Oral-History:Pierre Lapostolle

About Pierre Lapostolle

Pierre Lapostolle studied at the École des Telecommunications before his military service with the Allied forces during World War II. After the war, Lapostolle joined the Paris laboratory of the French Centre National d'Études des Telecommunications (CNET), researching microwave amplification. His work on the traveling wave tube and particle acceleration shaped his 1947 Ph.D. thesis, directed by A. Blanc LaPierre. In the mid 1950s, Lapostolle joined CERN, the Conseil Européen pour la Recherche Nucléaire, and worked on the design and installation of its proton synchrotron in Geneva. Around 1960, Lapostolle's work at CERN transitioned to particle separation and then to machine improvements. He developed a new, longer accelerator, still in use by CERN at the time of this 1996 interview. The equations Lapostolle developed for this project became universal, incorporated into codes for machine design.

Lapostolle recounts the initial period of European nuclear research in the 1950s, describing CERN's development of gradient focusing using a quadrapolar field for linear acceleration. He details his work on this project and his management role in a team of about forty researchers. Lapostolle explains the technical challenges that shaped his 1960s design improvements, describing his use of simulation to model the "space charge" effect, the repulsion of accelerated particles. Collaborating with researchers at the Brookhaven National Laboratory in the United States, Lapostolle developed the RMS (Root Mean Square) equations for accelerator design. At the conclusion of his comments on CERN, Lapostolle assesses the improvements that shaped linear acceleration in the early 1970s.

Moving from Geneva to Paris for personal reasons in 1971, Lapostolle joined the Telecommunications Research Center in the chiefly administrative capacity of Scientific Director. During this period, he continued his work in the particle acceleration field, participating in the design of a small new linear accelerator (LINAC) and of a heavier circular accelerator installed for physics experimentation in Normandy.

Lapostolle retired in 1985 but continued contributing to projects at CERN and at Los Alamos National Lab in New Mexico. His post-retirement work focused on industrial applications of linear accelerators. He has considered linear accelerators as potential replacements for nuclear power and has continued to develop new equations to improve accelerator efficiency. In the concluding section of the interview, Lapostolle considers the evolution of particle acceleration and telecommunications as fields.

About the Interview

PIERRE LAPOSTOLLE: An Interview Conducted by Janet Abbate, The Center for the History of Electrical Engineering, July 22, 1996

Interview #296 for the The Center for the History of Electrical Engineering, The Institute of Electrical and Electronics Engineers, Inc.

Copyright Statement

This manuscript is being made available for research purposes only. All literary rights in the manuscript, including the right to publish, are reserved to the IEEE History Center. No part of the manuscript may be quoted for publication without the written permission of the Director of IEEE History Center.

Request for permission to quote for publication should be addressed to the IEEE History Center Oral History Program, IEEE History Center, 445 Hoes Lane, Piscataway, NJ 08854 USA or ieee-history@ieee.org. It should include identification of the specific passages to be quoted, anticipated use of the passages, and identification of the user.

It is recommended that this oral history be cited as follows:

PIERRE LAPOSTOLLE, an oral history conducted in 1996 by Janet Abbate, IEEE History Center, Piscataway, NJ, USA.

Interview

INTERVIEW: PIERRE LAPOSTOLLE

INTERVIEWER: JANET ABBATE

PLACE: FRANCE

DATE: JULY 22,1996

Education; CNET

Lapostolle:

After school, since I was not bad in Mathematics and Physics I went to École Polytechnique. A bad period, because it was during the war, it was in Lyon not in Paris. But it was a normal course. Afterwards, as still usual now, I joined a specialized school. I was by chance led to telecommunications, which was not disagreeable to me because I had liked electricity and Physics. Then I went to École des Telecommunications, which I did not finish because I joined the Allied Army. After the end of the war, I joined a laboratory in telecommunications, the French Centre National d'Études des Telecommunications, which was a new laboratory because there had not been not much activity during the War. One of the big subjects, a new thing at the time was the development of microwaves, which were used during the war for radar, but not much for telecommunications. No one knew that it would be so important for telecommunications. There was a new laboratory to work on the microwaves. I myself worked on the amplification of microwaves.

Abbate:

Where was this laboratory?

Lapostolle:

It was in Paris.

Abbate:

Is it still used for microwaves? Does it still exist?

Lapostolle:

No, it doesn't. It was a small place. Everything has been moved in another place of Paris, and in various places in France. New buildings have been built since, and the field of activities is now very large, but at that time there were only small places to house research laboratories. Then I worked on amplification.

