The central message of Walter McDougall’s long, long history of the space age concerns the danger of seeing space exploration primarily as a symbol for something else. Yes, space will always be symbolic in the grandest sense: in attempting to understand it, we demonstrate man’s “questing spirit,” our demand to know the unknown, our curiosity about our place in the infinite universe, and other impulses that may sound ironic but are the real reasons we continue to fire rockets into the sky.
But when the US began space exploration in the 1950s, McDougall says, it was in thrall to a narrower and less worthy symbolism. The “space race” was just one more way to battle the Russians in the momentous struggle for international “prestige.” The Soviet Union had scored a shocking early victory by launching Sputnik in 1957. From that moment until Neil Armstrong’s first footstep on the moon twelve years later, American policy was less concerned with space itself than with the space race against the Russians.
For this approach America paid several penalties, McDougall writes. It permitted itself to be worked up into a wartime crisis mentality, ratcheting the central government’s control of scientific research to a level unprecedented even during the Second World War. And it guaranteed that once the race to the moon was over and the Russians had clearly lost, space would stop being as interesting, like last year’s pennant race. (This was also the theme of the only book about space even longer and denser than McDougall’s: James Michener’s elephantine Space, published in 1982, which lamented America’s eagerness to duck out of the space race once it started racing only against itself.) It was as if the court of Aragon and Castille, having “beaten” the Portuguese with Columbus’s landing in the West Indies, declared the “America race” over and its interest in exploration at an end.
The analogy is not perfect—the conquistadores had few purely scientific concerns in mind as they looted and proselytized their way from Florida to Patagonia, and there were more obvious riches to be had in Peru than on the moon—but the importance of their New World colonies to Spain and Portugal is a reminder that exploration can have tangible consequences. The exploration of space has, so far, had nothing like the practical effect of the opening of the Americas. How it will look a thousand years from now, or even a hundred, it is impossible to say. But even now, space provides a powerful, illuminating example, not just a “symbol,” of why America has faltered in the more down-to-earth forms of technical and economic competition.
For forty years, the US has tried to compete simultaneously with the Soviet Union and with Japan—“Japan” being today’s shorthand for other technically advanced market economies. For the last fifteen years, America has found this double race increasingly onerous, and in the last five years it has all but collapsed from the effort, as the defense budget and the trade deficit have chased each other to record sizes. Usually the two forms of competition occur in separate spheres, but space is the rare field in which the US competes with market economies and the Soviet command economy at the same time.
For several years, but especially since the Challenger explosion, the Reagan administration, has pushed hard to “privatize” space exploration—to get NASA, with its subsidized payloads, out of the way so that new, more flexible companies can emerge and thrive. But this new industry, like many old ones, is being “freed” from one kind of government control only to be dominated by another, that of the military budget. If the US can rely on military spending to stimulate true industrial advances, then our space plans may well succeed. If not, space will become a large-scale illustration of habits we must change to survive. The steel and auto industries are the most familiar illustrations of businesses in which the United States has lost its accustomed lead. Space may be a more important case, because it more directly demonstrates what happens when military spending becomes the dominant form of “industrial policy.”
In the months since the Challenger explosion last January, the familiar, stylized drama of previous manned flights—pre-launch jitters, midmission crisis, hair-breadth escape, and the inevitable happy ending—has been succeeded by the increasingly common pattern of social and industrial disarray. There have been the discoveries of shoddy workmanship and inattention to detail, the air of generalized low-level dishonesty, the rapid transition from seemingly unquestioned leadership to a scramble to catch up. Finally there have been the lawyers and politicians who try to sort out the spoils while industrious foreigners keep plugging ahead. (The families of several of the Challenger astronauts are filing negligence suits against NASA and/or Morton Thiokol, manufacturer of the defective solid-rocket booster that caused the Challenger to blow up. One crucial point of fact in the suits is how long the crew lived, consciously anticipating death, before the cabin hit the water at two hundred miles per hour. The emergency air pack of the pilot, Michael Smith, whose family is suing NASA for negligence, was reportedly three-fourths used up when discovered deep undersea.)
Even as the Rogers Commission—the thirteen-member commission appointed by President Reagan to investigate the shuttle explosion—was preparing its report on Challenger this summer, the Russians were completing the first-ever crew exchange between one space station and another, the Japanese conducted a flawless test of their first mainly made-in-Japan rocket, and a flotilla of European, Soviet, and Japanese spacecraft held a rendezvous with Halley’s Comet, without any participation from the United States. Even the Chinese, last heard from on the world-technology front with the bare-foot-doctor concept, have been booking commercial satellites onto their Long March 2 and 3 rockets,1 which now have more confirmed launch reservations than all fledgling American private-launch companies combined. (Not to overstate things: the Chinese have received deposits from two customers, American companies from one.) The director of the Great Wall Industry Corporation, promoter of the Long March rockets, has been trying to drum up customers in America and Europe, telling one and all that “everything is negotiable.” (If this had been reported in The New York Times and in a Doonesbury cartoon, in which place would it have seemed more appropriate?)
The prelude to the current disheartening situation is the subject of McDougall’s book. He concentrates mainly on the 1950s, showing how the increasingly desperate “race” against the Soviet Union drove America toward a more centrally controlled, “technocratic” society, emulating the Soviet Union in mild degree in order to compete with it. During World War II the government had requisitioned and deployed money, talent, and scientific effort—but that was all-out war. After the launch of Sputnik, McDougall says, many of the same traits became part of normal life. During World War II, the cost of government-sponsored research and development remained below 1 percent of the gross national product—despite the Manhattan Project, despite all the scientists drafted to work on radar and bombs. During the 1950s it climbed rapidly, largely driven by the space program, until it reached 3 percent of the GNP in 1964, the same time NASA’s budget peaked. Since then, government-financed research has fallen back to about 2.3 percent of the GNP.
