TAPE 1, SIDE 1
MR. COLLINS: After reading through the analyses that you prepared on both North American's role in the Saturn II program, and the Apollo Command System Module program, there was a sense of the unresolved in those documents. I wasn't clear whether or not after reading those that you felt that North American was either in the right or the wrong in the way that they handled those programs, or whether the types of basic criticisms that were leveled by NASA were actually just. I know that you felt clearly that North American was either not performing in good faith or that their motivation wasn't sufficient were wrong-headed. But on questions about the management capability and the engineering capability, I wasn't as clear whether you felt there was substance to those concerns.
MR. ATWOOD: Well, of course, it's difficult to hold categorical views--certainly, nobody could expect anything near perfection the first time that these structures were designed or during the first models, when there's no chance to check them out against actual service. So totally, I guess I felt that we did a good engineering job. All the components were made as expeditiously as possible. However, the nature of the engineering made it pretty hard to judge that with any fairness, because the factory had to do their work over again so many times, because of new ideas, changes, and adjustments to other parts of the project and other elements of the hardware.
I do feel quite strongly that the allegation that the poor workmanship caused the Apollo fire is completely wrong. I think it's demonstrated that the fire could not have started in a reasonable cabin atmosphere. This was proved literally when we put 40 percent nitrogen into the oxygen atmosphere, that's going halfway up to standard air which is 80 percent nitrogen, and then we laid incandescent wires across the inside of the spacecraft, and the fire would not propagate. So that's almost prima facie evidence to me that we did not literally cause the fire.
The big mistake was in assuming that 100 percent oxygen was even needed, when people were in their pressure suits and getting their own breathing air and discharging their own nitrogen into the oxygen ventilated pressure suits. The idea that on top of that was needed a high pressure of oxygen was literally wrong. I don't mind saying categorically, that was wrong. That and that alone was the factor that caused the fire, and the death of these people. You cannot reasonably expect a first article, particularly a first article, with test equipment wired into it, to be completely 100 percent free of any possible heat concentration or possible spark. That apparently was what was presumed. So for that incident, I feel that the company took the blame for something it wasn't responsible for.
COLLINS: In your extended discussion of this point, you feel that pretty much from the top of the North American organization down to the engineers working on the CSM, that they were not aware of this test regimen that involved oxygen pressurized?
ATWOOD: Well, I think the people at the Cape that saw it being inflated with oxygen, were aware that is the technicians there that were working in support of NASA. I've got a paper which Dale Myers gave me just a month or so ago. It's a speech he gave at the University of Michigan this year or last year, covering almost the same points I made in my paper on the Apollo fire. He cited other incidents also that supported his point that better systems engineering might have prevented these things. But he made the point that this was something that was unexpected. I'll give you a copy of it if you want it.
COLLINS: That would be interesting. I guess maybe it speaks to one of the problems that perhaps Sam Phillips was pointing to, and that is a lack of an effective understanding of the whole job by the key people involved. That seems to be really his most fundamental criticism, when you read through the documents. That's really what I think he's referring to, when he refers to systems engineering as an area of inadequacy and ties that to the problems in cost and schedule that--[were encountered?]
ATWOOD: Yes. Well, systems engineering is really tying together other engineering elements, and the objective of course is to make sure they're compatible, work together, and secondly, try to make sure nothing is forgotten. Lots of things are forgotten. I'm afraid that will never be completely eliminated. But I also pointed out in my discussion with you the elements of the Apollo program, and components were given to various companies, and each of us was a component manufacturer. At no point was there an overall performance responsibility in any single company.
Boeing could only be responsible for the stage of their boost, and then, that wasn't enough to get people to the moon so the next responsibility was North American, second stage. It had its performance parameters, as I pointed out, and so did the Douglas S-4 stage, and then naturally the Service Module and the Command Module. But it's pretty hard to say that one element of the booster had more responsibility than another. The North American part of it turned out to be the most efficient part, by a big fraction, but--I don't know how you close that loop. I really don't. I know if it were our airplane, speed, and range and performance would be our responsibility. But the Apollo was not any company's responsibility. The systems engineering that involved the oxygen inflation was NASA's responsibility without question.
