Oral History Transcript — Dr. Robert Frosch
This transcript may not be quoted, reproduced or redistributed in whole or in part by any means except with the written permission of the American Institute of Physics.
This transcript is based on a tape-recorded interview deposited at the Center for History of Physics of the American Institute of Physics. The AIP's interviews have generally been transcribed from tape, edited by the interviewer for clarity, and then further edited by the interviewee. If this interview is important to you, you should consult earlier versions of the transcript or listen to the original tape. For many interviews, the AIP retains substantial files with further information about the interviewee and the interview itself. Please contact us for information about accessing these materials.
Please bear in mind that: 1) This material is a transcript of the spoken word rather than a literary product; 2) An interview must be read with the awareness that different people's memories about an event will often differ, and that memories can change with time for many reasons including subsequent experiences, interactions with others, and one's feelings about an event. Disclaimer: This transcript was scanned from a typescript, introducing occasional spelling errors. The original typescript is available.
See the catalog record for this interview and search for other interviews in our collection
Interview with Dr. Robert Frosch
Robert Frosch; August 19, 1981
ABSTRACT: Reviews Frosch's (b. May 22, 1928) education at Columbia University (PhD, 1952, theoretical physics) and, in detail, his varied career as a physicist and a science manager, beginning with his work as a research scientist at Hudson Laboratory (1951–3) and then as Asst. Director and Director of the Theoretical Division (1953–63). In 1963 he became Director of Nuclear Test Detection, Advance Research Project Agency, Office of the Secretary of Defense; in 1966 he was appointed Assistant Secretary of the Navy, Research and Development; from 1973 he served as Assistant Executive Director in the U.N. Environmental Programme; from 1975 he served as Associate Director of Woods Hole Oceanographic Institute; and from 1977 to 1980 he served as Administrator of NASA.
DeVorkin:I believe that you have some comments on previous questions that we had left open.
Frosch:Yes, a couple of points. I remember the name associated with the origin of Hudson Labs that we discussed. I couldn't remember the name. What I do remember is the Hartwell Report. I believe Hartwell was the name of the Admiral who ran a committee to look at the question of convoy defense against submarines. Now, you asked me in a previous interview, I think the last one, about an Airlie House meeting concerning Navy laboratories, and I had very fuzzy recollections. But I went back to look for another speech that we had also discussed, and found what it appears to be — it's labeled as such — my address at that laboratory workshop. It looks as though it really is my address at that laboratory workshop, rather than something that was canned and not given, so you may take that. It will describe my views on the in-house laboratories at that time, and I can jog my recollections.
DeVorkin:I see, okay. This is an address by the Honorable Robert A. Frosch, Assistant Secretary of the Navy at the Airlie House Laboratory Workshop, Warrenton, Virginia, 24th April 1968.
Frosch:Now, I went back to look for a speech I gave in June of '67 to the Defense Project Managers Course at Wright Patterson. I did not find that speech so labeled. That's somewhere in my papers, but I haven't found it. I did find one which is an address by me as Assistant Secretary of the Navy at the Systems Orientation presentation, Fort Wayne Chapter of the IEEE in November of '69. But I see that what I have labeled on in pencil, this being a xerox was previously "June '67, Defense Project Managers Course, Wright Patterson." and then I just apparently — it is, as my recollection goes, precisely that speech, and I've written in pencil before the beginning of it "as a literary device, imagine you are project managers" and then duplicate. I also seem to have put a preface on it, something from Carl Sandburg's introduction to his "Complete Poems. I see now that I should have used that in some other places (laughs). Now, I have two other things that I think come in the department of curiosity, both of which fell out of papers which I was moving. One of them seems to be a college freshman paper, which clearly confused the instructor, who, as I remember, was a very nice English literature guy. He must have requested an expository paper on some subject. I wrote an introduction to hyperspace, and I see I got an "A-minus" with the comment: "For all I know. this is a good paper conscientiously done." I thought you might be curious about this. The fourth piece of paper is much more recent, and deals with subjects we haven't discussed yet. I just labeled it, "Probably fall of 1980," but I'm quite sure that's the case. It is handwritten, although I think somewhere there is a typewritten copy. It's entitled "Ballad of the Certification Review". I wrote it during a moment of frustration in the course of the Shuttle discussions. Comes under the heading of "Original Doggerel."
DeVorkin:In the course of the shuttle discussions?
Frosch:Yes, the question under discussion was: were we ready to certify the Shuttle for flight?
DeVorkin:Would you care reciting some of it? That would be marvelous.
Frosch:Okay, you can have a recitation. It's entitled: Ballad of the Certification Review Decisions here, decisions there, Shall we fly it? Do we dare? Are we ready? Are we sure? Are the certs and quals all pure? In spite of several program howls, Let's seek advice from knowing owls, To help us settle what to do, And when to fly our shuttle new; Alpha Beta delta T Does the vehicle fly free, Is it stable? Will it yaw? Is it safe for my grandpaw? If Beta goes beyond the edge, Do we fall right off the ledge? Or is the temperature okay, So all our problems go away? Push LRU's to .04? Or must we dare to shake them more? If nothing breaks should we be happy, Or go on farther 'til they're crappy; Just cycle T to cycles 5. And hope the boxes stay alive. Will stress and strain prove systems good, As ELV men say they should? And as for fragile thermal tiles (as big a pain as having piles), We'll push and pull and test with CLOT and hope the tiles don't go to pot! Or should we find a linear sip, (for some of them think that's the tip), Or let non-linear take a ride, And hope our bird can keep its hide? Decisions here, decisions there; Shall we fly it? Do we dare? Are we ready, are we sure; Are the certs and quals all pure? In any case, we're getting there, (although the process tears my hair), We'll push the stone right up the hill, For fly we must, and soon we will.
Frosch:I can interpret all that, but there are lots of people who know what those initials and stuff mean. Want me to do that? I can do it.
DeVorkin:Well, I think we should. Can you give me the events that led you up to this?
Frosch:In 1980, Al Lovelace and I decided that we would get a group of outsiders to look at the process of flight certification and qualification. They were there not to certify that the vehicle was all right, but to tell us whether they thought the NASA decision-making process was all right. Meanwhile, we were following what they were doing, and reviewing the whole process of certifying the thing, system by system. That got complicated and frustrating, because all sorts of issues came up that had been around, some of which we knew about. But there got to be very strongly polarized opinions about what you should do.
DeVorkin:Who chose the outside people?
Frosch:Walt Williams was the Chief Engineer of NASA, and he went out through the aerospace community and tried to find people who were expert in various areas. He set up a very elaborate system of committees and subcommittees, all of which interlocked to cover the whole Shuttle system. It was frustrating and it was intensely interesting as well. People got to arguing, not so much about the Shuttle, but how you dealt with such an engineering problem. For example, we wanted people to look at the certification of avionics computers, electronics and black boxes generally, people not involved in the Shuttle, but who knew about that kind of thing for launch vehicle use. We pulled a lot of people who had spent their lives in the expendable launch vehicle business, the ELV business. They were, I should say, rubbing their hands with glee at a chance to look over the shoulders of the manned space flight people and make their comments. It turned out, among other things, that philosophy of life and reliability among the manned people and the ELV people was different, because they faced different problems. The manned people tended to rely heavily on redundancy and redundant systems. They were more cagey about the extent to which they tested individual boxes, whereas the ELV people were essentially single string. They tended to beat the boxes to death before they certified them. Now this opens some interesting philosophical questions. When you beat a box to death and find out when it breaks, does that, in fact, tell you something about all the other boxes, if you have not done this with the proper sampling process out of a large run of objects? But since you never have a large run of objects in this space business, it's a little hard to tell. We were not really arguing provable engineering, but suppositions about engineering, which had a philosophical base and lots of theoretical questions. It got difficult.
DeVorkin:Who were the people on each side? You said that this got very complex?
Frosch:Well, there were the in-NASA manned spaceflight people. There were the in-NASA expendable launch vehicle people. There were people from industry who had been involved before, but weren't directly involved in Shuttle, and people from academia, some of whom had been involved in similar things, some of whom had not. It was a display, not only of advice on the Shuttle and advice on certification and qualification, but on philosophies of how you go about making that kind of an engineering prediction. There must have been 150 people involved as consultants. It was a big operation, went on all year.
DeVorkin:Yes. Who were Walt Williams' chief subordinates in this, especially in NASA?
Frosch:Oh. you'll have to get somebody who was involved to work their way through it. I know all the guys, but I can't conjure up the sequence of names at this point.
DeVorkin:The consultants themselves disagreed about this themselves, these different philosophies.
Oh yes, happily. It was very interesting. We were picking our way through it and deciding whose philosophy to believe, when to decide we were all right, and when to order more tests. For example, the chairman of the group that was looking at the question of tiles was Holt Ashley, who I think is from Berkeley. He's a very good guy in the materials and ceramics kind of thing. His view was that we should not fly the Shuttle unless and until we were prepared to repair tiles in orbit. He felt that strongly. The panel apparently felt that strongly. They made it as a strong recommendation. Even though it was not quite a comment on certification or qualification, it was a comment on their level of confidence in qualification and certification. Al Lovelace and I finally made the decision — and I felt it very strongly — that we were not going to do that. We simply weighed the odds. You see, there was opposition to Holt Ashley. The strongest guy involved was Chris Kraft, but also John Young and Bob Crippen felt somewhat this way. The question was, do you have enough confidence in the tile system so that you're going to use it? And what is the risk involved in sending the astronauts out to repair something, or even to inspect it? You really have two risks. It tended to be viewed as: if you have a problem with the tiles, you should be prepared to repair it.