Traveling wave tube; Ph.D. thesis

Lapostolle:

I started in '45, and in '46 came the idea of the traveling wave tube, by R. Kompfner in England. I decided to study this idea and try to understand it and see what could be done and to see what best to do with it. I derived a theory of the interaction between an electron beam and an electromagnetic wave showing that there could be waves, including particles where some of them could be amplified, some of them not. I got into certain conditions between the velocity of the particles and the focusing properties to such an extent that it was thought that it would be a good subject for a doctoral thesis. So, I wrote a doctoral thesis in 1947 under the direction of A. Blanc LaPierre. In fact, at that time, it was a tradition that the president, Louis De Broglie in this case, would propose a second subject and ask the man who was presenting his thesis to do some study on another kind of research, just to try to understand it and give a presentation. And the second subject was particle acceleration.

Abbate:

It was the president, Louis De Broglie who said, "Why don't you do this?"?

Lapostolle:

Yes. I did not have any knowledge of particle acceleration at that time.

Abbate:

But he saw the relation between it and the subject of your thesis.

Lapostolle:

Yes. The traveling wave tube has a lot do with telecommunication. Particle acceleration has no relation with telecommunication. But in the traveling wave tube, you have an electron beam, which you focus on an axis, and then you introduce an electro-magnetic wave around it, and in interaction with it. What you try to do is to transfer the energy from the electrons to the electro-magnetic wave, which is in interaction with it in such a way that the electro-magnetic wave is amplified. So that is what is done in the traveling wave tube. In a particle accelerator, you just do the opposite. In an accelerator, a linear accelerator, you have a machine on the axis of which you put a beam. In the machine, you put an electro-magnetic wave. But what you try to do, is to make it such, that the electro-magnetic energy from the electro-magnetic wave is transferred to the particles, in such a way that the particles gain energy and go faster and faster, get more and more energy. There's still an interaction between a beam and an electro-magnetic wave, but the transfer is done in the opposite direction. So that is why I guess Louis De Broglie proposed to me to look at this question and give some short presentation of it. But it turned out that I found it quite interesting. I didn't know it at all, I'd never thought about it before. But at the time, I found it interesting.

Abbate:

May I ask what happened to the traveling wave tube? Are they ever used?

Lapostolle:

Oh, yes they are still used. But of course there have been a lot of developments, and I don't know how the present traveling wave tubes are built. In fact, since solid state devices have been introduced, you can do a lot with solid state devices today. Traveling wave tubes are still useful for giving more energy, more power, than solid state devices. But today solid state devices can reach quite high power, and traveling wave tubes are less used today. They have been used even in satellites.

Abbate:

So they were used to amplify microwaves?

Lapostolle:

Yes. To amplify microwaves.

Abbate:

For repeaters or something?

Lapostolle:

Yes. Of course solid state devices are more interesting today because there is less noise in them than in traveling wave tubes. But they had been used for decades in this field.

So, that is what I did and then after having made my thesis I worked on the development of traveling wave tubes and did a lot of work on this subject.

CERN

Linear accelerator

Lapostolle:

But then in 1954, many institutions were joining in Europe on a new subject where Europe was late, compared to the United States. This was nuclear physics. Not only nuclear physics but high energy particle physics; and while there were many accelerators built or being built in the States, there was hardly anything in Europe. So all the countries decided to join to build a big accelerator and it took years to decide. Eventually it was decided to build a big machine, a proton synchrotron, in Geneva. That set up CERN, the European Organization for Nuclear Research. Louis Leprince Ringuet, a well known physicist who had been my professor at École Polytechnique encouraged me to join that venture. Before the official start, there was already a group of people working on that project and I joined them in Geneva one month before the final decision was ratified. The new machine of CERN was a big circular machine of 200 meters diameter. The beam was to be injected from a linear proton accelerator. From what I had done, I was asked to take part in the installation and running in of this injector. So I was asked to work on the linear accelerator. In fact, as I told you, there had been a lot of discussions before. In England, a linear accelerator of the kind which seemed to be convenient for that had been designed and was in construction. And so it was decided that a second, similar machine should be built. However, it was made slightly differently, using a new type of focusing. In order to accelerate the beam, you must keep the particles together because you can produce the electro-magnetic in some place. Of course you let that the particles move about the field, it's strong and it's convenient for acceleration. If you let the particles diverge, it is hopeless. So you have to focus. The smaller it is, the better. So you have to focus quite well. There was a doubt. This was not clear how to focus the particles well. There were ideas that were not very good. Sometime, not much before, CERN officially decided, and there came a new idea about focusing which was using a quadrapolar field. Should I explain?