COLLINS: Just to carry us a little farther along, a quote from this memo that you've seen that Sam Phillips prepared in December of 1965. As follows: "One of the fundamental difficulties adversely affecting progress of both programs is in the area of planning and control. There are over 2000 people and some 40 organizational entities. Integration of detailed planning, scheduling and budgeting is woefully deficient, and ability to track detailed progress at successive levels all the way to the top is either non-existent or totally ineffective."1 He then goes on to outline some specific recommendations for improving that situation.
ATWOOD: I don't believe that.
COLLINS: It seems to touch on something we were discussing earlier, and that is the expectations about how you express, in some format that both the corporation and the contracting agency can agree on, for understanding what's happening. Your sense is that the systems that North American had in place to track the progress of the program, its relationship to schedule and cost, were adequate.
ATWOOD: The system was certainly adequate. The planning dept. personnel were competent and experienced and orders were properly prepared for material tooling, parts and assembly. The most difficult part of the S-II, the common bulkhead was successfully mated by October 1964.
The thing that people don't understand, I guess, and never will, is that until something is pretty well engineered, the rest of the system won't work. You give various orders to buy some material, and then you put a Stop order on it because you've changed your drawings. Immediately you start to screw up the system. When the planners have planned the tooling and manufacturing sequences of a part, and that part is changed, the engineering of it is changed, of course everything else is changed.
These changes were in large part weight related. If we could have could have laid down that S-II stage with each piece of material 20 percent thicker than it was, most of these changes wouldn't have had to be made. Twenty percent more weight added to it would have made the engineering pretty straightforward. Then when the drawings are set, then your system will work. When drawings are not available, and not stabilized, then you're stirring up, you're churning your system, and a lot of that took place, primarly to reduce weight. Certainly it did also on the CSM.
In the Rocketdyne, on the other hand, there were a lot of changes but they were self-generated, within the Rocketdyne Division. They had control of all the engineering. In fact, they controlled the interfaces because the booster stage designers were the people who had to adapt to the engine, the way it came out. Rocketdyne handled their changes and kept them within their planning, material ordering, tooling planning, toolmaking, manufacturing and finishing-the sequences, you see. So it does make a difference.
Take the CSM. I didn't consider that any stretch in technology, never have, and nobody else did. We didn't have to invent any real techniques either to build it. We built several of them. They'd been launched and even come back from space at full lunar transfer velocities. But the arrangements inside and the instruments and wiring and all that were in a state of flux for a long, long time. That's where the astronauts had to consider what would be the most suitable for them. They'd never been in this kind of situation before, and they'd change things a lot. So that was the nature of the program we had.
I think the Grumman Lunar Module had very much the same thing. They weren't reaching for any technology extensions either, as far as I could tell. But they had an estimate and budget weight on that vehicle of 25,000 pounds. That's what the entire system was being designed for. In his book WINGLESS ON LUNA2 Neil Armstrong said the weight was 32,500 pounds. I guess he's about right.
But Grumman went through a very, very severe traumatic period, because they were trying to make their specification weight and they couldn't do it. They went through, as nearly as I can read their record, an awful lot of, oh, self-criticism, worry and ostensible effort, that is, visible effort to try to make the weight, and they couldn't do it. Finally NASA sympathized with them, and worked with them, and they finally got a design which was 32,500 pounds, as I say, and NASA was convinced that Grumman had done everything they could to bring the thing in at minimum weight. But it was the kind of thing, visible effort from every side. Astronauts were there. They could feel the tinniness of the structure when they got into it and readily believed it was as light as possible.
COLLINS: You're talking about the CSM or?
ATWOOD: No, I'm talking about the Grumman, the Lunar Lander. I'm just making a comparison. They had the same problem, and they came out to what did I just say, 30 percent overweight or so. In some way it became visible enough and understandable enough to everyone working on it that this was acceptable. Now, if one of the Rocketdyne engines had been 30 percent off, you see the difference? That would have just been irresponsible bidding, to bid on something that it couldn't make by 30 percent. But the engine isn't a case that you understand well enough. I think Grumman did a good job. I think they did everything they could to make that weight. Dale Myers on the Command Module had the same problem, although the percentage wasn't as big. He came out, I've got the figures on it, roughly, close to 90,000 pounds, the CSM and Command Module. But he was getting more understanding for his problem than the boys working on the second stage were getting. It was just there, just make the weight. That was awfully hard to do.
