DY Pegasi

Transcription

>> Dr. Elias: Good afternoon. As Mike says, my name is Antonio Elias. They’ve coached me so that I should tell you a little bit about my experience. Let me start by saying that if I could have it my way when somebody asked me, “What you do for a living?” I would love to answer, “I’m a systems engineer.” The only problem with that is that people look at you weirdly and say, “What?” I demote my status and say I’m an engineer. Even that is an exaggeration because I’m now in management puke, but I like to believe I’m still an engineer and furthermore like to believe I’m still a systems engineer. The fact is, when I went to school, as Mike mentioned, I came this close to being a professional student. I could just not get the school out of my blood. I did choose my academic career with a purpose of being a systems engineer. Now, at the time, in the early ‘70s there was no systems engineering curriculum. There was no discipline called systems engineering, even though some of the elements that today are labeled “Systems Engineering” with a capital “S” capital “E” started to develop then. So what did I do? I sampled – the Department of Aeronautics and Astronautics at MIT is divided into five main disciplines. You have the classical ones: structures, aerodynamics, propulsion, guidance and control, and so on. So what I did is I chose my eight years as an undergraduate/graduate student at MIT to sample all of the disciplines and become as equally proficient in as many of them as I could. Not that I quite achieved it, but one thing I must say is that that has helped me more in my professional career than almost anything short of deciding to go to MIT. As Mike mentioned I joined the faculty for several years until finally they just decided I wasn’t good enough to be given tenure and then I got a phone call from a younger co-student of mine. Younger because when he was an undergrad I was already a grad. He said that he and a few other colleagues were here in Washington and they had started a space company. Of course, that was David Thompson, it was Orbital. I had the advantage of not having been given tenure, which means that you also feel like taking more risks. So I’m terrible with dates, but there’s a few that I remember. One of them is September 2, 1986, which was the day I joined Dave Thompson and nineteen other people. The day I joined the company there were twenty of us. From Dave Thompson to the receptionist, and then me next to her on the ladder. Today, Orbital, I’m not sure how familiar you are, but Orbital is about 4,000 people. We make one to one to one-and-a-half billion dollars a year in business. We build small launch vehicles and we’re now developing our first medium launch vehicle. A plethora of spacecraft including one which will rendezvous with space station probably late this year or early next year as a cargo resupply. We have also made very large inroads into national security space. If you count the number of satellites and rockets Orbital produces and you divide it by the number of employees, you’ll find an absurdly high ratio of vehicles and programs per employee. Well, the secret is that’s all that Orbital does. Orbital doesn’t have ship yards or SITA support contracts or weapons systems or stuff like that. All we do is space. How did we get there from twenty people? Actually three people in a spare bedroom in Dave Thompson’s house in Thousand Oaks California in 1982 to where we are today? One of the elements that achieved the Orbital of today was a particular flagship product. That’s the one that I chose today to give a little bit of body to my remarks and my personal experience on systems engineering and how it affects both the results, my own professionally as well as the organization I happen to be working in. Now, let me advance one of my punch lines. I’m also a pilot, a very avid pilot. I have an ATP, type ratings, instructor certificate. I love flying. I find a lot of analogies between the discipline of flying and a lot of the disciplines in systems engineering. There’s one flying element that I’m going to punch line as perhaps the single most important aspect of systems engineering. In flying, it’s called “situation awareness” or sometimes “being ahead of the airplane”; knowing or feeling, intuiting what is going to happen so nothing surprises you. Situation awareness is probably the single most important element of systems engineering. Chris was kind enough to share with me some of the terminology that you’ve been using in this course and I think what I’m going to talk about really translates in your vocabulary as the alignment element or especially the vision element. In order to do that, I have to tell you a little bit about what Orbital looked like in 1986. As I mentioned, twenty people. The first product that Orbital produced wasn’t even built by Orbital. It was an upper stage for the space shuttle. Those of you who are history buffs will recognize the Transfer Orbit Stage. Only two were built. One actually flew on the shuttle. The other flew on a Titan III. This one was Mars Observer, the other one was Active Advanced Communication Technology satellite for NASA. What Orbital did at the time wasn’t even technical or industrial. It was just the financing and then they subbed with a small aerospace company that was very hungry and was going to lose their launch vehicle product because the shuttle was going to eliminate all of the EELVs. That small company was