The hero of McDougall’s tale is Dwight Eisenhower, who saw no reason to get so worked up about beating the Russians into space. McDougall argues, persuasively, that the US could easily have spared itself the shock of Sputnik by sending up a satellite well before the Russians did. That would have required merely unleashing Wernher von Braun and his associates at the Army’s Redstone arsenal in Huntsville, Alabama, who by the early 1950s had a rocket poised for rapid development. Instead, Eisenhower deliberately chose the Navy’s Viking rocket for satellites, even though it was smaller, less completely developed, and bound to take longer before a first launch. In part, McDougall says, the administration’s desire to be first in space was over-shadowed by other strategic concerns. It was eager to establish the “freedom of space” principle, so that the US could conduct aerial reconnaissance of Soviet installations. (“Freedom of space” was less vital to the Russians, who had other ways to find out about America’s military activities.) The odds that the Soviet Union would object would be lower if the Russians themselves flew first than if a US Army missile, designed largely by rehabilitated Nazis, sent the first satellite into space.
Eisenhower had more basic hesitations about rushing into space, McDougall says. He was deeply fearful of too rapidly mobilizing the economy, of seeming to put the economy on a wartime basis, and of encouraging a wartime mentality. As it turned out, his relaxed, laissez-faire approach backfired: after Sputnik Americans were panicked and embraced precisely the steps Eisenhower hoped to forestall. McDougall says:
From 1945 to 1960 Truman and Eisenhower searched for a means of deterring the Cold War enemy without the United States itself becoming another garrison state…. Ironically, the presidential efforts to keep a lid on spending, to keep open the option of a negotiated arms freeze, to preserve the United States as a symbol of free inquiry and international cooperation, and to cope with a secret, singleminded technocratic adversary without giving in to paranoia all contributed to the failure of the United States to be first into space. And that celebrated failure did more than anything to defeat Truman’s and Eisenhower’s hopes of adjusting to the Cold War without transforming American government.
The villain, by corresponding logic, was John F. Kennedy, who in staking national prestige on a race to the moon naturally increased the government’s role as central research planner. “The Kennedy call to arms amounted to a plea that Americans, while retaining their free institutions, bow to a far more pervasive mobilization by government, in the name of progress.” According to McDougall, the position of James Webb, director of NASA under Kennedy, was that “the space program required nothing less than the mobilization of the nation to a war footing in peacetime.”
McDougall argues, in short, that the space race gave emotional and political legitimacy to a militarization of American life that was already impending because of other cold war pressures. But while militarization of any sort increases the government’s power—which is McDougall’s fundamental objection to it—not all military efforts are the same. Some of them succeed. The Second World War was a concentrated total effort with a specific goal, which it achieved. So was the space undertaking that most resembled it, the Apollo program, with its clear timetable, objective, and success in reaching the moon. But America’s post-Apollo space efforts, to which McDougall devotes only the last couple of dozen pages of his book, have been different, as the postwar American military establishment has differed from the force that won World War II. The post-Apollo space program—which is to say, the space shuttle—has resembled the “peacetime” American military of the last fifteen years. Its missions have been unclear, its public and political support inconsistent, its standards of success uncertain.
In part a resemblance between the space program and the military is unavoidable and natural: during the last fifteen years, military programs have made up an even larger share of America’s efforts in space. The American military relies on satellites for communication, navigation, and worldwide surveillance. Last year, Paul B. Stares of the Brookings Institution published The Militarization of Space: US Policy, 1945–1984,2 which described in detail the evolution of military space programs. Department of Defense spending for space programs started slowly, and during the Apollo years was less than a third of NASA’s. Just after Ronald Reagan’s inauguration the Department of Defense’s space budget exceeded NASA’s for the first time. Under the influence of the “Star Wars” Strategic Defense Initiative, the military now spends twice as much on space as NASA does, and the margin is growing. NASA’s budget is about $7.5 billion—the military’s budget for space is about $15 billion.
The indirectly caused resemblance between space and the military is more interesting and important, because it parallels the effect of the military on other technical and economic activities. Without clear, World War II or Apollo-type external measures of success, organizations often drift into a concentration on their own internal convenience and concerns. They become intent on ways of spending money and of making life easier for themselves. This pattern has reached advanced stages in the peacetime military, and it also helps to explain why the Challenger blew up.
As government documents go, the five volumes released by the Rogers Commission are of remarkably high quality. The summary report is clear and tightly written, with only rare descents into space-ese. (“The team assumes responsibility…through post-landing crew egress and safing of the orbiter.”) The what-went-wrong analysis sustains a kind of detective-story tension as it moves through the possible sources of trouble and eliminates all suspects except the right-hand solid rocket booster. At several points in the transcript of the hearings, commission members vow not to let themselves become the modern counterparts to the Warren Commission, whose hasty report on the Kennedy assassination has nourished a generation of conspiracy buffs. They withheld nothing: for instance, before reproducing countless photos of the launch, they offer fifteen pages of diagrams showing where all the cameras were. No one will fault them for covering up.
From the one-thousand-odd pages of hearing transcript, distinct dramatic characters began to emerge. The Nobel prize-winning physicist Richard Feynman, confident in his reputation, is willing to ask elementary questions (“Excuse me, but what’s the difference between a motor and an engine?”) and bring things down to practical demonstrations, such as putting a sample of O-ring into his drinking glass to see what happened when it got cold. (“I took this stuff that you gave me and put it in ice water, and I discovered that when you put some pressure on it for a while and then undo it, it doesn’t stretch back.”) The commission’s chairman, William Rogers, is the unchanging voice of probity, sighing his disappointment with errant NASA officials and trying to keep Feynman in line.
The hearings revolved around a narrow question of technology and a broader one of organizational pathology. The technical question, of course, was why the exhaust gases from the solid-rocket booster blasted their way through one joint in its metal case, despite the rubbery O-rings that were supposed to seal the joint. As stories in The New York Times and other newspapers have made clear, the O-rings had caused worry—though evidently not enough—long before one set of them catastrophically gave way.
In order to contain the stupendous explosive forces of the solid booster, the rings had to be pushed into a sealing position by the first wave of pressure, milliseconds after ignition. Once firmly sealed, they could prevent a flow of exhaust gas from developing—but if they did not seal immediately, the flow would slip past them, gather speed, burn the rings away like a blowtorch, and eventually blast right through the metal case. NASA’s way of testing the rings involved blowing them in the wrong direction, which increased the odds against proper sealing.
