Warning! Extreme dullness follows. This is my only serious magazine piece. Here I am, a 24 year-old know-it-all, yammering about scramjets and aerospace policy.
This article, from the April 1987 issue of Technology Review, was adapted from my master’s thesis, the only collegiate thesis ever reviewed by the New York Review of Books. I was quoted in everything from the New York Times to scientific journals to Congressional testimony. Mostly I think people just latched onto my catchphrase “hypersonic hyperbole.”
As it turned out, my skeptical, emperor-has-no-clothes view of the project was justified, but that is an awful way to begin one’s professional career. Youth should be a time for starry-eyed dreaming, not dismissive nitpicking.
Will the Aerospace Plane Work?
In his State of the Union Address, President Reagan spoke of an aerospace plane that could, “by the end of the next decade…fly [from Washington] to Tokyo within two hours.” Former presidential science advisor George A. Keyworth II told a congressional subcommittee that ticket prices for a civilian version, the hypersonic transport (HST), could compete with today’s ticket prices.
Best of all, according to Keyworth, the aerospace plane would not require massive development efforts. “This technology is here…It is simply waiting to be used and put together and assembled in an innovative fashion.” Official Pentagon estimates of the R&D costs are about $3 billion, a pittance for space endeavors.
A Department of Defense (DOD) press release envisions that “a future aerospace plane would be able to operate as an airplane at hypersonic velocities (4,000 to 8,000 per hour)…or as a space launch vehicle capable of accelerating directly into orbit.” Military leaders boast that it could reduce launch costs from about $2,000 per pound to $20 per pound.
Says General Lawrence A. Skance, commander of the Air Force Systems Command, “We are talking about the speed of response of an ICBM and the flexibility and recallability of a bomber, packaged together in a plane that can scramble, get into orbit, and change orbit so that the Soviets can’t get a reading accurate enough to shoot at it.”
Much of this talk is hypersonic hyperbole. The technological challenges have been downplayed, the development costs are grossly underestimated, and the utility of the aircraft is vastly exaggerated. This vehicle would be the most complex aircraft ever attempted, and its true development cost is likely to be six times the official estimate. The aerospace plane might be a viable alternative to the space shuttle, but its value as a civilian hypersonic transport or as a military aircraft is dubious.
Nevertheless, in January 1986, a three-year, $700 million program to design an aerospace plane in detail began. Participating government agencies include the Defense Advanced Research Projects Agency (DARPA), NASA, the Air Force, the Navy, and the Strategic Defense Initiative Organization (SDIO). In April DARPA awarded two $27 million contracts for the preliminary design of a full-size revolutionary engine, a crucial element of the aerospace plane. One went to Pratt & Whitney, the other to General Electric. On the same day, DARPA awarded $7 million contracts for conceptual design of an airframe to Boeing, General Dynamics, Lockheed, McDonnell Douglas, and Rockwell International. The plan is that in 1989, after studying various applications and technology areas, the DOD will decide whether to procure the plane.
What makes the aerospace plan special is its intended use of fuel-efficient, air-breathing engines instead of relying exclusively on rockets to reach orbit. Seventy-five percent of the space shuttle’s lift-off weight of 4.4 million pounds is propellant, and 83 percent of that propellant is oxygen. A launch vehicle carrying oxygen as it flies through the oxygen-rich atmosphere is like a fish carrying a canteen of drinking water. The aerospace plane would rely on oxygen from the air up to an altitude of about 40 miles. Small rocket engines would take over for the final acceleration into orbit. The rockets would also be used for orbital maneuvering and for initiating re-entry.
The concept does impose some penalties. Designed to take advantage of atmospheric oxygen, the complex air-breathing engines would be heavier per pound of thrust than conventional rocket engines. One set of engines, either the air-breathing engines or the rocket engines, would always be dead weight. Furthermore, an aerospace plane would take off horizontally like an airplane, so to gain lift it would need a fuselage and wings, components that a vertically-launched rocket does not have. Heavy-duty landing gear would be necessary to support the vehicle at take-off, when its weight would be nearly five times as great as at landing.
Propulsion is the technology that will make or break the plane. The entire underside of the airframe must serve as an extension of the air intake and nozzle of the engine because at hypersonic speeds (speeds higher than Mach 5, five times the speed of sound) very large engine airflows are required for adequate thrust. Eight to twelve engines, each about three feet across and ten to fifteen feet long, would be slung beneath the fuselage to ingest the compressed air of the shock wave the vehicle’s nose would produce. Design changes in any part of the engine or airframe could require changes in all the other parts.
