This is lecture 2 of the Systems Engineering curriculum from MIT and edX, Engineering the Space Shuttle.
Links to the lectures:
- Origins of the Space Shuttle or The Making of a new Program
- Development of the Space Shuttle
- Bureaucratic Space War
- Political History of the Space Shuttle
- Space Shuttle Orbiter Subsystems
This lecture has a very strong focus on correct requirements. It’s not surprising as Aaron Cohen was a program manager and having correct requirements was his responsibility. In the lecture he points that it is important to challenge requirements. That’s where technical leadership, personal courage, and wisdom are necessary. I remember, my mentor told me once: You can give the requirements and even a design specification of a fighter jet to a bright young man who just graduated from an engineering school with honors and he should be able to build an airplane but it will not do anything interesting. Good engineers are able to create a special system while driving the requirements offered to them and still accomplish original goals set for the system. We told that such people were ‘touched by God.’ There are very few people who have this quality.
Introduction by Jeff Hoffman
This is Aaron Cohen’s first lecture and a few minutes into the lecture Aaron reviews his personal history, both before, during, and after his NASA career, so I won’t give any more information about him here except to say that he was a wonderful man and a superb systems engineer. Aaron could be tough as a program manager and many engineers were certainly frustrated by his continuously having to reject their demands for changes in the systems that would increase cost or delay the schedule, but personally, everyone liked and respected him.
I was present at his NASA retirement party, and the warmth and affection felt for him by all the employees at NASA’s Johnson Space Center was really apparent. Aaron was a personal friend, and without him, this course would not have been created.
Aaron died in 2010 and we all miss him, but we’re lucky to have him on video and I hope you are going to learn a lot from him, both about the Space Shuttle in particular, and about general principles of systems engineering of which he was a master.
Let me now just make a few comments by way of explaining some of the things that Aaron refers to. At the beginning of his lecture Aaron talks about students working on the design of new space craft. Redesigning the Space Shuttle using 21st century technology was part of the assignment to students taking the course for credit back in 2005. In general I have tried to edit out references to this assignment but I leave this initial mention in just as a reminder to our online students that hearing lectures is only part of the education process. Ultimately, you have to use the lecture material to do projects and solve actual problems in order to cement your knowledge. You won’t be doing a design project as part of this online course, but I hope that at least some of you engineering students out there will be able to make use of what you learn here in your engineering careers.
Aaron mentions early work with the MIT instrumentation lab, and then he refers to the Draper Lab. Charles Stark Draper or doc Draper, as he was always referred to, became famous for his work in the development of inertial navigation systems. He was head of MIT’s Aeronautics and Astronautics department, and he created the instrumentation lab which at that time was part of our department. The first contract that NASA made as part of the Apollo program was with the instrumentation lab to develop the computer in inertial guidance systems for Apollo. In the early 1970s, the instrumentation lab became an independent non-profit entity renamed as the Draper Lab.
Draper formally separated from MIT, although we still have close ties and many faculty and students in the Aero/Astro Department at MIT work closely with Draper scientists and engineers.
At the very end of Aaron’s lecture we had to edit out some irrelevant comments, so the lecture presented here ends quite abruptly with a few comments on it by Aaron on various aspects of systems engineering as they applied to the design of the Space Shuttle.
So, here’s Aaron Cohen’s first lecture about the early development of the Space Shuttle.
Program Management Decisions
You are going to have a very unique opportunity, you’re going to have people speak, lecture you on the various Shuttle sub-systems in quite a bit of detail, you’ll have some people that will be very, very positive about the Shuttle, that think it’s a great design, a great operation, you’ll have others that will not think it’s so good, and you’ll have some that will give you just a detail technical approach of what happened.
What I would like to suggest to you for your good and for the good of the future – future students, future designers, – that you ought to come up with your own decision: Was it the correct design? Should it have been done differently? And if so, why and how would you do it? I think that’ll be good for you when you get to work on future projects. It will also be good for NASA because this is something we could turn over to NASA. So I think it’s a very valuable thing to do. I would suggest you do that as you go through the course, and I will be happy, very happy to talk to you about any of your ideas that you have through the internet, through e-mails, or in personal discussions, and you might want to do this later on as a semester develops.
