in Germany at that time. It is not a point, as Millikan says,of nationality.Let's look at some of the important developments of thepostwar period. First, there is the so-called sound barrier(Figure 3). I object to the word "barrier" because it impliesprecisely a kind of collisional approach. It has nothing todo with barrier; there is no barrier, there is nothing there.There is just air, like anywhere else. The point is, that if youget near the speed of sound to about 0.7 Mach, then underthose circumstances the drag coefficient on the airfoil increasesvery steeply, exponentially, until you in fact reachthe speed of sound.The reason for this is that through the development ofshock waves, which affect the airflow over the airfoil, acertain amount of the lift energy is converted into shockformation. That energy is taken away from the lift capabilityof the plane, and under those circumstances you experiencevarious kinds of instabilities and difficulties with theplane itself, which have to be countered simply from thestandpoint of understanding the problem. You have to makethe kind of geometrical adjustments in wing design, or anythingelse, that are necessary to do that.One of the principal adjustments in wing design that canbe made was invented by Busemann, the so-called sweptwingdesign, the arrow design. You can see (Figure 4) howthe critical zone for the development of shock waves thatinfluence flight and lift negatively is at the 0° angle; that is,if the wing is at right angles with the fuselage, you get theonset of the critical area at 0.7 Mach and then the dragcoefficient declines afterwards.If you have a 60° angle of the wings, then not even halfthe drag coefficient develops and you get it also at a muchlater point; namely, beyond Mach 1. <strong>And</strong> if you have a 70°inclination with the fuselage of the wings, then you get to apoint where you get a very low, very late onset of the criticalphase. Also, the amount of reduction in lift or the amountof increase in the drag coefficient is not very substantial. Itis there, it will always be there, because Shockwaves form.Shock waves are real, as was certainly determined bythese methods of research in aerodynamics carried out inthe 1930s and 1940s in Germany primarily under Busemann'sdirection in Braunschweig. They are not what Rayleighhad said, when he criticized Riemann's 1859 paper.He said, shock waves do not exist. What exist, are singularitiesin the mathematical formulation of the wave equations,but we cannot assign any reality to such singularities.All it means, is, that we have failed to come up with a solution.As Riemann said, these things are real, and he said it 50years before Rayleigh made that idiotic criticism. It wasprecisely because of the realization that the shock wavesare real that they were taken into account when supersonicFigure 3THE SO-CALLED SOUND BARRIERThe sound barrier has nothing to dowith a barrier. When a plane gets nearthe speed of sound to about 0.7 Mach,the drag coefficient on the airfoil increasesvery steeply because shockwaves develop that affect airflow overthe airfoil.Figure 4SHOCK WAVES THAT INFLUENCEFLIGHTThe critical zone for the developmentof shock waves that negatively influenceflight and lift is at the 0° angle. Ifthe wing of the plane is at right angleswith the fuselage, you get the onset ofthe critical area at 0.7 Mach. But if thewings are at a 60° angle, then not evenhalf the drag coefficient develops. Witha 70° inclination, there is a very low,very late onset of the critical phase.Figure 5THE D-558 DESIGN BEFOREAND AFTER BUSEMANNThe Douglas D-558, which wasdeveloped simultaneously withthe Bell X-7 as a supersonic designin 1945, had a conventionalstraight wing (a). After von Karmanand others visited Germanyand interviewed Adolf Busemann,their design was modifiedto his swept-wing design (b).48 January-February 1986 FUSION
flight was studied in supersonic wind tunnels in Gottingenand Braunschweig, and later on in Munich and Lake Kochel.An interesting example is the Douglas D-558, which wasdeveloped simultaneously with the Bell X-1 as a supersonicdesign in 1945 before the von Karman mission went to Germanyand interviewed Busemann and others. Figure 5(a)shows their design before the trip to Germany: a straightwing sticking out, so you have the 0° angle situation ofbefore and a tail end that sticks up, just as in the old designsof aircraft in the subsonic range.Then von Karman and others came back to the UnitedStates in the summer of 1945, and after that the D-558 lookedlike Figure 5(b). All of a sudden it became a swept-wingmodel with a swept-tail configuration.Several years ago when I was visiting a scientific conferencein Moscow, a Russian researcher showed me a pictureof one of the models for a supersonic passenger jet typethat the Russians had acquired when they moved into theeastern part of Germany. "What do you think that is? hesaid. I said, "Everybody knows, that's the Concorde." But itwas not the Concorde, it was a model built by Busemannfor a supersonic jet—in the late 1930s—to which the Concordedesign