Martin Harwit and Henry Kondracki

DeVorkin:

This is an interview for purposes of understanding an infrared Aerobee rocket payload.

Possibly we can start with Mr. Kondracki. Would you identify yourself; then we will go to Martin.

Kondracki:

Yes. My name is Henry Kondracki. I was the engineer for this particular experiment.

Harwit:

I am Martin Harwit, and I was the principal investigator at Cornell University on this payload.

DeVorkin:

Let’s start by asking the obvious kind of question. How did you two get together to build this? What was the decision process?

Harwit:

I had been a visitor at the Naval Research Laboratory for a year on a National Science Foundation-sponsored Fellowship that was set up in honor of E. O. Hulburt, who had been a former director of research at NRL; or at least, of the division of astrophysics. I think Friedman then became his successor.

The purpose of my fellowship was to start doing infrared rocketry. I was interested in doing infrared astronomy and the Naval Research Lab people were interested in getting a rocket program in that area. They already were strong in ultraviolet and x-ray astronomy, and wanted to round out their expertise.

DeVorkin:

Is there any reason why NRL had not done infrared before that time?

Harwit:

They didn’t have anybody on their staff who could do it. They had good people in the solid state division who had built infrared detectors, and we consulted with them a lot. There also was a young man there by the name of Blair Zajac, whom they wanted to train to take over a project of this kind. There were all kinds of personality conflicts, and eventually he left; and the project was taken over after I came back to Cornell from NRL by Douglas McNutt, and by Kandiah Shivandan.

The understanding was that during my fellowship we would jointly build at least one infrared payload for some rocket or other; and that we would fly it, share the results, and then after a year, I would go back; and then somewhat later, as soon as we had succeeded, we would sever the relationship. Then I would see what I could do at Cornell.

DeVorkin:

Yes. Is this how you met Mr. Kondracki?

Harwit:

Yes. Now what happened when I got down there was that, although the money was available (I mentioned earlier that whenever there was a problem, Herb Friedman, the division superintendent, would always say, well, it’s only money, and we could get the money), there was a limitation on ceiling points, that is, Government positions that Friedman could make available. So we didn't have the manpower, but we had the funds. Zajac, who had been there a long time, knew his way around NRL, and started getting together a team that he constructed by choosing the best people he had run across in the machine shop area and in the electronics model shop. Hank had been working with the machine shop as an engineer. I guess you should say what you had been doing.

Kondracki:

I remember the first time an order came over to Engineering Services Division requesting an engineer to work in cryogenics in the infrared field. Of course, at that particular time I didn't know what it meant (laugh). But anyhow, I was picked for the job, and had several meetings with Blair Zajac. Then Martin Harwit came aboard, and I had a few meetings with him. And I thought he was crazy.

Harwit:

I didn’t know that (laugh).

Kondracki:

It was something that I thought was almost impossible to do. But then, coming back and looking at the problem, I saw that it wasn’t that complicated. I guess that’s how our friendship began. We eventually got an experiment built for Martin Harwit at NRL, and we successfully flew it from White Sands Missile Range.

DeVorkin:

What was the crazy aspect of this? Why did you think it was crazv?

Kondracki:

The crazy aspect of this was that somebody wanted to cool something down to about -352 degrees Fahrenheit, plus to make it fly in space. I don’t think at that particular time anybody had done anything like that; or knew what kind of materials to use. And on top of that, you had to have a vacuum, or pump out the atmosphere where the system was going to work. So one of the things that bothered me was: what kind of lubricants; what kind of materials should be used. Because when you have things under a vacuum, things outgas. So I thought it was a big problem; and at that particular time it was. Today it isn’t, because we have materials, and lubricants that we didn't have back in those days. So it was a pretty big hurdle to get over.

DeVorkin:

Were you responsible for those sorts of problems, such as the lubricants?

Kondracki:

My responsibility at that particular time was project engineer for the experiment for Prof. Harwit.

DeVorkin:

When you had a problem like a lubrication problem, in what direction did you go to solve the problem?

Kondracki:

We went to several companies in the United States to see if they dealt with the problem; and most of them did not. So we went to high-temperature people, and we found out that they coated certain surfaces for lubrication at high temperatures. So what we did was apply that to cryogenics, where we would bake certain materials on the base material. And we found out that it would work. (crosstalk)

Harwit:

This was for bearings, particularly.

DeVorkin:

So to get the structure straight, you were ready to do technical research, applied to the design of this machine to make it work. But were you also in a position, if you found an outside commercial contractor, to contract with them to take care of that?

Kondracki:

Yes.

DeVorkin:

But you didn’t find one.

Kondracki:

Yes, we did. I don’t know the name of the company, but there was a small company in New Jersey, where they would coat bearings for high temperature purposes. They were coating them at that particular time with molybdenum disulfide; and we used that technique on the bearings and it was very successful at cryogenic temperatures.

Harwit:

Didn’t someone advise us at the Naval Station in Philadelphia?

Kondracki:

How we got onto this for solving that particular problem is that on a jet engine, when they throw in the afterburners, they condense a cone. The cone opened and closed; and it was galling at high temperatures. The materials would fuse together.

Harwit:

If you have two pieces of aluminum, a nut and a bolt, and you work it awhile it seizes.

Kondracki:

So anyhow, they would coat this cone with molybdenum disulfide; and we said, let’s apply this to bearings. The company in New Jersey would take the bearings apart for us, we would have them coated with this molybdenum disulfide, and put back together. We found out that we could get four times the life of a bearing operating at liquid nitrogen temperature, and in a vacuum.

DeVorkin:

Yes. Was this the standard procedure to look for outside commercial sources? You weren’t trying to develop new technologies internally, if they already existed?

Kondracki:

That’s correct.

Harwit:

We figured we had a big enough problem of our own; and anything that we could buy, we bought.

