anyway, they had troubles of the same kind. But there’s another instability that takes place at higher temperatures, and we haven’t understood the basic mechanism of that yet. It also causes things to blow up. Hydrogen gas trapped in the methane comes out when the temperature rises above 65K. If that happens rapidly, the pressure rises and breaks the container. I find it a challenge. I hope I’ll have time to get into that one sometime. Mike Rowe and I went there a year ago to hear them out, and about nine months ago, to suggest what they might do. They’ve been working on that, but still don’t have the whole works. That experience, by the way, inspired a whole research effort in Dubna where Evgeny Shabalin picked up on this. Now he has a colleague working on him, Sergei Kulakov. He and others on his team set up an elaborate experiment to build a cryogenic moderator at the IBR-2 pulsed reactor. They learned a great deal more about this behavior. They’re very impressed with it, and they’ve made a great contribution as to general understanding, at least, of these things. We worked together, Shabalin and I. We were designing a test rig to put in the Penn State reactor, when Paul Sokol was still at Penn State. We had Paul’s collaboration, and so on. That never got built there, because Shabalin actually had only a year’s appointment at Penn State to do this. He made great headway designing things, doing safety studies, and so on, but the reactor people just dragged their heels, and so it never happened. Shabalin went back to Russia. Anyway, it was a good try, but he never got anywhere with the test rig. Instead Shabalin went to work at the pulsed reactor, his home reactor. There, he and his team found that mesitylene, a methylated benzene molecule, is both a good moderator and much less sensitive to the burping phenomenon. And they have recently tested and installed a system to manufacture and transport mesitylene pellets through the IBR-2 reactor. Now, a long time ago when I conceived of the accelerator-driven pulsed source, even at the time at the 1968 CINS Committee discussions, that what we might have as target material for the pulsed source a subcritical, enriched-uranium system. That was a recommendation that came out that the best looking new project would be pulsed spallation neutron source with a uranium target and with possibly eventually some critical multiplying system. That was always in the back of our minds. In the mid 1980s we started designing a multiplying target. That would require, as it came out with our design calculations, 78% U235 in an assembly with what the reactor people call the effective multiplication constant of .8. So this, in rough terms, would give you a gain factor, 1/(1-0.8), which is which is 5. In principle, you get a gain factor of five. I know now and I knew at the time you don’t get the full factor of five because of the basis for that — you don’t really have that basis. We did calculate that we’d gain a factor of 2.5, which is a big step. The trouble at the time in the mid-1980s was that we didn’t have codes to calculate things that included fission and neutron multiplication, which is what you’d need for calculating this. Although reactor people did this all the time. The thing is, the reactor people don’t calculate for systems that are far from critical. Anything with k around 0.99 is fine for them, but k at 0.8 is not, and we had to teach them that this is a subcritical multiplying system. You need to use the codes differently. Well, I finally got Roger Blomquist from the Nuclear Engineering Division to work with me. He was expert of these codes. He was able to coax the codes into doing multiplication in subcritical multiplying systems. He’s the one that helped work out the design and determine the operating parameters. Somehow, a very effective guy named Dick Prael, was also working on this. He was using the codes for neutron and proton transport that had been just under development at Los Alamos. So somehow between Blomquist and Prael we got the figures worked out for the IPNS Booster Target. The factor of two and a half is what we estimated. So we started out with the practical preparations for this. We needed to make discs out of uranium 78% enriched in U-235. There was no trouble getting this material at that time, but we had to redo all of the metallurgical work. We couldn’t do that one and risk uranium here at Argonne, so we did that development at the Y-12 Facility in Oak Ridge. Al Hins, who had been a principal of all of this work, spent a lot of time working with them to refine the processes for casting, machining discs, doing heat treatment, the cladding, and so on. The lesson, hard to transfer to the Y-12 effort, was how careful one must be to maintain clean surfaces to produce good clad-to-metal HIP bonds So we made the target discs of enriched uranium, and put them into IPNS. We carried out all the safety studies we had to do to get this all done, put the Booster Target into IPNS, and we got a factor of two and a half — more intensity. So it worked. Of course you had to rebuild the whole target