In addition, just before the shuttle lifted off, its “main engines”—the three liquid-fuel motors underneath the shuttle orbiter itself—ignited first, a few seconds before the solid motors. (One reason for this sequence is that the main engines could be turned off if something went wrong, but once the solid rockets are ignited, they are about as controllable as a lit stick of dynamite.) This forced the whole booster-orbiter assembly to lean away from the vertical, which “rotated” the solid rockets—a misleading term meaning that the booster’s joints flexed and opened. This occurred at just the time the O-rings had to seal, and made it harder for them to do so.3
The system had its drawbacks even in the best of circumstances, but it proved to be fatally dependent on weather. Since the rings had to move in order to function, they had to be flexible. In laboratory tests conducted at 100 degrees Fahrenheit, the rings were so flexible and resilient that they never lost contact with surrounding surfaces through a series of compressions and expansions. But at 50 degrees, they were so much stiffer that, at the end of one ten-minute test, they still had not regained their original shape. At the time of the Challenger launch, the rings had sat through a night in which the temperature reached 26 degrees.
The scandal of the O-rings is that their failure was not an unforeseen calamity, a chance-in-a-million misadventure, but a major defect noticed and complained about for more than a year. Because they were never intended to hold back the hot exhaust gases by brute force, but rather to keep a flow path from developing and thereby prevent the worst heat the pressure from ever reaching them, any sign of charring or “thermal erosion” of the rings would mean something was not working as planned. By mid-1985, postflight inspections of the boosters recovered from the sea (in keeping with the shuttle’s claim to reusable efficiency), showed deepening damage to the O-rings, correlated with cold temperature at launch. (Of the twenty shuttle flights launched at temperatures above 65 degrees, only four showed evidence of O-ring “thermal distress.” Of the four pre-Challenger launches at 65 degrees or below, all revealed damage.) Six months before the Challenger explosion, one primary seal failed altogether, and only the second, “redundant,” O-ring kept a disaster from occurring then.
Since the O-rings were classified as “criticality one”—if they go, so does everything else—the manager of the solid-rocket booster program at Marshall Space Flight Center in Alabama, Lawrence B. Mulloy, applied a “launch constraint” because of increasingly frequent O-ring problems. But he did not tell higher-ups about the constraint, and routinely waived it for the next six flights.
By the night before the Challenger launch NASA’s concerns had grown sufficiently dull that Mulloy and other officials smothered the objections of Morton Thiokol engineers, who begged for delay because of what cold might do to the O-rings. George Hardy, deputy director of science and engineering at Marshall, said he was “appalled” by the thought of another postponement. After a few minutes’ reconsideration, Thiokol’s managers agreed that they could not prove the launch would be unsafe. (When called before the Rogers Commission, Hardy claimed that he’d only said he was “somewhat appalled.”)4
The Rogers Commission devoted much of its time to analyzing why, being aware of the O-ring danger, NASA did nothing to eliminate it. It said in its summary:
The space shuttle’s solid rocket booster problem began with the faulty design of its joint and increased as both NASA and contractor management first failed to recognize it as a problem, then failed to fix it and finally treated it as an acceptable risk.
Was the problem simply overwork? Before the go/no-go meeting for Challenger, most of the participants had been getting by on minimal sleep—but that would not explain the preceding months of neglect. Did the rushed launch reflect a crass, Deaveresque desire for a live TV hookup with the shuttle during President Reagan’s State of the Union speech? The commission said no—NASA normally ruled out any broadcasts during the first twenty-four hours of a mission, and the speech was scheduled for the very night of the launch. Was there too much pressure on the whole shuttle system? Yes, since the program was creaking and groaning in its attempt to meet an unrealistic target of twenty-four launches per year. Spare parts were running short, training time was going down, the entire system (according to the commission) probably could not handle more than twelve to fifteen launches per year. Was there too much economic pressure on Thiokol? Probably so, since its one important customer was unmistakably asking to be spared any awkward truths and told only good news.
Over-ambitious schedules, problems born of too-complex design, shortages of spare parts, a “can do” attitude that stifles embarrassing truths (“No problem, Mr. President, we can lick those Viet Cong”), and total collapse when one component unexpectedly fails—does any of this ring a bell? Of course it does, for it greatly resembles parts or all of the life histories of many major, modern military programs: the AMRAAM and Maverick missiles, Trident submarines and Aegis cruisers, the Divad anti-aircraft gun and B-1 bomber. Indeed, once the shuttle is seen as an application, outside the military, of modern military procurement principles, many of its other characteristics make sense.
The most important question about the shuttle, transcending the widespread internal silence about the O-rings, is (as John Lodgson of George Washington University put it in a recent article), “How did the United States get itself into a situation in which a single accident, however tragic its toll in human lives, could bring the vitally important national space program, with its array of critical scientific, commercial, military, and intelligence missions ready for launch, to a halt?”5 The answer is found in the original design of the shuttle and the increasingly military-like nature of the civilian space program.
The crucial decisions came during Richard Nixon’s first year in office, just after the first landing on the moon. The excitement of the Apollo program was past—the race was over and we’d won—and NASA’s budget had been declining for several years. Nixon appointed a task force on future space options, headed by Spiro Agnew. Its report offered him a variety of choices, ranging in ambitiousness from a permanently manned space station and an Apollo-type manned mission to Mars, all of which would cost atleast 10 billion a year, to a much more modest shuttle and space station program, at half the price. Nixon chose the most limited option, and the shuttle was born.
The guiding principle behind the shuttle’s design was to hold immediate development costs down—not eventual operating costs, or the total costs of the program eventually, but money spent right now, on Nixon’s watch. For instance, a shuttle using new liquid-fuel boosters would have offered potentially lower operating costs. But since it would have cost more to develop, solid boosters won out.