COLLINS: Why was it more difficult, for NASA to appreciate the kinds of problems that North American was confronting in these programs?
ATWOOD: I don't know that it was. I think that von Braun understood it very well. He was a very intelligent guy. I think some of the people at Huntsville understood it. But people like Eberhard Rees, I think they just thought they'd take it out of somebody's hide, you know, with forcing it. We did, I think, get the weight out in a very effective way. Now, that's what caused much of our problem, and I guess you'd say we were out of control, and you know, companies can falter in a lot of ways, of course.
I think the way I would have done it if I'd been a project engineer on one of these things would have been to relate everything I did somehow in the minds of the top NASA people to the total performance of the stack. I'd have found a way to do that. I don't think, as I say in my supplementary note here, I don't think Phillips had the advice of a really competent structural engineer.3 the best advice we could get from experts that we employed, and I think we extracted everything there was in that design possibility. But we didn't make it very well understood.
COLLINS: Earlier you had indicated the centrality of getting specifications firm, of being able to get drawings that you can rely on. In a situation like this, where there's some back and forth about the nature of the final product, where does the responsibility lie to achieve that firming up of the specifications?
ATWOOD: Well, the chief engineer has the direct responsibility. I suppose the whole company has a responsibility there. When I first began to realize the implications of that S-II design was back in about '63, '64, and it was really quite a shocker to me in a lot of ways. We'd never had anything really to compare, nothing that we'd ever built that had just that form of a problem with it. The scale of the thing is unprecedented in size, as you know. And the scale factor is an important aspect of the weight, the strength issues. I know even Harrison Storms didn't understand it. I did because I was structural engineer. But he was a little too optimistic in what he promised as an engineer. He promised, as near as I can tell, 93 percent fuel/mass fraction, and we came out at 92.5 or 92.6, and the S4B as I pointed out, which is only one-fourth the weight, had an 89.6 mass fraction, and the difference, if related to the S-II, would have been some 25,000 pounds, which, as I say, would have amounted to thickening everything out about 20 percent, with the corresponding increase in weight.
COLLINS: Resuming after a brief pause.
ATWOOD: Yes. To summarize, I don't believe there was anything wrong with our system. It had been tried and tested over all kinds of experimental aircraft, and production aircraft, and I don't think that was a valid point of view. Of course, nothing's perfect, and a system is no better than the people that are operating it, and make all the qualifications you like, and we did make undoubtedly many mistakes and backtracked perhaps four times. We had to.
But the management of NASA was for a long time, at least, in the early stages when designing was being done, it was not a coordinated set of instructions to the manufacturers. It was a continual sort of a rumble of desires of various elements of NASA's technical staff, information on history, things that had worked and things that hadn't worked, and resolution of arguments among the NASA people themselves in many cases took quite a lot of time.
So all those things are true, and I suppose you could easily imagine, starting with a clean piece of paper and making something quicker and simpler, but my experience doesn't lead me to feel that it's really likely. I've seen many, many a design go through experimental production, all these things, and in the last analysis, the work we did for NASA was no different in principle than what we'd been doing on aircraft.
The structures, the metal work, the finishing, the sealing, all those things embedded in the structure itself are very well known to all our people. The instruments, the instrument mountings, control functions, all those things were nothing really new. The heat shield never gave us any real trouble. It might have been a little heavier than it would have been if we'd had more time to experiment gradually. It's pretty empirical how that heat shield is going to char and burn away. It's also fairly empirical about where hot spots are going to occur on re-entry, and how the air flow is going to behave itself exactly.
But those uncertainties were pretty much submerged in the general margins of safety that we had in the basic design, and so, to me it's just a matter of, well, you take a big airplane, whether Boeing builds it or North American or whoever. It usually takes four to five years on a new design to bring it through, and make a first experimental test flight. A lot of things have to be tested and re-tested and re-designed, and much is done towards optimization, rather than just a something that will meet the form, fit and function of a requirement. There's no use building these things unless they're better, more efficient than something that goes before. So they all take about that same kind of a cycle.