called Martin Marietta. They were the ones who actually built the transfer orbital stage. Then Challenger happened. With that, the presidential directive that the shuttle shall not carry commercial spacecraft anymore. So the need for a commercial-cost, commercial-priced transfer stage to carry commercial spacecraft from the shuttle Earth orbit to geosynchronous orbit vanished in the terrible fireball that also killed six of our great colleagues. What to do next? Oddly enough, our next idea was not a launch vehicle. We never saw ourselves being into launch vehicles. As a matter of fact, there had been many attempts. I don’t know if you’ve noticed, but every seven years, as regular as the cicadas come out of the ground, every seven years, somebody decides that what this country needs is a better $29.95 launch vehicle. Some succeed. Some don’t. We certainly weren’t the first ones. As a matter of fact, we had seen the battlefield of new low-cost vehicles littered with corpses. We didn’t want to be one of them. But we had the idea of building a constellation of small - let me not call it communication - messaging spacecraft to provide things such as remote asset monitoring and so on and so forth. One of my first jobs as the company chief engineer – twenty person company remember – was to find avenues to launch thirty or forty under 100kg, under 50kg spacecraft into low-Earth orbit. It turned out that the opportunities for secondary payloads on launches were nonexistent. We would need seven big launches that all happened to go to the orbit that we wanted and all would be able to host five or six little satellites. Forget it. We went to see a private company called American Rocket Company, one of those seven-year-itch organizations. Oddly enough, instead of welcoming us with open arms as potential customers they thought we were spies trying to find out their secrets and they kicked us out of their office. Had they said, “Oh please, sit down. Have a cup of coffee. Here’s our fine rocket.” We would have bought it. Instead we decided it was time we developed our own small rocket. That small rocket ended up being Pegasus, as some of you may know. Of course, the unusual aspect of Pegasus is that it is air launch. Here begins the systems engineering saga of the Pegasus. My role was the chief designer/systems engineer. There was a program manager, former NASA employee by the name of Bob Lovell. You are too young to have ever met Bob and then there was Dave Thompson who was essentially managing the business side, the money side. So what was unusual about this program? First it was done by a group of people who didn’t know anything about launch vehicles, starting with myself. It was a very small group of people. I have in my office a picture of about twenty people - fewer than half of the group here. That is half of the team that designed, built, and flew the first three or four Pegasi. We were very short in money and we didn’t have a lot of time. Now, fast forward and what were the results of that effort? Some were good. Some were disappointing. On the good side, we have flown since 1990, the date of the first flight. Forty, four zero Pegasi. We had a few problems initially, but the last 26 have been 100 percent successful. In terms of track record, 26 in a row is not bad. When we tallied up all of the cost of developing the vehicle, and there were a couple of short cuts that I’ll mention, it added up to in 1990 dollars, 42 million dollars. We had initially budgeted forty. So wow, we ended up within five percent of our expected non-recurring budget. On the other hand we had initially targeted a price, not cost, of six million dollars. The first sale of the Pegasus was close to 12.5 million dollars and I won’t even mention how much we charged NASA to pay for a Pegasus because it’s ridiculous. So definitely on the recurring cost standpoint, we didn’t fare as well as we did on the non-recurring. Also, when we embarked on this endeavor we had these incredible hopes for how many flights we will fly. I remember a view graph that said we’ll start with six the first year and twenty the second year to a steady state of about thirty five a year. I just mentioned 40 flights in 22 years. This is reality. Now, Pegasus today is a very small fraction of the business at Orbital, a couple of percent. But, Pegasus did something for Orbital, which is like that ad for this credit card, this so much, this so much, priceless. There’s a priceless element to what Pegasus did, at least for the company, surely for me. I wouldn’t be as bold to say from the space community or for NASA. For Orbital, it was the flagship product. It was the first product the company did. It proved that the company could do something quite spectacular. What was spectacular about Pegasus? It was developed with private money. It was developed in three years and it was air launch. Now let’s go back to the beginning of systems engineering. I could bore you with a long list of trades that embody the realization of what alternatives we had in the production or the development of this particular product or vehicle and all the math that went through it. But I will not do that. I would rather I will point out two or three characteristic decisions and what systems engineering, with a lowercase “s” and a lowercase “e” has to do with it. The first one is air launch. Why air launch? Now, this question is gaining a second life because the same as seven years a small company starts to develop