Kondracki:

Yes. You have to understand, in the Engineering Services Division we would do the technical stuff. What we would do was look on the outside and determine who had done it. If no one, then we’d do it inside. And basically, that’s what we were set up for.

Harwit:

You were short on manpower, also. They serviced the entire NRL.

Kondracki:

At that particular time there were about nine divisions at the Naval Research Laboratory, and one engineering services division. We supplied the project engineering aspect for each division, and of course, I was chosen for the Atmosphere and Astrophysics Division where Martin Harwit was located.

DeVorkin:

So then, lubrication was one problem. What was another problem?

Kondracki:

The other problem was to get to the altitude that Martin Harwit wanted for his data; in other words, to get above the (atmospheric) water vapor; is that correct?

Harwit:

Yes, and also the OH layers, and the ozonosphere. We wanted to get up to about 150 kilometers or higher.

Kondracki:

So at that particular time we went and talked to the NASA people who had very good data on Aerobee rockets, and for us to reach that altitude, we had to have a certain payload weight. And so with that in mind, we had to design this thing, that was supposed to operate in a vacuum at very low temperatures, to be light enough so that we could reach the proper altitude to take data. Well, I think we successfully did that, because we flew the experiment.

DeVorkin:

Is this why your initial reaction was to say, it’s crazy to do that? That it would be impossible?

Kondracki:

The answer is yes (chuckles). I looked at him for a while, as a matter of fact, during the first meeting; and I just sat there, and thought, well, either this guy is a genius, or he’s crazy. But anyhow, after looking into the problem, and when the experiment was finally built, it really wasn’t that hard, because I think we accomplished everything in a period of under 12 months.

DeVorkin:

Let me ask Martin a question. To what degree were you aware of the technical problems, and how to overcome them when you proposed this experiment?

Harwit:

I think I probably wasn’t aware of the bearing problems. I thought everything else could work. I initially had thought of making it rather smaller, more compact perhaps. I previously had sketched up a conceptual drawing with dimensions for the optics and so forth. And I knew what we wanted to do; namely, to build a telescope that had a wide field of view, and a mirror six inches in size, something of that order. I thought that the whole thing could perhaps be made somewhat smaller. I think, if I remember correctly, that Blair Zajac was more ambitious. He wanted to have a somewhat bigger instrument, and that made it, I think, slightly more difficult to engineer. But on the other hand, we had a fall-back position to something which could have been quite a bit simpler.

I had built small instruments at MIT, which I cooled with carbon dioxide-alcohol mixtures down to -80 degrees Centrigrade. And so, I didn’t feel particularly upset about the cryogenics. Liquid nitrogen didn’t seem to be a big problem. We talked about going to liquid helium temperatures right away, and very quickly gave that up, figuring that liquid nitrogen was going to be hard enough, and that was all we needed for the near infrared to cool the telescope down so we wouldn't see the side walls.

DeVorkin:

Let me clarify then. Was the package that we are talking about, or eventually will be talking about, was it initially designed for liquid helium?

Harwit:

This one here was, yes. But we didn’t operate that way because we didn’t have it sufficiently stress-free. It got leaks in it, which would allow the helium liquid to evaporate into the vacuum chamber, and that would give us a lot of heat conduction. And we would immediately lose the liquid out of the liquid container.

DeVorkin:

Right. I just want to identify this, though; the first instrument that you are talking about that had these design problems, was not this one.

Harwit:

That’s right.

DeVorkin:

It was a previous instrument.

Kondracki:

At that partiuclar time we didn’t know enough about liquid helium to actually use it in the correct form, as you are speaking about liquid helium today.

Let me explain another thing about the particular size of this experiment we're talking about here.

DeVorkin:

This particular one, not your first project.

Kondracki:

Yes, this particular one here; and the reason why this size was used was due to the thermal properties of the material. When you use cryogenics, you have to think of a thermos jug; and the best material to use for a thermos jug is glass. Of course, you can’t use it in a rocket, due to the forces; so we had to go to stainless steel. We were working with temperature gradients from room temperature down to liquid nitrogen temperature. We had to have certain distances, such that given the boil off rate of liquid nitrogen, it would not be gone when it was time to take data. So this means that certain parts of the experiment had to be built so as to make it act as an efficient thermos jug. This means that the amount of heat going into the system had to be controlled so that it wouldn't boil off all the liquid nitrogen, so that the detectors would not warm up and could actually get the data that Martin was looking for. So the configuration has to do with several things: the boil off rate — how strong you want it to be — and the endurance. That's how the system came about.

DeVorkin:

Okay.

Harwit:

Let me refresh your memory on something Hank, because the first payload we actually built at NRL was not the Aerobee. You remember the Atlas?

Kondracki:

Yes.

Harwit:

We built a payload that was supposed to fly on an Atlas missile piggyback in what was called a pod. Blair Zajac had heard about this offer from the Air Force for experiments to be flown piggyback on an Atlas. That meant you had something like a two-and-one-half foot diameter metal tube that fitted onto the side of the Atlas, and would go up with it. This was an Atlas that was to be test fired for defense purposes. We never found exactly what they were doing. Blair Zajac had found out that this was available, and we had talked with Friedman about taking advantage of that opportunity, because it would allow us to build something where we didn't have to skimp too much on weight.

In fact, we had three telescopes built into one enclosure, one vacuum housing. They looked in three somewhat different directions out the side of the pod. Each of them was separately cooled from a central liquid nitrogen reservoir. Rather than having the telescope built into the reservoir directly, I think the liquid was somehow circulated through it.

Kondracki:

We had temperature sensors in the photometer.

Harwit:

Yes, but I don’t know how we circulated the liquid. Was it gravity-fed?

Kondracki:

No, it was pressurized in the central container, then as liquid was required in the photometer, it would flow through.