Abbate:

You could just do it briefly.

Lapostolle:

Usually, in all machines, in order to focus one used an electromagnetic field. But usually one likes to have a thin beam, but a circular section, which is satisfactory. So one uses lenses having circular symmetry around the axis and electro-magnetic focusing, and that is what is done with electron microscopes for instance. One is using electron lenses. But just due to the properties of the electron and the electro-magnetic fields, there is a limit to the possibilities of focusing. Especially for heavy particles, the electrons focus quite well, but for heavier particles, like protons, which are two thousand times heavier, the particles are accelerated slowly. There's a strong limit and not much possibility of focusing. So the machine could not be very good. Then the new idea was to do something different. Instead of having a circular symmetry in the lenses, the electro-magnetic field can have various symmetries. It can be circular symmetry, but it can also be quadrapolar symmetry, there is some relation, focusing in one transfer direction, and focusing in the other one. Of course the sum is zero, but one might have expected that it was what one was expecting beforehand. But if one was made changing successively plus-minus, plus-minus, the average would be zero and of no interest. But that was a new theory, a wonderful theory, it was called A.G. Focusing, [unintelligible] Gradient Focusing. AG Focusing can produce very strong focusing, much more than one have ever suspected...

Abbate:

That was first used at CERN?

Lapostolle:

Yes, yes. That was first used at CERN for this new linear accelerator. So it was built with, of course some wonder about efficiency and would not one have some bad surprise. But the computations were quite certain and those no reason, but still, as I told you, one never knows about the effects, because the field was an imperfect field. Linear, exactly linear, and exactly aligned, perfect. In practice, it is never like that. One knew that the linearities could produce effects which were not here at all. Of course it was clear that there would be some non-linearity. It was difficult to know because it was difficult to measure and more difficult to compute also. So one was not really afraid, but still one was wondering whether it would really work.

Abbate:

So part of your job was to see if the actual accelerator worked the way the equation said it would work?

Lapostolle:

Yes. The construction was underway. And it was not really possible to change unless it would really not work. But still one had to put the beam correctly at the entrance, take it out and change it to going to the circular machine. And that I had to design and build with some flexibility to take care of surprises. If some would occur. But no change inside the machine itself, except changing slightly the field level or things like that. But not much. No change in the construction itself which was in a way hard to change. Anyway, it worked quite well. Not perfectly of course, but still much better than expected. So, and the machine, the big machine was a success. It came out...There was in Poland, two machines were built in Poland in Europe and in Brookhaven in the United States. The United States also used a new focusing system. We succeeded in CERN to be several months before.

Work environment

Abbate:

Must have been exciting. What was it like working at CERN? Were you supervising a group of people?

Lapostolle:

Yes there was a group, not much, about forty people.

Abbate:

That you were in charge of?

Lapostolle:

Yes.

Abbate:

Were you deciding what experiments to do?

Lapostolle:

Yes and took part in the experiments of course. Of course, such experiments were usually doing at night?

Abbate:

Why?

Lapostolle:

Because there was still some installation work around, it was done during the day.

Abbate:

So that's the only time it was safe?

Lapostolle:

And quiet also.

Abbate:

So you were working around the clock?

Lapostolle:

It was not done every day. But it was an interesting time. So that is what I did and that was finished in 1959 or 60.

Particle separation; machine improvements

Lapostolle:

And then in 1960 of course one had to operate the big machine and think about improvement of it, but also think about new projects. And then a lot of ideas were under discussion and some of them in which I took part was at the end of the big machine, from interaction with the target you get several secondary particles of different masses. Then you like to separate them to measure each of them independently from the others. In order to separate them, you need what is called a particle separator. Then was the idea that our microwave electro-magnetic wave could be useful to separate the particles. So I was asked to think about that and make some proposals and think about that for one year or so. In a half year I derived some ideas, some of which were used later at CERN. But then I was asked to take charge of another accelerator at CERN which was a smaller machine, circular machine, run by a single section. One might perhaps improve it, find some ideas to understand better what were the limits of it. So I was asked to take charge of this single section, which I did for a few years. Then started a development of the motor, to study the central part of it and try to understand better. But I did not like so much circular machines.

Beam dynamics and equations of longer machine

Lapostolle:

So apart from these few ideas I mentioned, I started discussions about a new big machine, much bigger, not 100 meter radius, but a kilometer radius.

Abbate:

Was that the Saturn machine?

Lapostolle:

Sorry?

Abbate:

Was that Saturn?

Lapostolle:

No, Saturn was a small machine in France.

Abbate:

What was the new one?