But when the instructions for building it are not well coordinated, like when you're building for four or five airlines, they all have inspectors and operating people and pilots there--it's been a very disruptive thing in the past. I think these days they've gotten it all down where the airplane company has control. But there have been many times in the past when that isn't true. So I don't think there was anything really unusual or unexplained in North American's performance as compared to anyone else's.
For instance, we had worked out on the S-II a cruciform type of support for the five engines. It was a beam went this way and one across this way, one engine was mounted at the center of the crucifix, one on each beam arm. We had designed and developed an especially efficient and light beam of stainless steel, and it worked out fine, and we offered it to Boeing for the first stage, which had much bigger engines. Boeing looked at it and said, "That's fine, but we really don't need it. We don't need the weight reduction." So they didn't accept it, didn't use it. Maybe you can see the difference, to some extent.
Now, the Boeing S-I stage was really quite efficient. When you think about it, was probably over a 93 percent mass fraction, of fuel, but the fuel was JP, very dense compared to hydrogen, so for a given weight, they had only about l/11 the volume. Of course, volume is one of the factors that adds to weight. But they didn't have to stretch to get that efficiency. It came almost with the necessary physical design of the tanks. So we did have a harder job. But the way it worked out, there wasn't too much difference in the final development time. If we hadn't lost that one test unit, which was overpressurized with helium, after a firing test, we would have been about even with the other stages, that is, with the Boeing stage, which started at the same time. We'd have been about even on finishing.
Of course, the S-IV was done before for an earlier version of the Saturn, Saturn I-D, you know. But when we got all through and measured up, our cost was not high. It was, on a per pound basis, the way they sell airplanes, always have, it was half the cost compared to the S-IV B, which had been perfected some time before. So I don't think there was anything really unusual, is what I'm trying to say. It's almost SOP, Standard Operating Procedure. The S-II cost and production record clearly shows that the planning and control methods could not possibly have been as the Phillips report indicated.
COLLINS: But you didn't, if I read your essays correctly, have a grasp on those kinds of numbers and comparisons to make that judgment.
ATWOOD: No. I had no comparisons to make. We had nothing to compare it with, that I was aware of.
COLLINS: The situation you described at NASA, is that what you mean when you refer again in these essays to "NASA spoke with many voices"?
ATWOOD: Yes. Of course. Borman's4 book, interesting thing, he said he took charge of the spacecraft for the Apollo 8 flight, the first circumlunar, and he had an argument with George Mueller, and he claims that he went to NASA headquarters and got a decree that Mueller wasn't to have anything to do with his spacecraft, and he said Mueller didn't. Well, that's what I mean. Those are Borman's words. Of course I think Borman was wrong on a lot of things, but he wrote quite a book. That was one example. The others are, well, just the variety of people that would come in teams of 100 or 150 at one time to camp and go through everything. It was a highly supervised operation. That's putting it mildly, when I say they spoke with many voices.
COLLINS: This is something I shared with you previously, it's something that was prepared, the date is April 20, 1967, and it's a memorandum for Mr. Webb.5 The subject is negotiation with North American Aviation. It's part of the consideration of where things should stand in the aftermath of the fire. It's apparently prepared by one of their counsel staff, and it is an examination of where NASA thought North American might be able to level some criticism.
Let me just read a couple of things here because they tie into this discussion. The title of this outline is "Possible Criticism Which May Be Aimed at NASA by North American Aviation Inc." Under heading 1 they say, "There was no single face in our dealings with NASA. North American is accustomed to working with DOD where there was a single on-site manager, and a single project manager at a remote location." Under this they list several other things which I'll read as well A, "There was no lead center, headquarters NASA group or resident group that had clear authority to deal on the total NASA-NAA relationship problems. B, "The NASA work at Rocketdyne was under the technical direction of MSFC. Quality assurance and contract authority was split between the resident Air Force and MSFC people." Then next point, C, "MSFC spoke for the Saturn II vehicle at the space and information division, MSC spoke for the Command System Module at SI" and D, "Neither could speak for the total or coordinate differences. The result was that S and I D was directed to operate differently on the two programs at greater expense."