A little bit what I’m going to say today really starts off on where Dale Myers, the previous speaker left off, and it gets in a little bit more detail. As this course develops, you will get more and more and more detail. But let me start off again, just very much like they Dale did, and talk about the Shuttle history.
In 1952, a fully reusable launch vehicle concept was discussed. People are interested in that.
In 1962, fully reusable vehicles were seriously considered. The Air Force studied project Dynasoar, which was canceled in 1969.
In 1969 NASA adopted the idea of a fully reusable spaceship.
I became the orbiter project manager for NASA in august of 1972. At that time I also was a manager of the systems engineering organization for the first two years. I graduated from Texas A&M University in 1952 and went to the army, went to Korea. Then when I came back, I went to work for RCA where I worked on microwave tubes, the microwave oven. Everybody has a microwave oven now. In 1954/1955, when they came out, they were about 3,000 dollars a piece. So when I told my wife that I was working on something called a microwave oven and we were going to be able to cook a roast in a couple of minutes and a potato a couple of minutes and she looked at me and she said: That will never sell. Then I worked for General Dynamics on the Atlas and Centaur and then in 1962 I went to the Johnson Space Center and worked very closely with the MIT Instrumentation Lab, now the Draper Lab, on the guidance and navigation control system of the Apollo program. I became a head of systems engineering in Apollo and then the manager of the command and service module in Apollo. Then in august 1972 I became the manager of the Space Shuttle orbiter. Then I became director of Research and Engineering at Johnson Space Center and then I became director of the Johnson Space Center, and then for a while I was acting deputy administrator in Washington. And then I retired and went to Texas A&M to teach, and then Jeff asked me to come do this, and I’m very happy to be here.
There was specifically a systems engineering group at the center which was separate from the project offices. It’s something which is probably worth talking about at some point.
In Apollo we sat around the table for many days and months trying to figure out how you define systems engineering. We would not even know what systems engineering was. In fact, today I’m not sure you’ll get a clear definition of what it is. As we go through the lecture, I’m going to give you some examples of what I think systems engineering is and how it is used.
Something is very key in any design and something you really need to pursue whenever you do a design project is understand the requirements, because if you don’t understand the requirements you may get a very good product that’s useless. You got to understand what your customer wants, the top level requirements.
Top level Shuttle requirements
- Fully reusable
- 14 day turn around to next flight
- Deploy and retrieve payloads
- Design, development, & test phase estimated to be 5.1B in 1971 dollars
- Original cost per flight for 65,000 pounds was 10.5M per flight in 1971$ for a flight rate of 60 per year
One of the things that was handed down to us, the Shuttle was supposed to be fully reusable. You’re supposed to be able to turn the Shuttle around in 14 days, you have to deploy and you have to retrieve payloads, design, development, and test is estimated to be 5.1 billion in 1971 dollars.
As we started working on the Shuttle, the headquarters took away two years of inflation. Were they to give us those two years of inflation we would have met the 5.1 billion in 1971 dollars. Dale Myers fought for that but lost. The reason we missed this target was due to the calculation of the original cost per flight for 65,000 pounds which was 10.5 million dollars per flight in 1971 dollars, but for a flight rate of 60 flights per year.
When I told my wife I was doing that, and she’s been around a space program for a long time, she said: You never agreed to that, did you?
60 flights per year is pretty hard to do but that’s what we came up with at the time.
Dale mentioned the Air Force requirements but here are the studies.