the country’s lowest-cost launch vehicle with a certain periodicity everybody revisits air launch. They say, ‘Wait a minute. If air launch worked so well for Pegasus, why don’t we build a medium launch vehicle air launch?” And so on and so forth. Our initial intent for why air launch was as follows: a) for those of you who have done a little bit of modeling of launch vehicles and have observed the significant fixed mass items that any launch vehicle has to have. Let me be blunt and go back to the IMU and the computer and the batteries that the last stage has to have. Those are relatively size independent. So whereas that tax is trivial and insignificant for an EELV, not that trivial but still not that big of a deal in a medium launch vehicle as the vehicle becomes small it becomes a higher and higher tax. For those of you who are also students of history, look at the Scout launch vehicle. What a remarkable invention design that was. It did not have a flight control system. How they did it was remarkable. But that was the reason for it. Now, by air launching you gain a significant number of performance advantages. Where shall I start? Look at the pressure losses at sea level. As you know at sea level an engine has less thrust. The same engine that would have some thrust in a vacuum has less thrust. That is a product called exhaust area times the outside air pressure, which means that you don’t see for the first stages engines with very large expansion ratio because that big back pressure would kill the thrust. So you reduce your expansion ratio, which reduces your ISP, but it reduces your pressure losses so there’s this sort of jiggling systems engineering trade that determines or maybe just the propulsion trade that determines that. When Pegasus starts its life at 35,000 feet, where the ambient pressure is one fourth sea level. So two things happen: a) the total loss is lowered because there’s less back pressure. b) You now can get by with much larger expansion ratio than you would at sea level. So the overall specific impulse is higher. Number two, the aircraft is going at a pretty good clip, 770 feet per second. So take whatever delta V your rocket had to part and subtract 770 feet per second. As you know, delta V, size, mass, propellant you use whatever, there is no linear relationship, it’s a logarithmic relationship. So saving those last 750 feet per second are worth a lot more. When you lift off from the ground you can’t go horizontal. You could on the moon. Just clear the hills and you’re OK, but on the earth you have to get out of the atmosphere. So get all of this vertical velocity which ain’t going to help you get into your orbital vertical velocity, so in time you have to turn and you’re using thrust to rotate your velocity vector and your nice vertical velocity ends up being zero when you inject on orbit, so people call it turning loss. Well, Pegasus starts pretty horizontal, drops a little bit and then rises up. So the flight path angle never really deviates from minus five to plus thirty or forty and so on and so forth. Against that you have the weight of the wings, you have the aerodynamic drag of the lift and so on and so forth, but overall you have a 17 percent saving in delta V. Turn that into payload improvement or mass reduction. At this size vehicle it turns out you get twice the payload that you would if you took Pegasus, got rid of the fins, you have to add the thrust vector control first state. That’s another systems engineering trade I’m going to mention. You get about half a payload. I have actual physical proof of that because the boosters that we’ve built for the Missile Defense Agency’s ground-base mid-course interceptor are Pegasus without wings and with the TVC in the first stage. So I know exactly how much payload those puppies would give us if we were to use them as launchers and it’s half a payload. They’re about the same size, mass, cost, everything. So it is a big deal. Why else? Air-launch. Our thought was we could get away without a launch range. The thinking went something like this. Put it on an airplane. Fly the airplane 500 miles up to sea, point in the right azimuth, or whatever inclination you want to hit, push the button, the rocket goes, never comes and lands so it doesn’t have a flight termination system it doesn’t have a tracking system. It doesn’t have anything. Crew flies back home, has tea comfortably in their back yard that evening and that’s the end of the story. No pad, no rain, no nothing. Well, that didn’t quite turn out quite that way. Because the first thing everybody said is – and remember this is the days before GPS onboard rockets – wait a minute, what if something isn’t right? Wouldn’t you want to track the vehicle? Wouldn’t you want telemetry? Oh dear, you’re right. So we need to range assets, which means now we have the closer to the range. As a matter of fact, we’re going to need a flight termination system and so on and so forth. The end result was that we did not get away with flying without a range. What we did get away with, though, is a significant reduction in the complexity of going from launch location A to launch location B. Pegasus has been launched from the Western Test Range - I can’t say Vandenberg Air Force Base because of the first six flights that overflew Vandenberg Air Force Base, but it was their range assets. Eastern Test Range, yeah we did use the shuttle landing strip to take off, but by the way the flight control center