There were a number of problems. It is no mean trick to go out there with the equipment we were preparing but didn't have yet, equipment never tested in space. To put on the suits, do an EVA (an extra-vehicular activity) for the first time, and go wandering around on an umbilical. All that's very difficult. We had never really quite been happy with the umbilical in the previous manned space flights in Gemini and Apollo. People kept getting tangled up with them and cursing them and so on. Letting them go without an umbilical was dangerous. This rather complex job clearly would involve several hours EVA. I concluded that adding that kind of training and complication to the mission and the mission preparation, and telling them to go out there when it might not be necessary, increased the risk of the mission more than the risk associated with the tile problem. Our feeling was to make sure that the tiles don't get into trouble, rather than develop this other activity which carries its own risks, which seem to be as great or greater. That was really how the decision was made.
Of course, there were always people who would say that we hadn't looked at the quantitative assessment of these risks. You know, there are always people around who say they know how to make a quantitative risk assessment. I have been down that road, and that's just a complicated way of confusing yourself in complex situations. The usual technique is that you assess the probability of failure or success in every individual element and you pile them all up. But people forget that it's difficult to assess the individual ones, and you have no idea how to add those errors up, because they are not independent. Even if they were independent, you would worry about it, so what you get is a pile of meaningless numbers. And you're pretending that you're somehow making an objective decision based on analysis as opposed to an "intuitive" decision, whereas, in fact, you're not. What you're doing is abdicating a decision to some unknown noise level in the mathematics.
DeVorkin:Did Ashley feel it was possible to have acceptable reliability for the tile?
Frosch:He and his panel quite honestly did not feel that they could certify that the process we were going through was point 9999, or whatever one wanted.
DeVorkin:Why was that? Was it something in your manufacturing process?
Well, no. The point is, they were being very academic, very cautious, they were not making a perfectly clear statement, a perfectly definite statement, such as, "We can't be sure." Well, we knew they couldn't be sure. There is no way to be sure. But they didn't point to a problem. They just felt that it was a very tricky business. There are 30,000 tiles, so how low does the probability have to be, etc. Then there was the other tile issue which was: what happens if a tile comes off? So we looked at that. And the answer, of course, is that it depends on which tile and when. Some tiles are there to protect things in ascent. Some tiles are there to protect things in descent.
There are some placed on the bird, and if a tile falls off, you get a hole in the skin and nothing much happens, but you've got to repair that. There are other places where you'll get a hole in crucial hydraulics or something, and that's the end of that. So, there was a set of discussions suggesting that we analyze what happens if various tiles fall off, so we could answer the question. Well, we vetoed that, because again, we had 30,000 tiles. You'd have to look at the position of every tile, hypothesize the hole, look what's under the hole, and go through a whole fault tree. You look at the amount of engineering that that takes, and you ask where are those people going to come from? And you conclude, they've got to come from people who know the Shuttle. But those are the people who are building the Shuttle. You end up saying, "No, I don't care about analyzing what's going to happen. Go make sure it doesn't happen." Suppose I have the analysis. I've got 30,000 pages, each of which tells me what happens if a tile falls. So what do I do now? Well, I filter through it and I make damn sure that ones that are crucial really are all right. But the fact is we were already doing that for almost all of the tiles. Well, here we are, you see; so you go round and round. And there is no simple, analytical way out of this. Somebody takes the responsibility. I took the responsibility up to the point at which I left it, saying, "Yes, I think it's all right. This is the way we're going do it. We're not going to take all the tiles off and put them all back on again." It's the same kind of problem, you see. You come to a certain point with pulling tiles off and putting them on again where you say to yourself, "I can't test it that last time before flight."
Frosch:So, how do I know that in taking it off and putting it back on, I didn't do damage more than if I hadn't taken it off? You get to the point where you think your surety in the quality control and the process itself is all right, and then you quit. We made a kind of judgment on that and so far it has worked.
DeVorkin:The design that included the tiles was certainly set long before you came to NASA.
DeVorkin:Had there ever been any alternatives considered?
Frosch:There were a number of alternatives looked at. Some looked very good when we were having trouble with the tiles, but didn't look very good when you considered the problems associated with the alternatives. For example, there were some of the high temperature metals, not for body, but for skin, Columbium and so on. I don't kow much about this, but I looked at it some. Columbium has very bad oxidation properties, particularly at high temperature. Therefore, it would have to be covered with a layer of paint or covering to prevent if from deteriorating in the process of getting ready to fly. Al Lovelace put it this way. He said, "You know what we'd be doing now? We'd be inventing methods for detecting pinholes in paint." That's exactly right. Everything looks good until you look at what the real engineering is. We concluded we weren't going to play that game. We were just going to work the tile thing. Meanwhile, we restarted some work which had been stopped for lack of money, on some of the alternative schemes, but I don't know where that's gone subsequently.
DeVorkin:Okay. Could you interpret, then, some of the interesting cryptics? This is the poem we're talking about.
Frosch:Certs and quals are certifications and qualifications. "Alpha, beta, delta T, does the vehicle fly free, is it stable, will it yaw," etc. — here was the issue. Alpha and beta are, of course, the pitch and roll. Delta T is the heating, the temperature increase due to the friction on re-entering. You know what the stability and yaw are. The issue was, how much variation in the attitude of the vehicle around the nominal value is there going to be as it comes in? It had been pointed out that, for certain peculiar combinations of roll, pitch and yaw, there might be turbulent flow from forward parts of the body that would change the heating on afterparts of the body. So the hypothesis was that you could get hotspots that were very much hotter than you had predicted.
Frosch:All right, so that was an issue. Then that got tied up with the issue of the flight control properties of the bird. Nobody had, before the Shuttle flew, this time, flown a full-scale aerodynamic vehicle between Mach 5 or 6 or 7, and Mach 25! This was a first. We didn't have a wind tunnel in which we could do it full scale, so there was a mixture of theory, scaling up from other vehicles, sub-scale vehicles and so on, and wind tunnel tests on partial things. The difficulty is that nobody was quite confident that all of that was good enough so that you could understand the flight dynamics. The first problem you had was that you really didn't know for sure, from experiment, what the aerodynamic control equations for the full scale vehicle were. At the same time you were building a computer model of the control equations, which was a pile of software through hardware. It was supposed to take the air data, use what you thought were the control equations and a computer approximation, and run the vehicle. You had two sets of worries. One, is your understanding of the real control situation right? And two, if it is right, is the software right? The hardware of course, has its own filtering and oscillation properties. So you need to know if all these package together so that you have a stably-flying vehicle. Or, will you have a peculiar motion, maybe not damped enough, which will put you into a situation where you have control instability. You go through a regime where things fall apart, or just enough so that you have some of this turbulent shedding that overheats another part of the vehicle. And that gives you another problem. That's what all that is about, that was a central problem which we discussed. That was one of the things that this group of "owls" was discussing. That's Al Lovelace's term, "a tree full of owls," meaning the outside experts.
DeVorkin:Now, the group of "owls" are the 150 consultants?
Frosch:Yes, these are the consultants. It was a generic term, but in this case, it meant the 150. Now there was within this group a panel that was worrying precisely about flight control. That panel finally came through and said: "All things considered, it is our opinion that it should be all right. In any case, we can't think of anything else to do." That's roughly the summary of their report. They thought all the evidence showed that it was stable. You have a year's worth of this going on, eight or nine months at any rate. Occasionally, you get dispatches from the front. Somebody calls up, or there's a rumor that the panel thinks it's totally unstable, and everything is going to fall apart. That's always an exaggeration. There were periods when it was clear that the panel had found a new worry. They suddenly found something in the approximation equations that they thought was not a good enough approximation. There, too, there were philosophy questions. What do you allow the system to do? How do you deal with this? In fact, some of it was put in very nice formal mathematical language. At one point a panel member said, "This is complicated enough that even though I know where some of the zeros are, that is, the instability points and the stability plane, I can't guarantee I know where all of them are."
DeVorkin:Would you characterize these groups of "owls" and their interactions and their reporting structure as being both formal and something of a rumor mill? Were there rumors really flying around?
Frosch:There were always rumors flying around. Sure.
DeVorkin:How were they propagated?
Frosch:Oh. in the usual way. Somebody was at a meeting or at a consultation, and heard something, and told somebody else. You will see in one of the speeches I gave you today, the one for project managers, that I describe a process that I call amplification upwards.
Frosch:In fact, I describe two processes that might be rumor mill processes: amplification upwards and amplification downwards. Amplification upwards is when somebody finds a bad solder joint in the factory, and by the time the assistant secretary hears about it, the whole project is in trouble and has to be redone. Amplification downwards is the when an admiral asks his aide about a certain project, and the aide says he'll go find out. The next time the admiral turns around, there is a full-scale briefing with color slides to reassure him that the project is really all right, and all he really wanted to know is, is the project really all right? Those are generic in this system.
