"Alan Brown Part 4" Avstry #3d
Alan Brown, of the Lockheed Skunk Works concludes our conversation on the F-117 Nighthawk and some reflections on the Northrop B-2 Spirit and F-22 Raptor.
Published Date: Tue, 10 Jul 2012
Direct link to m4a audio file of show. Recommended (right-click to download/save).
Before we get to this episodes we ask that you consider making a small financial donation to the Aviation Story podcast. This donation will help us gather better stories, and share them with the Aviation Story podcast community.
Sincerely, The Avstry Team
"J.R. Warmkessel: Going back to one of the concepts, you said that the airplane was unstable at all three axises and-
Alan Brown: No, no, not the same time
J.R. Warmkessel: But it was also known as the Wobbly Goblin, right?
Alan Brown: It got known as the Wobbly Goblin for a different reason and that was, and this is another interesting aspect that I learned as a program manager, and that is, for heaven's sake, don't accept help from people who don't know quite what they're talking about, and in this case we had a very competent chief aerodynamicist, Dick Cantrell, and the fins on the F117 were really very, very small and I'm an aerodynamicist and it seemed to me like they were on the small size and, you know, I just didn't like it and I talked to Dick about it and he said "Well" he said "We've done wind tunnel tests and we think they're OK". And I said "Well, OK. You're the boss. We'll do it". We fly the airplane and even if you have a computer control if the control surfaces are not strong enough all the computers in the world won't fly the airplane right because the control surfaces just don't have enough meat to do the job and the airplane was, in fact, laterally waving about and it got, after the first flight, quickly got called the Wobbly Goblin by the Air Force guys there. So, of course, the obvious thing to do was to increase the size of the fins and we put them up to what was a reasonable size and I said "Dick, why did you, in fact, make these fins so marginal?". He said "I was trying to keep the radar cross-section down for you." I said "Dick, you don't understand how radar cross-section works. It doesn't work just by what you can see from the front, like airplane drag". He was trying to hide them behind the rest of the airplane. I said "In fact, if you have the surfaces at the right angle and the leading and trailing edges at the right angle, the radar cross-section goes down as the size of the article goes up because the longer it is, the smaller the spike narrows the spike width that you get with radar". So I said "What happened Dick is that you went ahead and did something which you knew wasn't the best thing aerodynamically because you thought you were helping me, but you never checked with us radar guys as to what, how it really works. So please do not help" and I was very adamant about that in future occasions, not just aerodynamics but with everybody, to make sure that this revolutionary airplane did not get screwed up by well-meaning people who were only trying to help because, you know, that's a very common disease in advanced design of any sort, and you probably know that in your computer experience.
J.R. Warmkessel: I do indeed. Any other good F117 stories that would be worth sharing with us?
Alan Brown: Well, the most interesting one, technically, to me is when a young Air Force Captain, John Beasely, was flying the airplane, and he had just taken off, flying along, and you should know that the airplane, you know, although my boss Ben Rich promised the Air Force it would be ready in 18 months, in fact, it did take two and a half years to first flight and I figured out later that Ben's technique is just to come up with a number which he feels is extremely optimistic and then expect you to at least get at close as you can. Two and half years for that airplane, compared with the more conventional F15 which took five years for exactly the same development, from pencil on paper to first flight. So we beat the F15 by a factor of two on an airplane which was dramatically revolutionary rather than just evolutionary.
OK. So continuing on, about that, what that means is that when you get an airplane in the air for the military you haven't gone through all the structural testing and fatigue testing and all that stuff that you would go through with a commercial airplane before it gets into service because you figure Air Force pilots can look after themselves.
John Beasley is flying the airplane and his cover is flying a T38, another Air Force major calls in and says "I think John, you had better return to base. Your right hand fin just broke off at the root and is not with you anymore." The interesting thing about that is that was a major part of the airplane that fell off. It fell off, we determined later, and in fact fairly quickly because it was obviously, by sonic fatigue from the very high-frequency energy that was coming from a letterbox shaped nozzle. The nozzle was only six inches high by six feet wide and generates a very high-frequency noise and that had broken the fin off, and, of course, we hadn't done all of the fatigue testing and sonic testing that you would normally do with a commercial vehicle because of the interest in getting the plane in the air as soon as possible.
