Joe Gavin, “The Lunar Module Design and the Apollo Program” - MIT Gardner Lecture

Search transcript...


MODERATOR: More or less annual event in the Department of Aeronautics and Astronomics, which was made possible by the [INAUDIBLE] Major Lester Gardner. It's an annual lecture on history of aerospace. Gardner was a graduate of the class of '98, which we now have to distinguish by saying 1898, as the Institute has now cycled through more than a century.

And as best I can figure out, he was basically a buddy of [? Hunziker's. ?] Is that right? Does anyone remember this far back? [? Hunziker ?] seemed to have this network of buddies who were, A, influential, and B, had money. That's the image that comes across. And Gardner was one of them.

He published magazine for a number of years called Aviation and Aeronautical Engineer. And he was a decorated aviation officer in World War I, and gave us the fund that allows the annual sponsorship of this lecture, which is a prestigious list of lectures. I'll just read a few off of the previous lectures. Sikorsky, Doolittle, Sperry, Otto Koppen, the two tailors, [? Abe Silverstein, ?] [? Doc Draper, ?] Hans [INAUDIBLE], Bill Sears, Gerhard Neumann, T. Wilson, Neil Armstrong, and Bob Seamans, and most recently, René Miller.

And we're indeed honored today to have one of our alumnus, Joe Gavin. Not only alumnus, a very active supporter in [INAUDIBLE] department. Joe received his SB in 1841. A member of the famous class of '41 that included Jimmy [? Marr ?] and a few other people who were quite active in the department. Who else was in the class of '41?


MODERATOR: It's a long time ago.

AUDIENCE: Long time ago.

MODERATOR: About 55 years ago. He was in the Naval Reserve and served in the Bureau of Aeronautics during World War II. He was a project officer for the Navy's first jet fighter. He joined Grumman after the war as a design engineer, and he served in many capacities there, including the one in which he will talk about [? us ?] today. That is, he was a vice president and director of the Lunar Module Program during the Apollo era. Served there until his retirement in 1985.

Outside of Grumman, he's had a lot of influence on the nation and the world. He's a member of the AIAA since 1946, which was before it was the AIAA. In 1982, he was the president of our professional society, and became the honorary fellow [INAUDIBLE] [? honorary ?] fellow in 1984.

He's received a NASA Distinguished Public Service Medal, and was a member of the National Academy. He's a member of the International Academy of Astronautics, and he was a director of the AAAS, and became a fellow of that in 1991.

And in his spare time, he's been active here at the Institute as a member of the Educational Council. Was the president of the Alumni Association. Has served on a number of visiting committees, including ours. Is a life member of the Corporation, and has served on its executive committee, which is the seat of real power in MIT, from 1984 to 1991.

So I would like to-- I would have the honor, in fact, to introduce Joe to give his address on the technology and programmatic development of the lunar module.

GAVIN: Thank you.


Well, it's a very considerable honor to join the group who have given this lecture. It also brings back a rather strong pang of nostalgia. Now, I have to explain that it's a little hard to feel nostalgic about 10 years of sheer desperation, because that's about what it was. And I'm just going to hit some of the high points here, and I'll skip some of the real difficult times.

But this program, the Apollo program, started as a result, I think, of the competition in the Cold War. There were many things that led up to it, and I would suggest that the beginnings of NASA and the beginnings of Apollo are worth reading up on, because they're quite complicated.

In my view, one or two major things. Mr. Eisenhower decided that NASA was going to be an open, unclassified agency. He managed to fend off the Air Force, who wanted the job. And then, of course, Mr. Kennedy made the commitment to do it. And at the time that he made that commitment, one of his advisors was saying that, with the reliability available [? and ?] components at that time, that it would take probably 40 landings-- 40 attempts to make a successful landing.

Well, it turned out otherwise, of course. And the credit really has to go to an awful lot of people. I'm here today to tell you, as I remember it, about the program. But there are an awful lot of people who were involved at Grumman and at subcontractors at the various NASA centers. And it's pretty hard to find any group, I think, that worked as diligently and in as focused a fashion for those 10 years.