NASA was facing so much pressure to cut its budget that it had to be happy with any shuttle it could get. It forged a crucial nonaggression pact with the Air Force, whose own longstanding desire for a manned orbiting laboratory had been frustrated by Nixon. The Air Force would not put up its own money for the shuttle, but it would plan to send up its satellites via shuttle, and would refrain from promoting any competing launch system. A NASA/Air Force design was approved in 1972, and during the fourteen years until the explosion it moved step by step through the stages of the military-procurement disease, which are as follows:
- The Vegematic promise. Viewers of late-night TV on non-network stations are familiar with that omni-capable product, the Vegematic. Its advertisements proclaim: “It slices! It dices! It cuts carrots into curlicues and radishes into rosettes!” A similar spirit seems to have influenced the design of many weapons. One reason modern weapons rise relentlessly in price is that they’re designed to do everything at once. Fighter planes that are supposed to prevail in all circumstances, tanks theoretically “optimized” for all terrains, ships that can hold off every threat—they become so complex and expensive that they have to do everything they promise, because even today’s amply funded military can’t afford a range of weapons designed for different needs.
So too with the shuttle. To justify its construction, as Lodgson shows, NASA, the Air Force, and the president’s science advisory commission advertised it as a replacement for twelve existing launch systems. In order to handle all those duties, especially carrying military cargoes, the shuttle became larger, more cumbersome, and more complex than anything else ever flown.
The Air Force, which had been thinking all along of a vehicle that could go up quickly, inspect a trouble spot, and get right back down to earth, insisted on “cross-range capability,” so the shuttle could maneuver as much as 1500 miles across its normal descent track to find a friendly airfield. The need to maneuver meant wings (which also gave the “pilots” something to “fly,” as opposed simply to riding along), and it exposed the shuttle to more reentry heat. The extra heat required a new thermal-protection system, the new system meant more weight, the added weight meant a smaller payload, worries about the reduced payload increased the pressure to devise an even higher-performance engine, and higher performance made the engine more delicate and temperamental.
The Air Force planned to send up a few very large cargoes, containing its biggest reconnaissance satellites. The shuttle therefore needed a sixty-foot-long cargo bay and a maximum 65,000 pound “payload,” or total cargo weight, even though most other payloads would fit in a small fraction of that space. (The speed with which most of the shuttle’s previously scheduled cargoes have switched to the much smaller French Ariane rocket is one indication.) Therefore, in almost all launches except those of the Air Force “big birds,” the shuttle would have to handle several cargoes at once—and the complexity of juggling and integrating these varied cargoes was, according to the Rogers Commission, a major headache and source of delay.
Above all the shuttle had to be manned, with all the attendant weight and equipment needed to keep people alive in space and on the way back through the atmosphere. Scientists barely bother to argue about the drawbacks of manned space flight anymore—the technical superiority of unmanned probes, which do not have to be kept within a narrow temperature band, shielded from shocks of more than a few G’s, or indeed ever brought back home, is so well understood and clear. But the political superiority of manned missions is even more obvious. Since the days of the original seven Mercury astronauts, converted by Life magazine into Capra-esque heroes from next door, the cult of the astronauts has given the space program its base of popular support.6 A shuttle relying on robots might have been less expensive, more flexible, and more valuable in its spillover effects to American industry as a whole—but it would never have been built. The strongest emotional argument for going ahead with the shuttle now is that seven attractive people have already died, and to quit would make their deaths a waste. This is the kind of heartstring Ronald Reagan was born to pluck.
- The rosy prospect. In the planning stages, advanced weapons are justified through impressive projections, taken from computer simulations rather than realistic tests, which show how much more efficient and deadly they will let our soldiers be. The classic case was a radar-guided missile advertised as ensuring American pilots a better than 900-to-1 kill ratio against their enemies.
Even though the shuttle’s designers consciously pushed expenses onto future users, in order to save immediate development costs, they soon got into the habit of promoting the shuttle mainly as a good business decision—not as a tool for advancing scientific knowledge, or as a vehicle of national prestige, but as an unbeatable deal. They relied on their version of the 900-to-1 kill ratio to prove their case. In 1972, according to Lodgson, NASA was estimating that each shuttle orbiter would cost $250 million, each flight would cost $10 million, and each pound of cargo would cost $160 to lift into orbit. Even after allowing for inflation, these estimates proved to understate real costs by factors of six to ten.
The key to these cheering predictions was a now-notorious analysis by the consulting firm, Mathematica. The Mathematica study, like many other military models, was based on the assumption that everything about the shuttle would go right. America would have more and more payloads to launch, the shuttle would carry all of them, the fleet would routinely manage fifty flights a year—one a week—without a hitch in supplies, maintenance, or spare parts. The shuttle’s main engines, despite being the most complex and high-pressure engines ever developed, were assumed to be virtually maintenance-free, going fifty-five launches apiece between major overhauls. If everything went as planned, then the shuttle could fulfill its promise of offering “routine” and inexpensive access to space. If it didn’t…. Well, we’ll get to that in a minute.
- The big technical leap. For years, the Soviet Union has pursued a plodding, evolutionary approach to weapons design while the US has tried less frequent, more dramatic steps upward in technological complexity. The same pattern applies in space flight. With the shuttle, American designers tried several leaps at once. The solid boosters were far more powerful than any others ever used; the liquid-fuel main engine, running on liquid hydrogen, was to be a phenomenal feat of engineering, with the highest combustion efficiency of any engine so far produced; and the heat-absorbing tiles were somehow to shield the craft from incineration on reentry without flaking off themselves, unlike the “ablative” coatings on previous one-use-only space capsules.
Also in common with many military projects, the technical leaps seemed to be made for their own sake, rather than to accomplish a specific mission. During the Apollo program, the mission was to get to the moon, as during the Second World War it had been to beat the Germans and the Japanese. Scientific breakthroughs and engineering innovations were used or discarded, depending on how well they served the central goal. But during the shuttle program, the mission was…to build the shuttle and see where that led.
Such an undirected approach is not on its face ridiculous: the interstate highway system, the great nineteenth-century canals, other open-ended investments by the federal government have had unforeseen, stimulating effects. But those investments were outstandingly flexible—vehicles large and small could use the highways. Because each flight costs so much, and so few of them are made, the shuttle has not been so flexible nor has it stimulated so much new activity. One “obvious” benefit of a manned shuttle, for instance, would seem to be its ability to retrieve and repair damaged satellites. But that is out of the question for the large number of communications satellites in geosynchronous orbit, fixed above one point on the ground. They are 23,000 miles up, or about 22,800 miles above the shuttle. The few that have been recovered from low-earth orbit have had negligible value on the scrap market.