Phase “A” studies were conducted to determine basic requirements and their effect on design in 1969
- Size and weight of payload
- Cross range of the orbiter
- Heat-resistant structure or reusable insulating material
- Hypergolic reaction control system or liquid oxygen/hydrogen
- Fly-by-wire flight control system
- Wind tunnel tests to determine wing size and configuration
- Air breathing engines were considered for fly back; later were determined to be too heavy
- Entry techniques
- Landing speed
- Approach pattern
Phase “B” studies were performed in mid-1970’s to determine a preliminary design
- Fully recoverable orbiter
- Disposable fuel tank
- Parachute-recoverable solid rocket boosters
- High performance hydrogen-oxygen engines placed in the orbiter to be recovered
The Phase “A” studies were conducted in terms of basic requirements and their effects on design in 1969. The principal issues were the size and weight of the payload, the cross range of the orbiter, and what kind of heat material were you going to use. You got to recognize that we were going to use the heat-resistant structure, a reusable insulating material. You got to recognize that our background was a Mercury Gemini Apollo, and they all use an ablative material. An ablative material cannot be reused because basically the surface changes. You had to have some kind of insulating material so the surface did not change.
The principal issues in the Shuttle studies were, should the reaction control system be liquid oxygen/liquid hydrogen or should you use a hypergolic system. The Reaction Control System is a propulsion system that controls the vehicle about its center of gravity, it’s for attitude control basically. It’s a storable system and presents a major decision. The hypergolic fuel is a fuel like hydrazine, it is an oxidizer and a propellant in the same fuel.
The other issue was a fly-by-wire flight control system. Now everything has a fly-by-wire, all the military jets have it. For us the decision was, was it going to be a digital computer control system or were we going to have cables to fly the machine.
Wind tunnel test determined wing size and configuration. That is a very difficult thing to do, but this is starting to get into what you might say is systems engineering.
Airbreathing engines were considered for fly back and later determined to be too heavy.
Still we had to complete studies of the entry techniques, landing speed, and the approach pattern.
The Phase “B” studies were performed in mid-1970s to determine the preliminary design. The results showed a fully recoverable orbiter, disposable fuel tank, parachute-recoverable solid rocket boosters, high performance hydrogen-oxygen engines placed in the orbiter to be recovered.
This was all systems engineering that led up to the design. So you have systems engineering in various phases of the program and usually systems engineering composes of an interdisciplinary team that has been given some assumptions, some constraints, they have some top level requirements, they do an iterative process with some tools such as computer tools for calculation of loads and flight mechanics and they come up with the iteration of what the design is going to be. So that’s basically how it was done.
Results of studies
- Fully reusable with fly-back booster was greater than 5.1B.
- Many configurations were studied (examples)
- Turnaround time of 14 days required landing a winged vehicle on a runway
- Payload deployment and retrieval requirement determined location of orbiter on launch configuration
Those were the ground rules. Once we started putting the total system together we showed that the fully reusable with a fly-back booster was greater than 5.1 billion dollars, so that was thrown out.
Now, there is a question, and that’s why I asked Dale, should we have said, “Hey, we need money to really have a fly-back booster.” But they gave us a constraint of 5.1 billion dollars in 1971 dollars and it didn’t make it.
Many configurations were studied and the turnaround time of 14 days dictated landing with a winged vehicle on a runway. You weren’t going to be able to land in the water or with a parachute. You would have to land it on the runway so you could quickly turn it around in 14 days.
The payload deployment and retrieval requirement determined the location of the orbiter on the launch configuration, because if you look at that large payload bay, it would be very difficult to put that on top of the vehicle.
Once that was done, we could see what the design was starting to look like. The first picture is the Agency Commitment in March of 1972, in May of 1972 we had the North American (NAR) proposal, and then I became orbiter project manager in August of 1972. PRR is the Preliminary Requirements Review. We made some changes and the production commitment was made in May of 1973.
This chart shows in 1971 dollars cost per flight for the Thor, the Atlas, the Titan-3iC, the Saturn 1B, and the Shuttle. That’s payload to orbit. You can see that the thing that was really missed in the Shuttle was the 10.5 M dollars cost per flight.