ATWOOD: That was a very small part of the problem. That was all true, but I wouldn't make a point of it really. In fact, that whole paper that you have there in April, 1967 seemed to be anticipating some kind of a legal challenge or some kind of a kickback as a result of their report on the fire. Of course, I had no such intention at all. But it seemed a little peculiar to prepare such a thing. I don't know why.
TAPE 1, SIDE 2
COLLINS: For my thinking, the import of this document is not so much its application to legal considerations, but it highlights some of the issues that we've been talking about, about sorting out the major relationship between North American and NASA. A couple of other things are mentioned here that I think are worth setting down. The Saturn-II contract contained a detailed model specification, while the MSC contract contained only a broad general model specifications. There was apparently a deliberate attempt by the MSC to keep the CSM specifications vague for the first two years of the contract. Then it goes on to say, "MSFC had a strong configuration management program, while the MSC placed little emphasis on configuration management at least through block 1." I think this sort of touches on our question about specifications.
ATWOOD: That's true.
COLLINS: But you feel these things are really kind of minor issues in all this?
ATWOOD: Well, they were complicating factors, but relatively small, the way I see it. One of the things that was a little more troublesome was the fact that the centers, that is, Marshall and Johnson or Houston, were being held accountable or rather responsible for all recorded changes. Maybe I should put it another way. The changes that they required, formally required were supposed to be controlled, so what the North American people, Stormy (Harrison Storms) and his people thought was that Houston wanted to avoid a detailed specification as long as they could so that they would have free license to make changes by waving their arms. Just by instruction. If there was a document against which the change had to be recorded, that was one thing. If the document didn't exist, changes could be made by informal verbal communication.
The people at the space division, S&I D, thought that both centers were trying to avoid the specification formalization because any departure from it would then go into the change traffic, which they tried to control very strongly, as you know. Now, to what extent that's true, I don't know, but it was a point of a lot of debate and argument. Are we going to have a specification or are we not? For a long time we didn't, you know. So that's an open point. It's just one of the things that goes with something as new as this type of project. I think if it had been an airplane with a lot of background, things would have stabilized a lot quicker, as far as requirements are concerned.
COLLINS: A point of comparison, where there was apparently a problem in the Autonetics Division when they were working on the guidance for the Minuteman Program and you mentioned in your essay that Sam Phillips or someone else had organized the so-called Tiger Team to come in.
ATWOOD: Yes, we had that same thing.
COLLINS: How did that situation compare with what was happening in the NASA case? What was the North American response in that particular case?
ATWOOD: Well, I'll tell you the general framework of what happened on the Minuteman guidance. First we had the Minuteman I. It came at a time when integrated circuits were not really ready to use, so each electronic action was controlled by a diode, a transistor, coupled with some kind of a transformer, a resistance, and a capacitor. Those are the main elements used in electronic circuitry at that time. Well, a diode's a discrete unit, so's a transistor, so are these other things. But they're common electrical circuit control devices.
Then we used a tape for a memory. It was a punched mylar tape on reels. Of course it had to be punched very accurately. The program was put on that punched tape and it would go through and control electronically the actions of the engines, the guidance, the gyroscope cutoff, all those things that go with the guidance of a missile or a spacecraft.
Well, we made an estimate of the cost. It was a very poor estimate, very low. We overran it very badly. But one of the requirements was that the missile operate in the hole, in its environment, running for about 8000 hours without service and replacement. Now, 8000 hours is about a year of time and when you have all those components, you've got the probability that one of them will fail.
Now, if they all hold up for 8000 hours, the probability equations are such and the realities are such that the average time for failure of each of those components has got to be much much higher than that. Like a million hours. A diode would operate a million hours without failing, so that by the odds of chance, one of them won't fail in 8000 hours or so. It's not a very unusual concept, and it's not very complicated, but a million hours is 100 years or so, or many years anyway, and getting that kind of testing is not practical. So what they do is simulate aging of these components, by over temperature, over stress electrically, over amperage and over voltage testing.