Well, now let's get into the slides. This particular slide is from Apollo 15, as I recall. And the reason for showing it is, first of all, it is on the lunar surface. And secondly, the rover is in sight. And the last three lunar modules had the ability to carry the rover. They had more oxygen, more battery power, and could stay longer.

But let's go back and see how it started. The command and service module got started, but it was many months later that the decision was finally made to embrace lunar orbit rendezvous. This was something that John Houbolt promoted. And looking back at it, I don't think we would have gotten there without that concept.

So that was finally embraced, I think, in July of '62. And then there was a competition. And it was an interesting competition. The center in Houston asked 20 questions. And you had to answer those 20 questions in not over 100 pages, and with print not below a certain size. And the idea was that if you knew what you were doing, it would show in the answers to those questions.

Well, we did submit a proposal. I'm not sure that our then-senior management expected that we would ever win anything. But we had to have a strong design in order to answer the questions.

And this is what we had at the time. Now, this was a vehicle that was going to be considerably lighter than what eventually came out of it. And it had, interestingly enough, five landing legs on the premise that if you broke one, it wouldn't tip over. The other thing that's interesting about this is that there were two docking hatches. And we had reason to believe that that was an advantage. We had a helicopter-like cabin. That didn't last long, but the idea of good visibility was there.

And I'll show you. And this shows that there's an aft equipment bay. Well, we did have an aft equipment bay. It didn't quite look like that. And then we'll jump right away to what finally came out.

And this was taken from Apollo 9, the flight test in Earth orbit. And it's really upside down, but it's easier to understand this way. And here, you can see the probes on the landing gear. There's great concern that the dust on the surface of the moon would obscure things to the point where you needed some mechanical indication of landing. This also shows the exhaust [? valve ?] of the descent engine, and the rather massive insulation around it.

Up at the top is the rendezvous radar. There was a flasher here, which was a backup means of rendezvous. And in this picture, you really can't see the two windows, because what we have done in developing the cockpit arrangement was to turn the front of the vehicle inside out and pull the windows right back to the astronaut's eye.

Now, we hadn't been into this contract for more than three months when we began to understand what some of the priorities were. Now, this was a "cost plus incentive fee" contract. And the incentives depended on performance-- which turned out to be largely weight-- schedule, and cost.

Well, I think in about three months, it came to us that those things were not equal. Performance was clearly number one. We do our best to meet the schedule. And I hate to say this, Joe, but we really left the cost to come third, as he well knows.

But that was an important idea. And it took us about two years to firm up the final design, because at the time we were in the competition, we did not know, for example, the details of the MIT-designed guidance system. And then, of course, there were requirements like a backup guidance system and so on.

Let's go on. I'm not sure which mission this came from, but this was the ascent stage. And I picked this slide particularly to show the rendezvous radar, the front door-- the front docking hatch that we proposed became a door, not a docking hatch. And then this was the ascent engine exhaust here.

And this gives you a pretty good idea of where the propellant tanks for the ascent engine were. It also shows clearly the quad reaction control thrusters. There were four on each of the four corners of the vehicle.

It's also pretty clear that this is not a very aerodynamic vehicle. To a group of airplane designers, this was a real awakening, that we did not have to worry about conventional aerodynamics. And it became very utilitarian in appearance, and only its designers could love it.


GAVIN: Now, at the time that we started, not much was known about the lunar surface. In fact, there was a well-known astrophysicist from Cornell who said, look, there's going to be 10 meters of impalpable dust, and not only will you sink into it, but electrostatically, it's going to swarm right up over the vehicle. Well, I think he told me that in 1965, by which time we were reasonably committed to the design, and we didn't buy it. And fortunately, he was wrong.

But we did have a pretty conservative approach to landing, as you can see by reading some of these things. And vertical 10 feet per second is a pretty good jolt. Now, if you land on an aircraft carrier, you worry about maybe 20 feet per second. But for a land-based airplane, that's not too unreasonable.