The interesting thing, of course, to John Beasley was that he had no indication whatsoever that anything had happened to his airplane because the computer automatically corrected for what he wanted because basically the way an unstable airplane works which is fly by wire is the pilot will move the controls to where he wants to have the airplane go if it were a conventional, stable airplane. So if he wants to do a right-climbing turn he will pull back on the stick, kick in some right rudder and some right aileron. The computer, however, says, looks at where the controls are put, says "I know what you want the airplane to do, turns out the controls are to do something entirely different, so I'll make them do what you want". So the computer system, fly by wire, moves the controls in such a way that the airplane still does what a stable airplane would do relative to the pilot but, in fact, the control surfaces are being operated dramatically differently. A good example would be the pilot wants to climb. He would pull back on the stick and then probably hold the stick as he's climbing If the airplane is unstable in pitch, once he starts into that curvature the airplane would continue and it would go over on its back as quickly as possible. So what the computer does is when he sees the stick go back it will start the elevators moving to make a climb and then the elevators will abruptly go in the opposite direction to hold it in the direction its going rather than keep the stick back to keep going up. So the computer knows better than the pilot does what it has to do.
So, we gave that sort of a lot of thought and thought "Well geez, suppose something does fall off the airplane. How are we going to notify the pilot?" This is about the time that Toyota came out with this nice soft-spoken lady to say "The passenger door is open" or something like that. You know, should we have somebody say "Pilot, your left wing just fell off"? You know, something of that sort, but decided something like that was a good idea because you couldn't cover every situation.
Anyway, so the obvious answer was to stiffen up the fins so they wouldn't fall off anyway which is what we did. But it was an interesting problem that we honestly hadn't thought about, the fact that something can happen to the airplane and the pilot may not be aware of it.
J.R. Warmkessel: The next airplane of the stealth interest is the F22.
Alan Brown: Right.
J.R. Warmkessel: Can you talk a little bit about that?
Alan Brown: Yeah, and in fact, I'll jump between that and I'll now mention the name that did escape me earlier, John Cashion, who was my opposite number on the B2 airplane and also on the DARPA XST. John Cashion was the principal radar cross-section designer on the B2 bomber which Northrop got a contract for about six or seven years after we did the F117. By that time, both Lockheed and Northrop had perfected curve surface analysis for radar cross-section whereas in the F117 days we only had flat surface analysis. So that was a big step forward.
Furthermore, the Air Force had got a lot more confident about how radar cross-section worked, and it wasn't just magic and we weren't just lucky, there was a lot more people in the Air Force now understand the mathematics of it. So understanding that, and the much-improved operational analysis capability showed that you didn't have to have the very high levels of sweep-back that the F-117 had, but you could do an operation analysis where the radar cross section might be different from an F-117 at different angles, but you design the airplane to be 100% survivable. So the name of the game is really survivability rather than radar cross section per say. Now of course you do certainly need a very low radar cross section but not necessarily identical to what the F-117 had, so the B2 the big development there was the curved surface and improved Ops analysis and the Air force's higher level of confidence that people actually knew what they were doing. And so the B2 was actually designed with a much smaller sweep angle, it has 35 degrees sweep on all its wing surfaces. It has curved surfaces which makes it more aerodynamic the less sweep back angle meant it could have much higher aspect ratio, so the B2 very quickly becomes almost comparable in capability to the B-52 in terms of the weapons it could carry and its range payload capability. Whereas the F-117 without any doubt paid a very significant aerodynamic penalty in order to achieve stealth. So we operated in the 70s where we paid we were able to design an airplane that could do the job but it used a lot more fuel to do it than say a conventional airplane like the A10 would have. So our basic efficiency was a lot less, perhaps about 30% less than a more conventional airplane. By the time you get to the B-2 it's just about breaking even with what a B-52 could do in terms of acronymic efficiency and range payload and it's 100% survivable. By the time you get to the F-22 which is the 1990's it was the late 1989 when it won the prototype competition against Northrop. It did its test flying in the 90's and went into service in early 2000's. That airplane now was not only 100% survivable like the F-117 and the B2 but also was 20% better than any contemporary fighter of its time. And I think it still holds that situation. So in other words you've gone through from paying a penalty for stealth to breaking even to getting stealth and being a better airplane as well, so that's the path that we've followed over the last 20 25 years.