Lapostolle:

At CERN? That is the present accelerator at CERN, which is thirty kilometers in circumference. Which means about ten kilometers in diameter. That took time and a lot of discussion, political discussions about where and which country should it be. But still, work was going on and I was asked to look into it. There was still a linear accelerator to inject particles into this big machine which was of course much bigger than the one which was built. The one which I'd been working on was thirty meters long. The one which was in discussion was one kilometer long, or at least almost one kilometer long. There were a lot interesting problems. Among the problems was to be more sure about the beam dynamics in such a long machine. Because it was clear that the CERN was almost losing, not half of the particles, but a big fraction during acceleration.

Abbate:

So with a longer machine, the beam would disperse more?

Lapostolle:

Audio File

MP3 Audio
(296 - lapostolle - clip 1.mp3)

That was the fear. One would like it to not only not to disperse, but to lose no particles or hardly any particles. But it was clear that if one was losing ten percent or twenty percent it would still be acceptable. But one didn't like to lose much. But less than at CERN? And then there were all of the discussions for not so big machines, but still quite big. In the United States at Brookhaven and in Los Alamos where a big machine was in discussion not for injecting into a [unintelligible], but just to do physics with an accelerator. Then it was understood that the way the machines were designed and computed were rough, but good enough to produce what had to be done. If one was multiplying by more than ten the dimensions, one might like to have some better knowledge, so it was clear that the equations were not very satisfactory and there were a lot of discussions about another one. There were conferences in which there were discussions about it. So I thought well, this still is obviously wrong. It doesn't satisfy the electromagnetic field exactly. Since I was quite familiar with electromagnetic field, I was doing innovations on the relation between electromagnetic fields and other things and other waves, because I wanted to know what to do. So I derived a new set of equations. There were other people in the United States doing similar work but in fact my equations were adopted everywhere. So nowadays codes have been developed to make the computation for the design of a machine [unintelligible]. But the equations are what I made at that time.

"Space charge" problem and simulation

Lapostolle:

But another problem is what is called "space charge" problem. When you try to accelerate as many particles as possible, you have together, close together, many particles, billions and billions of particles in a small [unintelligible] of a few millimeters in dimensions. All of the same sign. So, particles of the same signs repulse each other. So the beam tends to diverge of course. So you have to focus it to keep it focused. That is the "space charge" effect, which is a divergence, repulsion effect. It was very well known. The trouble with that is that the particles repulse each other, but the field between them depends on the distribution of the particles. If the distribution was perfectly uniform, then the field would be linear. But particles don't like to keep a uniform distribution. Of course they don't like discontinuities. So outside the lens, there are no particles, so it must go down smoothly. As soon as there is anything like that, non-linearities appear. The "space charge" field is almost strong as the accelerating field, and the electro-magnetic field and the focusing field. So the non-linear field is also quite strong. I mentioned this AG Focusing [unintelligible] was good and was perfect, under the linear approximation. But when there are non-linear terms, it becomes much more tricky, and difficult to compute. But when [unintelligible] becomes the case, it becomes what is called nowadays chaos. In Physics a chaos is something that is very well known. This requires some non-linearities and the stronger the non-linearities, the stronger is the chaos. The chaos is such, as one can guess, is such that funny things appear. Particles may go anywhere, not knowing where.

Abbate:

So how were you able to simulate that?

Lapostolle:

So one knew that there could be something like that. One was not speaking about chaos at that time, but was speaking about non-linearities. One didn't know very well what to do, because one knew that particles don't like a linear situation and in fact it was not physical even. So there were high non-linearities. But which ones? Difficult to guess. One had to make some assumptions. One knew that first time would give some [unintelligible] of second order, third order. But what would be the order in the beam itself? So, what was done at that time was to try to do simulation. To do simulation, what one does, is that I was stating that in one bunch, one will have billions of particles, so it's not possible even on a computer to represent billions of computations. So one takes a few thousands or ten thousands. What one does is group thousands or millions of particles in one place. So instead of billions you have only a few thousands. It's not the same, but still...

Abbate:

You could use them.

Lapostolle:

So use that to do simulation. And that is what I did, in various ways at that time, was to treat a continuous beam, not a bunched beam, it was more complicated. To start with that, is what I did. Later on, I went to three-dimensions, bunched beams. But still, my....was that one started simulation, then there was not so much urgency because there were already codes, I had modified beam dynamics. So for the "space charge" ones I was studying simply not to be able compute, but one will already use a linear approximation, to see what it gives. I was not happy, but there was something, partly working. So, I tried to go further, do simulation, but started with a linear [unintelligible] continuous beam, two dimensions only, transverse dimensions, and do simulation. And study the "space charge" distributions very specifically. I was lucky, I was not allowed to do that and I was in close contact with people in the United States in Brookhaven.