The engine that would propel the aerospace plane at hypersonic speeds is still experimental. Moreover, this engine—the supersonic combustion ramjet or scramjet—is the only possible one for hypersonic flight. The reason is clear from a review of its technological ancestors, all of which share a basic principle: they produce thrust by expelling gases faster than they come in.
Turbojet engines are found on most airliners. Turbojets use a fan-like compressor to raise the temperature and pressure of the air before combustion. The burning gases expand, and as they escape, they power the turbine that drives the compressor. Turbojets reach their operating limit at about Mach 3. That’s when the temperature of the gases turning the turbine reaches about 2,300 degrees F, the maximum today’s turbine blades can withstand.
A ramjet is the simplest of aircraft engines. For compression, it relies on the so-called ram-pressure of the incoming air stream in a moving aircraft. Air enters the engine, slows down to subsonic speeds, mixes with the fuel, ignites, and escapes through the rear nozzle. Ramjets can’t operate until the aircraft reaches about Mach 1, so some other form of propulsion brings the vehicle up to that speed. At about Mach 6, ramjets lose their effectiveness because the temperature in the combustion chamber becomes so high (nearly 2,700 degrees F) that the fuel is expelled before it can be completely burned. The partially burned fragments do not contribute their full energy to producing forward thrust.
The air turbo-ramjet (ATR) marries turbojet and ramjet technology. The Aerojet Company conceived the ATR in 1949, but so far has only tested it in the laboratory. Up to Mach 2 this engine compresses incoming air mechanically, but between Mach 2 and Mach 6 it relies on ram-pressure. The crucial feature of the ATR is that the compressor turbine is driven by the fuel in gaseous form, not by the hot inlet airstream. The hot air is diverted around the turbine to mix with the fuel coming from the turbine. Then the mixture of fuel and air burns in a ramjet-style combustion chamber. However, at about Mach 6 the air entering the chamber slows down so rapidly that its temperature soars to about 3,000 degrees F, much as in a ramjet, so that even non-moving engine parts may fail from overheating.
A New Engine
A scramjet is very similar to a ramjet. The major difference is that while air slows down to subsonic speeds inside a ramjet, air maintains supersonic speeds throughout a scramjet. Engine heating is minimized since less of the air molecules’ kinetic energy is converted to thermal energy.
The greatest problem is that burning fuel in a supersonic flow is equivalent to lighting a match in a hurricane. At Mach 7 a fuel particle has less than a millisecond to ignite before it is swept out the exhaust. Traditional jet fuels, like kerosene, do not ignite quickly enough. Only hydrogen will work, and hydrogen’s low density necessitates fuel tanks five times as large as hydrocarbon fuel tanks.
Scramjets would circulate liquid hydrogen through the engine and airframe to help dissipate the intense heat generated by hypersonic flight. Conventional ramjets inject fuel from the engine walls, but a scramjet’s short mixing time requires fuel injectors on struts that span the air inlet. Cooling these struts would be especially important. According to NASA researchers, the scramjet fuel-injection struts also present the most formidable structural problems of the aerospace plane. The struts must withstand both shock waves within the engine and the thermal stresses resulting from two temperature extremes (-375 degrees F inside the struts and 1,700 degrees F outside).
Scramjets would begin to work only at speeds above Mach 4, twice as fast as the Concorde flies. Air turbo-ramjets would get the aerospace plane up to Mach 4. Estimates of the top speed these engines could attain vary from Mach 10 to orbital velocity—Mach 25. Wind tunnel experiments show that scramjets can operate to at least Mach 7, the limit of NASA’s current test facilities.
As speeds increase, the percentage difference between the inlet airstream velocity and the exhaust gas velocity becomes smaller, so even slight engine inefficiencies in the scramjet inlet or nozzle would have major consequences. Also, even if efficiencies could be completely eliminated, the scramjet’s performance would decline significantly as speeds increase. In contrast, a rocket engine’s performance remains constant. By Mach 20, a scramjet would have only a slight advantage over a rocket engine.
The exact speed at which scramjets quit may determine whether the aerospace plane is feasible. The earlier the switch from scramjet to rocket thrust, the more liquid oxygen the aerospace plane would have to carry. Supporters of the aerospace plane claim that it would be a launch vehicle the size and weight of a conventional aircraft, but my analysis says that the plane will live up to this claim only if scramjets can function up to around Mach 17.