By approximation and estimating and general experience, they can relate the million hours that you need for this component to the thousand hours you tested it, under much more stressful conditions. That's generally turned out to be a pretty good method. Now, we asked for 30 million dollars, finally got it, to develop the reliability of these components. Now, we didn't spend it in our labs but we contracted with other suppliers to do this, to burn in, to artificially age their components and design to these artificial requirements. Well, the upshot of it was that after the first few weeks or months of troubles, with the Minuteman, they began to run over 8000 hours in the hole, on standby, on duty really, and we met the requirement.
Then came the Minuteman-II. Of course they wanted better performance, more accuracy, more capability, all those things, and it was decided to go to integrated circuits--you know, they're the type of circuits that are done by chemically and photographically making the circuits on a chip. Well, the reliability on those things wasn't very well known, so we asked for a large amount of money again to get these things worked out the same way. Well, the negotiating people would not come up with that amount of money. They said, no, you overestimated that in the Minuteman I, you had no trouble reaching 8000 hours, consequently, we'll give you some fraction of that amount of money, which they did.
The upshot of it was that we put those Minutemen II out into the field, into the silos and they started to fail. We had at one time over 150 of them down for repair, and they were supposed to be the big deterrent for the country. It was almost a national security crisis, at least on paper. Well, we had a chemical problem. They called it the Purple Plague. The chemistry some way would cause a change in color and deterioration of the chip. So we fought that thing real hard and spent a lot more money than we should have and finally got it fixed up. We and our suppliers did. The Minuteman II was very successful. Then of course they went to Minuteman III and used the same techniques, and it's what we've got standing by in the holes right now. It all goes back 25 years.
COLLINS: Where does this Tiger Team review fit into that?
ATWOOD: During the periods of time when we were not making ourschedules, overrunning in costs, the same type of thing came up. So the Air Force sent these teams in and they made these reports and pointed out every deficiency we had in the whole place. We had plenty, and of course it was up to us to work off this sort of thing. We finally did. But not necessarily to the criticisms that the Tiger Team made, but straightened out the whole problem, got the engineering right, got the components satisfactory, and then of course we completed the job. There was a delay. There was a big overrun. But the parallel is there, even though the technology part of it is different.
COLLINS: Were there similar pressures to change personnel in the Autonetics program?
ATWOOD: Well, I finally let go Fred Eyestone, the president of the division, on the urging of the Air Force and Sam Phillips. I just had to replace him. He was a good man, a good engineer, but I guess some way he wasn't making it with the Air Force. He went on to work at Varian Associates. He was a top executive there, and he worked under Norm Parker. He also used to be head of the Autonetics Division, but Parker had gone to be president of Bendix and he became, after that, president of Varian Associates. It's up near Palo Alto.
So I had to do that. I guess I did. I don't know whether I did or not. But it was the same cycle, in a way, and of course, these people are all good people, highly talented people. It's just a matter of degree, you know, and I know in the military, of course, they relieve people without second thought as to who's at fault. It's just that things have to go, and it's a crisis and people are dying and the regiment is at risk and that's the way they have to do it, but it isn't quite the same in industry.
COLLINS: To switch gears a little bit, one thing I'm unclear on and I don't think you've put down on the record anyway is the sorts of things that you personally were involved in and that you had North American do in the immediate aftermath of the fire. You see things that are happening, you know, when it gets to April and May, but what were the sorts of things that were happening in February and March? Do you recall some of the more immediate actions that took place, or was there a period of waiting while the review took place?
ATWOOD: Well, of course we had the investigating team, the Thompson Committee, which had been appointed, and of course Johnson had agreed with Webb the fire should be investigated internally. I thought they were putting one or two of our people on the committee, including George Jeffs, but of course, at the very time of the fire he was chief engineer of the CSM, and so I think they relegated him to a kind of an information supporting type of person.
It turned out finally we had nobody on the team from the company. We had people working with them, giving theminformation, data, parts, samples, everything that they wanted, but I guess I thought that there'd be really a--I'm searching for words, I hate to use the word impartial--but I guess I thought there would be, like the military accident reports, a really impartial report on it. I was waiting for it. I don't know, maybe I was wrong, but felt that to start our own separate investigating team in a competing or interfering way certainly was not the thing to do. I've always relied on military accident reports when we had a failure or crash or something like that in the past, and they've always, to me, been a report that was concise and clear, allocated responsibility and recommended the solution. I guess that was my mentality, because I did not start any counter-investigation.