system in order to incorporate the new target. So we did that. We put this in, and the intensity went up a factor of two and a half. Jim Jorgenson immediately began to complain, “You ruined my instrument! The resolution is just gone. It’s bad. Something went wrong here. Your multiplying target makes broad pulses.” The calculations didn’t say that you would make broad pulses, so I didn’t believe it. So what’s the trouble? We need a little bit more information about the broadening of these pulses. We kept trying to get that information, and we never did. They were never able to get definitive numbers on this. I convened regular meetings to discuss this, the Quidnunc committee (“busybody”, Latin for “what now?”) I was trying to get a handle on this, and finally I said, well, I’m not so sure, but bring the reflector assembly up. Take the target out. I think I know what the trouble might be. We had pieces of cadmium in here to prevent neutrons from coming back from the reflector, and making trouble broadening the pulses. So this is a good part of my patent, actually, a beryllium reflected source decoupled with cadmium sheets. So that was always the situation. I thought, well, maybe there’s something missing. We brought the reflector up. I dressed up in funny clothes, and reached down in where the target had been brought up. There were high radiation levels, but I expected that and I knew just want to look for. I reached down inside underneath in this cavity, and I could touch no cadmium in one place where it should be. Terry Scott, who had built this thing, would swear he had put the cadmium in. Of course he thought he did, but there was one piece he didn’t put in. You couldn’t see it because you had to look up from underneath to see it. So that piece was missing. And when we rebuilt the assembly with the cadmium in, the resolution problem was gone. So that was sort of an interesting little side light on the installation of the multiplying target. Another thing that I already knew would happen. Now we had a delayed neutron background. I’d recognized it in the time of ZING-P. Measured it actually. But I knew that that would go up with the multiplying target, and not just a factor of two and a half; it would be a considerable factor upon the two and a half. So this is a background, and the people doing sensitive measurements don’t like it. With Bob Kleb I designed the chopper that would knock down this background. We built it, put it in, it worked, the background was better than it had been with the non-multiplying target, so we cleaned up the reflectometer and increased the intensity by a factor of two and a half. People were happy that the pulses were not broadened for this effect, the delayed neutron background was suppressed, and a factor of two and a half intensity was there, so we were happy… Until about three years round numbers after we began running the booster, kaflewy! We detected fission products in the cooling water. What happened? We shut down the operation immediately. Then we thought the water might corrode the uranium if the cladding broke. What broke? Why did the cladding break? We didn’t know. We had a lot of alarm about having this enriched uranium object, and fission products in cooling water, and it looked bad. It looked really bad. So the laboratory appointed Harold MacFarlane, who would eventually end up as President of the American Nuclear Society for a time, but anyway, he was working on safety questions in Argonne West at the time. He came for the purpose of having to sort all of this through. I worked with him, I’d give him all the data that I could possibly feed him, and we could see that there was no hazard to the outside world, and there had not been any radioactive release, and there never were any critical hazards, and all of these questions were answered. Then in the end, he wrote a nice report that basically left us off the hook for having done anything untoward, but we never did get to put the enriched uranium back in because the DC people got all upset about it. People were worried more and more those days about controlling the use of enriched uranium, and efforts were going on to replace enriched uranium in all the instances where they could replace it with low-enriched, 20% enriched uranium. That wouldn’t do us much good, so we eventually gave up on the booster target. A little past 1985 a pretty interesting series of events took place. Outside of IPNS, I was out promoting IPNS and IPNS II, or working around the world in different places where people were pursuing these ideas. But when Cliff Shull retired at MIT, there was a celebration. I appointed myself as a laboratory representative to go to this celebration. At that time, Oak Ridge was promoting an advanced neutron source reactor, the ANS. Alvin Trivelpiece was at DOE at the time, and later became Director of Oak Ridge laboratory. The multipurpose DOE laboratories were contending for big projects, and he had made peace by getting everybody to agree that each multi-purpose laboratory would get a big project. Some people refer to tis as the Trivelpiece Treaty.