Harwit:

That’s right. It would flow the liquid through and cool it off, so that the photometer itself didn't have to also be the container.

DeVorkin:

Yes.

Harwit:

We built all these things in tremendous haste, because we were supposed to get the whole thing built in four months, and fly it. It then turned out that they waited another 21 months before they ever fired it. First, they decided it wasn’t going to get fired from Cape Kennedy, but from Vandenburg, which I think was just getting opened up at the time. And they would have to ship the Atlas back to Vandenburg. Here we were, halfway through my year there. We had built this payload, really working incredibly hard. I remember, one day when we were there, all of us were working feverishly; and outside, the whole city was standing still because of the funeral of President Kennedy.

DeVorkin:

That certainly dates that.

Harwit:

Yes, November 1963. We were going night and day there. It didn’t matter whether the whole rest of the world stood still, we were going to get this off the ground. Then it was delayed, and we went in to Friedman and said, look, we’ve finished this off. We don’t know when the hell we’ll fly this, and we’d really like to do something. So he said, okay, even though this hasn’t flown, we’ll put it aside and you can start in on an Aerobee.

Now the weight consideration came in that Hank was talking about, but we had already solved some of the subsidiary cryogenic problems before.

DeVorkin:

First of all, did the Atlas ever fly? What were its results? And then, in your Aerobee package, what were the results leading up to the eventual design that we have in front of use in this ‘67? Can I identify this as the 1967 payload?

Harwit:

Yes. The Atlas did fly. It flew after we had already had some Aerobee shots under our belts. It was a complete disaster. We had been asked to build a payload that would have a hold time on the ground of at least four hours before flight. And we had been assured that this was going to be a nighttime launch. When it actually took off, it had been held up for six hours on the ground, and dawn was breaking. In the middle of the night, we asked whether they could take the instrument off, so we could at least fly it on some other mission. They said, no, they needed it for the ballast. The rocket was balanced; they couldn’t take it off any more. So this thing got launched. It still had liquid nitrogen in it, and after about a minute, it lost that also, just before the vacuum covers on the telescope were blown off, or maybe a few seconds after the covers were blown off. So we got absolutely no data after all.

Now, on the early launches, on the Aerobee, we also had a number of disasters. The first one, I think, the rocket didn’t despin. It was a rocket that had weights that were to be released, and it was supposed to have a slow roll.

DeVorkin:

Was this as early as 1963?

Harwit:

This was, I think in the fall of 1964 or early in 1965. By that time I was back at Cornell. We never did launch an Aerobee while I was still at NRL, but shortly thereafter.

Kondracki:

I think it was maybe a couple of months, or a month. You left and went back to Cornell, because you had to teach.

Harwit:

I think it was maybe in October or November. I could look it up.

DeVorkin:

I could ask you later in another session what Cornell thought of all of this. I am interested in that, but not for now.

Harwit:

Okay. We made a common mistake. You are new to a field. You are given standards parts: a standard parachute package, a standard despin package, and attitude control system. You always have this confidence that the people who hand you this have had lots and lots of engineering experience. And these things never fail. In fact, they always tell you they never fail when they hand them to you.

Well, this particular despin mechanism had weights in it on a string that were supposed to spiral out and those weights were held in by some explosively-retracted pins. We didn’t realize that there were different sizes of explosively-retracted pins and put in ones that we got from some place. Sure enough, they retracted, but they didn’t retract quite far enough. There was something like, maybe a 32nd of an inch hang-up. Things didn’t despin, and because the rocket was now spinning very rapidly still from the spin stabilization it had received at launch, the chopper motor couldn’t spin up. We had had a chopper that was along a radial axis. In the payload we are talking about here, the chopper is on a vertical axis. In the payload we are talking about here, the chopper is on a vertical axis.

DeVorkin:

In the ‘67 payload.

Harwit:

It’s on a vertical axis, and all of our rotating choppers thereafter always were on a vertical axis, because, when you have a gyroscopic force on a horizontally rotating chopper, the centrifugal force is at odds with the spinning of the rocket.

So the motor couldn’t spin up. We didn’t get anything. And we had a number of failures until, I think, the third or the fourth Aerobee finally worked properly, and then we started getting data.

Now, this payload — the ‘67 one came after that.

DeVorkin:

So you are saying that you had gone through this disaster with Atlas and three or four different Aerobee payloads that didn’t work for one reason or another. Were these all Aerobee problems, or was it ever instrumental?

Kondracki:

No. First you have to understand that this was maybe the fifth or the fourth year, I believe, for the Aerobee rockets to fly, where the Government said the scientists can use this as a tool to collect data.

Harwit:

Is that right? I don’t know when it started.

Kondracki:

It hadn’t been very long.

DeVorkin:

Well, I’m confused. The Aerobees did send up payloads for NRL people as early as 1949.

Kondracki:

True. But they didn’t have to slow them down, they just had to fly them; get them up there. Friedman was one of the people, Dr. Chubb, was one and they were looking for x-rays. They didn’t have to point in any particular part of the sky. They just wanted to see if they were out there.

DeVorkin:

Tousey’s did.

Kondracki:

Tousey was after us.

Harwit:

Well, he did something before, but maybe not much. The attitude control systems were relatively new, I think. Aerobees had been used for all kinds of atmospheric purposes, also. I don’t know what the history is.

Kondracki:

They had some successful shots with the despin. This is the yoyo despin that Martin was talking about. They had several teams at White Sands. And if you got an inexperienced team to check things out, while you did the horizontal, obviously then, it would be a failure. I think what happened is that we were a new group going to White Sands. We weren’t identified at that particular time with the Touseys or the Friedmans. You know, who’s Martin Harwit, he’s not a famous scientist coming down. He was starting off young; and I think, probably that Murphy happened to get into the system, and we had a couple of failures.

DeVorkin:

Murphy’s Law.

Kondracki:

Murphy’s Law.

