History Home | Book Catalog | International Catalog of Sources | Visual Archives | Contact Us

Oral History Transcript — Dr. Robert L. Kustom

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.

Access form   |   Project support   |   How to cite   |   Print this page


See the catalog record for this interview and search for other interviews in our collection


Interview with Dr. Robert L. Kustom
By Catherine Westfall
At Argonne National Laboratory
May 6, 2009

 
open tab View abstract

Robert L. Kustom; May 6, 2009

ABSTRACT: Robert Kustom, who holds a Ph.D. in electrical engineering from the University of Wisconsin at Madison, is an expert accelerator builder. He has designed and built several accelerators at Argonne National Laboratory, including the Zero Gradient Synchrotron, the Intense Pulsed Neutron Source, and the Advanced Photon Source. He has helped build accelerators elsewhere, including the Spallation Neutron Source at Oakridge National Laboratory. This interview focuses on his work on designing and building the Advanced Photon Source, a materials science accelerator that was the most expensive Department of Energy accelerator of the 1990s. Kustomís chief responsibility was the RF system.

Transcript

Westfall:

This is Catherine Westfall, and Iím speaking with Robert Kustom. Itís the 6th of May, 2009 and weíre at in building 401 at Argonne. So you came to Argonne in 1958 from an industrial research laboratory that specialized in developing protective equipment for lightning and transience on power line transmission lines. What was its name?

Kustom:

JOSLYN.

Westfall:

Okay. As I understand, the group you joined in the physics division became the charter group of the high energy physics division. First you concentrated on developing high field magnets for experiments at Brookhavenís Cosmotron; later you worked on magnetic spark chambers spectrometers for an experiment at CERNís Proton Synchrotron PS in Geneva, Switzerland. Then you became a member of the spark chamber neutrino experiment at the Zero Gradient Synchrotron, ZGS, at Argonne. This work included developing large spark chambers, control systems, film scan tables, and digital track measuring equipment. In the mid í60s, you took an academic leave. This so at that point, you had a bachelorís degree?

Kustom:

Masterís from the Illinois Institute of Technology.

Westfall:

Okay. So you were from this area or relocated here?

Kustom:

From this area.

Westfall:

So you took an academic leave to pursue a PhD at the Electrical Engineering Department at the University of Wisconsin Madison. After your return from the university, you joined Argonneís particle accelerator division, working on electrostatic particle separators, RF traveling wave and superconducting separators, and particle track devices. You also participated in some of the early research on the resistive wall instability and the use of vertical damper systems. You became a group leader in the ZGS operation in late 1971, associate division director of the accelerated research facilities division in October 1973, deputy division director in March 1978, and director in August 1979, which was when the ZGS closed.

Kustom:

Iíd have to look at my records, but I think thatís true.

Westfall:

OK. I created this advanced photon source time line, and your name comes up when Yang Cho is talking about gathering a group to work on early APS designs. In between you worked on the IPNS. Right?

Kustom:

Right.

Westfall:

So, after the closure of the ZGS, you worked on the IPNS. I donít remember what you did at the IPNS.

Kustom:

Well, I think I was called project director. I have to look up what the title was, but basically I was responsible for the development and operation of the machine. By the time that we had a change in leadership on the program, the machine was not operating well. It had not performed to the expectations of the users and required a lot of attention, so more resources were going in to help the program to get on track and I was the one who was in charge of doing that.

Westfall:

Yes, I remember reading something that Ron Martin wrote about the difficulty. He referred to it as kind of a clash of cultures, that neutron scattering people demanded very high efficiency. But the accelerator builders were used to building state-of-the-art machines that had some start-up problems.