But since we didn't know about the surface, we made some very broad assumptions. We assumed everything from landing on the equivalent of ice to landing in four small craters ideally located to trap all four landing gears. And we landed with various pitching motions, rolling motions, yawing motions, uphill, downhill. And 400 runs later, we thought we had conservatively surrounded this problem. And as it turned out much later, we probably had over-designed.

Now, one of the things that occurred to us-- not at the very beginning, but into it a few months-- was that this vehicle only has to make one landing. Now, that's a revelation too to somebody who's been designing airplanes. And so we decided that-- oh, well, this is the tip-over condition that we worried about.

So the shock-absorbing material is aluminum honeycomb in a billet. And it's good for one landing. It will take the shock and eat up the energy. It does not rebound, and you can't use it a second time, but of course, we're not using it a second time.

Now, we-- in conventional aircraft fashion-- ran some drop tests as soon as we had a vehicle to drop. And we hoisted up in the hangar and got all the conditions just right, and let it go. And it came down with a crash that was absolutely frightening, because all of the gauges in this machine were much lighter than anything we had experienced in aircraft. And I'm going to say a little bit about that later.

But it made a terrible noise. And it made us so worried at a later date about the flexibility that, much later, we insisted on a drop test where we had all of the wiring and electronics in the vehicle to see whether the flexibility of the vehicle would interfere with the electrical systems. Actually, we didn't find any electrical system problems, but we did find one small mechanical problem that we had to fix.

Now, I'll add a few more points about the landing gear. We provided the ability to absorb something like 30 inches of [? travel ?] in the landing gear. I don't think any of the actual landings used more than 3 inches. And by the time we got to this point [? that ?] we knew a little bit more about the lunar surface, the astronauts had spent a lot of time in the simulator figuring out how to fly it. And what they really wound up doing was to set a known rate of descent on the autopilot, and then fly the machine manually in the other planes. And that worked very well.

And I think, as a conclusion, we can say the landing gear was somewhat over-designed. But this also shows one other thing. It shows that there's ample clearance between the descent rocket engine exhaust bell and the surface.

And this became critical later when we added the lunar rover and the extended stay time provisions, because we then had a heavier vehicle. And in order to get there with the same amount of propellant, we had to extend this bell a few inches. And it turned out that that would work, and was acceptable, and that's what we did.

Now, the other thing that came out of these landings is that, unlike what happens on the ground, on Earth here, when a vertical landing machine comes down, you have dust that starts out sideways, and then pretty soon, you have billows coming up because of the interaction with the atmosphere. Well, on the moon, there is no atmosphere. And what seems to happen is that the dust takes off horizontally, and just goes out of the picture, so that the dust that we worried about earlier wasn't that much of a problem.

Now, one of our worst problems-- it was with us every day for 10 years-- was the tendency of the vehicle to grow. And I think the significant point is from about here, when the design was finally pretty much finalized. And then we had the weight growth. And then we had a huge effort-- huge in numbers of man hours-- to squeeze the weight out of it.

And then over here in early '67 was the fire at the Cape, and that led to some tests with an iron monster cabin and a variety of other things to ensure that fire would not occur inside the reduced pressure of a pure oxygen atmosphere in the cabin. And so the weight went back up again. And then, of course, this represents the addition of the rover and the longer stay time provisions.

To give you some idea of what the fireproofing amounted to, this is the back of a circuit breaker panel. And we have little baggies-- well, we found, of course, that the circuit breaker housings themselves would burn. So then, what to do?

So we created these little baggies that went around that circuit breaker body. And then all of the other electrical connections were covered with this mud-like material. I thought I'd never forget its name, but I have. And it was a very extensive redesign, and I would say that we lost about a year at that point.

To continue with the weight story, this is a picture of the ascent stage tanks that are spherical. They're very thin. Proportionally, they're thinner than egg shells. And they were designed on the basis of fracture mechanics to be good for no more than 100 pressurizations.