J.R. Warmkessel: Absolutely amazing, absolutely amazing. Take a moment and reflect that you started your career at the very beginning of radar and now you've defeated radar.
Alan Brown: First of all in the 1950's and 60's people certainly understood the basic mechanics of how radar worked. But there was no good quantitative way of calculating what the radar cross section of an airplane would be. Simple shapes like cylinders and cones were amenable to calculation and in fact books were written about radar in the 1960's which covered those very simple shapes. But it took the development of what we called physical optics and then later physical theory diffraction in the early 1970s combined with very much improved computer capability because the size of the calculation was so enormous that even if people understood them back in the 50's there was no way that they could get an answer within a hundred years. So the combination of high powered computing , developing the calculations which were then good for a complete airplane was the big jump and for the F-117 to be successful that had to be combined with two other things. One is successful fly by wire of a totally unstable airplane and also precision weaponry which allows you to use that airplane in the most effective way. So the three things all came together with the F-117 which was to some extent fortuitous that because in early days you maybe had one out of the three but not all three together. The F-16 was the first airplane, military airplane, to do successful flyby wire and that was still just in the longitude direction and not in lateral. And precision weaponry really came along in parallel with the development of the 117 in the late 70's and early 80's.
J.R. Warmkessel: Is there any other good stories that I should ask you about now?
Alan Brown: Yeah I've got a good one for you. After the airplane was officially released I mentioned I saw a picture of it in Brazil. It wasn't long after that before people in electrical engineering departments at major universities who were teaching electromagnetics came to the Air force and said you know we've been teaching electromagnetics but we don't know a darn thing about this radar cross section stuff. And we in fact were hiring people who were PhDs coming right out of school very bright people and we'd have to spend about 6 months going through the rudiments about what radar cross section was all about and what kind of work they'd be doing. Because clearly it was a top secret thing then and they would know nothing about it in school, but they were very well schooled so they picked it up fast enough but still it was a new thing. So what they asked the Air force, the university people said, is it possible that we can have somebody come and give us some initial briefings just to get us started on the right track of what we should be doing. So we don't spend a lot of time going over stuff which you guys have obviously all done. So they sent out, the Air force asked two people to go out and do this. One was John Cashon from Northrop who I mentioned earlier and the other was myself. And we went to a variety of universities and one of them that I went to was UCLA. Now I've got to back up a little bit on the theory of that we developed on the theory of that we developed for calculating radar cross section. As I mentioned it was all based on the radar return from a flat plate. If you illuminate a flat plate at any arbitrary angle with a radar which is coming in with a uniform electromagnetic distribution then any element on that flat plate will re radiate identically to a first approximation which is caused physical optics. Now in real life it's not quite that simple because the edges of the plate have the energy flow around them and those flowing around the corner generates another signal which we call diffraction. Just like diffraction optically. And that diffraction term then gives a stronger signal than the signal that comes from the main part of the plate. So all around the periphery of the plate if you were to illuminate it like a window frame you get a bigger return from the frame than you do from the body of the plate in the middle. But we didn't know how to calculate that. Now the Air force has a group of people called foreign technology division also at Wright Patterson Air force Base whose sole job is to look at foreign documents, determine if they're of interest, translate them, and get them out to our military and our military contractors. And one of the ones that they did, and they translated this in 1971 was the work by a Russian physicist predominately done initially for his PhD thesis had at I think the Institute of Theoretical Physics in Moscow, I'm not sure the exact title of the institution, and his name was Peter Yovincet. And he came up with the theory of physical diffraction which is one step beyond our physical optics. The Air force translated