Kustom:

Well, clearly it was a culture clash. Most of the accelerator people came out of high-end physics. High-energy physicists were inherently more integrated into the accelerator development than neutron scattering people were. High-energy physics is basically based on how fast and how good an accelerator is, how high energy an accelerator can get to. And the data was taken over a long period of time, typically. I mean many experiments spanned years. So the fact that you stumbled along at the beginning and made sure that the kind of experiments that started were experiments that could tolerate start-up problems, and only after a few years would there be the experiments that needed more reliable operation. Material science people, biologists, they do their samples and they come running in and they make an exposure over a short period of time, and theyíre gone. Many of their samples, it turned out, were unstable. You know, if someone is using a deuterated sample, it doesnít stay deuterated forever. So if the machine is down long enough the sample because hydrogen again. That means the experimenter has gone through a lot of trouble and canít make the scattering comparison between that deuterated sample and an ordinary hydrogen lattice as planned. So when that kind of situation occurred, it certainly created conflict between the accelerator builders and users. After all, materials scientists were used to using reactors. Well, you want neutrons? You raise the control lines. You donít want them anymore? You lower the control lines and thatís the end of it. So it took a little doing to get the accelerator to the level of reliability that materials science users were comfortable with.

Westfall:

What do you remember of these beginning ideas for the design of the 6 GeV source?

Kustom:

I will say it was exciting prospect, but my excitement frequently was moderated by the reality that DOE can be capricious. They had just finished awarding the design of an electronuclear machine to CEBAF. It was fundamentally flawed and yet it was chosen over the Argonne GEM design.

Westfall:

And then the winning pulsed stretcher ring design was changed to a superconducting design.

Kustom:

Yes, it ultimately was changed to the design of a machine there wasnít a demand for. So Iíd say yes, the 6 GeV was an interesting project, one that I certainly wanted to work on, but I was not convinced that I was necessarily going to actually build it.

Westfall:

At least at Argonne.

Kustom:

Yes, at Argonne. You know, there was lively competition.

Westfall:

Did you have the sense that it was a science-driven machine? You said you were excited about it.

Kustom:

I knew exactly what it was for. I had worked on the IPNS and was close to the users, and after all, many material scientists are users of synchrotron light sources and scattering light sources. They had different kinds of probes, but they provide information often of the same kind of materials.

Westfall:

Well, there was some discussion during that era that it might be too soon to make the jump to a third generation source, Aladdin had been having problems and so had NSLS. Were you at all concerned that it was maybe jumping the gun?

Kustom:

No.

Westfall:

Good. I like definitive statements. Explain to me why.

Kustom:

Well, first of all, there was no fundamental reason why it should not work. Secondly, one has to push these things starting at an early stage. It takes a long time for a project to get approved and to get funded. So you have to start at these programs early. Itís a lot like building a bridge. You have to be prepared to build it, and to start well in advance. Unfortunately, itís not guaranteed that if you do start long in advance that you will actually ever get the project or that the project will go ahead. Then after you get it thereís no guarantee that they wonít take it away, like they did with the SSC. But the fact of the matter is, if you canít start with enthusiasm, you shouldnít do it.

Westfall:

What advance made it possible to build a new generation light source?

Kustom:

Well, improvements in undulator technology were extremely important. The prior light sources had not had that. But the key to making undulators more successful was is to make the machine lattice stable enough and at the same time the beam emittance small enough to function well. Some thought that it wasnít possible. But basically all the elements of the machine were within reach — it might have been a matter of pushing the state of the art, but they were not far beyond the state of the art.

Westfall:

Okay. So as an experienced accelerator builder, you took a good look at the back of the envelope ideas that Moncton, Eisenberger and others had, and you said, yes, this is do-able.

Kustom:

Yes. But I didnít say there wouldnít be a challenge. And there were people in the community who felt that it wouldnít work.

Westfall:

They were wrong.

Kustom:

But they were wrong.

Westfall:

What did experts think about the Berkeley Advanced Light Source proposal?

Kustom:

Lattice theory is not my expertise, but you can look at them and form a reasonably intelligent judgment. But there were lattices, very sophisticated lattices, that had been built here and abroad that was close to what was the requirement. So I didnít think that was impossible.