Westfall:

It’s in my advanced light source paper, yes.

Carpenter:

The point is that each of the major laboratories was to get a major project. It ended up that Brookhaven would get the relativistic heavy ion collider accelerator RHIC, and Oak Ridge would get an advanced neutron source reactor ANS. Berkeley Laboratory would get an advanced light source ALS, and Argonne laboratory would get an advanced photon source APS. Everybody went to work. That was the back-story to the event at Cliff Shull’s retirement. Cliff was really a much beloved character, and he had done great fundamental work in neutron physics and established neutron diffraction. He shared the Nobel Prize in 1994 with Bertram Brockhouse who had done comparable things, developing neutron inelastic scattering methods. These two people had done their fundamental work in the 1940s and ’50s. Anyway, Shull retired in 1985, and I was at this retirement celebration for Cliff Shull. So now you have my pressure for a pulsed spallation source, which didn’t fit into the scheme. We had five projects and four laboratories. The pulsed source didn’t fit into the scheme, and there was competition between the reactor neutron source as the next step in neutron source development and IPNS II, or its more advanced design possibilities. And I was promoting the pulsed sources. Here we are at the cocktail session at the Cliff Shull retirement. Two very well known and trusted colleagues, Mike Rowe and Jack Rush came up to me and each took my by an elbow. They started out at Argonne but ended up at Bureau of Standards, and they were the central people in that operation for many years. So I knew them well, and they came up to me took me by the elbow and they said, “We’re going to buy you a drink.” I said, “OK, let’s have a drink.” So let’s have a shot of whiskey, which we did. Jack says, “You’ve got to do something. You’ve got to give up pushing on this spallation source idea.” He says, “If you keep doing that, nobody’s going to get a new neutron source. You’re stirring up the works. It’s making confusion. It won’t do. You’ve got to give this up.” After a brief pause, I said, “Fine, I will give that up for the time being. When a decision is made for the reactor source, then I’ll have another chance.” So I agreed to give it up, and I gave it up. In the process, I agreed to become a member of the scientific advisory committee for the advanced neutron source at Oak Ridge, the reactor. It was supposed to be five times better than the existing High-Flux Isotope Reactor HFIR, and therefore about five times better the French ILL reactor. That was 1985 or 1986. I joined that committee. We met often, I don’t remember how often but a lot. The advisory committee would hear the progress on the conceptual design, the evaluations of performance, and the costs, the scheduling, and everything. It went on and on and on and on. In the early 1990s it became clear that we were not going to make the power goals. It was going to be more expensive than they thought, and they needed to use enriched uranium. So at some point, it must have been about 1994, these things became clear. The cost had gone up to more than three billion dollars, while the budget limit mentioned for the DOE project at one billion dollars. It had gone to three billion dollars, and rising, with doubts about the performance. Eventually, this weighed the project down, and it collapsed. Oak Ridge gave it up. They had by that time, a conceptual design report of many, many volumes that spread maybe 20 feet across a bookshelf. It just was not looking good. John Hayter, one of the outstanding neutron physicists of Oak Ridge, who had spent some time at ILL, was so dedicated to ANS that he just worked himself, I think, literally to death to see it succeed. He collapsed about the same time that the project did. He died of a stroke.

Westfall:

Because I also had read that there were additional problems that had to do with criticisms towards reactors, because it had just become a time when reactors were in very bad political favor.

Carpenter:

That’s probably true, that there were more and more misgivings about reactors as a general thing. Whether that influenced the decision, we were going to put a reactor on the big reservation where they already have reactors and decades and decades of experience, so that couldn’t have been the definitive reason.

Westfall:

It wouldn’t have been the definitive reason, but it’s part of the background, because you were asking to spend this huge amount of money for something that has that bad flavor. For example, as I say, reactors are… Eventually, somewhat later the HFBR reactor at Brookhaven was shut down because of its tritium plume problem, and it’s part of a whole difficulty that reactors have in a political sense, which means that funding a really expensive reactor will just intrinsically be more difficult than switching to build a kind of accelerator, which is more politically correct.

Carpenter:

Yes, just to go further with that and the public outcry after HFBR shut down because of the tritium leak, probably they compounded the problem by not being upfront. There’s a lesson in transparency there. But then earlier there were problems because they had discovered problematic shut down conditions in the reactor, which was cooled by water forced to flow downward through the core. If they lost pumping power, then there would be nothing to drive the flow, but natural circulation would force the coolant to turn around and go upward. The problem was that when they looked into this in detail, they could not guarantee that the reactor would survive the flow reversal. That caused them to lower the maximum power of operation.

Westfall:

This is at Oak Ridge?