- The unpleasant “surprise.” A Pentagon analyst named Franklin “Chuck” Spinney, after analyzing the histories of scores of modern weapons, has described the inevitable collision between rosy projections and unforgiving reality. For example, the new airplane that is supposed to beat all comers: when it goes out to the field, “unexpected” maintenance problems predictably emerge. It predictably but “unexpectedly” gets low on spare parts; it cannot maintain its sortie rates; its pilots can’t train as much as they should to master the plane’s complicated systems. More money must be found to bail it out of its difficulties; other programs are unexpectedly shortchanged.
The shuttle was well into this cycle even before its explosion. Its heat-protective tiles—31,000 of them, each unique in size and shape—saved the craft from incineration but were almost unbelievably fragile and hard to handle. Installing the tiles on the first shuttle unexpectedly required 335 man years of work, or 2.8 days per tile. 7 The main engines, instead of going fifty-five missions between overhaul (as projected), have required repair after about one tenth that much time. Through the first twenty-four successful shuttle flights, “unexpected” repairs have taken more maintenance time than all the scheduled routine work. For example, after Challenger’s first flight in 1983, all three of its main engines had to be repaired or replaced, even though they had been used only in ground tests before. Because of unforeseen weight problems, the payload on the first orbiters is smaller than anyone anticipated—about 47,000 pounds rather than 65,000—and, even before the explosion, the 24-launch-per-year schedule was proving impossible to meet. (This has not deterred James Fletcher—who was director of NASA when the shuttle was designed and was called back by President Reagan after the explosion—from projecting a steady twenty-four flights a year, when flights resume in 1988.)
What has all this done to those reassuring, bargain-rate, Mathematica-induced cost projections? The initial price to commercial users was set at $22.4 million per launch, more than twice as much as NASA’s original target of $10 million. But even so the price was artificially held down. It was based on the most strictly defined marginal cost of a launch—fuel, crew time, and so forth—and frozen at 1975 prices. By 1986, the price had risen to $74 million per launch, but even that reflected a subsidy to compete with France’s Ariane. There are many ways to calculate the true cost of a shuttle flight, but the most realistic estimates start at $150 million and go up.8 The physicist James A. Van Allen pointed out earlier this year that “in 1985 only ten shuttle flights were carried out at a true launching cost of at least $5,000 per pound, or about $2,000 per pound in 1971 dollars, a figure twenty times greater than the original [Mathematica plus NASA] estimate.”9
The culmination of the “Spinney cycle” in weaponry is a perverse, backward progress, in which each new generation of “advanced” equipment costs more, and works less of the time, than its predecessors. Civilian technology, of course, works just the opposite way, as the real costs of manufactured goods are driven down, down, down. The shuttle follows the military, not the civilian, pattern. “Prices are actually going up, not down,” David Gump, formerly of Space Business News, has written. “The space shuttle costs roughly $6,000 per pound of cargo, while the Saturn rocket used for moon missions could boost cargo into orbit for only $3,800 a pound (adjusting for inflation to make the two prices comparable). It is astounding—costs went up on a vehicle NASA expressly designed for cheap flights.”10
- The house of cards. The final traits of the complex military project are its brittleness and vulnerability to surprise. If all goes according to plan, fine: but when one thing goes wrong—a sandstorm in the Iranian desert, guerrillas who won’t stay put to be bombed—many others do too. The classic example of military brittleness is the Maginot Line.
This year has seen a display of the astonishing brittleness of the American space program. With one explosion, a quarter of the main launch fleet is destroyed and major scientific, commercial, and military programs are delayed at least two years. By most estimates, the shuttle program was taking half of NASA’s total budget before the explosion, eliminating alternative launch systems and starving scientific missions as well. In the days of Mathematica’s rosy visions, the shuttle was advertised as carrying four vehicles per year on to planetary exploration—Mars, Jupiter, beyond. So far it has carried none. The remarkable images sent back to earth from Uranus, in the same week as the explosion, were provided by Voyager 2, propelled by a mere “expendable launch vehicle” nine years ago. James Fletcher has been telling congressional committees that he will keep cutting “anywhere” in NASA’s budget to protect money for the shuttle fleet and the proposed space station.
Since the shuttle explosion, the government has been producing a great many space-policy reports and proposals for new objectives in space. Yet the remarkable thing about the military-procurement pattern that led to the shuttle disaster is that it’s about to start all over again. The eggs-in-one-basket approach, the reliance on rosy estimates, the lack of interest in less glamorous forms of technology—these traits lie behind the three big steps President Reagan has taken to get the space program going again. They are construction of a new shuttle to replace the Challenger, at $4 billion to $5 billion; progress on a big, complex space station, at $8 billion or more, which creates the demand for the replacement shuttle to ferry up the parts; and, less publicized and scrutinized, development of the “aerospace plane” which President Reagan dubbed the “Orient Express” in his State of the Union speech just after the Challenger crash.
The aerospace plane is the kind of forward-looking, big-vision, let’s-not-sit-around-and-mope project that Ronald Reagan seems personally to love best. In principle, it also appeals to common sense. In a normal rocket launch, a tremendous amount of cost and energy seems “wasted,” since only a tiny fraction of the enormous mass sitting on the pad makes it into space. It’s especially frustrating to have to lift vast, heavy tanks of liquid oxygen up through the oxygen-rich atmosphere, “like a fish carrying a canteen of drinking water with him,” as a young engineer from MIT named Stephen Korthals-Altes has put it in his study, The Aerospace Plane: Technological Feasibility and Policy Implications.
With the aerospace plane, the wasted effort would be cut to a minimum. It would roll down the runway and fly like an airplane, using “air-breathing” engines, until it reached impressive (supersonic) speeds. Only then would it switch to rockets and bottled fuel for the brief, final ascent into space.