Westfall:

Okay. Thereís this interesting time — I spoke with Al Shriesheim. He says that when he made the deal with Al Trivelpiece, the Trivelpiece Plan, he felt like he had to act like Argonne had the new light source. But there were competitors, and if the leaders at other labs had gotten their congressional delegations behind them, someone could have taken the project away from Argonne. In other words, he didnít assume that because he had made this deal with Trivelpiece that it was a done deal.

Kustom:

Right.

Westfall:

What do you remember of the time when it was still possible that a competitor could take it away?

Kustom:

Well, I remember that SLAC was a competitor. People at SLAC were very aggressive about trying to get the machine. I think they were in a good position. They had a machine that was already doing light source studies. Thereís no reason to think that Brookhaven would not want the machine. They had a strong user base and therefore useful experience.

Westfall:

Brookhaven got the heavy ion machine RHIC in the Trivelpiece plan. And of course, Berkeley got the ALS. But there was Cornell in addition to SLAC.

Kustom:

Well, traditionally accelerator laboratories, like Cornell and SLAC both have had large, nationally funded facilities. They both had considerable experience. They both have synchrotron light source programs. So itís natural for them, I think, to want the APS.

Westfall:

But they didnít push it?

Kustom:

I wouldnít know how hard they pushed behind the scenes. Both of those laboratories are encumbered by the fact that they are high-energy physics laboratories. In both cases, materials scientists could have been stigmatized as being second-class citizens.

Westfall:

Moncton had very negative memories about being a user at SLAC.

Kustom:

Yes. I know of materials scientists who went through a long period of developing an experiment, and then the laboratory decided that they could not continue because the laboratory was a high-energy physics lab. So the laboratory shut off the materials science experiment and they had no recourse. And that could have very well have had an influence.

Westfall:

Well, by late 1985, the fix was in. By that time it was clear that Schriesheim had gutted it out, and that DOE was going to honor the Trivelpiece Plan. Schriesheim points out that various times the SSkC really helped.

Kustom:

Oh, yes.

Westfall:

Okay. Tell me why you think the fact that the SCC proposal was pending at this time really helped.

Kustom:

Well, itís somewhat speculative. But you start with the GEM project, which became CEBAF. Argonne lost that project. But towards the end, a lot of the Illinois politicians were brought in to support the program. I got the Illinois delegation was saying that rather than stop a project thatís been approved, letís try to make future projects more successful. I think that failure brought together the Illinois delegation. After that you have in this state two big projects proposed simultaneously, the SSC by Fermilab and the APS by Argonne. Nobody thought we would get the SSC at Fermilab and the APS here. I mean thatís just common sense. The taxpayer isnít going to put in money for two huge science facilities side by side.

Westfall:

Not at the same time, youíre absolutely right. Point well taken.

Kustom:

But then the SSC went down to Texas, opening the possibility that part of the Texas delegation would give us Argonne support. And I have the feeling that the Illinois delegation was a bit reconciled to the loss of the SSC.

Westfall:

Okay. I know from the documents that took a while to get funding for the Advanced Photon Source because DOE was preoccupied with the SSC. You get a little bit of funding, one million in fiscal year 1985, which comes from the LDRD funds, that is, laboratory discretionary funds. And then itís a while before you get the first chunk of construction funding. But there was this meeting of accelerator builders at Ames laboratory. At the meeting Yang Cho was urged to set the parameters of the 6 GeV. One of the things that happened at this time was that the machine was changed from 6 GeV to 7 GeV. Do you remember that meeting at Ames?

Kustom:

I didnít go to the meeting at Ames, but I remember it. When we really got down to designing the machine, it was designed it to go to 7, about 7.8 GeV, actually. That was the design energy that made sense with the undulators.

Westfall:

Thatís interesting. At the beginning, it didnít have another name, it was just the 6 GeV project.

Kustom:

Well, because people just kind of said, ďWell, itís about that energy.Ē Now, it sounds like if you make a change, itís something big and significant. It wasnít big and significant. It was just an actualization of overall performance. But certainly a 7 GeV machine costs more than a 6 GeV machine.