The propellants were fed to the engines by pressure. The pressure was helium. And so every tank had a pedigree. Every test that was made was recorded. And we made sure that before it went off on a mission, that there were an ample number of cycles left.

And this is a view of the descent stage tank. And this is similarly thin stuff. In fact, we did a lot of our testing with referee fluids-- in other words, fluids that had the characteristics of the nitrogen tetraoxide and the UDNH, but which were not nasty to handle.

And we did a lot of the testing with heavyweight tanks. When we received our first lightweight descent stage tank, of course, we filled it. Checked the volume. And consternation set in because it held 20 seconds more propellant than we had counted on.

And so then the question was, did somebody make a mistake? Well, no, they hadn't. What happened was that under the weight of the propellant, the tank stretched.

And we were very happy to have that extra 20 seconds' worth, because if you remember Armstrong's first landing, he hesitated a bit to clear some boulders that he was coming in on. And I think he was down to about 20 seconds at the time he touched down. The rest [? of ?] mission control were blue at that point.

And while we're talking about tanks, I'll talk a little bit about one of our other very major problems, which crept up on us so slowly that we didn't realize how bad it was going to be. But we had leaks. Now, nitrogen tetraoxide is a nasty material. If it finds a small leak, the small leak soon gets to be a bigger leak. And it's toxic, so you really don't want a leak.

And we had designed everything so light and slender that we had leaks around these ports at the bottom of the-- particularly the descent stage tanks. And we worked hard at this. In fact, we wound up changing all the seals. And in fact, one of the vehicles that was already at the Cape had to send back this whole lower structure so that we could redo the seals.

It was a terribly frustrating affair. And we were looking for leaks that were on the order of 1 cubic centimeter per year. Pretty small leak. With jet fuel, you would never worry about that.

Well, then this is a picture in our final assembly area, where we have an accent stage that's being supported independently of the descent stage. And it shows you some of the precautions we took with some of the thinner gauge materials. And some of these structural gauges-- say, in this shear web-- were really quite thin. We got down to ten thousandths. We would chem mill down to ten thousandths and keep the extra thickness around the edges where the connectors were.

And we tried going to 008. And then we counted the number of grains across a 008 section and chickened out. We went to the ten thousandths.

I might add that we chem milled almost everything you can think of. Propellant tubing was chem milled, keeping the original thickness where the connectors were. And I can't offhand say how many man hours were expended in trying to save weight, but it was a large number, and it was an endless process.

Now, this is a picture of the ascent stage in sort of a stripped condition. Can see the ascent propellant tank. You can see its support out here. This is a good picture of the window, which is just inches from the astronaut's eye.

And you can also see that we have the separate tanks for the reaction control system. And that was one system that was completely redundant. We had a set of tanks on either side of the machine. And these tanks were positive expulsion. In other words, there was a bladder in them that allowed the pressure to surely drive the propellant out. And of course, we used the reaction control system not only for maneuvering, but to settle the propellant in the main tanks before a main engine firing.

Now, this was the next step in the assembly of the asset stage. And for those of you who are familiar with airplanes, this is an inside-out machine. In other words, in an airplane, you get the sleek exterior, and then you package everything inside it as densely as you can, as some people say, so that even if you dropped a broken egg into it, it wouldn't drain through.

But here, we had a machine that didn't have to be aerodynamically slick, so we built it inside-out. Much of the wiring-- much of the equipment were outside the pressure vessel. And you can see some of the wiring here. And then, on the outside, we're going to have a thermal shield which also doubled as a micro meteoroid bumper. And that had to be supported somehow. And that led to this birdcage-like affair, which supported this skin.

Now, this-- you might ask, well, how come some parts are shiny and some parts are black? Well, let me just say that a very extensive thermal analysis was done, because we were trying to minimize the power necessary in this vehicle to cut down the amount of batteries that had to be carried, and also to cut down the amount of heat that had to be put overboard. And this is the result of some tests and both-- well, analysis as well as tests in a thermal vacuum chamber.