COLLINS: Did you have informal conversations with Harrison Storms and other people on the program?
ATWOOD: Yes. Oh, yes. As a matter of fact, I was given so many criticisms of Stormy over a period of time that I finally decided that he just couldn't make it with NASA at that time. So I did remove him, about the time of the investigation. But I hadn't made up my mind to do that until after we'd gone to Washington for the Congressional hearing. I'd gotten such strong signals from Webb and everybody else about Storms that I had Larry Greene, our Washington representative, in contact with Webb. I guess he convinced me that Webb was not going to allow Storms to continue. Greene came out and one Sunday we typed up that letter to Webb and Larry Greene carried it back. I said I would replace Storms, and Larry Greene took it back and took it to Webb that Monday morning I guess or Sunday night, I've forgotten which. I called Storms in and told him that I'd replaced him. I don't remember the date, but it was right after we'd been in the hearings of Olin Teague's committee. Olin Teague, incidentally, was, I thought, a very fair and supportive type of guy. He was not stampeded by any of this at all.
COLLINS: We talked previously about the role of North American's Washington office and their role in Congressional liaison. What kind of information were they getting back to you about the tenor of things in Congress, if you recall?
ATWOOD: Well, of course, Larry Greene was kind of hard-nosed himself. He didn't have much sympathy with Stormy either at the time. He said, "God damn it, I think you ought to judge people more on objective standards, and I don't think Storms is getting it done and I think you ought to get rid of him." He said, "If I'm not doing my job, I expect you to kick me out." I'll never forget that. I said, "Well, you think this is Webb's position?" "Well," he said, "I don't think the thing is going on any longer the way it is." In other words, I think Webb was ready to throw some kind of a contract stop or cancellation of some weapon like that at me, and I had every indication that he felt that way. When he got the letter he told Larry Greene, "It's a good letter." I didn't think it was a very good letter. But I could see that Stormy wasn't going to last with NASA any longer. I don't like to use the word "scapegoat." Of course I think Joe Shea was a kind of a scapegoat. But I do think his emotional personality was partly responsible.
COLLINS: You're referring to Shea or Storms?
ATWOOD: Yes, Shea. Shea couldn't, I mean, I think it broke him up. I think it was fairly evident. But Stormy never did break up. He did very well. It is reported, I believe reliably, that Joe Shea received a letter before the fire from a responsible source warning him of the danger of the high pressure oxygen inflation procedure and that he neglected the warning. This, of course, would be a severe burden.
COLLINS: I was trying to get at whether you were getting feedback from Congressional members about their feelings, about the event and North American's role.
ATWOOD: Well, it was a mixed bag. This guy Ryan from New York was very, very anxious to find something wrong. And I felt that Teague was--well, he was behind the space effort very strongly, and I think that feeling extended to some extent to the contractors. I think his sympathy for North American was fairly evident, although he was certainly, I thought, objective enough in his handling of the hearings. Fuqua from Florida was fairly supportive, Don Fuqua, and the Representative from either Pittsburgh or Philadelphia, Corman, I believe, I've forgotten which city, he died not too long afterwards, was quite supportive. It was mixed there too. Congressman Bell from California, he was trying to help but he didn't have much basis for helping. What's his name, Bell? Alonzo Bell, that's it, Alonzo Bell. I was going to show you one thing. Now, this is Senator Mondale--
COLLINS: I assume you're reading from the CONGRESSIONAL RECORD6 there.
ATWOOD: Yes.
COLLINS: Do you have a date for that?
ATWOOD: Yes. "Mondale: Would you comment on that? Is there a Phillips Report? Dr. Mueller: I answer this by trying to place it in perspective. Frank Borman and I were talking the other evening about the equipping of Gemini 7 spacecraft because at that time we had in operation the same set of standards, the same set of specifications, the same set of tests as we had used on the Mercury program. I believe that the use of velcro in Gemini 7 was considerably greater than in the Apollo 204. So that in terms of the application of our knowledge, we did fully use whatever knowledge we had gained about the flammability and flame propagation in Gemini in the Apollo program.