of the first Eastern Test Range flight was at Wallops. We’ve flown out of Kwaj [Kwajalein] very near equatorial locations. One of our Kwaj operations is something to behold. You go to Kwaj a month before launch and there is nothing there that says Orbital, Pegasus, NASA, nothing. Launch occurs a week later, there’s nothing there that says Orbital, Pegasus launch. So for one month before to one week after the launch, that’s it. Everything gets flown in, most of the stuff is flown on our current carrier aircraft. We’ve launched from Europe. I think we have the only space launch to orbit from the territory of a European country. The country is Spain, my ancestry country. The location is the Canary Islands, which some of you may say, well is that really Europe? Well it is European territory. Oh but Peru is French territory too. Anyway, so we’ve launched from there and you look at all of the places of Pegasus launch, Wallops is one of them. So five launch locations with zero permanent infrastructure at those launch locations. Interestingly enough that was not one of the original objectives of the program. The negative side is that we’re now the proud owners of a Lockheed 1011 that spends most of the year holding down the tarmac at Mojave Airport. But the first few launches were done not on that Lockheed 1011. They were done on an old, NASA B-52. The old Balls 8, one of two B-52s used for the X-15 program. I must say that Orbital is a company, Pegasus is a launch vehicle, and myself owe a debt of gratitude to the former Dryden Director Marty Knuzten, who, I’m not sure if he’s with us, who when we went to him in 1987 to ask for the use of the B-52 could have laughed us out of the place and instead of that he said, ‘Oh that’s very interesting, let me call a few of my guys. Would you like to tour the aircraft?’ And the rest is history. But I finagled my way going back to the situation awareness and breadth of experience that is systems engineering. I finagled my way into being the launch panel operator for the B-52. The B-52 carried two pilots and a launch panel operator. I won’t tell you exactly the why’s that I used to do that, but I ended up being the operator through another trade. The trade of when you drop this thing from the aircraft how long do you wait so that if there is a catastrophic failure at ignition you don’t imperil the aircraft and its crew? And yours truly did that trade calculation. I said, “I’m willing to put my money where my mouth is, I will be in the B-52 exposing myself in case my calculation is wrong.” I think that impressed them sufficiently to say, ‘Ah, OK, let’s do it.’ Now, since it involved the B-52 and dropping something very big, by the way, Pegasus is the largest thing ever air-dropped from an aircraft, period. It’s about 50,000 pounds. There’s a famous movie that involves dropping something out of a B-52. Anybody care to remember what that movie is? Dr. Strangelove. Do you remember how Major Kong ends his career? That is April 4, 1990. We flew on April 5. This is the live Pegasus. There’s 40,000 pounds of ammonium perchlorate and aluminum under me. If you look carefully, there’s an anti-static mat under my butt. If you look carefully, I have my flight boots tilted out. I didn’t want to touch the vehicle and the vehicle has been painted. It didn’t have the decals set up yet. Again, the lead systems engineer flies in the mission. Now, after the first six flights, NASA decided that they could not in all honesty support operational use of the vehicle with the B-52. Today they probably wouldn’t say that, but at the time, they did. So we looked around and found a very nice low-cost, large, commercial carrier aircraft that could carry the B-52 and that’s the Lockheed 1011. We bought Lockheed 1011, serial number 67 from Air Canada. Most large bodied aircraft structures have a keelson in the middle of the structure and the ribs are attached to the keelson as well as all the stringers in the longitudinal direction. Not the 1011. The 1011 has a dual keel running on both sides of the center line and it’s approximately 48 inches. Right here in the middle of the space between the landing gear wheel wells there’s a hydraulics service center which is a big room with some plumbing and so on and so forth. Well, let’s go back to systems engineering and how this ties in. The motors for Pegasus were the derivative of a technology that Hercules Aerospace developed in the late ‘80s for a project called the small ICBM. Sometimes it’s called a rail garrison. It was going to be an ICBM capable of reaching even higher from a mobile launch erector either on large trucks or on railroad cars. For a number of reasons, thank goodness, that program was never required, so it was cancelled. But Hercules had developed this very high efficiency high performance, small solid motor technology. It was a little bit small for our size. So Hercules Aerospace, now part of ATK developed three 48-inch, two 48-inch motors and a smaller 38-inch motor using that technology. The way the 40,000 to 50,000 pound vehicle is carried on the aircraft is through four hooks attached to the wing. By the way, why does it have a wing? Another systems engineering trade. Anybody care to speculate why? For take-off? No. The lift of this compared to the lift of the B-52 is peanuts. It’s when you drop how much negative flight path angle. If you just use a rocket thrust to modify your flight path angle then you lose so much energy that you negate a