Westfall:

Because of the magnets.

Kustom:

Yes, it also means you need a higher magnetic field, which results in higher costs. It could have changed. But in the end, the conclusion was that it would significantly improve the quality of the science to raise the energy to 7 GeV, and it was demonstrated to do that. Itís interesting to note that the first machine built was 6 GeV and that was at ESRL. We went from 6 to 7, and the Japanese went from 7 to 8.

Westfall:

Now, ESRL is where?

Kustom:

Thatís in Grenoble France.

Westfall:

What do you remember of the design of the design phase? Yang Cho speaks of it as though you gathered the resident group of accelerator experts that had moved together to work on the ZGS, GEM, and the IPNS. Was it pretty much the same crew?

Kustom:

Very much. Yes. There were several of us who had worked on a number of machines. We all started on ZGS and worked our way to IPNS. Many of us worked on GEM, too.

Westfall:

What do you remember about the design process, maybe in comparison with the design of other machines like the ZGS.

Kustom:

Well, I think in many respects, work on the various machines was almost all the same. They all have certain features that become challenging. Individuals have different ideas about how to face those challenges. Eventually, out of that sort of chaos, some common sense prevails. Or in some cases you end up at a fork in the road that everyone ends up coming back to the same place eventually.

Westfall:

What were the particular challenges of the Advanced Photon Source?

Kustom:

Well, the lattice was particularly challenging an issue because to get the achieved results, the design called for operating very close to instability. The first thing that happens when the very first lattices were designed was that we ran simulations for standard tolerances for components. You can only build the magnets so accurately. Iíd say, ďOkay, what are the average numbers for this?Ē And youíd run a Monte Carlo, statistical program, to find out that the properties were within a particular range and so on and so forth. And after the first hundred tries you would begin to have an idea what the properties were. But it is unnerving to build a machine with tolerances that are that sensitive. The second major item in the machine is that we built it with individual power supplies for every magnet in the ring, with the exception of the dipoles. That meant you were relying on the precision of all the power supplies, but also you had to worry whether the lattices were focused right. Several hundred magnets had been built at Cornell. So itís not like we were making a tremendous leap in the unknown, but we were certainly going a little bit beyond what people had done before.

Westfall:

You were pushing the envelope a bit.

Kustom:

CHECK Yes. And thatís the case in a lot of the elements. The lattice issue ultimately got resolved by Ed Crosbie. He was a clever guy who got help from Don Khoe. Both of these guys were really good. Don had very good theoretical knowledge, and also good practical knowledge. Ed and he kind of came up with resolving the issue by making some basic changes in the lattice, making it a little —The original lattice tried to accommodate two different kinds of dispersions in the string sections. Ed concluded that the symmetry was better in the condition that had better at symmetry. Also came in with a phase advance, which was slightly altered and considerably more safe. At first people had their doubts. Those are probably the two biggest issues I can think of. And, of course, the other thing was to worry about the coupling of beams between the beam and the wall. We put a lot of effort into making sure all elements were basically minimally interacting. Ultimately, Iíd say today you can do more kinds of things that people worried you couldnít do. We had this vague feeling that we wonít be able to get the beam around for a mile, and weíd getting things going better and better and better, and then weíd finally get the right balance. Even though it seemed we could succeed. Theyíve made a lot of changes since those days. Weíve got much better control now.

Westfall:

To go ahead, since it seems like youíve teed up the question, what was the start up like? Youíre talking about whatís possible today in 2009, but when the beam starts up in 1995, was it a tough start up? The first stored beam was on March 25, 1995.

Kustom:

No, not by my standards, it was not a tough start up. For all start ups youíve got thousands of components that have to work, and some of them donít work. We had some major problems, but I think that we had a set up that allowed enough flexibility to create a three-part solution. We had enough people so that one group could keep the day-to-day problems under control well enough to keep the machine running. We had a second set of people who were developing a whole new approach to solving problems, seeing what was going on and figuring out how to make the problem go away. Then you have a third group thatís sort of in between those two, making sure we were not using up certain key components — they were finding people who could find these key components on a timely basis. So certainly there were some real concerns and crisis moments, but by and large the users were unhappy. Nothing like they did for IPNS.