With respect to the question concerning the relative performance of contractors, in terms of carrying out the program, I remember my first experience before Congress about two or three days after I arrived on this task here, finding that there was a special investigation of a report on the Mercury program which highlighted some of the quality control problems that we had experienced in the Mercury program, and we had an investigation, in this case by the House, of that particular report. I think that our experience in fact with this spacecraft has led us to believe that the Apollo 204 for example was in fact a better quality article than the first Gemini spacecraft. That is the general opinion of the people that have worked on it down at the Cape when it came.
Now, I think on the other hand it's fair to say that on the first article of any of these more complex equipments, we all learn and we do have a great deal to learn. We have after all a block 2 coming along after block 1, and we have incorporated into those changes that we had learned from our experience in block 1. I do not regard the performance of the contractors in the Apollo program as being any better or any worse than the program performance of the contractors for the Gemini or Mercury program.
We do have and prepare regularly reports, at least of our contractors, concerning their progress. I am sure that General Phillips has on several occasions prepared reports at least of our contractors, I do recall, and not all of them have been completely complimentary. On the other hand, I know of no gross overruns in the Apollo program, and we are talking the budget and the expenditure that we had established for the program.
I think you recognize that in any R&D program a considerable amount of management attention has to be paid to areas of weakness, both internally in NASA and externally in our contract structure. We endeavor to create a system that highlights problems at an early enough time that appropriate action can be taken to solve them. Specifically I know of no occasion when we thought of a second source (such as) Douglas on the Apollo.
Senator Mondale: Is it your testimony that there was no such an unusual Phillips Report? Is that rumor unfounded? Dr. Mueller: I know of no unusual General Phillips Report. I do know that General Phillips has examined each of the contractors in the course of the program, and as a matter of fact, we have a yearly process of examining contract structure, to be sure, and we do from time to time have special reviews of contractor problems. I don't know of a specific report such as that. Senator Mondale: Was there a report in which General Phillips recommended looking for a second source? Dr. Mueller: I do not recall such a request."
I don't know what to make of that. I think that the Phillips Report, to Mueller, was a mystery for 20 years. I was accused by some people of perjury before Congress because I didn't know about it. Amazing. Webb said there wasn't any report.
COLLINS: Yes, well, that gets into a very sticky area.
ATWOOD: Well, it was.
COLLINS: Unless you want to add some things about that, I wanted to get just a sense of what really happened after the fire review, some of these things we discussed previously and get a sense of what the reaction of Congress was in your judgment.
ATWOOD: Well, you see, the Thompson Report, the only thing that stuck in my mind of course was that not only the report, which is a little more involved than that, but every news source in the country got the message that the fire was caused by deficient design, workmanship and quality control. That was the punch line. I guess you'd have to think at first that that's what happened, somebody had put something there that caused the fire. Of course, it turned out it was oxygen. But it was a very puzzling time to me, very. I couldn't understand it.
COLLINS: Did you ever consider trying to get a more balanced account of this analysis?
ATWOOD: No. I talked to our people, I told them just, forget about it, let's get to work. We'll take the blame, if any. There's a guy out at the company named Ron Swim. I saw him on some occasion a year or two ago. He said, "Well, I'll never forget the uproar but when you said that, everything kind of settled down and everybody went to work." That was his version.
COLLINS: Okay, why don't we stop there for today.
ATWOOD: All right.
1 Memorandum: S.C. Phillips to G.E. Mueller, December 18, 1965, Page 3 GWS Oral History Project working history files, National Air and Space Museum, Washington, DC.
2 Neil Armstrong,Wingless on Luna: presented at the Grand Hyatt Hotel, (New York: 1988).
3 J. Leland Atwood "Notes on Phillips report" January 12, 1990, GWS Oral History Project, Working history files, National Air and Space Museum, Washington, DC.
4 Frank Borman, Countdown: An Autobiography, (New York: W. Morrow, 1988).
5 Memorandum for: Mr. Webb from John Biggs, April 20, 1967, page 1, GWS Oral History Project, working history files, National Air and Space Museum, Washington, DC.
6 Congressional Record, 1967
Rev. 08/13/96