DeVorkin:

I wanted to know who Murphy was.

Kondracki:

Right. So we started to get experience. We know our package works. Now, we started looking into the other people; like for an example, does telemetry work. We actually went and sat, and yes, it works, because we could see our signals coming out. Does the despin work? We’d have that checked out: yes, it works. Are you using the same squib puller — it’s the explosive portion. So after this amount of experience, then you know what to look for. The rocket business was young. And obviously, a lot of mistakes were made.

Harwit:

This is not unusual, incidentally. Friedman was telling us at the time; don’t expect your first one to be successful. I think most new payloads took several tries; or most new groups in particular, took several tries before they succeeded. In our case, it was a makeshift group. We didn’t get the support of the regular launch crews that Friedman and Tousey were using, because they fired off shot after shot as time went on. In fact, there was initially a little friction between the support crew that Friedman and Tousey used and us, because they sort of figured we were a bunch of hacks.

DeVorkin:

In their terminology, what does a hack mean?

Harwit:

Well, I think we were inexperienced. They figured that we couldn’t get along without them, because they had all the experience. We’d have loved to have gone with them. On the other hand, their services were not made available to us.

DeVorkin:

Who made that decision?

Harwit:

Friedman. Friedman didn’t have the manpower to fly his own rockets, fly Tousey’s rockets and fly ours. So we had to come up and start a new team; and these more experienced people would look at us and clearly see that we weren’t as experienced as they were. They thought we had a lot of gall, I think. And then at the same time, there was a little jealousy, I think, because they felt we should have turned to them. Of course, we didn’t have that option.

DeVorkin:

You're talking about the White Sands group?

Harwit:

No, I’m talking about the NRL people. See, because Friedman had built a support group within the Astronomy and Astrophysics Division. But that was not available to us, because he didn’t have enough manpower there to service everybody. And we went then to the Engineering Support Division, to Hank, for electronics, John Reese.

Kondracki:

Let me try to say something here. Dr. Tousey was in the ultraviolet, and Chubb was in the soft x ray business. So now, we’re coming up with a new business, infrared. So they don’t want to spend their time on it, because it is out of their field. So when a new business came in, you tried to supply as much help as you could; but there were not enough experienced people in that division to say, all right, this is what you do when you go down to White Sands, because they were busy working for Tousey and Friedman.

DeVorkin:

But clearly, somebody at NRL wanted to bring in infrared.

Harwit:

It was Friedman, yes. And we had a lot of encouragement from Chubb. This was carried out in Chubb's branch. But forming a new group is difficult, and particularly difficult if you come in as an outsider. There was quite a lot of friction between Zajac and myself. I had a number of friends at NRL who had been working in other divisions, and when they heard I was coming down, they asked me whom I was going to work with. And I had been told that Blair was rather difficult to work with; and I in fact asked Friedman, as a result of that, whether he couldn’t suggest somebody else; and he said, no, this is the guy whom we have. He is very keen on this. He has made a bad start here; but it would be nice if he could inherit this business, and he’s energetic. He wants to do it very much.

So we started out, and it lasted for maybe six months. It just became impossible. I think it was partly because of the frenzy of the work, which always brings out some of the most intense feelings between people. We just weren't working well together at all.

DeVorkin:

Well, is it the thing you see in NASA today sometimes when there is friction between project scientist and project engineer over who's running the show, anyway? Is this what you’re talking about?

Harwit:

I had been appointed to run the show. But I think Blair resented it in part, because he had been there longer, and he tended to be somewhat secretive about things. We worked long hours, but I had a family, so I would go home for supper. And the crucial decisions would always come up when I was away. If I came in and objected, he would say, well, you wouldn’t want us to hold up the work while you were gone. It’s a well-known tactic; and we were in a very, very big rush, so it was difficult to say, yes, I would like you to hold it up, or you could have called me up at home, or something.

Eventually we got to a point where he made some decision which in fact led to the collapse of one flight dewar that we had. That in itself was not sufficient reason for anything. But I talked to Friedman and Chubb then, and said, look, it just doesn’t work. So then they assigned Kandiah Shivanandan and Doug McNutt to the project. Blair Zajac did other things, and eventually went out and joined a computer software programming firm, I believe.

DeVorkin:

Okay. But during this time you and Kondracki were in this all the way through.

Harwit:

Yes.

Kondracki:

Yes, but no comment. But from the engineering standpoint, about being green going down to White Sands, we learned a lot prior to lift off in 1964.

DeVorkin:

You’re on your second or third flight.

Harwit:

No, just the first flight.

Kondracki:

When we went down to White Sands, we didn’t get real good support like Tousey or Friedman got. But that helped us out later on in life. In other words, when Martin went back to Cornell University, we found out now that we could have a fill system on the tower.

We could fill the experiment with cryogenic fluids, either liquid nitrogen or liquid helium, and with the experience we had from previous mistakes working on the tower, we developed efficient systems to fill the dewar prior to liftoff. So that means we had 100% cryogenic fluid in the experiment. So I think that after looking at the problems we first had on the rockets down at White Sands, that it was helpful for us later on, on maybe the second, third and fourth generation of experiments.

Harwit:

In addition to that, also, I think the first time you work with a new facility, you have to find your way around. You generally have a feeling that these people know what they are doing. White Sands had been in the rocket business a long time. They’ve launched any number of Aerobees down there. We would concentrate on making our part work. They would concentrate on making their part work, and the whole thing will be a success.

As you get more experienced, there are two things that happen. One is that you become more familiar with your own payload. You bring the right tools for checking it out in the field with you. You know what you will need. You can do it much more rapidly. You have more confidence in it.

And you start looking at what the other people are doing, because you’ve lost a lot of confidence in them. They just are not supermen. And I think you get to have a certain, not cynicism, but questioning attitude, which when you first come out you don’t have, because you think these people must be experts. It’s the most common mistake, I think, that one makes.