is identical.There is no mystery of any kind involved here. It is asimple and straightforward story. It's a question of method,both scientific method and method of organization. It's aquestion of assembling the kind of scientific team that iscapable, on the basis of the right kind of methodologicalapproach, to find the mode of organization most appropiateto its goals. <strong>And</strong> these goals have to be set never withregard to so-called state-of-the-art designs, but in fact as farbeyond as you can possibly do.To the extent that you do that, you will be capable ofchanging this so-called state of the art rather than beingstuck with it. What we have to do in any program, whetherit is a crash technological development program or a basicresearch program, is to set our sight on the kind of goalsand tasks that are way beyond what we initially anticipatethe most immediate goal of the program to be. If that is notdone, then we will not confront ourselves with the type ofchallenge that is necessary in order for the scientific enterpriseto succeed.The lesson to be learned is that we do not need state-ofthe-artprograms; this is nonsense and leads to preciselythe wrong approach. The cheapest programs are not stateof-the-artassembly programs; the cheapest programs willalways prove to be those crash programs that look as farahead as possible in order to accomplish the immediatetask. This may appear to be quite expensive in the long run,bringing in basic research and technology and design togetherinto a program, rather than just doing the state ofthe art on the basis of what is on the shelf. The latter is goingto be the most expensive and the least workable approach,and I am afraid, to a significant extent, when we are talkingabout the SDI today, it is precisely that kind of approach tothe situation, that is most problematical.Von Neumann's Cost-Benefit NonsenseI have to mention one other villain who had somethingto do not so much with the scientific side of these developments,but had a tremendous influence on this organizationalside, John von Neumann. Von Neumann was anotherHungarian-born mathematician who studied at Gottingenand later came to the United States in the 1930s.In the minds of most, von Neumann is associated not somuch with his mathematics and physics, but rather with hisideas in economic theory. In particular he wrote a bookalong with Morgenstern called The Theory of Games andEconomic Behavior, viewing economic development essentiallyas a kind of competitive game between playersmuch like a poker game. In fact the first paper von Neumannwrote on so-called economics in 1928 was "The Theoryof Parlor Games."The next thing he studied in order to be able to modeleconomic development in the late 1920s was poker, and heinvented a simplified version of stud poker and abstractedfrom a simplified version of stud poker his basic ideas ofeconomic development. Don't underestimate the influenceof this nonsense. What has come out of that is theRand Corporation, the Air Force Systems Command, andevery single bit of so-called cost-benefit analysis optimizationnonsense that we are suffering from today. It is one ofthe principal problems that we have to solve in order todefine and push through the kind of crash program for theSDI that is desirable.The other thing that has come out of von Neumann'spoker theory is the famous Robert McNamara way of "winning"the Vietnam War. You remember what that was: thebody count method—cost benefit analysis applied to militarystrategy and tactics. Most of you were probably treatedto this every night on TV: You had a body count, so manyVietcongs, so many North Vietnamese killed, so manyAmericans killed; the ratio looks good.The McNamara crowd made detailed analyses of howmany people exist in each age group in Vietnam to see howmany people were being eliminated per day, and then askedthe question, how many troops do we have to put in to winon the basis of cost-benefit analysis? How much do we getout of it if we put so many soldiers, so many tanks, so manythis and the other things in. From the standpoint of linearprogramming and optimization analysis, how do we win?You can't win that way.The principal strategic problem in military terms and otherwisein politics is the principle of the flank. The principleof the flank defies by its very definition the idea of costbenefitanalysis, and this has precisely to do with the unexpected—toput a tremendous amount of cost into onearea where it is unexpected, in order to be able to thensucceed as quickly as possible. The very opposite of thekind of thinking so much associated with von Neumannand much of the Pentagon thinking today is what is calledfor under these circumstances.If we keep that in mind, and let that be reflected in ourpolitical approach to these questions, we may have a chance.Uwe Parpart Henke is the director of research for theFusion Energy Foundation. This article is adapted from hispresentation at the Krafft Ehricke Memorial Conference inJune 1985, sponsored by the FEF and the Schiller Institute.FUSION January-February 1986 49