DeVorkin:There was nothing special with this group of owls?
Frosch:I would say it was calmer than I might have expected. The consultants were very responsible. There was nobody around in the system trying to make any trouble. The spirit of it was, "Let's see what we can do to fix it." Some of the ELV people were saying, "We suspected for years these guys don't know how to do reliability. Now, we're going to show them." Then they began to understand there were a few problems there. If you like, I can talk more on this.
DeVorkin:Could I ask you one more question, if it's pertinent to this point?
DeVorkin:How did Shuttle production managers react to the outside consultants? How did staff react?
They were annoyed by it, and always very nervous about it. Now, part of that is the normal bureaucratic reaction — nobody trusts me, they're bringing in people to tell me my business, and so on. Part of it was a dual, valid worry. You can't have 150 people running around looking at the system without interfering with the work. One of the things you have to do when you conjure up something like this is to decide whether or not you can stand it. From time to time we had to cage the process a little bit. We had to make sure that some panel member didn't get so enthusiastic about some point, important or not, and require so much work to be done that the real work was stopped. There were some things connected with the engine examination where people wanted us to do something that looked as if it would be so much work that it wasn't worth results. We just said. "No, we won't let the panel do that." and got that untangled. Mostly, it was not a problem. There is another problem. There will always be some panel member who has some personal hobby, and the fear is that he'll build this hobby into the examination.
We didn't have much of that. There was one area in the course of the whole thing, mostly before this period, where we had that problem. I'll mention it. The question was how do you do the development testing on liquid engines like this? That was an issue. The choice made earlier in the process was that for this engine, which is tightly integrated, the engine itself will be the test rig for the components and the engine. Now, the previous engine developments had mostly been done by building a test rig. You test all the pieces separately over the right range of things and then put them together into the engine. All through the engine testing, whenever there was a problem, there were some people, some people that I never met, in fact, who would pop up and say, "I told you so; you never built a test rig and you never tested all the pieces." After watching this for awhile, and talking to the engine people, I concluded that it was a damn good thing that somebody had made that decision to begin with. Because if we'd built a test rig, we would have carefully tested everything under the wrong conditions. We didn't know what the integrated engine conditions were; the engine was much more tightly integrated than other engines. And we would have integrated the thing after making all sorts of fixes for the wrong conditions. We would have had exactly the same set of results, and would have lost three years in doing it.
Frosch:That's my guess, yes. The guys who wanted to do that did not appreciate how the business had been changed by the computerized engine control. In fact, you could play games with the engine that you could not have sensibly played on an individual component, test rig basis. It was just not like the old engines that had to work over too much range. This is really what I learned there in the course of things. At first, I wasn't sure. It sounded plausible to me that you could test all the pieces and put them together. Then it became clear that this was more nonlinear than that.
DeVorkin:How did you learn of such things? Who was the person who informed you most frequently? Was it Lovelace?
Frosch:Alan was one. But I would sit in long briefings on what was going on in the engine business from guys at Marshall, Rockwell and Johnson. Of course, some of these people were advisory ousiders. One of them was Gene Cavert (Eugene Cavert), a professor from MIT. He was asked by Adlai Stevenson's committee to look at the engine program; a very good guy. He started out with the same premise, that we should have taken it all apart and have tested all the pieces. Gradually, I think, he concluded that our way was perfectly reasonable.
DeVorkin:When did Stevenson's committee get interested in the engine?
Oh. I don't know. It must have been in '78 or '79, whenever there was some problem with the engine. There were always some problems with the engines. There was a curious thing with the engine problem. People were saying that we were having terrible trouble with the engines, that the engines kept blowing up. We didn't blow any engines up. We only melted the innards of one engine down. I finally got puzzled enough about this characterization, so I went back and asked for the statistics. "Give me some other engine programs and what happened with them, and then look at us." I discovered it was pretty straightforward. It was taking awhile, but we were having less trouble than with the Saturn engine, the F-1.
That engine really had a terrible history, and all the testing of components did not prevent them from having problems when they started to put them together. We were having problems, but not such big problems, only annoying ones. We couldn't fly until we fixed them, and it was a pain to fix them. We were also being dogged by the fact that for economic reasons, we really hadn't bought enough test engines. When something went wrong, we were trading engines and components around. At one point we decided we were going to buy five more engines. That helped, but they didn't come on-line soon enough. That was one set of issues. Back to the ballad. "If Beta goes beyond the edge, do we fall right off the ledge?" is the temperature etc. That is, does it take the wrong attitude, causing the aerodynamics of some parts to reflect on the others and change the temperatures. "Push LRU's to .04." LRU is line replaceable units, the individual black boxes in the electronics. The question was, to how many "g" do you vibrate them? I don't remember the numbers precisely, but whatever vibration level was specified, the ELV people used a higher number. It must have been .04 — it doesn't sound right to be .04 "g". It was .04 something. So there was an argument about whether you should, in effect, drop them.
Should they have certain kinds of shock testing? There was also a long argument whether the environment of ELVs was or was not different than the manned environment. I really don't even remember what we did with that particular one. I think we insisted that they be tested to a higher vibration level, and it was really all right. "Just cycle T to cycles 5, and hope the boxes stay alive," and so on. You have combined vibration and temperature cycling; the temperatures go from very cold to very hot. And you vibrate at the same time. The question was, how many cycles of that should you put it through in order to decide it was really all right? Some of this got a little bit theological, because five cycles isn't very much, and the number of recommended cycles wasn't very much, either. It's not like a fatigue problem where you cycle something five hundred thousand times or a million times. You are cycling it five times or ten times. You can't conceivably do it more than that. There was a question of just how hard do you try to break it before you decide it's really all right. That's what that was about. Now, the test with CLOT. CLOT is combined loads operational test. The problem with the tiles was this: what are the tiles subjected to? Well, you've got the thermal changes, so the body is changing in size and shape, and you've got temperature changes in material. You've got the aerodynamic forces that distort the body, at the first approximation, statically. Then you have flutter and aeroelasticity that is vibrating the body. Then you have flow, which is providing aerodynamic forces on the tiles. Then you have shock waves that are shed from other things. And then you have a phenomenon called the dancing shock. That means the position of the shock is unstable because of a discontinuity at the edge of the tile and may jump back and forth across the tile. So. you have all of these.
DeVorkin:At the edge of the tile, you mean at the joint?
Frosch:Well, there are places where you have a shock wave at the joint, or from some other change in the body. Sometimes the position of the shock wave doesn't appear to be stable. You could see this in the wind tunnel — the shock wave would jump back and forth, and you'd have a variable force on the tile. So you've got this combination. The real question was, how do you put this together into a theory and into a test? The problem with the theory was to determine the maximum stress. Maximum stress developed into a formula for adding all the forces together.
DeVorkin:Peak of all the peaks.
Frosch:Yes, you know, the peak thermal force and the peak elasticity, etc. You take the biggest peak, because those go up and down. And then you add those, making it the most conservative thing you can do, in terms of the extreme of forces. Of course, that produces a ridiculous result. I mean, you couldn't even hold the vehicle together if all that was true. Those peaks never actually pile up together. The peak of the aerodynamic drag forces is at a different point in the flight than the thermal forces, and at a different point in the flight than the shock forces, and so on. But you really don't have a good, detailed theory. You can't possibly build all this up together to make a realistic model. What you have to do is make some estimate of how to do this. And so there were lots of discussions of just how to put all these forces together, in absence of having flown any such vehicle. Now, the CLOT test arose because there were limitations in the test tunnels for vibration testing, acoustic testing, and thermal testing of the tiles. Some of the tests were separated out. The panel that was looking at this quite sensibly suggested that there has to be a better approximation to the combined test. Then we argued, what is a better approximation? Do you pile up all the peaks; what do you do? Finally, there was a lot of discussion and negotiation over what's a reasonable test, and what's a test we can actually do in the existing chambers? We didn't even know how to build chambers for some of these. Finally, a set of tests was, in fact, constructed. The first time it was tried, it was a disaster. I mean, there were tiles flying all over the test chamber.
Frosch:Yes. It turned out that there was some question as to whether the loads that were actually applied in the test were the same as the specified loads because of some problems with the chamber. Again, there had to be a theory of how the chamber behaved because of the walls of the chamber and the boundary conditions. It's not like a free space thing. That was a big issue of the clot test — partly because they weren't due to be finished testing until some time in February, and the flight was due to be in March.
DeVorkin:In March, yes.
DeVorkin:That was only one month.
Frosch:The problem was that nobody invented the CLOT test in this committee until sometime in the summer of 1980. This was a new idea which had not been planned. It was a perfectly good idea, but nobody had thought of that. This new invention was made, and it took time to untangle it. It just couldn't physically be done before that. Everybody just bet that when we finally came up to the final test, it would be all right. That's really what you were doing. If it had gone terribly awry, then the administrator would have had a considerable decision problem.