Westfall:

This was the commissioning period, but letís now take a step back to the construction period. One of the things that I understand is that apparently the APS was built in a new way. I have heard for example that there was a wailing wall, with the schedules outside Monctonís office, so if you were behind schedule, you knew that everyone would know it because the schedule was in public view.

Kustom:

Basically, all of the technical people hated the system. Itís a natural thing to dislike for them. They like to play with their toy. They like to build things and fiddle around. The last thing they want to do is fill out forms on a weekly basis. Can you estimate how much you spent, how much did you charge to your account? I think it gets more and more difficult to get people to seriously deal with the project management accounting the further they are into pure science. Theyíre less disciplined, generally speaking.

Westfall:

As was the case in most of the big machines, you have the preparations for the user program happening as youíre building the machine so that the users are telling you how to — youíre getting input from the users about what the machine needs to do for their needs.

Kustom:

Right. The difficulty is that the Department of Energy and the project management system requested that work be broken up into a very small units. So you have to report on a very short period of time and small elements, which means that you have a huge number of reports to file. To some extent, when youíre doing a project that has a strong development element in it, thatís not what youíre doing. And project management programs are essentially trying to, of course, get everything done on time, but also they also try to map what people are doing, to make sure, for example, that 16 people are not asking to use the same draftsman at one time because then the whole system could collapse because you canít get your drawings out. So you want to do effort leveling. But the problem is that if youíre dealing with engineers and scientists who are developing things, itís a little harder to know how much effort theyíre going to need on some elements, because you start to design something and then you realize that the way you were approaching it wasnít quite right and you have to change things around. So itís not as clear or well defined as, for instance, when weíre putting in several hundred yards of pipe. Weíll need pipe fitters and riggers, and you can almost sit down and do that. Then you can say, ďWell, we canít give you all the riggers and that —Ē And you can envision sending these people away and bringing them back, ďOkay, today I need 60 electricians. Next week Iíll need 20, so you send 40 back at that time.Ē They work that way. But thatís not the way it works with engineers. If they get something done, you donít send them away because if you did that, youíd have to get new engineers. They would have to re-educate themselves. Or in fact if you send them over to some other place, they have to reeducate themselves on what they were doing. That would not work very well. So to a great extent, for most engineers and scientists you level off the effort. And if they fall behind? Solution: you double the effort. A given engineers will work 80 hours this week instead of 40. Thatís the way we did it, you know. And if an engineer or a scientist wasnít getting things done, that person was excused from the project and we found somebody who was more aggressive and got it done. Because of the different way that engineers and scientists need to be treated, the project management system was sort of alien to people. They werenít used to that, and they didnít like. They also didnít like to have to break things down into such small packages. Thatís not always practical. And itís hard to get good information into the accounting system. For example, I get about one hour into my program, my phone rings and someone needs an immediate answer on something else. I shove my original task to the side and I start looking at answering the question. And then I call them back, and before I get very far on my original task someone else calls and asks another question. Unless a person sits there with a notebook and writes down exactly how, like a lawyer, each 45 minutes is spent talking to a particular client, it becomes extraordinarily complicated to keep track of how time is spent. So thereís a natural resistance to doing work that way.

Westfall:

What was Temple like to work with?

Kustom:

Temple is a strong project director. Heís very forceful. He will listen. Like all individuals, he can be erratic at times. We all have our strengths and weaknesses. Sometimes heís going to demand things that are unreasonable. But after a while of thinking about it, he would see the reality of things and calm down, become more reconciled.

Westfall:

He said, ďGalayda, Kustom, and Marty Knot took the whole thing seriously and they did what I said. They did watch the schedules.Ē Now, by the way, you were working on the RF, Galayda was working on...

Kustom:

He was division director.