Because, unlike the shuttle, it would be wholly reusable, the aerospace plane might be cheaper to operate. Naturally, the official cost estimates vary widely but—even more naturally—they’re all low. The National Commission on Space says the aerospace plane might lift payloads to low earth orbit for $400 a kilogram, or about one tenth of the shuttle’s current cost. (Some Air Force officials have said it would be only one tenth of that, perhaps $50 a kilogram). Although its cargo bay would be much smaller than the space shuttle’s, it could still handle most military loads and 90 percent of the commercial launch market. Taking one payload at a time would make the aerospace plane more flexible, like a taxi compared to a Greyhound bus. In theory it could cut the flight time between New York and Tokyo to two hours, or about as long as it takes to get from each airport to the city’s center. If all goes as planned, the first prototype will cost two to three billion dollars.11 Already the government has let out airframe and engine contracts whose potential value is $450 million.
Conceivably the aerospace plane will do everything its supporters say. But at the moment it is showing all the signs of the military/aerospace procurement disease, especially its reliance on everything working just right. The aerospace plane is an even bigger technical gamble than the shuttle. While the shuttle required familiar things to be done on a big scale—bigger boosters, more tightly wound engines—the entire promise of the aerospace plane rests on technology that has never been fully demonstrated, even in the lab.
The key to the aerospace plane is its engine, which must be designed to work in two phases: at relatively “slow” speeds (up to Mach 4) through relatively thick air as the plane is climbing and accelerating, and then at hypersonic speeds, through the thinnest outer wisps of the atmosphere, as it approaches orbital velocity and lets rockets kick in to carry it to Mach 24. The two standard forms of air-breathing jet engines, turbojets and ramjets, can’t come close to attaining orbital speed. In turbojet engines, used on most commercial jets and fighter planes, air is compressed inside the engine before combustion; in ramjets, the plane’s forward motion pushes the air through the engine. Both have severe upper limits on operating speed. The turbojet overheats at speeds above Mach 3, and the ramjet has different but also heat-related problems above Mach 6.
The one air-breathing engine that might conceivably go fast enough is the “scramjet,” which so far exists only in the lab. When the scramjet is running, air keeps rushing through at supersonic speeds, carrying off heat and preventing the engine from melting down. In principle, the engine should work at speeds up to Mach 24. It can’t work at all at speeds below Mach 4—twice as fast as the Concorde flies—so another new engine, the “air turbo-ramjet,” will be required to get the aerospace plane off the ground. Mainly because of physical limits in building wind tunnels, the scramjet has never been tested at speeds above Mach 7.
The crucial question is how fast the scramjet really can go, when it leaves the laboratory and is mounted on a plane. (So it might be more correct to say, “when a plane is mounted on it,” because the entire underside of the plane will consist of inlets and exhausts for the engine.) The greater its maximum speed, while still breathing air, the less dead weight the aerospace plane will have to carry for the final rocket burst. Under commission from NASA, Stephen Korthals-Altes carried out a feasibility study of this question. It showed that the plane might either fulfill its potential, or fail disastrously, depending on very small and now-unknown operating deviations.
The higher the plane’s speed, the harder it is for the engine to generate extra thrust. That is, the air is coming into the engine so fast it’s hard to make it go out even faster. The point at which the scramjet is overwhelmed by this challenge and reaches “zero net thrust” determines everything else about the plane’s feasibility. The “zero net thrust” point, in turn, depends on minute, untested variations in how the engine’s inlet and exhaust work at hypersonic speeds.
By cranking all the variables into a computer simulation (which he reproduces in his book, so his assumptions can be checked), Korthals-Altes shows that the aerospace plane can perform as advertised—only if the scramjet engines can take the plane to Mach 17, or well over 10,000 miles per hour.
If scramjets fall short of these performance goals, perhaps because of unexpected propulsive inefficiencies…then the vehicle weight rises exponentially [because of extra rocket fuel], making the concept untenable. In other words, scramjets are a very unforgiving technology. In contrast with rocket engines, which fail “gracefully” [by providing somewhat lower thrust], scramjet performance falls off precipitously. The overall technological riskiness of the ASP [aerospace plane] is primarily a result of this, the extreme sensitivity of scramjets to engine inefficiencies.
If the engines can reach Mach 21, the gross weight of the aerospace plane would be only one sixth that of the shuttle. But if they reach “only” Mach 12—three times as fast as any air-breathing plane has ever flown and almost twice as fast as it has gone in tests—it will weigh half again as much as the shuttle and carry a smaller load.
Because of all the technical unknowns, especially a motor tested at only one third its hoped-for speed, Korthals-Altes says, cost estimates for the aerospace plane must be treated with extreme caution. His own guess is that it would cost about $17 billion to develop. Analysts from the First Boston Bank have estimated $14 billion, and British aerospace is saying $6 billion to develop a smaller, unmanned plane called HOTOL. But the US Air Force estimates that the plane can be developed and built for $3 billion. On the strength of this figure President Reagan gave the go-ahead. Given the track record of the space shuttle and most of the American arsenal, the Air Force figures look the least believable of the bunch.
NASA plans to bear only 20 percent of the cost of the aerospace plane—the Pentagon would pick up the rest. But, as Korthals-Altes points out, 20 percent of even the low-ball $3 billion development budget “would be enough to fund three major planetary missions.” In pushing for the aerospace plane, instead of developing new and better rockets, Korthals Altes concludes, the US
may be making the same mistake we made with the shuttle, namely “putting all our eggs in one launch vehicle”…. While scramjet technology holds promise, it may be too much to expect that it will give us a vehicle able to combine the vantage point of a reconnaissance satellite with the maneuverability of an SR-71 [high-speed “Blackbird” reconnaissance plane], deliver the ordnance of a bomber with the speed of an ICBM, whisk passengers across oceans with flight times like that of the Eastern Shuttle and launch payloads into orbit with the ease of a DC-9.
In view of the brittleness of our current space program, is it wise to put more money into another “unforgiving” technology—or rather to look for a renaissance of basic propulsion, which would in a variety of ways let us go ahead with science, exploration, and commerce?