DeVorkin:What was your decision problem in that case? Could you see it going awry before February?
Frosch:No, because it wasn't going to be done before then.
DeVorkin:What would you have done? Did you have a contingency plan?
Frosch:You can't have a contingency plan, because the question is, how did it go awry?
Frosch:You have a contingency plan in some sense. Suppose a total catastrophe had been demonstrated. You believe the test. You think you did it right. You think the conditions were right, but it demonstrates your assumptions about the tiles are wrong. Well, you know what you do. You go back to the drawing board. You hope that you haven't collapsed the whole thing.
DeVorkin:There was no chief administrator, of course, during this critical period.
Frosch:Well, Al Lovelace was acting, and he made the decision.
Frosch:We ought to talk about that.
DeVorkin:That would be good.
Frosch:We're taking this all out of order, but I don't think it matters.
DeVorkin:No. it doesn't. as long as we get the good stuff down.
Let me go through this. The issue was, should we find a linear SIP, to let the nonlinear take a ride. SIP is S.I.P. It's the strain isolation pad. Now, remember, what you have here is a more or less elastic hull and skin, and a very brittle tile. You have to match the two. Even with the small tiles, you couldn't just glue the tiles to the skin, or it would crack, even with a little flexure over the area of the tile. So an elaborate process developed. Everybody talked about the glue failing. The glue never failed. Lots of other things failed. The glue never failed. It's a process of preparing the skin and putting on a suitable high-temperature epoxy, and then gluing the tile. You glue the tile to a high-temperature tolerant plastic felt to isolate it from the all the brittle tile. It's removed from the strain on the hull, and lets the tile ride slightly.
Now, there was a long series of problems. It turned out the strength of the SIP as glued was not the strength as designed. Nobody understood that at first, until we realized that this was the felt. Somebody did an analysis. What's the last thing you do when you make a felt? You go through a process that's called "felting". This consists of pushing needles through the material in order to give it just the right density. Each place where the needles have been pushed through is now a little column where the fibers are oriented differently than at every other place. One place is a little bit stiffer than the rest of the felt and becomes a stress concentration riser. In effect, the glue joint strength is not determined by the whole area of gluing of tile to felt. It's determined by the places of stress concentration.
That explained why, when a tile was pulled off and it failed, it didn't fail at the glue joint. It didn't fail in the middle of the tile where it should, if the problem was tile strength. It failed just inside the tile at the place it was glued to the SIP. The analysis demonstrated that it was because of these little stress risers. Once you understood that, it was clear what you did. You invented a method to make sure that the stress was spread over the whole area. And that turned out to be the business of painting the tiles. Somebody at Lockheed or Rockwell figured out a very ingenious way of increasing the density of the tile in a thin layer on the surface without adding the weight of densifying the whole tile. It was a very clever piece of technology. They made up a mixture of colloidal silica and finely powdered silica in some carrier or other, and painted on the tile. As the stuff penetrates the tile, the finely-powdered silica is just the right size to clog the pores and hold the colloidal silica in a thin layer without soaking all the way through. Then the colloidal silica has a habit of depositing itself, and it crystalizes out on the silica fibers, thereby increasing the density and strength, and providing a stress distribution layer.
DeVorkin:I would like a little clarification. This paint was applied to the outside of the tiles?
Frosch:Well, you have to take the tile off and apply it to the surface of the tile which will then be glued to the SIP.
DeVorkin:So, both sides are painted.
Frosch:In effect, but in different ways. The black stuff is a glazed glass paint, in essence, a silica glass, intended to give it the right absorptivity, reradiation coefficient, the right albedo. Okay, so Iguess that's all the points in this.
DeVorkin:That's the SIP.
Frosch:That's the SIP, yes. That's really all the technical points.
DeVorkin:Tell me, to whom did you show this ballad?
Frosch:I passed it around the office. And I was sternly advised by several people that it was not the sort of thing that ought to get into the press.
DeVorkin:Who was so conservative? I think it's marvelous.
Frosch:Well, several people. I think Bob Alnutt was bothered by it, and Neal Hasenball. Several others were just worried about what the press would do with the administrator making jokes about whether his machine was going to fly or not. They were right in the end, I think. I was going to comment about something earlier. Having mentioned some of these intricacies of the tiles and the engine, we kept trying to explain this to the press. But it is very difficult for them to deal with. How does a guy who writes a story for a newspaper, unless he's writing technical stuff, really deal with a question as intricate as, do you test the pieces separately, or is it so integrated a system, etc. etc. It is a validly difficult question. Unfortunately, many of the press people, rather than take that difficulty seriously, seem to say, "Oh well, you guys just can't explain what you're doing." And then they make a joke out of it. I was not terribly pleased with the performance of the press on that one. But they didn't do us any harm. They didn't really affect the program very much.
DeVorkin:There was a CBS Reports program on the progress of the Shuttle, where they highlighted the problems with the tiles and with the engines. I'm sure it was aired, either late in December or possibly by January, 19.
Frosch:Was this by Brinkley?
DeVorkin:No, this was Dan Rather, I think, or some of the others.
Frosch:I don't know. To be frank about it, I gave up even trying to watch them. There were enough of them, and it was an inconvenience. I would occasionally read the trariscripts afterwards, but I really couldn't get terribly excited about it. I would read what passed over my desk, but I would not pay enough attention so that I would go home and turn on the TV to see if it was on. People were perpetually calling me up and saying, "I saw you on TV." I'd say, "That's nice," and that was the end of it.
DeVorkin:You mentioned Brinkley. Was there some particular one you knew about?
Frosch:Yes, there was a thing, what her name? Betsy Aaron came in and did a thing for whatever the Brinkley-Betsy Aaron thing is called. It's not "20-20," is it?
DeVorkin:I don't know. It might be.
Frosch:It may be. That one annoyed me. Not that I saw it, but somebody on the staff got annoyed enough so they brought me a transcript of part of it. They chewed up a couple of hours of my time questioning about the Shuttle, and then the only way I can describe the way the transcript read was that it was a "ho, ho, ho" conversation about the shuttle between Betsy Aaron and Brinkley with occasional clips of me and others.
DeVorkin:"Ho, ho, ho," meaning?
Frosch:"Look at these idiots trying to put together this ridiculous machine. They can't hold the tiles together, and the engines keep blowing up. Isn't all this technology silly." That kind of a mindless thing. I have subsequently seen that program on other subjects, and my reaction was that it's all at about that level. It's not very good journalism. This was after I had said I was leaving on Inauguration Day, and the show ended with. "This guy isn't even going to be there to see the machine fail." It isn't that I mind the fact that they don't understand any of this stuff. God knows the poor things aren't educated well enough. It's the fact that they make no attempt to find out what it's about, no serious attempt. Now, there are reporters around who do, to be fair. This was a particularly bad one. Let me characterize this. The trade press, who follow technical issues, science reporters, guys like Dooling of the "Huntsville Times" are different.
Frosch:Dave Dooling, and some of the others, even when they are hostile, even when they think what you're doing is idiotic and say so, are very competent. They produce good, solid stuff. The problem is with the casual reporters, I mean the guy who is usually assigned to city hall, does political reporting or international affairs, who somehow gets wrung in to go deal with something like the Shuttle. He doesn't know the terminology and so on. He gets into trouble. The TV people are chronically this way. Most of them really have no specialized background. Cronkite does better, because he hung around the space program a lot. He's a fan. But a lot of the others do not. I remember one incident, and fortunately, I don't remember her name. A young lady who was the Capitol Hill correspondent for some Florida TV station came in one day; said she had been asked to do some stuff on the Shuttle. She was clearly a good reporter in the sense that somebody had given her some questions, and she could ask the questions and ask sensible follow-up questions based on my answers. But when she was through with this interview, she said, "By the way, do you have a picture or something of the Shuttle around? I've never seen it." Then I realized I knew what was going to happen with this story. That was winging it a bit too much.
DeVorkin:Yes. Tom Wolfe used the term "Victorian Gent" to represent the press in the Apollo era. Are you familiar with the term as he used it?
Frosch:No. I'm not even sure I understand what it means.
DeVorkin:This was in his book, The Right Stuff.
Frosch:Yes, but I don't remember.
DeVorkin:A press that was willing to go along with stories that were represented to them, understanding that it was a long and difficult process.
Frosch:There are reporters around who have understood that. I can't complain that we were too badly treated. It's just the annoyance that the intricacy, the "why it's so difficult" is not getting through the media so the people will understand that it's a very difficult problem. It's a sort of, "Gee, why can't you put it together and make it work the first time?"
DeVorkin:Did these things filter into the Stevenson committee and into other committee and give you any real headaches?
Frosch:No, no. Let me make two comments. You take somebody like Bill Hines, long-time guy in the space thing. Bill tends to be negative and sarcastic; you know, that's his style. But he's being negative and sarcastic about something which I think he understands, and the story conveys pretty well what it is he's negative and sarcastic about. I'd rather have that than a mindless optimism, in a sense. Now, let me say something about the committees. That's worth discussing. Who were the committees? Well, there's the House Authorization Committee, which is Don Fuqua. Larry Winn was the senior minority member. There's the Senate Authorization Committee which was then Adlai Stevenson, chairman, and Harrison Schmidt as senior minority member. And of course, one thinks of the committees in terms of the chairmen, because one deals with the chairmen. It is common when you have a hearing for the only continuous face in the meeting to be the chairman and the senior minority member. Now, those two committees were consistently supportive. They were also critical, beating on us, complaining. But, the message was "You guys are doing a tough job. And we are going to criticize you. We're not necessarily going to give you the money you want," and so on. It was always a feeling of considerable mutual respect.