With graduate students at Cornell, I find that that’s the last thing they always seem to learn, that other people who have been in the field many, many years make mistakes. They as graduate students ought to be able to catch them. They always think that, no, it’s the other way around, that they couldn’t possibly catch a mistake of anybody else.

DeVorkin:

That’s very fascinating. I can draw many parallels like that. Let’s move on to the actual decision process that designed this instrument then, if you can. In other words, where is its place in the evolution of infrared technology that you were flying?

Harwit:

Let me say something about the scientific side, and then have Hank say why we designed it this way. We wanted to build a liquid helium-cooled payload.

DeVorkin:

Let me interject one thing. Had you had, at the time, a successful nitrogen-cooled payload?

Harwit:

Yes, in fact, we had had a successful nitrogen-cooled payload just before we flew the Atlas. Maybe it was the third one. I think there was trouble with the second one still. I’m not sure. I remember not being that upset about the Atlas, because at least, we had had one decent flight.

I felt I wanted to keep the design as close as possible to the liquid nitrogen-cooled rocket. I wanted to get into the business quickly. By that time at Cornell we were really competing against the NRL team.

DeVorkin:

So when you went back to Cornell, an infrared team remained at NRL.

Harwit:

That’s right.

The arrangement that we made was that I would continue being in charge nominally, at least, of the team — although Doug McNutt actually was in charge on the spot — until we had had one successful liquid nitrogen-cooled shot; and that we would then part ways.

DeVorkin:

Was there an astronomer on the team then?

Harwit:

Doug McNutt had gotten his Ph.D. in physics at Wisconsin doing optical things, and Kandiah Shivanandan wanted to get his Ph.D. at Catholic University, I think it was, involved in an astronomical experiment in a rocket.

So when I got back to Cornell, I had a phone call one day from Nancy Roman, asking me whether I might like to start a rocket program, myself.

DeVorkin:

1965, 1964? Okay, I’d be interested in that.

Harwit:

It must have been early 1965, I think, very late 1964 or early 1965. I think I could probably find the letter. So I said, sure, I would want to, and we talked about the finances of it. The whole thing then eventually got started a few months later. It was just about the time when we were getting our bearings at NRL on the liquid nitrogen-cooled payloads. Doug McNutt was beginning to design a liquid helium-cooled one, also. Of course, we wanted to be first in flying a liquid helium-cooled one, as NRL did, also. And it was clear that this was where we were each going to have our own baby.

Thinking back now, I think I figured that by staying as close as possible to the earlier design that we had had at NRL we would do well in beating out the other group. We had a side-viewing system. Doug McNutt was in favor of an axially-viewing system. But that required blowing the nose cone off at altitude. I didn’t want to do that. I wanted to blow off just a smaller door which we knew how to do.

DeVorkin:

What was the danger in blowing off the nose cone?

Harwit:

It was just a much bigger piece, and by that time I was kind of suspicious of anything that we didn’t build ourselves, or had direct control over.

Kondracki:

There were no techniques at this time available for blowing off an Aerobee nose cone. Some studies would have to be done, meaning probably a year’s delay.

Harwit:

In addition to that, it had to carry the vacuum seal off with it. But above the atmosphere, there was no pressure on the rocket; nevertheless, you have to have a seal that was tight enough to prevent air from going in at any of the altitudes, and yet could be yanked out explosively.

DeVorkin:

Let me turn the tape over. This is Tape No. 1, Side No. 2.

You were talking about blowing off the side cover.

Harwit:

So, rather than having to blow off a nose cone and a telescope cover altogether, if I remember correctly, I thought we would just stay with the side cover, and we would try to build something very similar to what we had built with the liquid nitrogen temperature telescope.

DeVorkin:

Was there any consideration of the kind of measurements you actually wanted to make that dictated the design, whether the nose cone was to be blown off or the side blown off, or was this purely a technical problem?

Harwit:

It was mainly a technical problem, because the attitude control systems could have allowed us to point in any particular way we wanted to, whether we were pointing with the side of the rocket, or with the face of it.

Kondracki:

Martin, at this particular time with this experiment, weren’t we just scanning the night sky? You weren’t looking at any particular source.

Harwit:

No, by this time we did have a guiding system on the rocket, I’m pretty sure. I can take a look at the diagram we have in there which shows the overall stacking of the whole thing. (leafing through plans) ACS (attitude control system) Unit and Platform. So we did have that, yes.

DeVorkin:

So you were interested in looking at specific objects.

Harwit:

Yes.

DeVorkin:

Which were they?

Harwit:

I don’t remember. Orion was in there some place, I suspect. In fact, the publication attached to the brochure has it in there.

Kondracki:

Martin, I think you wanted to look at the zodiacal light early in the morning.

Harwit:

We may have wanted to do that. That's quite possible.

DeVorkin:

You mean the dust.

Harwit:

The zodiacal dust, yes, because we had been interested in that problem. I had done some theoretical work on it, so it is quite possible.

Kondracki:

With the front cover off, the cryogenic system, or the thermos jug, or the dewar, had to be designed such that we could look out from the side — let’s see, just refresh my memory now. This was done at Cornell?

Harwit:

No. We went back and forth and designed it. Then you had this place at Alexandria, the sheetmetal shop.

Kondracki:

What happened was that I didn’t have the facilities to build the dewar; neither did Cornell University. It was done by Sulfrian Cryogenics, a company located in Rahway, New Jersey.

Of course, the unit had a fill and vent line. It had a pump-out port for the vacuum system. It had an inner wall that holds the cryogenic fluid, and then an outer wall that held the vacuum. The inner chamber housed the stator, the baffles, the mirror and the detectors.

DeVorkin:

The stator, again, is?