Carpenter:

No, not Oak Ridge. This is Brookhaven. ANS had its problems, then the HFBR tritium problem, and all together the forces gravitated towards giving up.

Westfall:

So the Brookhaven reactor was shut down?

Carpenter:

Yes, the Brookhaven reactor was shut down. The tritium leak was a contributing factor. Then there were questions about transparency and openness about what to do. The past history of the flow reversal phenomenon is that they hadn’t appreciated it thoroughly until they looked into it later on. But they built the national synchrotron light source there, NSLS, which supported materials science research at that time. I can’t remember the sequence of the events. Maybe NSLS had already been built.

Westfall:

Actually, Brookhaven had both, plus they were trying to build the ISABELLE accelerator system. So they really had a lot of facilities there at the same time. But the point is that there were these problems going on in the background that were in addition to the issues that you were talking about to the ANS, which was the piece that was promised to Oak Ridge. To add sort of insult to injury, when the ANS was being cancelled, guess who was director of Oak Ridge?

Carpenter:

Al Trivelpiece.

Westfall:

[Laughs.] He had left DOE and had gone to be director of Oak Ridge at just the point that the project that Oak Ridge was supposed to get, they looked like they were not going to get. However…

Carpenter:

There was the Trivelpiece plan, yes. So I worked on the advisory committee for the ANS, keeping my mouth shut about the spallation source, for the better part of ten years.

Westfall:

But you had your day!

Carpenter:

I worked, and then the ANS project foundered, and Oak Ridge, who had been talking against the idea of spallation, because, they claimed, you couldn’t have a cold source and whatever else they could think of, all of the sudden they found that there was virtue in the accelerator-driven source. Meanwhile, working with many others, I had developed a proposal for Argonne, but not going out and promoting it in the world; I had just quietly done my work. We documented a proposal for an IPNS Upgrade based on a megawatt accelerator. It was a step beyond IPNS II. So that proposal was out there. We knew, in terms of the conceptual design evaluation, what we could do. Oh, and I have to tell, one of my best post-docs ever came aboard at that time when we were working on that design, about 1992, Dong Wook Jerng, a Korean. He came to me from MIT, where he had been a student of Neil Todreas, then chairmen of the Department, and he worked for me. He was supposed to come for a two-year post-doc, and I gave him the job of calculating the neutronics for the IPNS Upgrade project. He did it all in one year, finished the job, and went back to Korea. Now he has distinguished himself somewhere. He was fantastic, a fantastic post-doc.

Westfall:

So the ANS is going down.

Carpenter:

The ANS is going down.

Westfall:

You have in hand a conceptual design, which you had meant for Argonne?

Carpenter:

That’s right, it was meant for Argonne.

Westfall:

Okay, so tell us, then, how this proceeds?

Carpenter:

I don’t know exactly how it came about, but Oak Ridge inserted itself, said we’ll build a megawatt pulsed spallation source. Darned if I know what happened in the back rooms of the directorate.

Westfall:

Oak Ridge’s director decides to switch from ANS to a pulsed source.

Carpenter:

Yes, and they called the new project the National Spallation Neutron Source, NSNS. The director of the ANS reactor conceptual design project, Bill Appleton, continued over into the NSNS project.

Westfall:

You mean the SNS?

Carpenter:

The Oak Ridge version of the spallation source project was initially called the NSNS. There was a little bit of silliness going on here. After a few years of calling it the NSNS, somebody in Washington, a Congressperson or Congress itself, said, “You can’t call something the ‘national’ anything until we say so! You can’t call it the NSNS.” So let’s just call it the SNS. “You can’t call it the ‘national’ anything until we say so!” [Laughs.]

Westfall:

I thought it was too confusing because of the NSLS. But no, that never stopped anybody! [Laughs.]

Carpenter:

“No, you can’t call it national until we say so, and we didn’t say so.” Possible confusion with the British SNS, as it was called early on, but is now known ISIS.

Westfall:

So did this budding project draw in anyway on this?

Carpenter:

They went over to the spallation source I think in a rather orderly way. The same people as worked on ANS were now were doing SNS. I think there was dissatisfaction with Bill Appleton’s management, and he didn’t survive very long at NSNS. ORNL brought in Argonne’s successful APS project manager, Dave Moncton.

Westfall:

And this was after Moncton had successfully built and started the advanced photon source at Argonne, so it was up and running. And there was some kind of arrangement where I think he was on a leave of absence.