DeVorkin:There was no influence of the press?
Frosch:Well, they would read the stories. My impression of all that was that they didn't pay a whole lot of attention to the stories.
DeVorkin:The Stevenson committee, as you mentioned before, was particularly concerned with the engines.
Frosch:They wanted to reassure themselves by getting third party advice, that it really was all right, that we hadn't missed something. Maybe that came out of press accounts, although I don't think so. I think it was a reaction to the fact that we had some problems, a series of problems in a row. But they were also doing something else. I think it's true — they were protecting both themselves and us against the press and outside. They figured that it was probably all right. By getting a third party, they could point to him and say, "Look, we've got a third party, highly respected guy, first class guy. He's looked at it and he says it's all right, but he wants them to do thus and so, kind of thing."
DeVorkin:Is this your group of consultants?
Frosch:No. This was earlier than that, and this is the guy whose name suddenly escapes me.
Frosch:Yes, he's professor of aeronautical engineering at MIT, but you can find that in the committee prints.
Somebody in NASA will remember, or it will pop up properly. I know the guy. I like him. Well, anyway, just to finish. By doing this, they would be able to say, if there was any eruption in the press, "Look, we've looked at it. We have had this guy look at it. He thought it was all right, but they ought to do such and such, and they're doing such and such." There was a period when it looked like he was going to damn the whole engine design. It turned out that it was partly because he really hadn't looked deeply enough to understand the engine design. It was an early period and he was reacting badly, but he was a good guy and he was very careful. He would call up and let us know what he was thinking, and let the committee know what he was thinking, and that was very helpful. It was in the end, helpful. It looked like it was going to be an annoyance, but it was helpful. Now, let me come to the other two committees. The Appropriation Committee: The House Appropriation Subcommittee was Eddie Boland, who is a tough cookie, and a very nice guy, and always critical in a friendly sort of way.
He's sarcastic in what I think of as a Boston Irisher, although he's really from Springfield, sort of way. He always put you offguard, you know. He'd come bustling into the meeting room a few minutes late, sit down and make some remark like: "You guys really aren't going to waste that much money again next year, are you?" That sort of thing. At first, you say, "My god!" and then you realize that's his way of saying, "Hello, we're going to have another hearing today." He was critical of lots of things. Remember, this was a period when we went back to the Congress several times for supplementary money. NASA had never done that before, but we had to. They were all helpful. I'll add a little to this in a minute. The fourth committee was the Proxmire Subcommittee, and maybe at some point it's worth talking about Proxmire. He was hostile in an ambivalent sort of way. His attitude finally was, "I think what you guys are doing is a waste of time and money, but you certainly do it well, and I wish the rest of the Government would do its business that well." That was his finally-expressed attitude — great admiration for how we did it, but kind of wondering why anybody wanted to waste the taxpayer's money on that. He would always want to cut it. while expressing great respect for what it was we did. Pat Leahy on that committee was the same way. He was rather hostile, but very friendly. I got to be friendly with Pat Leahy. I have a great deal of respect for him; he's very straightforward. Proxmire, however, has some traits I am not so enamored of. He's working his own business, as much as anything else, and he was not well prepared for his hearings. His staff would prepare long stacks of questions on cards, and he'd go through them. It was pretty clear he had never seen them before most of the time. Occasionally, he'd come to one that didn't make any sense to him and get some staff consultation. He isn't so much interested in getting answers to questions as in probing for weak spots. If you answer the question, he'll change the subject, no matter how hostile the question was.
DeVorkin:Yes. At the end of the last session, when we had lunch, you mentioned your earlier contacts with Proxmire. I think he would be an important person to expand upon, when we finish this.
Frosch:Yes, okay. So, those were the four committees. Now, when we concluded the first time and the second time that we really had to have more money to carry on this work in an orderly way, we'd go up and talk first to the House Science and Technology Committee, Fuqua and Winn. Then we would talk with the Senate side, which would be Adlai Stevenson and Jack Schmidt. Then we'd talk to Eddie Boland, and then we tried to talk to Bill Proxmire, but he'd never see us, so we'd deal with that by the formal letter. We'd do all this informally before we put in a formal letter.
DeVorkin:You went to the Hill for that purpose?
Frosch:To cover the Hill, yes. When I put these names in tandem, I mean them in tandem. Normally, when one went to see Don Fuqua, he would have Larry Winn there, so we'd have both sides. And we'd discuss it.
DeVorkin:How many people would normally go? Or would it vary?
Frosch:Normally, three of us.
DeVorkin:Who were they?
Frosch:Al Lovelace and I, and Terry Finn, who was the legislative liaison, or somebody from Terry's staff. Now I learned, of course, a little bit more about the Congressional process than I had known. In effect, if they agreed to sponsor a big change like that, they would help us figure out who to talk to on the Hill. We would then start a series of pastoral visits to committee members, other important people. We got a lot of very good advice out of Danny Inouye on one of those. We just went to him for advice. It was his membership on the budget committee that was important, because there was a question of whether the budget ceiling for our piece, as modified for the next supplemental, would have enough room in it, so that we could get the appropriation passed. You know, it's gotten rather intricate now. Anyway, he gave us a lot of advice.
DeVorkin:To straighten out for my record, you went to the Hill two major times. Could you give me the approximate dates?
Frosch:I think there were three times.
DeVorkin:Three times? How would I find them out?
Frosch:Call Legislative Affairs people, because it is formally in the Congressional Record. They resulted in formal supplementals, and there were committee hearings. In fact, Congressional Affairs probably has a detailed chronological list of these records. Terry Finn would be worth talking to — if Terry is still there. Yes, he is still there. He sends me stuff. Terry Finn would be worth talking to about that history of that whole Shuttle thing.
DeVorkin:So, through the liaison office at NASA, we could find out over what period of time there were informal visits?
Frosch:I think you can find precise dates, and the files of what documents were sent up and passed, and when bills were introduced and when hearings were. You could track the whole thing in detail.
DeVorkin:Your daily memo pad is also at the NASA History office. Would that also record this?
Frosch:Yes, yes, it would. It may not record the subjects, but will record, presumably, all the appointments with people on the Hill.
Frosch:So you can track all of that.
DeVorkin:Between the two, we would get a record.
Frosch:You ought to be able to get a very good record of that.
Frosch:It would be worth talking to some of the staff people up there as well, who worked for these guys, if you can figure out exactly who they were.
DeVorkin:On the Congressional side?
Frosch:Yes, Terry would be the biggest help at that. Now, let's see. What else was I going to say about this?
DeVorkin:These were the times when you were asking for more money.
Frosch:Yes, or having problems, or reporting successes. we would keep them tracked in pretty well. Oh, I was going to make the point about the budget process in Congress. In addition to any walking around and explaining ourselves that we would do, those committee chairmen and minority senior members would also go and see their colleagues and try to get the votes. First they would talk to their colleagues on the committee, to make sure the committee came out right, and then with their colleagues on the committee around the Hill to make sure the votes came out right. I gather that this is what happens on a vote, particularly one that is sort of a technical vote buried in the bottom, which after all, these were. The members who are most concerned will stand by the entrance to the chamber and as people come in, they'll say to Don Fuqua, or Adlai, or Larry Winn or Jack Schmidt, "Well, this is your committee, what's it about?" And in essence, the senior members will say, "It's really all right. We've looked at it, you know; vote such and such." This was not a political partisan matter. That is, if they agreed, then we would not have an argument between the Democratic chairman and the Republican senior minority member. They would agree what they would do and they would both try to corral the vote, so it was not a political division. Although there was always some tension between the Democratic administration and the Republican opposition. But my impression was that it was more a guns or butter kind of thing. "Do we want to spend money for NASA, or do we want to do something else?" Now, the best index of the position of NASA in Congress in the period I was there is that all the crucial things we wanted and worked on went the way we wanted them. I think this was almost without exception. There may be a couple of exceptions — again, Terry would be a good one to talk to about that. If they didn't, it was because we had gotten advice on the Hill and decided to do something a little bit different. The votes usually came out the way we expected them to. On a number of occasions, they came out much more strongly our way than we thought they would. There was one vote which, by rights, we thought we would win, but we'd win it in the house by a 60-40 margin. It came out 350 to 50.
DeVorkin:How did that happen?
Frosch:Well. we miscalculated the amount of support we had. We figured that a lot of people realized it was a NASA issue, and they were for NASA, so they voted that way. The other index is that, if I am recalling correctly, the difference between the NASA appropriation as it came out of the Congress and the President's request was never more than about a one percent discount. Now, sometimes they'd do some damage inside by the way they shifted things around, and so on. But the general thing I found was that we could make the case. Basically, we got what was in the President's budget.