Westfall:

What was Knot doing?

Kustom:

Well, Marty Knot was working on control systems, but then he got into the personnel protection system. I think he drifted away from controls and did more on the personnel protection system.

Westfall:

Did you agree with sticking to schedules?

Kustom:

Yes, certainly I supported keeping schedules. Major projects that donít force you to keep schedules ultimately fall into disarray, and that can cost you a lot of money. Thereís this principle that says if you give an engineer any amount of time to do a project, he will fill that time. If you give him twice as long, he will fill that time. The problem is weíre trying to build the best possible product. You donít want it to fail and you donít want to make compromises. But in reality, no matter how long you take, you make compromises. So forcing people to make up their minds sooner and to take chances sooner — Iíll quote my friend, R. R. Wilson: ďIf in fact a thing works the first timeÖĒ

Westfall:

Itís over-designed.

Kustom:

Which I think is a stupid remark.

Westfall:

As an engineer! Responding to what a physicist says.

Kustom:

He almost didnít get the Fermilab machine to work. Itís a really complicated machine. You ought to try to get it close to right the first time.

Westfall:

So the APS was not built like the Fermilab main ring?

Kustom:

No, it was much more robustly designed. But aside from having said that, people do have to be pushed, because thereís always one more thing you can try. There are a lot of choices in some of these components when youíre designing something, and you have to take a chance. So some of that is necessary. I donít know if I totally agree with the DOEís desire to have things broken down into something like three-week packages. For some parts of the program, projects [inaudible] have to accept the fact of those you can and those who canít. Project management systems do have the problem if youíre buying some hugely expensive item, prior to the time it arrives on the project management system, everythingís fine. Youíre not falling behind because youíre not on a dollar basis. Thereís no way to say if youíre making it on schedule or off schedule. On the other hand, the day it doesnít arrive, your variance shoots out of sight. So you have to somehow accommodate those things, and they donít do that well, or at least they didnít at that time. But otherwise, having an integrated project management systems is required. Itís just a question of how you implement it. Now with regard to the power supplies, I actually told John Galayda, heís the division director and he was very nervous about the whole thing, I said to John that I expected that was an area we were going to have problems. He said, ďIf you expected it, why didnít you do something about it?Ē The first problem was I didnít know what would be the problem! [Laughter]

Westfall:

You just felt that there would be a problem with it.

Kustom:

No. What I will tell you is that power supplies are one of the components within the machine that is the highest stress, that in very high power, very high involvement, those are the things that a) are challenging, and b) when they fail, theyíre time-consuming to fix. Like they would take this whole room and blow it out the door, literally. When we had a failure, the doors were so warped from the explosion that no one could go in there in a day and fix them. But I had no doubt that, given time, we could solve all those problems. I only needed time to fix them. We had the reputation it wasnít a fly-by-day operation, so people believed us. But we still had problems. Some of the problems were remarkably intricate. For instance, there were the devices to protect the klystrons, a very expensive vacuum tube device. If in fact they were damaged inside, then you could lose a few hundred thousand dollars. So to protect that we have whatís called a crowbar, a device that shorts out he power supply, and so youíve got megawatts of power, it could only be shorted out in 10 microseconds. And in the process, it canít tear itself apart and it saves the tube, presumably it save itself. And you have a pool of mercury. There is a vacuum involved so itís insulating, and you stick a miter in this mercury, and when you detect that this fault is occurring, somethingís happening that you want to shut the power off, you give a command and there is a jolt of current through there which causes a little mercury vapor to pop up, and that shorts out the tube so that it shorts out the power supply. Now in the process of doing this, youíre going to have to know what to do with the stored energy. You have to do something with the energy thatís still coming at you, because that doesnít go away immediately. To recover, they put hollow pipes for insulators and they pump air through them and put the cold air on the bottom so the mercury can condense. And hotter air on the top to make sure the mercury can condense against the wall of the tube so when the discharge was over, it will all go back down into its nice little hole. Nice and simple. Except two columns didnít go hold up, then a third column: bum, bum, bum. So the third column, had the same pipes except they had no air flow. Well, whatís amazing is that chronometer would slowly build up inside the middle of these things where nobody could see in. The top part looks nice and clean, and if somebody pulled it up, boom! Some arc had taken place. Open it up and look for arc marks and couldnít find anything. And it wouldnít happen immediately. It would take an hour, two hours, a half a day. Well, sitting there, slowly the ozone builds up, until it finally hits the breakdown level. Thatís an incredibly challenging problem. I would like to say we were smart enough to figure it out. But what ultimately happened is one of the units finally had a problem and we had to take it apart. A guy looks inside and there were all these black tracks. It would be nice to have a working knowledge of whatís happening. There are a number of issues like that that were resolved and then the system worked fine. And the answer to Ed and also to my colleague and friend John Galayda,, I told him that we could put a crash effort on this particular problem with the power supplies. If they would give us machine time and let us work on it. I wanted two or four engineers concentrating on this single problem. And he clearly stated that no, no, no, the problem wasnít serious enough; we had much more serious problems at the time, orbit stability and some other things that were giving us trouble. Okay, John. Weíll continue to work on the schedule you gave us. So it took longer than expected.