In fact, a renaissance in basic rocket technology is underway. It is happening in three places in the non-Communist world, under different and illuminating circumstances: in Europe, mainly through progress on the French rocket, Ariane; in Japan, which has tested its H-1 rocket and is three years away from unveiling an advanced, truly competitive, totally made-in-Japan rocket called the H-2; and in the United States, which has proudly heralded the birth of a new, private launch industry but in fact is relying on the military for most of the funding and ideas.
New launchers are needed for military purposes, for scientific exploration, and for business use. The last of these has received a lot of publicity since the Challenger explosion, but in the immediate future it will be the least important. Indeed, the “space business” lacks many of the normal attributes of business, starting with people making a profit.
To date, only one part of the space business has proven commercially workable, in the sense that unsubsidized buyers and sellers willingly come to terms. That part involves satellite manufacturing—led by American firms such as Hughes Aircraft (a subsidiary of GM), RCA (a subsidiary of GE), and Ford (still on its own)—and the operation of satellite-communication networks.
Satellites now carry about two thirds of international telephone calls and more than 90 percent of data transmissions and TV broadcasts. But the future does not offer them limitless growth. For some uses, especially broadcasting to many points of reception (TV) or reaching customers dispersed across great distances (telephone and data networks in Indonesia or Brazil), satellites have an enormous edge over any available alternative. But for high-volume traffic among fixed, nearby points—along the East Coast of America, in the population centers of Western Europe or Japan—fiber-optic transmission is becoming more attractive. Messages are transmitted as pulses of light along the cable. Fiber optics are fast, reliable, compact, and becoming cheap. AT&T agreed this spring to lay a fiber-optic cable beneath the Atlantic Ocean.
Amid much hoopla in the trade press, Federal Express paid Martin Marietta $100,000 this summer, as a deposit for the launch of a Federal Express satellite aboard a Martin Marietta Titan rocket. This was the first contract for private American launch services ever. But soon afterward, Federal Express abandoned its “Zapmail” facsimile transmission system, which was what the satellites were for. (Martin Marietta has hinted it will soon have other contracts to announce.) A few companies that had satellites scheduled for immediate launch—notably Western Union—have suffered from the “launch crisis.”12 But in general, the satellite industry suffers from overcapacity more than shortage. Soon after the shuttle explosion and the failure of an Ariane rocket launched from West Africa, the president of ARIANESPACE said that his company was technically equipped to launch fifteen rockets per year by 1990. Nonetheless, he said, Ariane is planning for only half that level because “I don’t see the market.”
As if weak fundamental demand were not enough of a problem, the satellite business is also in the middle of an insurance crisis. Over the last twenty years, the performance of launchers, especially American, has been very good—fewer than 10 percent have failed. (Ariane, by comparison, has had four failures in eighteen launches, for a success rate of only 77 percent.) But during the last two years insurance companies have paid out more than $700 million in losses, against only $325 million collected in premiums. Back in the palmy days, a satellite company could get insurance for less than 10 percent of the satellite’s value; now the premiums are pushing 30 percent, and recently RCA gambled and sent one up uninsured. Below-market-rate insurance has become a major selling point for Chinese and even Russian launchers as they try to attract outside business.
One other commercial possibility—manufacturing in the weightlessness of space, perhaps even extracting ores from the moon or asteroids—might someday be very important. Some valuable materials become even more valuable when pure, and weightlessness can enhance purity. Everyone’s favorite illustration is gallium arsenide, a metallic crystal used to make semiconductor chips that run faster, on less power, in a wider range of conditions, than those made of silicon. But the pure crystalline structure of gallium arsenide is disturbed by convection currents, like those that make hot air rise, which develop in the molten material as it is cooling and forming crystals. On earth, the purest gallium arsenide crystals are very small. Convection is not a factor in space, but so far the cost of getting the material up and down far outweighs the benefits of making large, pure chips. “As soon as anybody has focused on one particular product, the ground is cut away from them,” Melinda Gipson, editor of Space Business News, has said. “Someone else focuses on it and tries to make it cheaper on the ground, and invariably they have.”
What this leaves for “space business,” then, is building the rockets and shuttles themselves, and hoping that uses will arise for them. Except in the US, this work is carried out as a straightforward and unembarrassed race by governments that grant subsidies to their own manufacturers. The Chinese, Russians, and French believe that advanced launching systems will serve their national interest in diverse ways, and their governments have ensured that they will develop.13 They make few pretenses that their plans must meet the harsh test of the market. (Ariane has been “privatized” and in theory earned a profit last year, but its research and development budget continues to be very heavily subsidized.) Now McDonnell Douglas, which previously licensed rocket systems to the Japanese, is considering buying the upper stage of Japan’s H-1 rocket to improve its own Delta. 14
Japan is not yet competing for anybody else’s launch business, and its space-agency officials are publicly humble and deferential—first modest steps, small island nation, hoping to learn from big strong America, et cetera. But they have relentlessly fulfilled their vision of technological independence—first licensing equipment from American companies, then building their own—and their domestically made H-2 rocket promises, unsurprisingly, to be a breakthrough in miniaturization and efficiency. If all goes as planned, the rocket will be able to lift as large a payload as the Ariane 4, even though it weighs only half as much. The Japanese say it will be able to lift more than an American Titan 34D, which weighs three times as much. No “market” principles are in evidence here, either: it’s much more costly for the Japanese to design their own rocket than just to buy services from somebody else.
Into this mercantile world of buyers more intent on making a sale than a profit comes the United States, with its brave talk about a booming new private launch business—and its huge budget for military space. This summer, President Reagan officially took NASA out of the commercial launch business. (Of the forty-four payloads booked onto the shuttle, only fifteen have been rescheduled. The rest are looking to Ariane, the Long March, or conceivably to American launchers.) When the shuttle starts flying again, it will concentrate on the space station, on some scientific projects (notably the enormous Hubble space telescope, one of the few payloads that only the shuttle can handle)—but above all on military cargoes. That will leave American companies “free” to compete with launchers of their own.