Major Shuttle configuration decisions
- Hydrogen/oxygen main engines
- This sized the liquid oxygen/hydrogen tank, which is not reusable
- Solid rocket boosters provided the additional propulsion required to get the orbiter into earth orbit
- Solid rocket boosters designed to be recovered and re-used
Here are some of the major decisions. We were going to go with a hydrogen/oxygen main engine. That’s one of the system problems, the decision what kind of engine you’re going to use sizes the external tank. Because of the equations of motion you can figure out how much propellant you need, you get the density of the propellant, and now you know what size of tank you need. Solid rocket boosters provided the additional propulsion required to get the orbiter in orbit, and the solid rocket boosters were designed to be recoverable and reused. Those were some of the system studies that led to the configuration.
- Orbiter entry cross range required delta wings
- Deletion of air breathing engines for moving orbiter required the Boeing 747 to carry the orbiter
- FO/FS guidance, navigation, and control system
- Fly-by-wire with a digital auto pilot
The orbiter entry cross-range maneuvering capability required delta wings, to go 1,100 nautical miles cross-range, you needed delta wings.
Deletion of air breathing engines for moving orbiter required the Boeing 747 to carry the orbiter. When we landed it at Edwards Air Force Base, we put the orbiter on top of the 747 and we flew it back to Kennedy. I became a project manager in August 1972 and I was having all sorts of problems. The first thing they did to me was they cut my budget in half. But I had worked on the Apollo program, I had a lot of friends in the organization. Three of my friends came into the office one afternoon maybe two or three months after we started, and said: We got a great idea! I said: What’s that?
– We could put the orbiter on top of a 747 or DC-10, and fly it back to Kennedy from Edwards Air Force Base. We may have to make one or two stops.
I looked at it for a moment, then I said: That is absolutely the dumbest idea I’ve heard in my life! I basically threw the people out of my office. They were friends. These people would not take no for an answer, that happened to be another very good quality, they were all world class model airplane builders and these guys had won competitions all over the world, three of them. They came back about 10 days later and they said: Come out, we want to show you something. They had built a radio-controlled model of the 747 and an orbiter and actually flew it for me and separated the orbiter from 747. That’s how it got started. And so we eliminated air breathing engines.
If you recall, when the orbiter lands, nose landing gear is very short. What we had to do, we had to actually replace the landing gear with a different landing gear that caused the orbiter to have a different attitude pointing nose up, and not nose down with its own gear. We put air breathing engines on, and we took off, and we had to have five in-flight refuelings to get from California to Florida.
And of course, we went with fly-by-wire with the digital auto pilot. This was a very fundamental change. The astronauts at that time did not like this very much. Fly-by-wire means that you have a computer that controls the surfaces where in past airplanes your stick actually had cables that controlled the surfaces. So, you get a lot more performance.
It’s a little bit confusing in the sense that a wire, you might think of it as a hard wire which is like the old type of an airplanes where there was a cable, so that when you pulled on the stick there was actually a cable, which went back to the ailerons and the rudder and everything. Everything now has a computer in the middle and what you’re really doing is flying the computer and the computer then issues the commands to the hydraulic system. The Shuttle was the first vehicle that really had that system. There were no commercial planes flying with that system. The real concern was with safety and reliability. Suppose you have a computer problem, what are you going to do?
Cross-range. The orbiter is a glider, not much of a glider but a glider nevertheless, and it will have a down-range and the cross-range would be out-of-plane, so you can actually maneuver out of plane. This was designed for the military requirements. They wanted to be able for reconnaissance satellites to launch into polar orbit. Because you’re going around over the earth poles as the earth turns underneath, you fly over all parts of the earth. They wanted to be able to launch out of Vandenberg on the West Coast into a polar orbit and because of security reasons, in a time of crisis, remember we were in the Cold War and everything, you wanted to be able to put a satellite up without necessarily giving the other side a chance to make all the radar measurements on the Shuttle and everything and figure out right away where the satellite is, and also there might be hostilities – the Shuttle was a strategic asset. They also wanted the Shuttle to be able to land the very next orbit. Well, alright, you take off from California, you fly over the pole, you deploy your satellite, and by the way, we have never ever with all the satellites we’ve deployed, we’ve never deployed a satellite on the first orbit, that would be an incredible feat, but that was the requirement. So you fly over the pole, you come back around, now you’re ready to land in California, but during that time the earth has turned by a 1000 miles – 24,000 mile circumference in 24 hours, actually 1,500 miles because in orbit is 90 minutes, one and a half hours. Your orbit now would put you right over the Pacific Ocean. If you just burn your engine, slow down and come down through the atmosphere, that’s where you’re going to land. Instead, as you’re flying through the atmosphere, you basically have to come down banked on your side and essentially you’re generating a lift vector, and instead of turning your lift vector up, you turn your lift vector to the side and that pushes you over, and delta wings can generate a higher lift vector than the straight wings and that was the determining factor.