lot of this 17 percent delta V you gained that I mentioned earlier. Not to mention the interesting 45 - 90 degree path angle. The wing itself is another systems engineering trade. Most aircraft wings, let’s see, you’re familiar with the problem, are optimized for a certain Mach number or a range of Mach numbers. Pegasus accelerates from Mach 0.8 to Mach 8 in about 58 seconds. It doesn’t stay at any Mach number more than one. This is the most pedestrian, basic, supersonic wing one could imagine. Very inefficient, but it does the trick. Interesting systems engineering trade and I use this as an example of non-numeric or extra-mathematical elements that got me into systems engineering. Ideally, the way to carry a vehicle like this is to have a long wheel-base attachment. Say a couple of hooks here and a hook routed towards the front and that’s the way the X-15 was carried. What happens when you don’t do that? If you only carry it as we ended up doing it from here. The whole rocket droops and there’s a lot of strained energy stored in that drooping vehicle. What happens when you release the hooks? Boing! The lateral acceleration the spacecraft sees when that happens is one of the design drivers for the spacecraft that rode on Pegasus. Why did we do that? What systems engineering logic, what normal flow- down of requirements and flow-up of verifications would lead you to do such a stupid thing? There’s a reason. Those 40 million dollars that we were going to spend were not all ours. Half of them were Hercules Aerospace. They were responsible for the design of the 48-incher and the 39-incher based on their previous experience with technology. But it was their money, their nickel, their design, their engineering. Had we had the logical two hooks here and one hook in the front, the load path that held the whole vehicle together would have gone from the hooks to the first stage, from the first stage to the second stage to the interested joint, from the second stage to the front hook. Hercules Aerospace said, “Time out. That’s too hard for us. Our experience has been a small tactical missile that are carried under a single point and one rocket motor body holds entire loads. We know how to do that. Don’t ask us to analyze, verify, and test the requirements associated with transferring the load through two of those.” Were we the Orbital of today and we were doing things today, what would have been our systems engineering reaction? “Well, Hercules, tough. Go ahead and do it. Cost is not an option. You’re on a cost-plus contract.” But that was not the case. We didn’t have a lot of money and they were not under a firm-fixed price contract. So we had to gulp and accept sub-optimal. So here’s an example of systems engineering going beyond the pure numerical set of requirements flowing down and flowing up. Now, when we transferred the program to the Lockheed 1011, we said, “Well now we know better. Hercules, don’t worry about your responsibility. You’ve already demonstrated. We’ve already flown six times.” And we added a third hook here and we only used two of the four hooks in the back. Now there’s a very interesting trade that happens here, which is the timing of the release of the front hook with respect to the rear hooks is kind of touchy. If you happen to hit the right timing and that timing is not just everybody at the same time. It has to do with how much elastic energy is still stored on this thing and what is the natural frequency of this vehicle as it flows. If you time it right, you can reduce that twang significantly. If you don’t time it right you can double it. So it took a little bit of experimentation. There’s this very complex mechanism going from the main hook release system here to the front that is actually adjustable. You can adjust the delay, the mechanical delay. I think we hit it right because it seems to be working. So these are a little bit of the examples of the systems engineering situation that happened during the development of Pegasus. Let me recap a few interesting items. A small company started with a limited amount of money and it went beyond a simple set of specs. We realized that this was going to be our flagship vehicle, so there was a lot of representation. If we were successful here we would be successful in other things. A very small team - forty people. Now sometimes, it is argued that systems engineering is required for the same reason that in software engineering you tend to decompose a very large program into isolated items and erect abstraction barriers between the various elements. That’s how large software programs can be developed. There is a price to pay, however, which is you’re setting up a number of interfaces between these various abstraction barriers because the idea is those abstraction barriers help you because you don’t have to look inside the other compartment. If you’re in compartment one you don’t have to look inside compartment two, so your job is a lot easier. On the other hand, what you lose is now the contract between compartment one and compartment two becomes a very critical contract. You will spend an incredible amount of money and effort and risk following that contract when it probably would have been a lot cheaper, easier, and faster and risk-free had the compartment that you can’t see change what they do just a little bit. That is the trade that we run every day in our jobs between taking advantage of compartmenting and being the victim of compartmenting. On