This is a better view of the reaction control thruster. You see, there are four-- a similar [? quad ?] on each corner of the vehicle. One of the things that was added very late in the game-- and which is not on my model here-- is this shield which protects the descent stage from the firing of the reaction control thruster that's pointing down. This was the result of some tests fairly late in the program that were done on the reaction control thrusters at the manned spaceflight center.

This is another view of the final assembly area. This is an ascent stage that is substantially complete at the front end, but hasn't been all put together on the back end. And the reason for showing this is to show you that the equipment-- batteries, radios-- were mounted on coal rails.

The coal rails had glycol circulating in them. The glycol heated water. The water went overboard through a porous plate so that, in effect, ice was formed, and it sublimed into space. And that was the only way we could get heat out of the vehicle if we had too much heat.

This also shows some of the paperwork. The weight of the vehicle-- I'm not sure whether it weighed more than the paper or not, but we had extensive pedigrees on just about everything you can think of. And that's a very important thing, because way back in the beginning, we worried about reliability.

At that time, aircraft experience was not particularly attractive. It didn't give you any great assurance. And one of our people came up with the idea that there should be no such thing as a random failure. If you have something that doesn't work, either a ground test or any test, you ought to be able to find out what it is and fix it.

And in the course of 10 years, we had something over 14,000 such anomaly reports, and I believe only 22 escaped satisfactory analysis. In the case of those 22, we changed something anyhow. And I think it was this attention to the nasty little details that made a difference.

I've mentioned tests. I should point out that we couldn't flight test this vehicle. It's made only to fly in space. We couldn't hot fire it because our experience with the propellants was that you really couldn't clean the system adequately if you used it. In other words, the propellers would tend to gum up the solenoid valves and so forth.

So each vehicle that was launched from Cape Kennedy was an unused, untested vehicle in every sense except electrically. We were able to do a rather thorough job in electrical checkout. But it did lead to the process of setting up a very complex-- and I should tell you-- steadily evolving ground test system so that we could assure ourselves that the equipment we were getting was good, that it would work after it was in the vehicle, and so on and so forth.

This is a photo from the final assembly area again, showing a complete vehicle except for the landing gear here. And same thing here. And the point here is simply that we were working toward a three-month on-center launch schedule. So we had to have these vehicles coming along fairly regularly.

This photograph shows a descent stage after the landing gear is on. The principal reason for showing this is that we had a rather elaborate blanket of insulation there to protect the descent stage when the ascent stage took off. We tried to run some tests in the vacuum chambers that NASA had at White Sands. And we could run short bursts at different levels of separation, but we had no way of running a truly dynamic test.

And in fact, we never really knew what happened until the very last mission, when the TV camera on the rover was aimed at the takeoff. And while this is a still, a lot of stuff went flying out sideways from that insulation blanket. But the vehicle just pops up. Ascent stage pops up.

Now, I should point out that we have a fly-by-wire vertical takeoff and vertical landing machine. In the landing condition, it's like a transport because it's heavy. And when it gets back to dock with the command module, it's like a [? fighter, ?] because it's gone through a 10-to-1 reduction in weight.

I've mentioned people earlier in what I've said, but it's impossible to do credit to all the people who were involved. Or for that matter, all the families that were involved. We worked a lot of people very, very hard.

And this is the check-out group at the Cape. And I'll read this for you. This is for the last mission. "This may be our last, but it will be our best." And by that, they meant fewer anomalies to clean up. And of course, the lunar module did work.

And I think that I'd like to use the view graph now for a minute. We could turn off the slides, and talk a little bit about something that I fell into by accident-- namely, management. This is a program organization chart, something that our customer really wanted. They wanted to know who was doing what. And we put this together.

Now, the trouble with this is that this is a static chart. It suggests that all of the people at one level are equal. And I've sometimes thought that an organization chart should look like a slowly-swirling galaxy of stars-- some that are brightening and some that are dimming-- because the roles tend to change as time goes by.