- Size of payload bay 60 feet long by 15 feet diameter
- Size of crew cabin defined to be over 2600 cubic feet
- Payload 65,000 pounds at lift off and 35,000 pounds at landing
- The orbiter is a launch vehicle, a space craft, and an aircraft
Here’s a very important requirement that you need to understand: the orbiter is a launch vehicle, it’s a space craft, and it’s an aircraft.
When you look at the two different systems, Apollo and the Shuttle, there’s no question that the Shuttle is much more complicated than Apollo. On the other hand, the Apollo mission is much more complicated than the Shuttle’s, but that’s something that makes it very, very interesting.
Let’s now talk a little bit about configuration. You might say, “How did you go from that conception of August of 1972 to something with all these numbers on it?
This is really a case in systems engineering. Let me explain to you what I mean by that. There were certain assumptions had to be made. One assumption was that the orbiter weigh was going to be about 175,000 pounds without payload and you were going to have a 65,000 pound payload. You had to make that assumption. You also had to make an assumption that you were going to use a liquid oxygen/liquid hydrogen engine. And why did you select that? Today I’m not quite sure it was a good decision, but we did decide. Why? Because of a specific impulse, the ISP, or the performance of the engine was the highest for the liquid oxygen/liquid hydrogen chemical propulsion. So we selected that.
When you use the equations of motion, you integrate the equations of motion, or you use any other techniques you may want to use, you then show how much propellant you’re going to need to get that into orbit. Then you know that liquid hydrogen has a density of about 4 pounds per cubic foot. Very low density. That means, this big external tank is mostly hydrogen. It’s a very, very big volume. Liquid oxygen is about 70 pounds per cubic foot. So that basically sizes your external tank.
That’s a very simple applied approach but that’s systems engineering. You do more than one iteration because there’s an old adage that the devil’s in the details. You keep iterating on that with this expert team and you then say, “Well that’s still not going to be the most efficient way to get you to orbit.” You really need a stage system. You put the solid rocket boosters on and now you don’t have a single stage, you have more like a two-stage to orbit. Then you size the solid rocket boosters and you determine when they have to come off.
Basically, that is how you go about doing the systems approach using systems engineering to come up with a configuration. That’s a very over-simplified way of doing it but that’s basically how you do it. And these are some of the dimensions that then come out of the vehicle. That shows you all the dimensions of the space Shuttle system, the solid rocket boosters the external tank, and the orbiter.
An interesting sidelight in this, which I got to thinking about after the Columbia accident where the foam came off. Of course, the reason why we put the foam on was because with liquid oxygen/liquid hydrogen at those temperatures you’re going to have a lot of ice form, and when ice forms and comes down and hits the orbiter you’re going to do some damage to the thermal protection system. One solution would be to put foam on the tank to eliminate the ice. Of course, you would assume that foam could stay on the tank. Another solution would have been to not go with a liquid oxygen/liquid hydrogen engine, go with what we call storables. Of course, they wouldn’t have as much performance and you probably couldn’t put 65,000 pounds of payload to orbit.
The point of bringing this to you, those are some of the decisions you have to make. As you go into various projects you need to challenge the requirements because the requirements are really going to decide what kind of system you’re going to have.