RMS equations

Abbate:

Was that very difficult? I mean, did it take you a long time to find the right equations?

Lapostolle:

One can always write equations, but when one writes equations one has to do some hypothesis, some simplification. One is never sure that it is the right one to do. One has to check whether it leads to interesting results or not. I was lucky doing these derivations based on the simulations I had done to find that when using the what we call RMS equations, which is Root Mean Square. When you have a distribution it's difficult to tell what is the size of the beam, because it may go quite far. But if you use Root Mean Square dimensions, it is not ambiguous. And it turned out that thinking about it, I recognized that it used second order properties and energies. So thinking about that you realize that maybe you would find some relations which are energy properties. Which will be good ones, solid ones. So I derived equations which turned out to be quite useful relating the RMS dimensions of the beam to the various properties which in fact are good because they are some expression of the energy. So, I write these equations, which have been used many places. From them it was found a good way to design the accelerator, using three-dimensional simulation one is not sure but still one can check with these RMS properties, and it is satisfactory. That is what is used universally in the world today, these equations.

Abbate:

When you figured out those equations with the RMS properties, could you tell right away that was correct or did you have to look at data from the accelerator and see if it actually fit?

Lapostolle:

See, it was difficult to relate that to the accelerator, because it was not possible to do changes in the machines. In fact, there are also many effects which are difficult to take into account. Misalignments, things which are not there when done in the machine as one would have liked, and things like that, at least in the beginning. The idea of RMS we shared with Bob [unintelligible] and Brookhaven. The idea of using RMS. But, I knew from the simulation that [unintelligible] because I clearly, want to see something.

Abbate:

So were you in correspondence with other people working on it? You sort of worked it out together?

Lapostolle:

Yes. So that was the situation.

State of linear acceleration in the 1970s

Lapostolle:

And that was during the Sixties and even until June of 1971 that I worked in this field. At that time, the linear accelerators seemed to be satisfactory, at least for what was requested by the project. In fact, there has been a new linear accelerator to replace the initial one was built at CERN, in the early Seventies, using all the theories which were no doubt the best in the world. Well then, 95% of the particles were accelerated. Oh it was just wonderful. That one can make comparison between the simulation and the machine, and a good one. Nowadays, as I shall tell you, one knows that there are some things which would have been better, but still... It was perfectly enough. It's still enough. For more than 200 [unintelligible], which is much more than any other machine in the world. When one reaches 100 [unintelligible], he finds that is wonderful. But that is more than 200.

Telecommunications Research Center

Lapostolle:

So at the end of 1971, for personal reasons, my sons wanted to go to high school, and it was more difficult in Geneva so I went back to Paris. When I came back I was supposed to join back the Telecommunications Research Center, where I was the Scientific Director. It was a big center at that time, more than three thousand people, not all in Paris, also in Britain and various places. Fortunately, my job was...though I could be in touch with the work of research, I could not take part in it. I had more administrative work.

Abbate:

Were you deciding which projects would get done?

Lapostolle:

Yes, but no big project at that time. It was a time where France was quite late, because there were not many telecommunication lines installed and possibilities were limited. The discussion was all of the new possibilities which exist now in telephone. That was big in the discussion of that time. But the discussion are not being done of course. So, position was interesting, but I prefer to really take part in the work.

LINAC; accelerator design

Lapostolle:

At that time, I found out about the French work on an accelerator, where a new linear accelerator was built for the Station Synchrotron. At the time when I was working on this set of equations on "space charge", I had been asked to take part in the design of a new LINAC, a new linear accelerator. And also using this new set of equations that the state was aware of at the time. So, I was in touch with the particle acceleration work in France.

Abbate:

So was LINAC the first one that was built using your equations?

Lapostolle:

Yes. That's true. Yes.

Abbate:

So was that exciting?

Lapostolle:

Yes it was. It was a small thin machine, ten meters long. It was not so exciting as a big one. Then, in France there was a discussion about a new, heavier accelerator. So I was approached to take part in the design and construction of this new French heavier accelerator. That of course was a circular machine, but still it was new and complicated type of machine, a new focusing system and that interested me. I was ready to take part in that. It's something new and it would really work. So I joined. That was virtually built in Normandy.

[end of tape one, side one]

Lapostolle:

We found ways to use these electrons to produce focusing of the beam and change the length of the bunches and many new things. They are quite interesting. So, I joined that. It was finished in 1981. I took part in the running of it again, mostly doing that, which is normal. In 85, I quit. It was less interesting, because it operated regularly for doing Physics experiments which are quite interesting and quite good, but no development on the machine.