Carpenter:

That’s right. So anyway, Dave took over the project. He could see that this was now not just a job for reactor engineers, and that the Oak Ridge laboratory itself was not likely to be able to develop a team of accelerator physicists with the different categories of accelerators that you needed for this project. And they didn’t have the skill and experience with pulsed source neutron scattering and instrumentation and source technology. It was David’s genius to realize that the laboratory had no real prospect for developing a team to build this source. After considerable discussion it was decided that the accelerator would be based on a partially superconducting linac delivering protons of full energy, with a storage ring to compress the proton pulses. Yanglai Cho played an important role in these discussions. It was pretty clear that people at Lawrence Berkeley Laboratory had the experience with the ion source and the low energy part of the accelerator, and that Los Alamos people had the experience with the intermediate accelerator. Los Alamos had experience because of the long years of operating the Los Alamos Meson Physics Facility, experience with normal conducting so they could do the normal conducting portion. People with experience in superconducting accelerators rested at Thomas Jefferson Laboratory, so J Labs did that portion. They assigned the conventional construction and target systems developed to Oak Ridge and instrumentation to the Argonne laboratory. I think that amounts to six different laboratories involved in different components of the project, so it was probably a hard job to figure out how to divide the budgets and the scheduling and so on, but Moncton did that. At some stage later on, I think, some mistakes were found in the costing estimates — I think they were in the conventional construction area — that had not been included in the budget and there was no way to go back and say, “Okay, let’s change this whole project,” or we’ll increase the budget to make this up. This was not a three million dollar sum; this was like a $100 million. So Dave Moncton lost popularity, although it was he who figured out how to reorganize the project and delay depletion of certain aspects of it so as to make up for the money shortfall yet bring on an operating facility. So Dave Moncton was worked out of the directorate and Thom Mason was put in his place. Thom is a bright and upcoming young man. He probably had arrived earlier, but he inherited the project from Dave Moncton. He has done a marvelous job. Anyway, I had given up on a large-scale promotion of the spallation source while I worked on ANS, and I eventually came to the project as an advisor on target systems and instruments. The Argonne portion of the effort was such that they sent maybe 20 people, engineers and scientists, to Argonne for four years to work on the instrumentation and develop detailed designs based on developed concepts that were later made in detailed designs for the SNS instruments. That more or less worked, although I would say it didn’t work in the respect that the people caught the IPNS culture. I don’t think that they caught up. These were people brought from different environments into the SNS. Some of them were Argonne people; some of them were brought from other environments to form this instrumentation team. But somehow the culture didn’t get transferred. Even though they worked right beside the IPNS scientists for four years, it somehow did not get across. I could just observe that that seems to be the case. And then in 2004, after we had been at it for four years, they all moved down to Oak Ridge. It was all prearranged, and they moved down there and continued their work at that point.

Westfall:

Right, because I think it started operating in 2006.

Carpenter:

That’s correct.

Westfall:

So by this difference in management, I think what you’re talking about is that whereas at Argonne there had always been this user focus, that that didn’t quite transfer.

Carpenter:

It didn’t completely transfer. One of IPNS’s best, Rick Goyette became assistant on the SNS reflectometers, and moved to Oak Ridge. He had a similar position at IPNS, but he’s only one of fifty. He went there, and he carries the culture with him, but they don’t really have enough people who had this culture. As I say, I don't know how to inculcate an organizational attitude.

Westfall:

And of course you now continue…

Carpenter:

I continue advising, yes.

Westfall:

But before you talk about that, I would like you to say a little bit about the very last days of the IPNS, which was shut down in 2008.

Carpenter:

The end of December. It was in an instant, somewhere between the end of December 2007 and the beginning of January 2008. Some instant in there.

Westfall:

So it was a rather abrupt ending.

Carpenter:

Yes, it was abrupt.

Westfall:

But they had been talking for a while that it would be — I went to a user’s meeting, and they were saying the IPNS will be shut some year soon.

Carpenter:

Yes, that’s right, so it’s not without announcement or a suggestion that it was going to happen. Everybody knew this was going to happen, but everybody imagined it would be about two years away from when it actually did take place. Then it was very suddenly done. I lost out on this, in terms of my retirement planning, but I’m in pretty good shape, but I did lose out. A lot of other people were put through much greater pain than I ever suffered, and this was the real threat because other people had trouble finding new positions, and they were without work, and had their career in front of them. It was hard for them.