That’s the basics of how the system evolves. I over-simplified it quite a bit, but you have realize there are a lot of iterations through this process.
One other key thing, when you have a systems team, when you do systems engineering, you’re doing it in a team, you usually have different capabilities on the team: mechanical engineers, electrical engineers, you have different types of disciplines on your systems engineering team that can help you do that and then you have a systems engineer that heads that all.
The other thing that’s interesting is what the mission profile essentially looks like. This happens for STS-5, but this is the basic mission profile.
One time, when I was director of the Johnson Space Center, I had Mr. James Baker who at that time was Secretary of State and Eduard Shevardnadze from Russia who had the same position in Russia, I took them over to Mission Control and put Mr. Shevardnadze into a flight controller seat and was going to let him talk to the crew. Not knowing that Jeff Hoffman was on duty at the time, Eduard Shevardnadze spoke Russian up to the crew, and before the interpreter could answer it, down comes this beautiful answer in Russian from Jeff Hoffman. That really floored Mr. Baker, Eduard Shevardnadze, and me.
The profile of the Shuttle mission is as following: You lift off and you’ve reached max dynamic pressure in about one minute or about 30,000 feet, you have the SRB separation in about two minutes, in the Challenger accident I believe it happened in about 60 seconds, but in two minutes you have the SRB Sep (separation) and the SRB landing is by parachutes. You have MICO, which is main engine cut-off, the main engine cuts off and that, to me, is the biggest issue. That main engine has to burn for over eight minutes, that’s taking all the propellant out of that big tank, the liquid oxygen/liquid hydrogen, firing the main engines to get you to orbit.
At the top of this diagram you see Main Engine Cutoff. Orbital velocity at the cutoff is with a few hundred feet per second.
And then you get External Tank (ET) separation. The external tank comes down about 9,000 miles downrange and lands in the Indian Ocean.
Then you use the Orbiter Maneuvering System, the OMS engine of pods on the back of the Orbiter, for a very short period of time to have a final tune to get into orbit.
Then you have your on-orbit operations, whatever you’re going to do, go out and service the Hubble or whatever.
You de-orbit with the OMS engine.
You have entry interface at 400,000 feet. This is the elevation at which you start sensing gravity, you get about 0.05 G’s at 400,000 feet, and then you reenter and land at Edwards or Kennedy.
The Shuttle turns during lift-off. This is because we used the Apollo launch pad and the ditches for the flame bucket where the engine goes. Because the Shuttle was not orientated correctly on pad and in order to get it to the right attitude we had to make that roll maneuver.
Russian Buran did the same roll maneuver at the lift-off. Dan Brandenstein, who one time was head of the astronaut office, saw that the Russians made this roll maneuver, and he asked the Russians: Why do you make that roll maneuver? They didn’t have to do it. They said, “Because you do.”
There may be one other aspect to it and that is the question of why does the Shuttle actually fly upside down on the way up. And the original decision was with aerodynamics. The thrust is asymmetric. You have the external tank, and the Shuttle is sitting on the external tank, and so the thrust actually has to be through the center of mass of the whole system, so it’s actually flying not straight but it’s a little bit skewed. And for aerodynamic purposes they figured there was less stress. Although recently they’ve started partway through the launch another roll maneuver, and that’s for communications. The early part of the launch when you’re riding the solids aerodynamically, when you go through Max Q, you’re in much better shape if you’re upside down.
When you’re doing flight design, depending on whether you’re going to launch east into a 28 degree orbit or a high inclination orbit, the placement of the external tank re-entry is a major, major flight design. Often the trajectory has to be shaped to be not quite as efficient as it might otherwise be because you have to control the landing of the external thing. The external tank breaks up, some of it burns in the atmosphere and some gets down back to the ground. The tank is mainly aluminum. Remember, the liquid hydrogen/liquid oxygen inside evaporate and they are non-toxic.
Solid rocket booster environmental problem is a real problem that had to be solved. I worked on you trying to get the Galileo cleared because of a radio nuclear thermal generator on it. To get that cleared going into orbit was a big job which I thought I’d never finish but we finally got it cleared. It’s a very tough job to have RTGs cleared to be launched.