Westfall:

I noticed that youíre not talking about Moncton very much. What do you have to say his management style? I heard that heís a delegator.

Kustom:

Yes. We hooked up together to work on the SNS. David is a really bright guy, excellent facility leader. And he does delegate. But heís smart enough to sit at the table of experts and to get the general gist of things and make key decisions as necessary. Now, heís basically pretty rational. At time he gets too fixated. Generally he was pretty good. Now mind you, when we down to SNS, I happily told him that I was happiest as an accelerator manager and he was busy designing the Guest House. I would just as soon not have him meddle in what we were doing. But he did design a nice Guest House.

Westfall:

Yes he did. What can you tell me about arrangements for users.

Kustom:

Well, early on Gopal Shenoy, who was responsible for the user end, felt that it was better if the users and the accelerator people were kind of kept apart. He felt that the users would end up giving a lot of extraneous and disruptive input. So generally, there wasnít a lot of fraternization. We had enough of our own to do anyhow. The accelerator people did have a lot of interaction with the XFD division, responsible for the beam lines and front ends and such.

Westfall:

So XFD stands for what?

Kustom:

Experimental Facilities Division. The problem with a machine like APS is that youíve got three thousand users, and not one of them has the same requirements, desires. And all of them think their program is more important than the other. So you can understand Shenoyís view. But you have to be careful. Some of them are extremely outspoken about what they want and you could end up with a hodgepodge accelerator design.

Westfall:

What do remember about problems with requirements?

Kustom:

There was an accelerator physics requirement. There is a phenomenon that can occur when you have electrons circulating where you ionize the gas, the residual gas. And you can have all these heavy ions being attracted to the beam, and conceivably causing dispersions and destroying the images. Now if you want positrons, you repel the ion beam, so the so-called ion cloud, doesnít occur. We were very much concerned about the originally, and we designed machine with positrons to prevent that.

Kustom:

It wasnít clear that the electrons could be a problem. Some machines didnít have that. But when youíre building these brand new huge machines, you want additional precautionary measures. The machine started with positrons, and then we decided to switch to electrons. But we had the facility go back and forth. When we designed the machine originally, we were prepared to go back and forth.

Westfall:

How much after commissioning before you went off to do other things?

Kustom:

Well, when the machine started to work stably, I took a leave of absence and went to Los Alamos to work on a machine that hadnít been working well for a couple of decades to see what was going on, just to see if I could find something to do.

Westfall:

The pulsed source, LANCE?

Kustom:

Yes. It has a peculiar instability which people didnít quite understand. It sounded like a fun thing to play with.

Westfall:

And then youíll get sucked back into the SNS.

Kustom:

Yes, I never finished my leave of absence. Beware of Monctons bearing gifts of wine!