Frosch:My problem was more with OMB than it was with Congress.
DeVorkin:In other words, convincing OMB —
Frosch:That it ought to be in the budget in the first place.
DeVorkin:That's what eventually became the President's request?
DeVorkin:Now, did you deal directly with anyone in the White House, or only through OMB? What contacts did you have with the White House?
Frosch:Well, a lot of contact with Frank Press and his staff. A lot of contact with, at first, Bert Lance, and then later, Elliot Cutler. Then there were a couple of successors of his. Then there was one level above that. Bowman Cutter was the deputy. And then we dealt with Jim McIntyre directly. The sequence was to talk it all over with the staff people who were assigned to us. You know, we'd have them over to lunch, or we'd go over there and discuss the budget and issues, and so on, systematically.
DeVorkin:Did you have any more trouble with them? That's interesting.
Frosch:Yes, Hugh Loweth was another guy. He was more or less on our side.
Frosch:Yes. Then we would discuss it with Bo Cutter. Any issues we couldn't work out with them, we'd work out with Jim McIntyre in a meeting with all those guys. We would generally agree on the disposition of almost all issues. I reserved the right, and they recognized that I reserved the right, that there might be some issues I insisted on going to the President on.
DeVorkin:Did that ever happen?
Frosch:Oh yes. It happened every year, I think, or every year but the last one.
DeVorkin:What were the issues?
Frosch:Well, there were some issues of Shuttle funding. There was an issue of getting money in the science budget for the gamma ray observatory, and for VOIR. That was in the spring of 1980. And the year before, the issue was getting money in the budget for GRO, I think.
Frosch:Yes, gamma ray observatory. And or, do we get money for the solar electric propulsion system as a requirement for a lot of things, but particularly for a good Halley mission.
DeVorkin:Right. Now, during the Carter Administration the gamma ray observatory, VOIR, solar polar, and Galileo, all of these missions, of course, were moving along. I'd be interested to know, even though we're jumping ahead, were they moving along as you would wish them? We might treat this solar electric propulsion differently, because there we know that there was a big cut.
Frosch:Well, let's see. They were all different. Galileo and solar polar were a running problem, a continuing problem. They were both a problem because they had weight problems, and of what the Shuttle performance would be.
DeVorkin:I'm not quite clear there. Could you explain?
Frosch:Yes, let me add one phrase and then come back. And those two in turn were tangled up with the upper stage question.
DeVorkin:You're talking about the Centaur problem?
Well, it later became the Centaur problem, but even earlier, it was just a plain weight problem. You do the problem backwards. The Shuttle can lift so much, and you need so much upper stage and fuel after you get it out of the Shuttle to get it where it's going. And that gives an upper mass for the total spacecraft that you can get to Jupiter. Some of that is maneuvering fuel on the way to and after there, and some of it is housekeeping. It's the hotel for the scientific instruments. And then at the bottom of the pile, you've got 200 kilos of scientific instruments. Well, all the obvious things can happen. You use all the weight you can get, all the mass you can get. Every time you have a weight or a propulsion or a thrust problem, or any other problem with the Shuttle, you wonder about whether the Shuttle that you will actually use in that year to launch that thing will be able to lift 60,000 pounds, or 55,000 pounds, or 57.5 or 54.
All that's going up and down. Meanwhile, the guys are designing the the instruments and the hotel — that is, spacecraft — and they are having their own problems, so they're falling out of bed by 50 pounds. They're using their reserve, and John Yardley's juggling his Shuttle reserve. Meanwhile, the Air Force is building what started life as the interim upper stage. When it became clear it wasn't going to be interim for anything, it was named the inertial upper stage. It's having its own performance problems. You've had about a six or eight variable problem, all of which points to the question: can you or can you not get the scientific instruments to Jupiter and do anything when you get there? Now, Galileo was complicated by the fact that you want to get two kinds of things there. You want to get a thing which will go into orbit around Jupiter and do remote sensing, and you want to get a probe into Jupiter that will make some chemical measurements. Ideally, you want to do the two of them at once, so that you're doing the remote sensing at the same time you get some ground truth, if I can put it in military sense. You can use the remote sensing, for instance, to get the ground truth in a place which is yellow, and looks like such and such, and therefore, you can generalize.
A whole bunch of issues like that. When you get into weight problems and interim upper stage problems and Shuttle performance problems, you wonder if you can get the performance up or the weight down. Or is there some way to take this all apart and repackage it, like sending the orbiter to Jupiter in one year, and separating the probe and sending it another year and having them coincide. Meanwhile, the astrologers (as I call them), the planetary dynamics people, are inventing new trajectories. "Maybe, if we fly it by Mars and give it a kick," so on and so forth. I got to be very suspicious. Every time the flight planning guys would come in and tell me that something was absolutely impossible. I knew in a week they'd be in with three new inventions of how to do it. I got sarcastic about it. It was helpful, you know. If something was impossible, they'd think of some way to make it possible.
DeVorkin:They were always very resourceful.
Frosch:Extremely resourceful, a very resourceful crew. I'm partly describing a bureaucratic annoyance, and partly expressing a lot of respect for these people who really worked the problem beautifully. Now, in the case of Galileo, there was a big political complication question you have to understand. Just as I was arriving, there was a big budget battle involving Eddie Boland in the House; on the Senate side, Bill Proxmire. The question was whether the telescope should be done, and/or whether Galileo should be done.
DeVorkin:They were pushing the two against one another?
Frosch:Well, they started out pushing the two against both of them, and then tried to bounce them off each other. In the course of NASA winning that battle, the scientific community really lobbied the hell out of Eddie Boland. He got to the point where he just didn't want to see another scientist.
DeVorkin:By NASA, or the scientific community, that means you got both?
Frosch:Well, they killed Galileo but they kept Space Telescope. In the process — I don't remember the details, but you can find this in the Legislative History — Eddie Boland lost a vote. He went in and tried to get something done as chairman of an appropriation subcommittee on the floor of the House and got voted down. He's a first class guy, but that kind of stuck in his craw, and so Galileo was something for which he had no use. We always felt that whenever there was a problem with Galileo, maybe this is the time Eddie's going to get us, and somehow knock that out of the budget.
DeVorkin:Oh. I see.
Frosch:And he knew we felt that way.
DeVorkin:That's interesting. Were all of these, then, the gamma ray observatory, Galileo, VOIR —
Frosch:GRO just got started toward the end of the Carter Administration. I think it was in the last budget.
DeVorkin:Right. They were all designed to go up in the Shuttle? Is that correct?
Frosch:They were all designed to go up in the Shuttle.
DeVorkin:Yes. Could you tell me a little more about it? I want to change the tape here, anyway. But after I change the tape, I'd like to know the decision structure in the Solar Electric Propulsion and the Halley Comet mission. That's the one thing I know that was really canned. I'm interested particularly in what happened to the Solar Electric Propulsion system — SEPS.
Frosch:I'm trying to conjure up the details. What I remember is the discussion we had with the President about GRO and VOIR. At any rate, I can tell you what the decision process was and how the discussion went. The decision process was roughly that we both wanted. Listen, Noel Hinners can tell you about some of this, obviously. He was the guy and he knows what happened with the others. His recollection will be a little bit sharper than mine on this one. Let me do the second one first. Let me tell you about the VOIR-GRO one, because it's an interesting sidelight on Carter. My recollection is that we were pushing for both GRO and VOIR. I guess we had settled all other issues, and Jim McIntyre at OMB was saying we could have one or the other. That's what it amounted to. I said, "No. I want to take that one to the President." By the time it went to the President, it must have been an alternative. Neither of them had been approved, and I was arguing for one of them. Maybe it had gotten down to that.
DeVorkin:That's even worse.
Frosch:I had asked for a private meeting with the president to bring him up-to-date on the Shuttle. This is also important.
DeVorkin:Did you have to ask through Frank Press?
Frosch:I asked through two people. I would discuss it with Frank, and I would discuss it with Jim McIntyre. In fact, I could call the President's appointment secretary and ask to see the President. But I knew that he would automatically call McIntyre and/or Press, asking what this was about. There was no reason to go around anybody. We would simply tell everybody, "This is what I want to do." I had several reasons for talking to the President; I asked to see him privately, and that was arranged. I wanted to do three things. Remember, this is the spring of '80. One: I wanted to tell him where we really were on the Shuttle, that in July we would decide when we would fly. When I told him before we were going to fly, we were going to fly, and this time I was telling him what I thought was really going to happen, and that I thought it was going to be in March of '. It was all pre-election obviously, so we're just making an assumption of continuity.
DeVorkin:Did you say March of '?
Frosch:The flight would be in '. The discussion is in the late budgetary spring of '80.
DeVorkin:All right, fine.
Frosch:The first thing I wanted to do was bring him up-to-date, tell him why I thought we were really getting ready to fly, why there was confidence, a sort of report to him. I did that. The second thing I wanted to do was say the following: "Look, I consider it my responsibility, the administrator's responsibility, to make the decisions with regard to flight, and to say we go or we don't go. And I don't think you should be involved in that." And he said, "I agree with that."