I don’t remember specifically all the noxious chemicals that come out of the solid rocket boosters but it is pretty nasty stuff. I remember one of my launches, the families and guests were taken to a launch site viewing area about three miles inland from the launch pad and it just so happened that the wind was blowing onshore that day. It was an afternoon launch and my brother is a real space nut and he always liked to watch as the rocket after about seven and a half minutes sort of disappears over the horizon, but after about five minutes the solid rocket exhaust was actually approaching the spectator area because of the wind blowing and they made everybody get on the buses and drive them away so that nobody got injured and my brother was really annoyed because he didn’t want to leave.
I’m assuming that everybody is aware that the Russians at one time had a space Shuttle program, and they have manufactured a vehicle which looked almost identical, not exactly, but almost identical and it was no accident because they used our plans and they only ended up flying it once, and they actually flew it unmanned and, despite a little bit of nail biting during the landing, they did recover it successfully and then they discovered just what we had discovered was that it is a lot more expensive to operate than they had anticipated, and they had a lot less money than we did so it basically never flew again. They had crews of cosmonauts who had been training to fly the Buran.
There were actually two differences. The first was that they put the engines on the external tank. I think that this did two things. First of all, it did improve the performance. And also the Russians turned out engines on assembly lines basically. They turn out a tremendous number of rocket engines. I don’t know the details of the performance of their engines, how their main engines compared to ours. I think they actually had four engines, if I remember.
The orbiter was a glider. There was no difference, it was an unpowered glider.
The other thing, they did learn one thing from us, when you have a delta wing vehicle, you’re very, very sensitive to the center of mass of the system, and it was a problem because the orbiter, when it flies back through the atmosphere, it hits the top of the atmosphere at Mach 25, and then it flies all the way down to subsonic, and the control characteristics and stability characteristics change throughout that flight envelope, and we will have a lecture specifically on the aerodynamics of the Shuttle, but it turns out that the Shuttle is extremely sensitive to the forward CG, and I think it was around the Mach 3 flight regime, if you’re just an inch or two forward of a critical area in the CG, you can lose control.
They always do a weight and balance on the Shuttle before launch, and on many flights we’ve actually had to put lead ballast in the Apt engine compartment of the Shuttle just to get the CG far enough back to get Mach 3 stability. I hate to tell you how many tons of lead we’ve launched over the course of the Shuttle program because of the CG. And if you look at the delta wing profile of Buran, it is slightly different from the orbiter because I guess they learned the lesson and so they were not as sensitive to the CG. But of course, once you build an orbiter, you can’t really change it. And so, we were sort of stuck with it.
The other concern was that with the thrust behind the orbiter and large mass in the tank you get what you call Pogo oscillation effect. One concept that Max Faget had early in the program was to have a swing engine, have the engines fire on the back of the tank and then, when you get ready to separate the tanks, swing the engines back in the orbiter and bring them back in the orbiter. We threw that out because that was a complicated mechanism.
You’re going to have a discussion on mechanisms and mechanical systems. Everybody could design them, but everybody has a hard time making them work. When you put up a mechanism, when you put up electrical schematic, not everybody accepts. When you put up a mechanical drawing of a mechanical system, everybody’s got an idea how it is supposed to work.
Here’s the picture of a vehicle when it rolls-out of the assembly building, there it’s stacked on a crawler which takes it out to the pad. This crawler was used for Apollo.
Is anyone familiar enough with Apollo to be able to say what the fundamental difference is with the crawler mechanism here and the crawler in the past? If you look at the picture of a Saturn rocket being rolled out on the crawler, the whole launch tower was on the crawler, and so they rolled the whole thing out. With the Shuttle it’s a little different. I’ll be bringing in some pictures at some point to show you some of these details.
They cut off the top part of the Saturn launch tower because the Shuttle stack isn’t quite as tall, and they added a movable what is called a payload change-out fixture.