The largest competitors will be the firms that won rights to “privatize” military rockets: General Dynamics with the Atlas-Centaur, McDonnell Douglas with the Delta, and Martin Marietta with the Titan. And what kind of business will sustain this newly freed enterprise? There is obviously not enough satellite or space-manufacturing demand to go around. That leaves, as the source of nearly all foreseeable private demand…the US Air Force. Over the next decade, the Air Force plans a tremendous volume of launches—a new global-positioning satellite system, better communications networks, and of course all those payloads for SDI. The Air Force’s forthcoming purchase of twelve to eighteen rockets for its new “medium launch vehicle” is the biggest single sale on the horizon, and the companies are competing fiercely for it.
This is not the first time America has used the military budget as a backdoor way of accomplishing social objectives—think of the National Defense Education Act, which beefed up foreign-language programs in the 1950s. But in this case, it seems a particularly clumsy and stupid way to accomplish a goal, if the goal is an adaptable, competitive rocket industry. Practically nothing in recent industrial history suggests that military contracts promote efficiency, inventiveness, or true commercial vigor. (One exception should prove the point: Northrup developed the lightweight fighter called the F-20 with its own money, making huge strides in reliability and manufacturing technique, only to have the Air Force turn its back on a plane developed outside its own system.)
Soon after the shuttle explosion, Hughes Aircraft and Boeing came up with an innovative design for the “Jarvis Rocket,” named for the Hughes payload specialist, Gregory Jarvis, who died aboard the Challenger. The Jarvis would fly on kerosene—not liquid hydrogen—and it offered tremendous lift power at relatively low cost. But once the Air Force indicated its possible interest in the Jarvis for the “Medium Launch Vehicle” competition, the design changed. In came the cryogenic liquid-oxygen engines and solid-rocket boosters and up went the price, to $150 million per launch. The president of the Hughes Space and Communication Group, Anthony Iorillo, was quoted in Aviation Week and Space Technology as saying that it would take $1 billion in subsidized development to get the Jarvis going. In November, the Air Force said it would not select the Jarvis, so its future looks bleak.
Is there an alternative to letting the Air Force determine what the American space system will be? If, through wheedling, treaties, or threats, the US could persuade other countries to stop subsidizing their launchers, open commercial competition might make more sense. But that is unlikely any time before the next appearance of Halley’s Comet. The other choice would be to declare that a robust, varied space program is important to America’s future—and for reasons other than providing a nuclear “shield.” Our national honor is at stake, not in the Sputnik sense of a “prestige” race against the Russians or Japanese but in the deeper sense of the duty to explore and understand. Even though the market may not immediately recognize it, there is a broad national interest in pursuing space science, in preparing for future commercial possibilities, and for providing services that are properly public, from weather forecasting to remote-sensing. The rush to “privatize” the launch business is not likely to fulfill these objectives—nor is the current NASA emphasis on such big-ticket extravaganzas as the space shuttle and the aerospace plane.
“In the end the United States got the worst of both worlds,” Walter McDougall says near the close of his book:
“a free market” twisted at every turn by intervention and a technocratic state incapable of managing the change it provoked. The Soviet State, by contrast, could make up in tyranny and disregard for deleterious side effects what its centralized technocracy lost in creativity, while France and Japan could plan with some success for specific commercial goals because their societies were relatively small, relatively free from military complications…. The United States was obliged to fight a total cold war—like the USSR—yet retain democratic institutions—like France or Japan. It didn’t work.
December 18, 1986
Space Business News, (April 21, 1986). ↩
Cornell University Press, 1985. ↩
The commission report said: “Contrary to design expectations, the joint tang and inside clevis bent away from each other [at ignition] instead of toward each other and by doing so reduced—instead of increased—pressure on the O-ring in the milliseconds after ignition.” Reduced pressure was bad because it postponed the ring’s movement into sealing position, and the main flow of hot, high-speed gas started roaring in. ↩
One of several would-be heroes of this story was Roger Boisjoly, a Thiokol engineer. He gave the commission this account of his last attempt at rebuttal, after Thiokol’s managers acquiesced in the launch: ↩
John M. Lodgson, “The Space Shuttle Program: A Policy Failure?,” Science (May 30, 1986), pp. 1099–1105, quote on p. 1099. This is a valuable retrospective look at how the program went wrong. The best before-the-explosion account is in Gregg Easterbrook, “The Spruce Goose of Outer Space,” The Washington Monthly, (April 1980). Easterbrook had this to say about the solid boosters: “During blast-off, unlike those capsules and modules with escape rockets to pull the pilots free in case of trouble, there is no way out of the shuttle . Here’s the plan. Suppose one of the solid-rocket boosters fails. The plan is, you die.” ↩
Michener’s Space is marvelous on this theme, with its lampooning of Life’s exclusive rights to the astronauts’ inside stories and resultant concern that they comport themselves in Jimmy Stewart-like fashion at all times. ↩
Stephen Korthals-Altes provides these figures on page 25 of his book about the aerospace plane. Several of the other facts I cite below also come from his book. ↩
Last year the Congressional Budget Office listed the different ways of calculating the shuttle’s cost and derived a range of estimates. The CBO pointed out that “if 18 rather than 24 flights were flown in 1989, the highest full-cost price would increase from $150 million per flight to $186 million; with only 12 flights it would increase to $258 million” (Space Business News, March 11, 1985). ↩
James A. Van Allen, “Space Science, Space Technology, and the Space Station,” Scientific American (January 1986), pp. 32–39, quote on p. 36. ↩
The Wall Street Journal, (August 5, 1986). ↩
Space Business News (June 2, 1986). ↩
Space Business News (May 5, 1986). ↩
The drollest “commercial” competitors are undoubtedly the Russians. Soviet “marketing” specialists have been advertising the reliability and reasonable cost of their Proton rocket and have offered great deals on insurance. They have told foreign satellite companies that they can physically stay with their satellites from the moment of arrival in Russia until launch. The Russians have also offered to take a British astronaut on a joint mission and have proposed handling some big launches for the Japanese “on a commercial basis,” according to deputy foreign trade minister V.L. Malkevich. See Space Business News (May 5, 1986). ↩
Space Business News (July 14, 1986). ↩