Frosch:The President. "But, we both understand there is a risk. There is a possibility of great success or terrible catastrophe, and probably not a hell of a lot in between."
Frosch:"And therefore, I think it is important that you, Mr. President, know precisely what's going on. There may be circumstances where I will want to have your ear in a hurry, and not through the system. Can we arrange that?" Finally, because of this circumstance, I haven't asked for this before, I do not have a direct route to you. "There is no single person in the White House staff that I know is my contact with you. If I call X he will immediately deal with you, other than for this special issue which I want."
DeVorkin:Frank Press didn't act in that capacity?
Frosch:Well, he did, but you know, I could call Frank, and Frank could call the President, or it would go to Jim McIntyre. On a couple of occasions we had had what I can only regard as a revolving-door thing. I wrote Frank a memo and got an answer back from Brezinski. Or I wrote Frank a memo, or McIntyre a memo and got an answer back from Frank, so that there is something in that black box that's bothering me.
DeVorkin:Obviously, you needed some contact.
Frosch:So at this point, I said, "It hasn't mattered before, but now there is a lot at stake. I want to be sure there is somebody who knows that he or she is the person between you and me." And he said, "Okay, I will consider that." We had some discussion of who it ought to be, Frank Press or Jim McIntyre.
DeVorkin:It had to be one of those two?
Frosch:Well, they were the logical ones. I was not going to ask for Brezinski — that was a loop that would be useless for me and for most other purposes, but that's another matter. I don't have a high opinion of him. That's a political loop of wild characteristics. It was really those two. The final conclusion was that the matters concerning me were either going to be neither technical nor money, but have to do with presidential risk, or they were going to be money.
Frosch:Yes, public risk because of the Shuttle. In which case, I was asking for our contact to be direct. I knew one other thing. Frank Press could get to see the President when he wanted, but not necessarily instantly. But Jim McIntyre, both because he was OMB and because he was an old Georgia colleague, could see the President instantly. The President said, "Which would you prefer?" And I said, "Well, I will take — if it's all right with you — I prefer Jim McIntyre. because I think you and he have more direct contact."
DeVorkin:And what was his reaction?
Frosch:He said, "Well. I'll think about that. I'll think about all these issues. Let's decide them in the course of the meeting."
Frosch:That was about to happen in the Cabinet Room, having to do with the GRO, VOIR issue.
DeVorkin:Which did include Frank Press and McIntyre?
Frosch:Which included Frank Press and McIntyre, and so on.
DeVorkin:And who else?
Frosch:Oh. various other staff people, several people from OMB, a couple of Frank's staff people, somebody from DOD, but not Harold Brown on this occasion, somebody from NSC, not Brezinski or Aaron. I forget who exactly. Then we went in to discuss GRO-VOIR. Tim Mutch, who was Noel's a successor at NASA, came with me on this one. We had previously agreed that if it came to the crunch — which do you hold in higher priority — that we would have to choose VOIR. In the meeting I started to outline the case. I think I have this tangled up. What you may have to do is go back and find the budgetary sequence, and we'll have to go through it again with you jogging me from what year, actually what happened then, when GRO went into the budget and when VOIR did, and so on. But the vignettes are slightly independent of that detail, although they are relevant. It seems to me GRO was already in the budget from the previous discussion.
DeVorkin:Yes, let me put it on pause.
Frosch:Leave it on. We had the discussion of SEPS and Halley versus GRO in the spring of '79. That's when we got GRO. In the spring of '80 it was VOIR and Halley again. The point is, we had agreed that VOIR was our choice, if we had to make a choice. Okay, so then we had this discussion.
DeVorkin:In the President's office?
Frosch:In the Cabinet Office. I don't remember the details of the science discussion. Anyway, the President made it clear that we could only have one or the other. Al Lovelace will remember these. At the end of the meeting, the President summed up the issues that I had raised with him privately, namely, this business of direct contact near the time of flight. I had made it clear that's what I wanted, and the question of a direct contact person. He said, "Not now, but when we get close to flight, within a month or so, I want to be reminded. And, Jim (McIntyre) you make a note that I want to leave instructions that I am to be reached at any time by Bob Frosch." He said, "In order to make clear what's happening in this period, Jim, you ought to take the lead in being the direct contact." So he made those designations. Now, let me jump back a year earlier, and assuming that I've remembered correctly that it was GRO and SEPS before Halley. It was SEPS before Halley, because we weren't ready to start on Halley. What we wanted to do was get the SEPS thing started.
Frosch:That was an interesting discussion, and there as we came in to sit down for the discussion, the President got called out for a phone call.
DeVorkin:This is a year previous.
Frosch:Yes. I may be having this all tangled up, but just think of them as anecdotes for the moment. They probably have more significance that way than necessarily tied to anything. As he got up, the President said, "Well, I gather we are going to discuss GRO and SEPS. Now, they're both good projects, but I have to admit I'm partial to the gamma ray thing because of its connection with the black hole problem. I've been reading Walter Sullivan's book. As a matter of fact," he says, "I can't find the book. Chip must have taken it (laughs)." It was clear from the subsequent discussion that he understood that. In the end, I was asked what I thought about the GRO and Halley. And I came down saying, "Look. Halley is very spectacular and, in many ways, important. But if you were to ask me where I think there's a Nobel Prize, it's probably more likely in the data from GRO." which was what we thought at the time. Current ideas may be a little different, but I still think that's probably right. The Halley thing is worth doing, very much worth doing. It will be very important, but personally. I have some doubts as to whether the data will be all that definitive, if we get to do it.
DeVorkin:To clarify the Halley mission, is this the one where, as they are talking now, the only valuable mission that NASA sees is to actually go and sample it, and bring back the samples? Or is it the remote sensing?
Frosch:No. this was a more elaborate mission. You see, if we had SEPS we could do a dual comet mission. That was the Halley-Tempel IT.
DeVorkin:That's right — the one to actually catch up with it?
Frosch:The idea was to make an encounter of Halley which was close enough probe into the coma.
Frosch:Plus remote sensing. Then you intercepted Tempel II and flew along with it for awhile. The key point was that people thought that Tempel II was less interesting than Halley, but worth looking at. It's less interesting because it goes around the sun regularly, and it's probably been much more changed by the solar wind than Halley.
DeVorkin:Yes. Halley's more representative of the primordial soup.
Frosch:Yes. Well, I thought it would be an interesting set of side things. I don't know whether you are going to talk with Al Lovelace, but you should. His recollections may be clearer, or distorted in a slightly different way than mine, or fading out in a slightly different way.
DeVorkin:I would like to get more clarified budget sequences. I think that would be very helpful.
Frosch:What you want to find out on these is exactly what happened in the years I was there with the project starts for the budgets for GRO, VOIR, and SEPS. What you need really is the project breakdown of the budgets available.
DeVorkin:Yes, those should certainly still be available from NASA.
Frosch:Oh absolutely. Yes, somebody will have them, the comptroller's office. In fact, you probably can find them in the Congressional Record in the Committees I mentioned. If I had them in front of me, I could disentangle some of this. This is why I'll never write memoirs. My memory isn't good enough. I hate keeping the papers (laughs).
DeVorkin:If we can establish the structure, we can always fill in the data.
Frosch:We can probably fill it in, yes.
DeVorkin:Right. But there are two things that we would not be able to fill in. First of all, these were all very different projects, and the science community must have lined up behind one or another; planetary people behind one, etc.
Frosch:Somebody lined up behind all of them.
DeVorkin:Of course, but then there was the age-old, nagging question that somehow was realized in the problem of weight and the problem of feasibility.
Frosch:"Can you do it." yes.
DeVorkin:The "can you do it" with the Shuttle, as opposed to the fact that it was proven that one could get these things up there with expendable launch vehicles. Now, that decision had already been made before you got to NASA?
DeVorkin:But, I'd be very interested to know about this juggling of the scientific missions. What went through your mind in this order, with discussions of the scientific community as to how it could have been? What were the criticisms?
Frosch:JPL, particularly, kept suggesting that we buy some Titan-Centaurs, and just go and do it. We discussed that formally on a number of occasions. To begin with, we didn't have the money to go buy Titan-Centaurs. That would be another requirement, a new budgetary requirement. In the second place, you couldn't get the Titan-Centaurs on time, just because of the sequence of manufacture and where the whole Titan line was.
DeVorkin:This was for the Halley mission?
Frosch:Well, for any of these.
DeVorkin:For any of them.