The other thing is that I would point out is that these people never admitted that they were on the third level. Grumman had, I think, a peculiar culture-- certainly in the time that I was there-- in that everybody looked at the company as "my company," and they were not adverse to calling anybody up to tell them what should be done differently, or what was going on. And it led to a certain amount of confusion sometimes, but the word got around very quickly. And I will say that, for many years, the company never had an organization chart. Roy Grumman, who founded the company, was [? against ?] it, so we didn't have it for many years.

The other thing that I would point out here is how the manpower was expended. And there are a couple of interesting points here. We were always under criticism for running too much overtime. But it's much more effective-- in my view-- to run overtime than to try to add people, in certain cases.

The other thing we ran into was that these slopes right in here, I think, were maximum achievable. You can bring people on board a major program just so fast. Otherwise, you wind up with people not being productive. In fact, a good argument, what happens in major programs frequently is that the end date-- or in this case, the lunar landing date-- was fixed. And there's a tendency for the start date to creep. And of course, the effort is to build up as fast as you can in order to get the job done.

This slope right about here is 1,750 people per year. Now, that's an instantaneous number. But at the real peak, we're adding about 3,000 people a year. And that comes about because there's an acceleration involved. And by the time we got here into '65, we thought we knew what we were doing, and could bring people onboard. And this is particularly true in bringing subcontractors onboard. About half of the effort went to major subcontractors, and some smaller subcontractors.

This gap right here represents the effort to fireproof everything that came after the fire at the Cape. And then you see we resumed. What happens here, of course, is that there are some things that you can stop, and there are other things that just have to keep going. So you can't really just come way down here somewhere. And then, of course, once you have reached an operational stage, things go downhill pretty rapidly.

Well, I'd like to think a couple of summary remarks. We had the advantage of a long-term commitment. There are not many programs that have had that kind of advantage. And I give Mr. Kennedy great credit for having the gumption to stand up and say, "before this decade is out," and so forth.

I think it was important that we recognized our priorities. And in a highly-technical undertaking, I think there's no question in my mind that [? what ?] you have to do the job right. If you take shortcuts, you pay more-- you pay a larger penalty later.

I mentioned the idea of no such thing as a random failure. It worked for us. Some of our people likened it to turning every stone on the beach over to see what was under it. It felt that way at times.

We invested a tremendous amount of effort on the ground simply because we couldn't flight test. And some of those ideas went over into our aircraft experience later on. We found we could save time and effort in developing aircraft if we did more work on the ground and didn't try to fly at the earliest possible date.

It also taught us what I call the basic program dilemma. And that is that if you're going to do something truly innovative, you can't possibly know what the schedule is and what the cost is going to be. And if, conversely, you know precisely what the schedule is, and what the cost is, you ought to stop and think whether it's worth doing, because it may be obsolete before you start. Now, this is something that I think engineers understand quite well. I've had a terrible time with some congressmen on that subject.


GAVIN: The other thing I would mention is that in being in Europe and in Japan-- and more recently, in China-- the impact of Apollo has been greater overseas than it has been in this country. The American public is pretty blasé about these things. It's assumed that the rocket scientists can do anything. Well, they almost can. But the American public passes on to something else.

But we created a reputation overseas that people are still trying to catch up with. And I think this is something that-- particularly those of you who are students should understand. That the Apollo program was a tremendous sales effort for American technology.

And finally, last comment. A lot of people talk about systems engineering. It means a lot of things to a lot of people. In some cases, it's just a matter of, do you get all the interface signals correct, and so on.

But I think that anybody who's associated with designing flying machines-- whether they're flying in the atmosphere, or in orbit, or outer space-- really is faced with doing a superior job of systems engineering. And the reason is that if you don't do a superior job, the results are catastrophic, and wind up on page one. And you really can't afford to have that happen.

I grew up as an engineer in a company that was founded by a pilot, and [? so ?] some years were some very senior pilots in the management. In the airplane days, it was simply "bring the pilot back one way or another." And we did that. And I think that if you have that in mind when you're involved with designing flying machines, that you will do a superior job of systems engineering.