Retirement; industrial applications of linear accelerators

Lapostolle:

So I decided to retire. So I retired in 1985. But still I've kept a lot of contacts with CERN and with Los Alamos. And especially linear accelerators were still what was interesting me most. So I kept in touch with CERN, going there not every month, but one week every second or things like that. And going to Los Alamos three or four times for several months. So I'm working...but nowadays it is clear that what is interesting in linear accelerators is not so much injection into a big machine but use of linear accelerators for industrial applications. Actually the particles themselves are used for energy production. That is an idea which is not much developed so far, but which compared to a reactor would have a great advantage. Fission is something which happens in the reactor, you have to wait until it stops. But then you switch off and everything stops.

Abbate:

So someday you can replace a nuclear power plant with an accelerator?

Lapostolle:

One day. But also it concerns the transportation of wastes. The waste from the nuclear reactors is a big problem.

Abbate:

Right. This wouldn't be hazardous if you just had an accelerator?

Lapostolle:

Yes, but if you bombard the waste properly you can transfuse them in such a way that there's some reaction. There's still some radioactivity which exists, but only for a short period.

Abbate:

So you could put the waste in the accelerator and it would take high grade waste and sort of turn it into low grade waste?

Lapostolle:

Yes.

Abbate:

Now that's very interesting.

Lapostolle:

It would also produce [unintelligible], which I know is of some interest for the applications. Anyway, there are many ideas which are in discussion.

Abbate:

So is this theoretical or is there anything actually...

Lapostolle:

No, no. That is the ideas, and then one thinks about the Physics about it. But of course, in order to do that one would need bursts of particles from time to time to inject into a circular machine. But you have an intense beam. I was mentioning at CERN that one can reach two hundred [unintelligible] or more. But that is for a hundred microsecond, every second. So, the average current remains small.

Abbate:

So you'd need a continuous beam?

Lapostolle:

But then one would need a continuous beam.

Abbate:

Has that been done?

Lapostolle:

No. But that was done in some way in Los Alamos with this big linear accelerator which I mentioned. But just to the level of one millionth [unintelligible]. It is not one hundred percent satisfactory. Because if one accelerates such an intense beam, it is not ten percent one is ready to lose but one part in [unintelligible], because the particles lost tend to produce nuclear reactions on the way, and make transmutations on the machine itself. Something which cannot be approached.

Abbate:

Right.

Lapostolle:

Which is dangerous. One has to limit to one part in ten to the fourth. It's rather like ten meter when one must not lose more than a certain amount. But it's a completely new problem. So, being in touch with the people and having derived the equations and the theory, and I know that even the equations were wrong to some extent, but they were much better than the previous one. But still, not very good. So I derived new equations.

Abbate:

When was this? Or are you still doing that now?

Lapostolle:

No. That is done. New equations at least for the big dynamics. Also, taking into account the type of structure one is using nowadays for producing the accelerations and in the accelerator. The accelerator is made of...you have gaps of acceleration, and then the small cavities around. But then in the past, one tried to have regular gaps, a well-placed field which accelerates the particles and as regular as possible. Still one had to change the dimensions from time to time. That one never knew what to do. It was accepted that it did not matter, since it was not so frequent. So what I derived was assuming that it was perfectly symmetrical regular. So non-regular ones were not treated. In addition, nowadays one is often using cavities with two or three gaps, which are unequal. But it turns out to be more efficient. So one did not know how to treat that. So now there are new equations to treat anything. They are now being transferred to the syllabus and they are now at CERN and...well, it's going around the world, these new codes which I derived with the new set of equations. But now, it's a question of "space charge". We were always assuming that the beam was of circular symmetry, or some particular symmetry. But it is well-known that in practice there're always misalignments; it's not exactly symmetrical. What is also known nowadays is that these [unintelligible] defects, misalignment or anything like that, produce non-linearities. And that produces chaos. And chaos produces lots of particles. It becomes very important to try to master these questions. So there are theories which are developed about that. But what I'm working on is the way to simulate real problems, the effect of misalignment or any defect.

Abbate:

So you're working on new "space charge" equations that take that into account?

Lapostolle:

That is not finished. That is underway now.

Abbate:

So you've already worked out the acceleration equations and now the other half of the problem?

Lapostolle:

Yes. I shall present in Geneva at the end of next month at a conference the state of the ideas which...using mathematical properties which are almost forgotten, turn out to be extremely just what is needed in this case.