it in 1971 because this material had never been kept classified by the Soviet Union. The most likely reason was twofold. One is I mentioned earlier you have to drop the radar cross section by at least a 1,000 to maybe 10,000 to be really effective against modern radars compared to conventional aircraft. That will probably be seen to be a ridiculous impossibility at best. So that was probably the main reason. The second reason, of course, is that Russia's predominant military posture was defense of mother Russia more than designing airplanes to be able to penetrate American air space. They were thinking more in terms of ICBMs for that job anyway. So there's probably less interest in the subject than we had, and also a general feeling that it wasn't practical anyway. If I put into a factor of 10,000 and into perspective, the one that I used when I would talk about this to students is imagine you work at the Ford Motor Company and your boss says "The world is running out of gasoline. We're going to have to do something dramatic. Currently our best gasoline cars are doing about 30 miles to the gallon. I want you to multiply that by a factor of 10,000 to come up with 300,000 miles to the gallon." That amounts to going around the world about 12 times on a gallon of gas. Well clearly they'd drop out the guy in a white suit and take him off to the local hospital because that sounds ludicrous and that's really what we were asked to be doing in radar cross-section, a factor of 10,000 over conventional airplanes. I'm not at liberty to say what we did to say what we achieved, but, I mean, clearly we must have done quite well. OK. So Peter Yovincet come up with his physical theory diffraction and it was published, and the reason I wrote this long description about the Soviets, it was published in their equivalent of the monthly unclassified IEEE magazine. It stood for electrical engineers. Given that, our people were able to find it fairly readily, translated it. Dennis Overholser, our guy who had developed, or done a good part of the development as a young person, young computer engineer was an older man called Bill Schroeder. Bill Schroeder was the guy who had the initial ideas. Dennis put them into practice on the computer and Dennis realized that he could take that physical diffraction, add it into his physical optics and get a much more precise answer for the flat plates that he was dealing with. It turned out in practice that that wasn't quite as important as he thought at first because we were putting absorbing material around the edges of all our plates anyway which tend to knock out the diffraction term and bring it back to the physical optics thing, but nevertheless it was very good to be able to have this exact theory and that became what we called our Echo 2 Program and that essentially has been used as the basis of design for all our airplanes through the F22 and the F35 and there's not many changes.
OK. The university professors here are asking for help and I'm at UCLA to give a talk. The guy who was the technical editor, west coast, for Aviation Week, which is our major, professional aviation magazine calls me and says "I hear on the grapevine that you're going to be giving a talk at UCLA on stuff. Is it OK if I show up?" So I said "Sure, that's fine, but just bear in mind I may not be able to answer all your questions because I'm very limited in what I'm allowed to say. I've got a group of subjects set aside and this is all I'm telling the people because clearly a lot of the stuff is still classified." So I got to UCLA and it's a Friday afternoon to a typical bunch of graduate students, coffee and donuts, and there's the head of the department, there's the Aviation Week tech editor and there's another guy who is about the same age as myself, at his early 60s at that time with a foreign accent and I'm introduced him and it's Peter Yovincet, this same guy who's written a PhD thesis on physical theory diffraction back in the 1960s. So, of course, I had to modify my talk and recognize that the guy who was responsible for about 1/3 of our analysis work was sitting right here in the audience and it was interesting to see his reaction. At first glance, you know, it was pretty clear that he thought he was thinking to himself that the enemy is using my stuff and as the Cold War was now over, this is 1991, and he's over at UCLA on a sabbatical, his second thought was "Well at least somebody is using my stuff" and it turns out that Peter and I become firm friends. He and his family have stayed in his house for a weekend and so we're, he's a good friend of mine and of course he never went back to Russia. The Russian scientific scene in the early 90s was a disaster and he had a wife and two young boys and his family, so he stayed here and of course he had a struggle financially but he's still surviving and we both keep in touch with each other in our 80s.