Westfall:

So before that happened, you had been going down to SNS.

Carpenter:

Oh, yes.

Westfall:

And now you continued to?

Carpenter:

Yes, the SNS paid half of my rate for some years during the time that the Oak Ridge people were here. I was supposed to work with them, and I did. So SNS paid half of my costs and Argonne the rest. Then when the SNS people moved back, I don’t know when they switched over or they ceased paying my costs. As long as I got my paycheck, I wasn’t too concerned with where it came from, and besides, I was always going to do what I was going to do, anyway.

Westfall:

[Laughs] Just like you have your whole career.

Carpenter:

So I worked a lot with the SNS people on target questions when the cavitation problem came to the fore and required a lot of thinking. Not that I solved the problem because I never did. In the early time there was a question of what kind of a target to make for the SNS. My design for the IPNS upgrade megawatt level was a water coolant system with tungsten metal plates — utterly conventional thing. But my friend and colleague, Guenter Bauer from Munich, with whom I had worked for years and years, he was promoting mercury. Liquid mercury had some clear values. It’s both a target material because of the high mass number of its atoms, its high mass density, and it’s liquid at room temperature, so you can just push it around. It carried off the heat itself; therefore mercury was very attractive from both points of view. But a little bit miserable in terms of chemical toxicity. It also is not such a widely used material that we had full knowledge of the behavior of structural materials in contact with mercury, so there were chemical mass transport questions, and radiation enhanced corrosion questions as categories of questions that needed to be addressed in addition to the detailed design of the system, like what alloys to choose. Then there was the further question people knew about was the mercury response to bringing a big pulse of protons. This produces a pressure pulse and there were concerns about the impact of the pressure waves inducing stresses in the containment shells for analysis of all the stresses, fatigue, other interactions. This occupied a lot of people’s attention. We knew about these things; we just needed people to work on them. In 1996, there was a meeting with Schruns, Austria called Liquid Metal Targets for Spallation Neutron Sources. The focus was on the mercury, or maybe liquid lead. I went to this meeting and offered a paper called “Prospects for Cavitation Damage.” I have a copy of this presentation, but the proceedings were never published, and almost nobody remembers this paper because that wasn’t the topic of interest at the time. But it was clear to me that when you get high pressures not only will that stress the container, but they resulted in a reflected wave or a rarefaction (negative pressure) wave, and cavitation bubbles will grow if the initiating pressure is high enough, and they were way over high enough. And I told what cavitation damage looks like if you exceed the threshold. But nobody paid any attention. Maybe because I have no official credentials in this field, which is true, but I happen to know stuff. While at Michigan, because of my early engineering background, I served on a number of doctoral committees on cavitation research carried out by F. G. Hammitt’s students: cavitation in flowing liquid metal systems. And my first published paper came out of summer work at Westinghouse Bettis Atomic Power Systems, surveying two-phase flow correlations. At that time, they were interested in reflected pressure pulses and the stresses and strains on the vessel, and so on. We worked on that for two years until the year 2000. In the year 2000 I was consulting with them about all of these questions. So about 2000, John Haynes, who was then head of the target design effort, came to me and said, “Say, we have this paper from Japan in which they’re hammering on bars of steel called a split Hopkinson bar, and they looked at the results of the pressures in the test region in the center of these bars, and they found the pits in the material. Does this have anything to do with us.” I said, “Well, give me a while to see.” So I did my best and calculated a comparison of SNS and the Japanese test conditions and I came back and said, “John, this has everything to do with us. We cannot ignore this. We have to work on this.” That started a new look at target problems, now going on for more than ten years. Finally, they still haven’t solved the problem. So that was my involvement. Maybe I don’t want to put in this paragraph that I told you so!

Westfall:

[Laughs] Okay, I think that’s up to now. Are we right up to…?

Carpenter:

We’re now somewhere in the middle of the ’90s, or in there, already to the year 2000 or so. We had to shut down IPNS in 2008, so we’re somewhere between two decades. But we can get back to this, okay? Is that what you’re suggesting? I see my voice is failing, so I can’t talk as much.