Abbate:

What do you mean "almost forgotten"?

Lapostolle:

I can tell you it's using [unintelligible] expansions. You've probably heard about Fourier series...

Abbate:

Right.

Lapostolle:

There's also Hermit, or Hermite I don't know, which are not used. But when you look in books you find one. Well, what is that for? Just for that. And you can make a three dimensional expansion, which nobody knew was possible. With no symmetry. It's not finished, but still it's quite promising.

Abbate:

So how did you figure out to use these Hermite equations?

Lapostolle:

By accident, by chance. Somebody mentioned it to me and I thought "What is it?" So I looked in books, but then I was thinking, well it's quite wonderful. [unintelligible] which are faster than any other ones. And maybe giving the possibility of studying by simulation of course. But if it is fast, you can use many representative particles. And see what are the [unintelligible] which are needed.

Computers and research

Abbate:

Did you find that your work was helped a lot by getting faster computers in? Were you always wanting a faster computer?

Lapostolle:

Well, of course you have to wait a night, 'til the next day to get results. Sometimes you feel like you would like to have it better. On the other hand, when you have to wait and you have the night, well, you can think about lots of things...

Abbate:

That's interesting.

Lapostolle:

But still, it is clear that if you have the possibility, even if it is not essential, the computation is fast, it is not a reason why you will do a new computation immediately after the end of the first one. You can still take time to think about it.

Abbate:

I was wondering if you could do more particles?

Lapostolle:

Yes.

Abbate:

Can you do millions instead of thousands?

Lapostolle:

Yes, yes. And especially what I mentioned, the possibility of doing some tests to try to reduce some...tests of seeing what happens when one does something either to fix [unintelligible] or to guide some theory, because it should be done also with theory. Simulation will still have its limitation. With a theory you can see what is the limit exactly and understand what happens. A simulation, you have some ideas but it's not the same. You need some theory too. But, simulation can guide the theory. Especially if you can do many tests. If any computation requires one week of computation, it's hopeless.

Abbate:

So did that change the way you worked?

Lapostolle:

Yes...well, you see I am not doing the theory on that. For the time being, I am trying to derive a good [unintelligible]. Afterwards, it will be necessary to use it. Then I think that the most active may be people in Los Alamos. Because at CERN, they are still interested in accelerators...accelerators for Physics. But for industrial applications...Los Alamos have been interested in applications or industrial applications, medical applications, so I don't know what...

Comparison of CERN and Los Alamos research strategies

Abbate:

So what were you doing at Los Alamos?

Lapostolle:

When I was there, I was discussing the development I was doing on this new beam dynamics or "space charge" problems. Having discussions and giving lectures and so on.

Abbate:

Was it very different from CERN?

Lapostolle:

No, it's not. It's a nice place too.

Abbate:

But it sounds like maybe they were working on different things?

Lapostolle:

No, but you see I was in a group which was mainly interested in theory, and the design of the machines, or in cavities or things like that. Also, simulation, beam dynamics, computations and design...no, it was...the final page is different. But this part was very similar. But there are more people still in Los Alamos than there are at CERN because at CERN nowadays, the machine is quite satisfactory. They like to be aware of new developments, but not direct, immediate applications.

Abbate:

You mean they're not working on new accelerators?

Lapostolle:

No they...well, they have a built a [unintelligible] LINAC to be injected in the big machine. So these new equations, well of course they are used for it. It was quite simple. They were quite interested to see, at that time, and to use it. But now it's in operation.

Abbate:

There is nothing new going on. Did you have anything to do with the super-conducting super-collider that didn't get built?

Lapostolle:

No. No, I was of course interested in it, but never worked on it. It is a new problem with big accelerators, when one thinks about using super-conducting accelerators which turns out to lead to somewhat more difficult problems in terms of "space charge." Well, not necessarily on "space charge", although no one never knows exactly, but in terms of intensity. Because when you accelerate a beam, you put high power into a cavity and try to transfer the power to the beam. But, for a long time, when I was considering that if fifty percent of the power goes to the beam, that would be wonderful. Nowadays at CERN, a particle can go up to seventy-five percent. But if you go higher, if most of the power goes to the beam, and not so much is lost in the cavity, if the beam changes a little, either source intensity fluctuates or anything happens, if most of the power is taken by the beam, then the extra power will go to the cavity.

Abbate:

Okay.

Lapostolle:

So they come into some instability. If you use a super-conducting cavity, you hardly lose anything in the cavity. So you have to find a way to avoid the difficulty. It is...[unintelligible] a risk, because the beam changes...

Abbate:

Because otherwise you've got all those powers going somewhere else?