OK. Another story for you, the CIA was who, of course, we worked quite closely, even though it was basically an Air Force Program, the CIA came to us in the mid-80s and said "We've got something that we want you to look at and we're asking John Cashion in as well", the same guy who'd done the university work and the B2 work, "We're asking him as well. We have been sent over a European full-scale model airplane to be tested at our range at White Sands Missile Range because there's nowhere in Europe that's capable of doing these kinds of tests with that kind of sensitivity and they want to know what it's about And they have made us swear on a stack of Bibles that we will keep this absolutely under cover, so I'm only telling you two guys, three milliseconds after we were asked about it, we want you guys to come to White Sands and have you guys take a look at this airplane, tell us all about what you think about it".
So we went over. John Cashion's background, he's a PhD in electrical engineering and he's much stronger in electromagnetics and radar design than I am, but I'm probably broader because I've got the airplane design background as well. So between us we made a pretty good team to go over the airplane. And the interesting thing about it, about its design, was that first of all, it had some very interesting special features in the inlet system based on what we called helmholtz resonator which are techniques where you use a resonance approach to essentially cancel out the return, much as we often do in acoustics. Helicopter blades are often quietened down or repeatedly quietened down by having an acoustic signal generator that will send something that is out of phase with the the blade's revolution and the helmholtz resonator inside the inlets, did this kind of thing. It's somewhat limited in frequency but there are a number of items on the airplane that certainly made me think, and I had a pretty good experience at both university and the practical business of airplane manufacture. I said, "You know, this has been done in a university." The design is basically a university design and furthermore it's interesting to note that the university people very typically of Europe have not been in close communication with the manufacturing people in building it because there's some things in the manufacturing which, you know, were just not up to the mark that I would have expected if the two people had been in close cooperation which, at the Skunk Works , that was one of the ways in which we operated. The design department in the Skunk Works was adjacent to the manufacturing floor. You just went through one door and people came back and forth, manufacturing guys would complain to the designers, designers would be on the floor seeing how their stuff was being built, very close.
This airplane, as I say, had some interesting academic features, but also had some very bad construction features that indicated that the guys who knew about radar cross-section were not on the shop floor talking to the workers. I said "This has got to be being down at some place like Gottingen university in Germany. This is typical, you know, German university attitude" and sure enough I found out that was my guess. I wasn't told whether I was right or wrong, but I said "This is being done by, its designed by a German university. It's probably being built by some company like Messerschmitt or whatever and they haven't talked to each other as much as they should."
Many years later I found out that was absolutely true and that same airplane is now in one of the science museums in Munich as an exhibit. It never came to anything and presumably when they actually tried to fly it and test it, it probably wasn't up to the mark that the original people expected.
So those are two stories about the F117 background which is quite interesting.
J.R. Warmkessel: Those are fantastic stories. One more question for you. As an aerodynamicist, it has to do with the wing shape on a stunt plane where you have a symmetrical wing, forward and back and there's always the question-
Alan Brown: You're talking about modern aerobatic aero planes.
J.R. Warmkessel: Affirmative, yeah.
Alan Brown: Like Extra 300s or things like that.
J.R. Warmkessel: So how do they get left? I mean-
Alan Brown: Oh, OK. First of all, it's a long story and it will take a lot of time and I've actually written several articles on exactly that subject, published, and the reason I've written them is that there have been some absolutely ludicrous articles written in magazines you'd expect to have some validity like the EAA Journal and I've seen things like PhDs, probably not PhDs in aeronautical engineering, written, which are awfully stupid, and what they tend to say is there's two things that have been rather interesting. One is a guy came up with an article that proved conclusively that current aerodynamic theory predicts the Cessna 170 cannot generate enough lift to get it off the ground. He made the mistake, as do a lot of people, that airflow on a wing separates right at the leading edge and goes round the top and bottom of the wing and then goes back and the molecules somehow know each other very well, have very great insight and meet each other at the trading edge. Now the correct answer is a symmetrical airflow at 0 degrees does, in fact, not generate any lift. As soon as I pop that up to angle like, say 10 degrees, the airflow that comes towards it, because its flying at subsonic speeds, the presence of the wing sends sound signals forward faster than the airplanes flying and so the air is aware that something's coming towards, because acoustic waves being propagated from the air foil. It actually stops turning before it reaches the air foil and the point at which it divides is not at the leading age, but somewhere further back underneath. So, that air now has to go that hits underneath side, has to now turn around, go very quickly round the leading edge at very high speed. The streamlines narrow down tremendously and by mass conservation if the area gets less, the velocity has to go up by the bernoulli theorem, if the velocity goes up the pressure has to go down so you et a very high lift generated on that top surface when the air goes round that corner. Now, you can quickly recognize that what I'm saying is true about not separating the leading edge but at a point underneath by thinking about how store warning devices on a lot of general aviation aircraft. There's a little sticking out flap down below the airplane, maybe around 10% of the way back on the air foil underneath.