Once the Shuttle gets on the pad, the whole enclosure roles over and forms a kind of a hermetic seal around the Shuttle so that you can open the cargo bay doors which, by the way, can’t support their own weight in one G, so you have to put an external strong back on them.
You put the payload in this payload canister here that’s installed in the change-out room and then it’s swung over, the doors are opened, the payload is installed in the bay, and then this forms a protective enclosure over the Shuttle, and it’s not swung back usually until the day before launch.
This is another detail in systems engineering. You go in from that schematic or that little plan of the orbiter, and you did iterations of the size, and then you do the iterations of the launch complex, how it’s going to be put payloads in, and that’s all systems engineering, levels of systems engineering.
It’s very easy to say: Well, we’ll just put the payload in on the pad. But the actual development of the mechanisms to do that is extremely complicated. What amazes me is that it has to be done essentially in a clean room environment. You are up there crawling around the pad inside the payload change-out room in white coats, buddy suits, and gloves on, and outside the wind is blowing and it could be raining and sand blowing by. And the whole thing has to be done in a clean room environment but it’s on the scale of a Naval Ship Yard. This is a huge vehicle. It’s really a challenge.
In all honesty it’s one of the reasons why the cost per fly has gone up. The original concept, when Dale Myers was Associate Administrator for manned space flight, we concluded that this will be very simple: we were going to make very standard payloads, you went and put the payload in, fixed it in, we were going to launch and deploy it, and then come back. Of course, that all changed, the missions became very complicated and we didn’t do that.
That’s another important point though. You better be sure when you get to be a project manager or manager – understand your requirements, understand your customer’s need. As the ground rules change your performance is going to change. You need to understand that.
We had cases where we had a bird on launch. One of the things you do, you fire dead chickens into the window to see if you could break the window and the system. You do that for an aircraft also. I remember one simulation where in the simulator right at lift off the instructor came in and threw a rubber bird in the commander’s lap and said: You’ve just been incapacitated by a bird strike. And so the co-pilot had to take over and fly it.
You should realize the Kennedy Space Center is a wildlife sanctuary. There’s a lot of birds, there’s a National Wildlife Sanctuary Park Rangers, there’s eagles in the nest, there’s a lot of turkey buzzards. There’s a landing strip right in the middle of the bird sanctuary and so, before the Shuttle gets ready to come in, they send plains to buzz the runway and scare the birds away.
They also had a problem at one point with woodpeckers. Woodpeckers decided that the external tank insulation was a good place to find bugs. It sounds funny but think of what that potentially could mean. They had loud speakers and stuffed owls and everything which they ended up putting around to scare away the woodpeckers. These are things which the original systems engineering never took into account.
Let me talk a little bit more about Systems Engineering. As Dr. Hoffman mentioned, you have the three-legged stool of cost, schedule, and performance. And that is a continual trade-off as you go through a system engineering program. The things you have to look at, the things that cause program managers and project managers to lose their job, first, you’re going to recognize that you have a weight increase in the vehicle, whatever you build is going to cause weight increase. The next thing you’re going to have is a schedule slip. And then when you put those together you’re going to have a cost increase. And then you’re going to have technical problems. And those are disastrous cases for a project manager. Those are the reasons to fire somebody. Luckily I made it. I made it because of a man that you’re going to hear talk, Chris Craft. He was my immediate boss and he ran interference for me and he’s a good man to have on your side running interference for you.
We were trying to get the cockpit laid out. It was very important to get the cockpit laid out and I wanted the crew to have their inputs. You want the crew to be a part of laying out the cockpit. I called the head of the astronaut office over and I said: We need to lay out this cockpit and you need to come back to me with a decision of how you want the cockpit. He said, “I can’t do it. I’ve got 100 astronauts over there and they won’t agree.”
I said, “I’ll tell you what you do. You tell them that you give them two weeks and then I’m going to do it.”
And that really scared him because I didn’t know anything about laying out a cockpit. And we had a cockpit laid out.
So sometimes you have to use management techniques.