Lapostolle:

But how to do that? Because it must be in [unintelligible]. And what may happen is that the field level will increase.

If you put too much power in the cavity, the stored energy will drop if it is not taken out from the beam. The stored energy will drop. Unless the beam comes back, ans does not find the right field, you come to some problem. So "space charge" is not well understood.

Abbate:

Did they figure out what to do about that, or do you think was still under discussion?

Lapostolle:

No it's under discussion.

Abbate:

Do you think something like that will ever be built?

Lapostolle:

Yes, someday it will be built. There are already some small machines which are built. Electron accelerators, super-conducting, but of small intensity so it's not very serious.

Abbate:

It's interesting, it sounds like it was very useful to CERN to have somebody like you with the microwave background. I wouldn't have thought someone from telecommunications would go work at CERN, but it sounds like that was a very good match.

Lapostolle:

Yes. In fact, I was not the only one experienced in microwaves.

Abbate:

That's what I was wondering. Were there a lot of microwave people going to work on that?

Lapostolle:

Yes, especially on the linear accelerator.

Abbate:

Are there a lot of people with telecom background there?

Lapostolle:

Telecommunication? Not really. But microwave, yes. There can be people expert in microwaves outside from telecommunication.

Abbate:

That's what seems unusual. Do you think that's been a change? Is it mostly physicists now working on that?

Lapostolle:

Yes, but for the time being it seems that microwaves were not very up to date subject in physics. But still in practice, it is of some interest.

Abbate:

So it wasn't fashionable for physicists to do that back then?

Lapostolle:

It was not, but there are still some.

Evolution of particle acceleration and telecommunications fields

Abbate:

That's an interesting combination. So would you say...looking back over your work in particle accelerators, has the field changed a lot in your experience?

Lapostolle:

You mean in particle acceleration? Yes. Sure, but the work has changed quite a lot since the beginning of CERN. That's for several reasons. One was that it was quite new, and there was not much reference...one had to exchange with hundred of other people. Nowadays there are more laboratories. But, still there are good contacts. Another is that many machines nowadays are much bigger than they were at that time. But the machines at that time were not homemade. There were small countries and small organizations involved, so there were good contacts with the people that were making the equipment. Nowadays, well, those two get contracts all over the world.

Abbate:

It sounds like early on you had to make some of the components yourself.

Lapostolle:

Oh, yes.

Abbate:

Is that different now?

Lapostolle:

There are still some components which are made on the spot, but not so many. So that makes a difference.

Abbate:

If you were a young student just going into this field now, what do you think would be the most interesting thing they could do?

Lapostolle:

You mean in general?

Abbate:

What do you think the exciting areas are today?

Lapostolle:

The ones in which I worked, yes. It's difficult to say. Well, mainly there are two fields in which I worked, telecommunications and particle acceleration. A young physicis today must think not only about today but also later on, decades from now. So it is clear that telecommunications is certainly part of electronics and computers, and no doubt it is a field in which there will still be a lot of development for years and years. And this is true also if one thinks about electromagnetic fields. One can also think about astrophysics or things like that, which is not completely unrelated. Particle acceleration is [unintelligible]. The atmosphere is different, because it's quieter, smaller than telecommunications which of course now is very big. It is an extremely large field. But the question is how long will it last? Because unfortunately, particle accelerators are bigger and bigger and more and more expensive. Someday there must be a limit. Europe is building a new big machine. But United States decided to stop, some five years ago, I don' t remember when exactly. They will not do anything more. Still, one wonders about what is the future? So, it is quite interesting now knowing this, but how will it be in five or ten years?

Abbate:

No one even can guess.

Lapostolle:

I would not give the direction to any young physicist.

Abbate:

Were you ever advising students or young people?

Lapostolle:

Well, of course I've supervised many theses. But the students came to me.

Abbate:

So they already knew what they wanted to do? But it sounds like if you, from your story, that if you have the right interest and expertise in a field, area like....

Lapostolle:

Well, of course I'm not the only one. I was lucky to arrive at the right time and the chance that I'd find the good ideas and be in touch with the right people.

Abbate:

It also sounds like your ideas were useful in a lot of areas in telecommunications and particle acceleration and energy production.

Lapostolle:

In telecommunications, you see, my thesis was on a new subject, and it was one of the first theories. The work of John Robinson Pierce was finished when I finished my thesis and it was clearly much more extensive than what I had done. My contribution was limited. But it was an opportunity for me to learn a lot of things. You mentioned energy production and things like that. Clearly, a lot has to be done.

Abbate:

Thank you very much.