Frosch:That's my recollection. And finally, there was a lot of political jeopardy for everything, not only the JPL guys. I'm just symbolizing them as the science community. They would accuse us of protecting the Shuttle at the expense of the important stuff, not realizing that when you open up that kind of an issue, it isn't the Shuttle in jeopardy — you're jeopardizing the whole program. You're also running a terrible risk of somebody stepping in —Eddie Boland in the case of the Galileo — and saying, "Look, you guys don't even know whether you can launch it. You haven't even decided on a vehicle. Let's cancel it, or let's put it off until you make up your minds." Or if not Eddie, somebody else would come in. They had very little perception of what the real political risk is. It was not all that one-sided. In fact, the risk was more to the science than to the Shuttle. You know, there were sharks out there who would like to get the Shuttle — not particularly on Capitol Hill, but elsewhere. Although Bill Proxmire would have liked to. The Shuttle was more firm politically than the science was. There were more people who would say, "Oh well, so you don't get the data on Jupiter," than there were people who would say, "Okay, so you don't fly the Shuttle." There was a misjudgment of the jeopardies involved. There was a tendency to suggest that it was just manned space flight guys who were protecting this stuff, and so on and so forth. But there was not a real perception of the political jeopardies, and, by the way, not a real perception, of how budgets are made. There are a lot of people in the science community who think there's a kind of conservation of budget involved. That is, the administrator gets X billion dollars a year, he then decides whether he's going to put it all on the Shuttle or put it all on science. Therefore, the Shuttle decreases the amount of money for science. It is nothing like that.
DeVorkin:What is it like? How does it work?
Frosch:The budget is built up from the pieces. There is a little bit of conservation of budget in the sense that, if we have to go back for extra money, can we go back for this or that, or only just for this? The interaction is not that strong. In fact, I would say that if the Shuttle were not in the NASA program and had never been in the NASA program, it would not necessarily be the case that there would have been more money for science. What would have happened is that the whole NASA organization and budget would have been smaller and weaker, and it would have been extremely difficult to do things like Voyager and Galileo because they looked so large in comparison to a smaller budget. I think the politics is entirely different than they think it was. It's as if the presence of the Shuttle, for many people on Capitol Hill, is the real thing, partly because of the military connection, but partly because of the whole manned space program. Given that you're doing that, it is reasonable to do science, too. This may not be rational in some sense, but it is, I think, a real set of political considerations. It's almost as though for some people the science is a sort of sop to the people who are doing this other stuff.
DeVorkin:This brings up an interesting point, as far as the goals of NASA are concerned. First of all, to what extent do you think you made the reality of this situation apparent to the scientific community, and did it get across? I'll ask that question first.
Frosch:By and large, I think most of the scientific community really doesn't see it that way.
DeVorkin:Who was most vocal against it?
Frosch:The members of the scientific community who worked closely in and around NASA area. Noel Hinners understands this. I think he probably even agrees at this point with this idea that the Shuttle was a benefit to the scientific budget. I don't know — it is worth asking him. Tim Mutch came to understand it. The point is that most of the members of the outside community — it doesn't matter whether it's scientific or not — who have not dabbled in the real construction of the budget, and dabbled with the problems with OMB, the White house, and Congress, really don't have a clear idea of what the structure of the thing is. This may just be ignorance, because I don't read that literature so much. I've never seen in any popular press or even in any foreign affairs magazines or in political or economic literature an article that really discusses how the budget business is really done. You'll find things which give you exhaustive detail of the formal process: who writes what document, to whom and so on. You'll find a lot anecdotal stuff about horse trading on Capitol Hill. I've never seen something that describes the kind of political-technical trade-off that I have just been talking about. There's room for a book for somebody. I don't know whether it's a history book or a political science book.
DeVorkin:In many ways, it's policy and politics.
Frosch:It's policy, politics, technical decisions, scientific decisions and what will the market bear, what will the budget bear. It's like many other kinds of business decisions, played with a lot of variables. Now, I would say the most vociferous critic is probably Bruce Murray. The planetary guys feel that they were sacrificed on the altar of the Shuttle, and what a glorious thing was shot out from under them. There are lots of things I would have liked to have done, but I don't think we did very badly. In fact, it looks a lot worse at the moment than it did before. I feel that the apparent expectation that we were going to do everything that could be done was just not realistic. I will admit that getting the planetary science budget was the weakest part of what we were able to do.
DeVorkin:I don't follow that.
Frosch:I would say I failed more to get a planetary program and budget than I failed to get other things.
Frosch:We did better in other areas. Everybody gets terribly excited about Voyager, for example, as do I. It is elegant science and quite important. It comes at a very high price tag. It's expensive. It's just terribly expensive, and there are some people on Capitol Hill and elsewhere who think it's lovely and so on, but 600 million dollars is a lot of money to go to Jupiter. I remember the first time that Eddie Boland and I met. I made some calls before I was even confirmed. I guess we met down in the Senate cafeteria, or the dining room. The only time he could meet me was at breakfast. We sat and had a cup of coffee. We talked and it was very pleasant. One of his opening gamuts was, "You scientists don't know where the money comes from, but you sure know where to put it!"
DeVorkin:(laughs) Interesting. What about the question of space policy? I know that some of the Congressmen, Harrison Schmidt for instance, were asking you for a space policy early in your career.
Frosch:I must say, in a certain sense, I didn't know what the hell they were talking about. I never thought there was any question about what the policy was. What they really wanted was a document that one could squeeze out of the President, that could then be waved on Capitol Hill and used to bludgeon OMB. What everybody really wanted was another Jack Kennedy standing up and saying, "We are going to get to the moon by the end of the decade." so that the next time the OMB guys tried to cut something, you could wave the President's speech at them. That's what I really think.
DeVorkin:People on the Hill?
Frosch:People on the Hill wanted that; that is, the people who advocated increasing NASA. I think that's what Jack Schmidt wanted.
DeVorkin:What was Frank Press's feeling about that? I mean, was there an idea that there should be a structure to the long-range programs that NASA was developing?
Frosch:You see, what bothers me about this is that there was a structure. There was an obvious structure. I could explain what the structure was, but everybody wanted a policy. Jack had this long-range plan — he still does. He's got a bill in again, I guess. The plan is to declare the decade of planetary exploration, the decade of planetary colonization, and so on out for three or four decades. Maybe I just don't understand the political process. But this seemed to be something that would take an awful lot of work getting, and I couldn't understand what you had when you got it. Okay, suppose you had a bill passed, and signed by the President that says that the 80s are the decade of planetary exploration, or the decade of manned low earth orbital construction. Now, maybe that's worth something, but my reaction is, that and fifty cents will get you a cup of coffee. It didn't seem to me that was the place to spend political energy, as long as we had in mind what we were going to do. Now, one thing that suggests that they may be right, however, is the extent to which the idea of producing a space policy did terrify the OMB people.
DeVorkin:In other words, to make a Kennedy-type statement: "We'll land on Mars by —"
Frosch:Well, that would have terrified them completely.
Frosch:But even the kind of thing that we finally wrote. Incidentally, it was better than the press document, not the Frank Press document, but the media document that was written to advertise it. OMB (Elliot Cutler) threw in totally useless comments like, "This gives us the freedom to adjust budgets when necessary." Well, everybody in town knows you adjust budgets when necessary. If you put that in, you are making a gratuitous comment that doesn't do any good. I thought that whole thing was a futile exercise.
DeVorkin:How much time did it take?
Frosch:Oh. it took months. This happened in the course of '77-78.
DeVorkin:In '78 you put out a policy paper.
Frosch:Yes, that's the one, that was the President's policy paper.
DeVorkin:Right. But then you had another one on in-house science at NASA that was not related to this?
Frosch:No, that was an in-house thing. This was called the Carter space policy.
Frosch:The one that said that it was not the time to do large scale this and that, which was a gratuitous comment. It was probably true, mostly because we had sopped up all our effort with the Shuttle, and weren't going to do anything until we got that flying.
DeVorkin:That's right. Frank Press' feeling about this? Did he agree with you that you did have goals?
Frosch:Yes, he agreed that we had goals, but he thought more than I did that it was worthwhile writing a policy paper.
DeVorkin:He did think this?
Frosch:Yes, I think he did. As I recall, he did.
DeVorkin:He and Schmidt were on the same wavelength?
Frosch:Well, I don't think they agreed on what the policy should be, but they agreed that there should be a policy paper. There was a general agreement there should be. I think it must have been Frank who got the President to agree that we ought to do a policy paper, and then we structured some kind of a process to do that policy paper. It involved our writing stuff, an interminable process of trading papers back and forth. I can tell you who is available — I hope he's available. He was in the middle of that process and can tell you more about it than anybody in the world ought to know, and that's David Williamson. I think he's worth talking to extensively, because he's a mine of inside history of NASA for 20 years; because he's more or less been in the administrative office for a long period of time.
DeVorkin:Right. There's very little information on him in the various information agencies in NASA, but I'd like to be able to talk to him.
Frosch:Yes, you should. He is an historian by the way, by trade, so he has a good sense of that kind of thing.
DeVorkin:That's a good point, okay.
Frosch:So I think it would be worth talking to him on all of this.
DeVorkin:Yes. Well, we've talked a good bit about NASA itself, but we really haven't formally gotten to NASA (laugh).
Frosch:Yes, that's all right. Catch them when they're jumping.
DeVorkin:Oh yes. absolutely (laughs). We haven't, in fact, even gotten to —
DeVorkin:Navy, yes. I hate to leave NASA. As you well know, this is really what we want to talk about, but I think we should go back and pick up your later period. Thank you very much for this session.
Address by the Honorable Robert A. Frosch, Assistant Secretary of the Navy for Research and Development, To the Graduating Class, Weapons Systems Management Class, Wright Patterson Air Force Base, 15 June 1967.
Venus Orbiting Imaging Radar