Now, sometimes, you wonder, well, can you apply that to other effort? Well, in a selective way, you can. Those little US Postal Service trucks you see running around are the result of applying some of the lessons that came out of this program. And that's a whole 'nother story, so I'll stop there, and I'll be happy to take any questions.



AUDIENCE: When you get to the end of a project like this, I think you often learn a lot along the way, and sometimes think you could have done something differently. Was there anything you learned at the end that you wish you knew at the beginning, and you would've done anything differently because of that?

GAVIN: Well, I can certainly tell you that if I were doing it today, there are some things I'd do differently. At the time that we designed the lunar module, we were very restricted on where we could use solid-state devices. In fact, all of the communication gear was made up of discrete elements. And I would certainly avoid that. That was a nightmare, getting enough units that would pass test.

I would also have a digital data bus, just to reduce the miles and miles of wire that we had. We deliberately went to a skinnier gauge of wire than is commonly used in aircraft in order to save weight. And we paid for it, because we had ever so many faults that came about from just plain handling the bundles of wire. So anything you could do to get rid of the old-fashioned way of wiring would be an advantage.

I think that it's also quite likely that a lot of the equipment that we had-- well, supposing we take the MIT-designed guidance and navigation equipment. I think that can all be done in a miniaturized version today that would weigh a lot less. I'm not sure we'd have GPS in lunar orbit, but I'm sure that what we had, while it was great for its time, is something we could do much better today. And that's probably true of almost all equipment. Larry?

AUDIENCE: Were you made aware of any of the parallel developments in the [INAUDIBLE] Russia, the [INAUDIBLE] LM?

GAVIN: Not a word. No, I knew nothing about it. And I think I can safely speak for the rest of the crew that we knew nothing about it.

AUDIENCE: If I could have-- [? and ?] another question. The success of the LM, among other things, was the watershed in flight simulation. Does it work the first time it was used in its environment based entirely upon simulated training? Can you tell us a little bit about how that worked, and how you had confidence that this could [INAUDIBLE] first time?

GAVIN: Well, you have to recognize that 40 years ago, I was a lot bolder than I am today. And so were we all. But there were some lessons learned in simulation. In fact, when we built the X-29A, the little airplane that had the forward short wings and the huge negative stability, we treated that in simulation much as we did the lunar module. And the pilot came back from the first flight and said, gee, it flies just like the simulator.

So there were some lessons learned during the lunar module. I crashed one of the simulators several times for a very simple reason. If you try to translate along the surface of the moon, there's no air resistance. So in order to stop, you had to fire in the opposite direction. You don't automatically slow down. Now, that's a thing that is so counter-intuitive that it takes training to do well. Yes, Bob.

AUDIENCE: Joe, you were dealing with a lot of different people within Grumman and within NASA. First of all, how did you spend your time? Who did you work with the most, and what kind of problems did you get into from a management [? stance? ?]

GAVIN: Well, I'll answer the last one first. Every problem you could possibly imagine. In the very beginning, there was a certain resentment among the dyed-in-the-wool aircraft people that this strange group had brought home a contract that was going to take a lot of the company's attention, and people, and reserves. So there was a matter initially of getting the right people. And fortunately, I had some excellent support from people you know, like Grant Hedrick, and from then-senior management. And so we did get the right people.

And of course, we had to manage quite a number of somewhat difficult subcontractors. Joe Shea remembers some of those days. He and I had a great debate one time walking along the street down at the Cape as to how much visibility the NASA people were going to have into our subcontractors. And prior policy had been-- in dealing with other government agencies, is that we would manage the subcontractors, and that was that.

But he won his point, finally, and we came down to the idea that NASA would have visibility as to our relationships with the subcontractors, but they would communicate to the subcontractors only through us, since we were the prime contractor, and we felt very strongly that we had the responsibility. I'm not sure I've answered that question adequately, but that gives you some idea. Yes, Todd.