Kondracki:

It’s the stationary part of the chopper which cyclically interrupts the light to give an AC signal to the detectors. And that basically took care of the cryogenic package. It had to fit in the housing of the Aerobee system. And in this particular case, I believe we are using the 20-inch can, as they call it, or a 20-inch extension. And we had to cut holes out such that we could mount the door, have the cover mounted, and have vent ports that are also on the skin of the rocket.

DeVorkin:

The extension itself, the vent ports, all of the outside material; is that aluminum?

Kondracki:

No, all the outside surfaces that are supplied by Aerojet General, that makes the Aerobee rocket, is made out of magnesium. All the parts that we use on the skin are aluminum. The Dewar itself — if I didn't mention this — is made out of stainless steel.

DeVorkin:

That is quite clear. So, I’m just looking at this diagram.

Harwit:

Now, there’s one part that’s not shown here. There are two things that Hank mentioned that I just might mention. The way that we chopped the radiation was by having a stator, which was a stationary part with radial spokes in it, and a negative of that which was a chopper, so that when the chopper was in one position, it could exactly cover the holes in the stator. And in the other position when it was completely open, it exactly was shadowed by the stator. Half of the radiation could then enter. So we immediately threw away half the light, but it was a simple way of chopping the radiation simultaneously over the entire primary mirror.

DeVorkin:

Right. Did you have an internal source that was calibrating in that chopping cycle?

Harwit:

Yes, there’s a pick-up that tells you where the chopper is at any given instant, and it provides a phasing signal for what is called synchronous demodulation in the electronic processing of the signal.

The second thing which Hank mentions is the pop-out baffle. We had a set of beryllium copper strips, a number of which I have brought down. In fact, enough I think, so that one could reconstruct one. The beryllium copper was bent into a circular form, somewhat like louvers on a window, which are not flat, but are bent. These were designed in such a way that one could fold them up, place the vacuum door over the entire system, and then when the vacuum door was ejected, these pop-out baffles, spring by spring, would jump out and form an entire cylindrical extension to the telescope.

DeVorkin:

Like a light baffle.

Harwit:

It’s a heat baffle that cuts out radiation from the fin and from the warm parts around the aperture. So these baffles were kept at liquid nitrogen — we had hoped at liquid helium temperature. They would warm up, maybe slightly, in flight, but not very much. We had calculated what those effects would be, and the detectors would be seeing an entirely liquid helium-cooled telescope.

That was the whole point, because we wanted to see very weak infrared signals from the outside. Now, the problem with this design was the door. And Hank, maybe you can tell us a little more about it.

Kondracki:

Any time you have a cylinder and then try and cut another cylinder through there, it’s very hard to do; and especially make the unit vacuum tight. So it was very important on how the groove was cut, so we could place an O-ring in this area here. Then when you applied the door to it, it had to be precisely the same, with maybe around five or ten-thousandths of an inch clearance. In this way it would make the seal by an elastomer, an O-ring, which was put in a groove here. The type of O-ring we were using here was a silicone, which is red in color. This means that it can operate at lower temperatures.

DeVorkin:

The idea was that you would maintain the vacuum inside the chamber until you were in the vacuum in space, where you could blow this off.

Kondracki:

Yes.

Harwit:

That’s right. Yes.

Kondracki:

It would be equalized pressure. Then you would blow the door off.

DeVorkin:

Was there a sensor on the instrument that knew when the pressure was equalized, and said you could blow it off? Or did you just calculate?

Harwit:

It was a timing operation.

Kondracki:

It was calculated timing, because the people that would operate the Aerobee would say that at a certain time after lift-off with a certain weight, with proper burning, the rocket would be at a certain altitude. If they missed that by, say, 20 or 30 miles, it wouldn’t make a difference; and so everything seemed to work out pretty well that way.

On the external part of the rocket, the things that would fit below our experiment were the telemetry and the attitude control system. Then on top of our experiment would be the electronics that would tell us what the functions were inside the dewar.

DeVorkin:

This is the rack that would fit in the upper cone as shown in the diagram.

Harwit:

That’s right in the ogive cone.

Kondracki:

Also, the military people would require certain things to fit on the nose cone, or the rack, and that was a beacon plus — what was the safety thing called, Martin, where they had a destruct mechanism that would cut off fuel?

Harwit:

I’ve forgotten.

DeVorkin:

But this was merely military interest in housekeeping, not any military interest in your data?

Kondracki:

For safety. They were involved with the safety part of the rocket.

Harwit:

White Sands is close to Alamogordo, and they always were afraid that rockets could be blown by the winds, or because they misfired, toward Alamogordo, and so they could cut the fuel and prevent that. Purely a safety factor.

Kondracki:

They had a small receiver located in here so they could hit a signal, give a frequency, and shut fuel off.

DeVorkin:

Yes, okay.

Kondracki:

So basically that was the experiment in itself.

Let me just go back again to the dewar, or the thermos bottle. If you take note that at the opening the material goes straight across, then goes down, goes over, comes back up, then back in. This is liquid nitrogen temperature, and this is room temperature. And one of the things that would determine how big this experiment would be when we first started to talk about this was that we had to get the length of delta T. This was a thermos problem such that, if you are sitting on earth and, if this outer surface got cold, moisture would condense on it such as when you put ice cubes in a glass.

Harwit:

Or the O-ring would freeze and leak. That would also be a possibility.

Kondracki:

And so this is the reason why we had to go down this way, and it sort of looks funny; but that is a thermal problem and it was the only way to get around it.

Harwit:

You wanted to have a long heat path, so that there would be a long distance that the heat would have to travel before it ever got down to the cold portions; or actually we tended to think of it as the cold coming up along the heat path.

DeVorkin:

You weren’t getting radiation convection, this was conduction.

Harwit:

That’s right.

DeVorkin:

So it had to go all the way down and back before it touched anything.