That little flap has got a hinge on it, and when it's blown backwards because the airflow separation is forward at that point it’s blown backwards and there's no stall warning. When the airflow, as you go to higher and higher angles the separation point, dividing point, I should say, moves further back. clearly if you're at 90 degrees a dividing point would be in the middle of the air foil.
So as it moves back it gets to a point where it gets behind this little flap and the flap gets blown forward. When it blows forward it closes a switch which tells the pilot he's within a couple of degrees of stall and it either sounds a buzzer or a red light or whatever, but the store warning device is a pivoting device underneath the air foil which works on the principal that is you get the higher angles the dividing streamline moves backward under the air foil and is not starting at the leading edge. Does that sound like it answers the question?
J.R. Warmkessel: That's a fantastic answer.
Alan Brown: Another F117 story, I was in Lancaster in the Antelope Valley for an Air Force celebration. I think it was so many, like 100,000 hours of flying of the F117 and it was quite a group of us, both Lockheed and US Air Force folks there, and at the end of dinner, just about finishing off dinner, and a young woman came up to me, put her arms around me and said "I want to thank you for designing the airplane that brought my husband safely home from the Gulf War in Iraq" and that, of course, is when it really strikes you about the, you know, how important the US air Force is to us and by comparison how very simple my job is, just me sitting at a desk and working on airplane design. It turned out that her husband, whom I met later in the evening, was the man who famously dropped the bomb down the air conditioning shaft of the basic systems control unit in Baghdad and I met, I have had the good fortune to meet quite a number of the US Air Force pilots who have seen active service with the airplane.
J.R. Warmkessel: As a pilot myself, we always appreciate the efforts of all of our armed forces and people who built the tools that they need to do their jobs.
Alan Brown: Absolutely. Yes, very, very important. The other story is on the same lines and that is that on a major anniversary of the F117, I think it was the 20th anniversary of it being in service so it would be around about 2004, I went out to Holloman Air Force base where the main squadron's F117 was based, and was fortunate enough to be one of the very few civilians who went to a full scale US Air Force dress dinner at the end of a couple of days proceedings which included a flyover of 25 of the F117s in formation and it's very interesting that we kept so many airplanes always available in an operational readiness when in fact I think the F117s operational readiness was equal or better to that of any of the other more conventional units in the US Air Force and that, of course, is a little bit different from what some of the pessimists were first forecasting and they felt that it was such an exotic airplane it would be spending 90% of its time grounded being constantly maintained and that was not the case. It's been a very good operational airplane and it was a pleasure to be one of the very small number of civilians who shared in that Air Force reunion with several generals including those who were involved in the Gulf War and those who were involved in the early days of the first operational flying and so on.
J.R. Warmkessel: That's good. And with that I think we're going to call it a night Alan.
Alan Brown: Good. OK. What time-
J.R. Warmkessel: Fantastic conversation.
Alan Brown: It's 10 past 10 now. It's only been three hours.
J.R. Warmkessel: Well, we'll break it up into a couple of episodes, but Alan thank you so much once again.
Alan Brown: OK. Pleasure.
Direct link to mp3 audio file of show (right-click to download/save).