Pegasus we did not have that problem. There was one guy who was a systems engineer for the whole program and that guy had the authority to say, “Well it’s still hard to put three hooks because Hercules wouldn’t play ball. Let’s do it this way.” But the most important thing is that there were a number of us that had situation awareness of the program. Situation awareness of why we were doing this. Situation awareness of what our real resource limitations were. Situation awareness of how to bypass those issues and get the job done. At the same time we did have the support of those who were in authority and the financial support. Nobody interfered with us. That’s a luxury that not every program has. So these are some of my personal experiences, I’d be glad to open the floor to any questions that you may have. >>Question 1: You mentioned the four hooks, I worked on Hyper X and we had fun with those four hooks as well. Was that a selection by Orbital to go with the four hooks or was that a previous design that you guys had to deal with? >>Elias: Actually the previous design was the three-hook. It was the X-15. And we had to build this device called a pylon adapter to go from the X-15 pylon three-point system - a very nice, robust, long wheelbase - to this tiny thing. There’s a little anecdote I’m going to share with you guys. We initially designed a pylon adaptor using welded tubes and as we looked at the structural dynamics of this bouncing and rocking system we realized that we required higher and higher stiffness from that pylon adaptor. The way to do it, since we had very little depth to get our moment arm from was to go to increasingly higher and higher heat treatments of the steel of the tubes to get the E of the modulus of elasticity that we wanted. As we did that, all sorts of horrible things happened. Wells popped out, the whole thing warped like a potato chip. So I went to my friend and buddy Burt Rutan, who, by the way, did design the structure and build the wing and fins for Pegasus. We were driving from the Marie Callendar’s in Lancaster towards his shop in Mojave. In the car, a one-hour-and-fifteen-minute ride we specced the product, we agreed on a schedule, and a price and we essentially shook hands on the contract. The project within scaled composites was called the Pylon Adaptor Now In Composites or PANIC. He started the process by instead of using the fancy aerospace core to set the composite, he used that stuff that cheap furniture is made out of…particle board, thank you. He came in, instead of 28 days, it was delivered in 21, 22 days. It was the first structure, he said, that I’ve ever designed that I didn’t have to worry about the weight. And if you guys go to Dryden, it’s still in some bone yard somewhere. The old-timers will point it out to you and say “Yeah, that’s the Pegasus pylon adaptor.” [Inaudible Question] >> Elias: Oh, very good question. You’re asking me about the dynamics of the original design. Actually, I didn’t mention the second trade immediately after the idea of air launch. The second trade was which launch aircraft to use. We looked at a variety of aircraft, but let me point out three that we looked in-depth at. We did look at the B-52 because of the X-15 experience. We looked at the Hercules because the version of the Hercules that carries these drones. They’re much smaller aircraft. They also wing carry like the B-52. The third aircraft we looked at in detail was the SR-71. In general we looked at the trade between small but fast release or slow release but bigger. Let me backtrack one more piece of trivia. If you go to the Udvar-Hazy museum on Route 28, the second largest artifact of the space wing of that museum is a Pegasus. It’s a Pegasus XL, which is a little bit longer than the original, but a Pegasus nevertheless. What visitors to that wing of the museum don’t realize is two curious coincidences about the location of that artifact in that museum. The idea for air launch after having decided, as you pointed out, that we couldn’t find a ride for a small messaging spacecraft. By the way, those of you who know the [----] Constellation that’s what that initiative eventually morphed into. The idea that came to me according to somebody who was there on April 8, 1987, at an incredibly boring meeting that an organization called a Center for Innovative Technologies, CIT, that we’re upside down in building the corner of 28 on the access road. In ’87 they were trying to pull together a Center for Commercial Development of Space Initiative sponsored by NASA with the cooperation of industry and academia. Unfortunately that meeting which three of us attended to represent Orbital was very poorly organized. The people who were supposed to speak didn’t arrive, the agenda was wrong, the timing was wrong. So we ended up sitting at a table very much like these tables here bored to tears for a long time. Now, the previous year, the Air Force had shut down a perfectly functioning spacecraft with the small rocket launch from an F-15. It barely had enough oomph to get orbital altitude and just before it fell down to Earth, the spacecraft hit it. But nevertheless it was an air launch orbit something. So I sketched on a piece of yellow-ruled paper on the back of the envelope a kind of poor F-15 climbing and a little rocket and then it was a kind of expanded view remember in our world we don’t say “exploded view”, you say “expanded view.” It carried small spacecraft. So I showed it to one of