Harwit:

That’s right. That led to this sort of funny convoluted shape in the system where there are the two intersecting cylinders.

Kondracki:

Yes. And that really makes it hard to manufacture.

DeVorkin:

Fascinating.

Harwit:

That was a design problem with this.

DeVorkin:

Who solved that design problem?

Harwit:

Well, we didn’t. We just gave up on it, and went to the next generation, which looked out the nose cone. There we decided we would do the same thing that Doug McNutt was thinking of doing. And at the time I had a German co-worker who had been sent over for two years on some sort of a fellowship by the name of Klaus Fuhrmann.

Kondracki:

You’re bringing back memories, Martin. (laugh)

Harwit:

He was very fascinated by this idea of McNutt’s and kept hopping on me that it was much, much better, and eventually I had to agree with him that he was right, and that that was the way to go. In fact, we then built one of these payloads, and still got it launched before the NRL people got theirs launched by quite some time.

DeVorkin:

Let me just interject and ask a question of both of you? This is the second time you have mentioned that. How much of a driver was it at this time that you be first? Was this something that was important for funding, or just your own satisfaction.

Harwit:

Both.

Kondracki:

One of the reasons was that was the first time anybody was flying cryogenic fluids in outer space, this series starting in 1963 64. One of the things that people were concerned with at NASA at that time was that we were just a young group, that the thing was going to explode. We asked, why. There really were no answers.

DeVorkin:

Who said that? Who were?

Kondracki:

Oh, I can’t remember.

Harwit:

I remember, I think it was Nancy Roman once telling me that, of course, the Stanford gravity experiment would fly before we would get this up into space. This was about a year before we actually flew a liquid helium-cooled payload, or maybe two years. And as you know, the Stanford group still hasn't flown one. This is almost 20 years later.

Someone else once told me (and this may be a complete fabrication) that he had been told by Nancy Roman that she really didn’t think we’d ever get this thing to work, that this liquid helium business was a shot in the dark, and she wasn’t at all sure of it. I don’t think that’s true. It’s unlikely that she would ever have said this to someone.

Kondracki:

But, let’s get back to the engineering problem on that specific case. I’ll tell you how Martin Harwit’s group solved it. We said, all right, let’s make some tests to actually identify, or at least approach that problem, or address it, so that we could solve it. I’m going to have to go back to NRL, because this is one of the original problems that was put to me, and I didn’t have an answer.

DeVorkin:

You were up at Cornell at this time?

Harwit:

No. You had better go into how that was.

Kondracki:

Okay. Let me give the engineering first. So we made some tests. The test was that we would fill this half full of cryogenic fluid. What happens: anything below the fluid line is at liquid-nitrogen temperature. Anything above that is warmer. So physics says that as long as it’s warm, that means the molecules are expanding; and when it gets cool, the molecules are contracting. We put this on a vibrating table with several gauges — a vacuum gauge, a pressure gauge — then we were going to run like hell if the thing was going to explode, or close the door. We started to vibrate this, and we were all looking through this big thick glass, and we see the vacuum gauge now, mind you, this is all at atmospheric pressure — and we were creating a vacuum inside the cryogenic reservoir, which is supposed to be at atmospheric conditions. And we said: “Why is that happening.” As the cryogenic fluid was splashing around and getting this surface cold, it was condensing the air molecules in there, and was creating a vacuum. So we had no problems with exploding. As a matter of fact, it would be better for us, because now it’s reducing the pressure.

So that’s how simply that problem was solved, when somebody said that it was going to explode when you go out into a vacuum. What happens is that this is going to be at atmospheric conditions, the liquid inside, so when you remove the one atmosphere on the outside, that means the liquid will want to get larger; but we build this strong enough to take care of that. By reducing the pressure inside the liquid, you could get the detectors colder and much more efficient. So that’s how that part ended. That was one of the original problems, so by flying first, this meant that you had all the engineering data in hand. You could publish many papers because of this; now everybody has to quote you. I think this is very important.

DeVorkin:

Was this important for future funding?

Kondracki:

Yes.

Harwit:

And also, the chairman of our department, Tommy Gold, was telling me at the time that I couldn’t get promoted until I had a successful rocket shot, more or less, you know. So there were a number of things like that. (laugh)

DeVorkin:

That’s beautiful: Did he say that with a smile, or was he dead serious?

Harwit:

No, he was serious. I hadn’t been an associate professor very long; but I had one of the biggest contracts at Cornell. And I thought I deserved their confidence and support.

DeVorkin:

But the contract was with outside money.

Harwit:

Yes.

DeVorkin:

Was it Navy money?

Harwit:

No, NASA, through Nancy Roman.

DeVorkin:

Yet, you had these very strong ties with the NRL.

Harwit:

Well, no. What happened was, we had agreed that we would part ways, and that I would try to do whatever I could at Cornell. At that time it wasn’t clear at all whether I would be able to start anything up at Cornell. So they had a head start on me at NRL. So when I got to Cornell, I didn’t have a lab or anything suitable. That had to be built up first. And I didn’t have any people. Then I asked Friedman whether it was all right with him if I used Hank as a consultant; because it was a potential conflict of interests. I think John Reese was also working with us initially for a while.

Kondracki:

No, he didn't.

Harwit:

I thought, for a little while, and then he decided he didn’t like the extra work. And then we got Bill Wernsing to do the electronics.

At any rate, Hank enjoyed working with us; and it was clear that we needed a person full time. So he came up and started his own firm. I didn’t want to have somebody who worked at Cornell for us, because at Cornell we couldn’t offer any kind of attractive salaries for engineers. The university just had technicians and then it has engineers on the faculty who are interested more in applied physics problems usually, not in actual engineering.