AUDIENCE: Joe, you talked about 10 years of nostalgia, and the nostalgia over 10 years of frenzy and stress. Can you share with us the memory or two of a [? horror ?] story?

GAVIN: Well-- oh, Clyde, before I get to that, I want to go back to Bob. There's one thing I forgot to say, and that is that I was very fortunate in having a cadre of people with whom I had worked for 10 or 15 years. So we were pretty well calibrated one to the other. And not only in the engineering and the things, but I had a number of good acquaintances amongst the technicians and the foremen.

And every day that I was in Bethpage and not traveling to Houston or some other [? garden ?] spot, I'd make it my business to walk through the manufacturing area or the test area and talk to the people who were busy there. And it's amazing what you can learn just by taking an hour's walk and seeing firsthand what people are doing, and what problems they're having.

And some of the problems are simple. I recall one where the thermostat on the heating system was stuck, and it was too hot. And apparently, the foreman had complained. Nothing had happened. Well, I made a phone call. It got fixed.

So the problems range all over the place. I'll tell you that we did occasionally have to send people home, because we discovered that some of them were punching out at the end of their 10- or 12-hour shift, and then coming right back in and working another few hours. And we had some people who actually made themselves ill in the process.

But everybody knew what the name of the game was. And NASA did a great thing in having the astronauts visit the plant, oh, I think about every second to third week. So it became a question of, we are building this vehicle for someone we know, not for some agency that is a bunch of bureaucrats somewhere.

Now, back to Todd. Well, the low point in some days was going to Houston to make a report on how we were doing and having to get up there and say, well, we've lost another couple of weeks. And it's not easy, but that's the way things were. And we did what we had to.

Speaking of Houston, I remember one such meeting where one of the fellows who was involved in a navigation guidance group was suffering from an earache. And I thought I had arranged it so that he wouldn't go on that trip, but he went anyhow. And he was under medication. And he got up to the podium to talk about what they were doing, and he was so relaxed from the medication that he created a sensational impression.


GAVIN: But he knew what he was talking about. He got away with that. And afterwards, somebody asked me, what is he on?


GAVIN: But that was the kind of spirit that we had in the group, and gives you a couple of ideas of what went on. Yes, [? Errol? ?]

AUDIENCE: Joe, if you had to reach back to your undergraduate years and think about what you learned in MIT, what carried you through your career?

GAVIN: Well, that's a hard question to answer, [? Errol, ?] because you see, I never became an expert at anything. But I do remember very clearly Professor Joe Newell, who taught structures, and who was a very conservative man. And when he went to Washington, he took the train. And that made a great impression.


GAVIN: Made a great impression. Yes, Stan.

AUDIENCE: When you commented on some of the things that might have transitioned into your airplane activities, can you comment on maybe the success and failures of translating the systems engineering approaches that you used into the aircraft business?

GAVIN: Well, the question-- in case you didn't hear it back there-- was, were you able to translate some of the systems engineering aspects of handling the LM back into the aircraft business? Well, you have to realize that we took what we had learned in the aircraft business and used it in the LM. And then, of course, we had on-the-job training because we ran into so many difficulties.

And it is true that we did sharpen up the aircraft business a bit after we space cadets took it over. We instituted formal pre-flight reviews. Now, there always had been some sort of review prior to flying a new airplane, but nothing like what NASA had taught us to do. So we introduced that.

That stands out in my-- and the simulation. That and the simulation are the two big things. You see, at the time, we were involved in some aircraft that were very highly-- well, had a lot of electronics in them, and where things like crosstalk and mutual interference had really been worked on very, very hard.

So-- well, we did also introduce the matter of certifying the people who could open a connection. With all those prongs in a plug, it takes real skill to avoid damaging it. And that helped.

But I think that some of the aircraft we were building at the time were just as complicated-- in fact, some of them were more complicated than the lunar module. As I recall, let's see, the EA-6B had something like 80 antennas on it. So that was not all that easy either.