my two colleagues who raised his eye brows and gave to my other colleague who raised his eyebrows and we started talking about it and we didn’t realize that the meeting had started. We had started talking about it and within an hour we said, “Hey, why don’ we get back to the office and start talking about it?” We stood up in the middle of the meeting and disappeared. That meeting was at the hotel across from Route 28, from the entrance of the Udvar-Hazy. So the germ of the idea for Pegasus occurred less than a mile geographically from where the artifact actually is. And twenty meters from where the Pegasus is an F-15 launched interceptor rocket. So the problem that led to the idea of Pegasus is twenty meters away and the location where the idea came from was less than a mile away. But the initial thought was to use the Hercules. So the initial Pegasus which was a much smaller vehicle than what you see today. It was about 8,000 pounds in weight and we thought we had a payload capability of about 50 pounds. Now, if you do the math associated with the sensitivity, partial derivative of payload to structural mass fraction uncertainties and specific impulse uncertainties, you come to the inescapable conclusion that the smaller the vehicle, the higher the sensitivity. So your design is not very robust. You make a small mistake on your mass fraction and kaboom! There goes your payload. Plus that would have meant, even in the unlikely case we would have hit that 50 pounds of payload that we would have to essentially launch one satellite with one rocket. So the next thought was how about if we go for 200 pounds of payload? So we could launch 400 pounds of payloads so we can launch eight. So from there we had to kick out the Hercules. The trade between speed and altitude and one more thing, flight path angle, is still being revisited by some people today. Burt Rutan before he retired – well that’s a different story – that within – I’d say that within three months of April 10, 1987. It was this size, this diameter, it was three stages. It didn’t have thrust vector control on the first stage. It uses the fins for aerodynamic control. Originally it had four fins instead of three. Three came about from ground clearance on the B-52. We didn’t have a B-52 yet, but we knew we wanted to use an aircraft. So I’d say within the first three months of a three year project. By the way the first flight was April 5, 1990. So from that event in the hotel to first flight, three days short of three years. I’d say that within the first three months. Ninety percent - 95 percent of the sizing and 80 percent of the design details including the outer mold line of the wing and the outer mold line of the nose. Anybody thought on as to why Pegasus has a rounded as opposed to a biconical nose? Again, another systems engineering trade. With a biconical nose, you get the strong attached shocks. So if you know what angle of attack you’re going to fly at, you can design your biconic to adapt to that. But Pegasus goes through these wild excursions, same as it goes through a range of Mach numbers so you can’t optimize the wings to a particular Mach number. I was concerned about the shocks in the biconic detaching at a particular set and then having these big upsets in pitch as the shocks attach and reattach. So what do you do? You put a smooth nose as you can to try and make any shock shenanigans be more continuous than discrete. Those items were all fixed within three months on the project. >> But my question was not so much the technical perspective about Pegasus. This group is a system is a systems engineering development program and one of the aspects that we’ve been toying around with is how does one grow and develop as a systems engineer? At a company like Orbital, obviously you can go out to organizations like NASA and obtain senior systems engineers, but do you have any training program in-house to develop your systems engineering? >>Elias: We have a training program for the functional elements of Systems Engineering, with the big “S” and the big “E”. The idea of flow: how you identify and define requirements, flow down. You know the classical down and up V. We have training modules that address that, but that does not address what I think is the most important systems engineering, the one with lowercase “s”, lowercase “e”. Our approach to that has two components. First we build satellites. We build GEO satellites and LEO satellites, now human rated spacecraft. We build launch vehicles, air launch, ground launch, small, medium, now with Taurus II solid and liquid. We build some electronics and payloads. We have a broad range of products. Number two, our typical program cycle length is three years or less. Unlike ---- or Hubble where somebody can make a career and the doctoral dissertation that started with the program, well the guy is retiring at the end just before it launches. We can afford two things that very few organizations can. We can hop people across programs and across disciplines and across products and so on with relative ease. And somebody within working ten years at Orbital has probably has probably worked cradle-to-grave three programs. That’s part of the that’s $29.95, that’s $200, this is priceless. The experience of somebody within ten years being able to cradle-to-grave three programs is our secret to systems engineering with a lowercase. I’m sorry I don’t have any magic bullets to offer. Well, shall we get back to real work?