The other thing was that by having somebody outside the university, we had absolute control. We had a contract. We could say what we wanted. There was no problem with somebody else at Cornell coming in — as we would have had in the workshops — and saying, well, I have priority over you. We had enough problems, so that I felt we needed to have people who were directly responsible to us full time. And that meant having a very small company working for us as subcontractors. Hank could make a reasonable amount of money that way, and it would still come out cost-effective for us, because his overhead was law.

Kondracki:

Yes.

Harwit:

And it met everybody’s requirements.

DeVorkin:

What did you do when you needed vacuum chambers or shake stands? Was this all provided by NASA?

Harwit:

Yes.

DeVorkin:

So you brought the instrument down to Goddard?

Harwit:

Right.

Kondracki:

Let me clarify something here about H. C. Kondracki working for Cornell University. When I was at Engineering Services Division I was a young engineer there; and most of the other people were relatively older. So when they had young scientists come in, like Martin Harwit on a sabbatical, all that work was pushed onto me, because they didn’t want these crazy people coming in, and saying, do this new work. I looked at it with great enthusiasm, because I said, gees, you know, this is really neat. So there were several other scientists in his particular category that I was working with at the same time that I was supporting Martin.

DeVorkin:

Were they in space science?

Kondracki:

They were in the same group; but one was under Friedman and one was under Chubb.

DeVorkin:

Do you remember their names?

Kondracki:

Yes. One was a Dr. Richard Blake, and one was Dr. Stuart Bowyer.

Harwit:

He wasn’t even a doctor yet at the time. He was working on his thesis.

DeVorkin:

Was this a typical thing at NRL, to bring in people?

Kondracki:

No, they had just started it. This had only been happening for about three years.

DeVorkin:

So Friedman must have started that.

Kondracki:

Yes, Friedman started that.

Harwit:

He was short-handed, and so he went over to get more manpower from the services division. But essentially, he expanded his empire this way.

DeVorkin:

I mean, the sabbatical program, bringing scientists in to become indoctrinated in space techniques.

Harwit:

Yes, that was Friedman’s, but Stuart Bowyer was there as a student. Richard Blake had been there for quite some time as a student, also.

DeVorkin:

Were people like Kupperian and Boggess also in that position earlier on, or did they really have line positions?

Harwit:

They had line positions, I think. They were quite senior when they went to NASA.

DeVorkin:

Okay. I can cover that later.

Kondracki:

So getting back to me, I saw this great opportunity of going into business for myself. Blake left NRL, Martin left NRL, and Stu Bowyer left NRL; and they all had money in hand. But they didn’t have the expertise to build the experiments. So I formed the company in Washington, D.C. prior to my leaving NRL.

DeVorkin:

How did you do that? A) How did you do it? and B) What’s the company’s name?

Kondracki:

How did I do that? I just went through normal channels of forming a company in Washington, D.C. Anybody can do that.

DeVorkin:

Even though it paralleled the work you were doing at NRL?

Kondracki:

I just started the company because I was anticipating something at a particular time; and I didn’t know what it was. The name of the company is Pan Monitor, Incorporated. The way that name was picked: the first name identification means “all”; monitor means what monitor is, incorporated. So I didn’t want to identify it with anything.

Stu Bowyer was at Catholic University, and he got a grant and didn’t have the people to do the work. But at this time I made a few trips to Cornell University to take a look at what Martin was doing, made some suggestions to him, and then came back down. I think several months went by, and then you (Harwit) asked me if I would go ahead and build the experiment. And I said, wow, that's a task that I will have to take a look at. So in the meantime a contract was drawn up by a guy named Rogers from Purchasing, at Cornell.

Harwit:

Yes. He was head of Purchasing, at Cornell.

Kondracki:

One of the things that happened there was that I had to live within a 50-mile radius of Cornell University.

DeVorkin:

Oh yes. I’ve heard of those sorts of things, not Cornell particularly, but this is usual.

Kondracki:

So anyhow, I says, gees, if I’m going to commit myself, let’s see what the market bears. At this time Stu Bowyer left for Berkeley University. You didn’t know this at the time, Martin. It’s probably all new to you.

Harwit:

It is, yes.

Kondracki:

So anyhow, I went out to Berkeley and see what kind of offer I would get there, and at this time I had a little girl about this big, five years old. So I already made a trip up to Ithaca, New York and loved it. It was very pretty up there. At that particular time in Berkeley and San Francisco, the Haight-Ashbury people started to go on dope and things like that, so I was looking at my daughter, and I said I think Ithaca’s better. Even though there was probably more money at Berkeley. So that’s how I ended up coming to Ithaca, New York.

But what happened in the meantime was that, while I was doing this work, running up to Ithaca, I was also running over to Catholic University to help Stu. At that particular time I said, gees, what an opportunity to set up a business to use my expertise to help these young scientists in the university to build their experiments. So I formed this company in Washington, D.C. I never operated in Washington, D.C., but when I moved to New York — this was about 1966 — then I transferred the company from Washington, D.C. to Ithaca, New York; and we are still doing the same business. As a matter of fact, we have five experiments with the School of Electrical Engineering. They are on their way down to Chile. We are going to fly the experiments for the next two months.

DeVorkin:

So this is how you made yourself available for the Cornell group, and the Center for Radio Physics and Space Research. I know that center had been doing radio astronomy, but when did it become Radio Physics and Space Research?

Harwit:

It always was. No, it always was that, ever since I have been there. Instead of astronomy, I think at that time space research was a golden expression. If you wanted to get NASA money, it sounded like the right kind of a thing. I don’t know whether Tommy Gold had coined the expression, but he was director from the beginning.

Kondracki:

Martin, wasn’t that building funded by NASA?

Harwit:

Our building later on way, yes. We weren’t in that building yet, at that time.

Kondracki:

You have to quit now.

DeVorkin:

Yes, I have to at least find out what’s going on. I can leave this in here, and just leave it running. (laughs)

Harwit:

No, that’s fine.

DeVorkin:

Thank you very much for the first session.