The government does not seem to have a reliable basis for the statement that it would cost $3 billion to develop an aerospace plane. Christopher Demisch, an aerospace analyst at First Boston Corporation, estimates the financial requirements to be roughly the same as the shuttle’s: about $14 billion.
British Aerospace, Ltd. is working on its own aerospace plane called HOTOL (short for horizontal take-off and landing). They estimate the development costs to be about $6 billion, even though HOTOL’s baseline design calls for a smaller, less complex, unmanned vehicle.
Even a subsonic aircraft costs between $2 billion and $3 billion to develop. The congressional Office of Technology Assessment has said that an American supersonic transport (SST) would cost $6 billion to $8 billion. Even Ford spent $3 billion to develop the Taurus.
My own analysis of aerospace plane development, fabrication, and operating costs uses a cost-estimating model developed by the European Space Agency. I concluded that the total cost would come to about $17 billion. The cost-estimating equations, based primarily on weight, are derived from component costs of past U.S. and European launch vehicles, as well as from proposed designs for advanced launch vehicles. Scramjet and ATR cost estimates use a method developed for the Federal Aviation Administration to analyze an SST.
The hard sell by aerospace plane advocates includes the similarly improbable prediction that launch costs would be reduced 100-fold. My analysis suggests that it is more realistic to expect a ten-fold reduction.
Furthermore, any reduction in operating costs is contingent on drastically reducing the number of technicians required for launch. A space shuttle launch is very labor-intensive, with launch personnel, maintenance, and crew training accounting for 46 percent of operating costs. All told, a shuttle launch requires some 6,000 people. Even if the aerospace plane really could reduce launch costs 100-fold, it would be because of revolutionary progress in collateral technologies such as highly automated systems for checking out the readiness of a vehicle for launch; long-life or maintenance-free subsystems; and facilities for servicing a launch vehicle horizontally rather than vertically.
Such improvements could be applied to any new launch vehicle to make it more attractive. This illustrates one of the deficiencies in arguments for the aerospace plane: comparing the vehicle with the space shuttle. The aerospace plane appears superior, but any competition between 1973 technology and projected 1990s technology is useless. A space shuttle built today with the latest composite materials would save over 15,000 pounds.
The Aerospace Plane as a Launch Vehicle
Because three separate applications are proposed for the aerospace plane, the feasibility of each must be considered.
The vehicle may be best suited to place payloads into a low-earth orbit at relatively low cost. But the idea of replacing the shuttle with the aerospace plane puts NASA in a delicate situation. The prospect of such a plane excites many NASA engineers, and since it is Reagan Administration policy to push the plane, NASA supports it publicly. At the same time, NASA officials fear that too much talk about the aerospace plane would suggest they were giving up on the shuttle. When the shuttle program is back on track, the agency may be less reticent about discussing its plans and more free to promote the aerospace plane.
However, NASA is not monolithic. The internal debate within the agency about the best option for a next-generation launch vehicle is not over. Ivan Bekey, NASA’s director of advanced programs, has said his studies indicate that a new kind of vertically launched rocket might be just as promising as the aerospace plane. This rocket would carry its own oxygen, but it would employ dual-fuel engines that burn kerosene in the lower atmosphere and hydrogen in the upper atmosphere and outer space. Like the aerospace plane, the dual-fuel rocket would need advanced, lightweight structural materials as well as a vastly simplified launch environment.
Charles Eldred, assistant head of the vehicle analysis branch of NASA-Langley’s Space Systems Division, has evaluated both conventional (vertical take-off, all-rocket propulsion) and unconventional (horizontal take-off, air-breathing propulsion) launch vehicle concepts. He concludes that advanced rockets would be smaller, lighter, less complex, cheaper, and lower in technological risks.
Of course, considering the secrecy shrouding the aerospace plane, it is possible that DoD researchers have made key advances, unknown to their civilian counterparts, that make the concept superior to its rivals. Rumors have circulated that DoD is on the verge of producing a scramjet able to operate in extremely thin air. If true, this would extend the top speed at which a scramjet could operate.
The Aerospace Plane as a Hypersonic Transport
Government and industry planners—including officials from NASA-Langley, the Commerce Department, and McDonnell Douglas—are investigating the feasibility of applying the air turbo-ramjets and scramjets to civilian aircraft. The intent is to improve air transportation, especially to nations in the Pacific Rim—hence the nickname “Orient Express.” The Orient Express would be a 300-to-500 passenger, Mach 5 HST able to fly from New York to Beijing in about two hours. The vehicle, which would probably be powered by a methane-fueled ATR, would benefit from aerospace plane technology.
The logical way to begin appraising an HST is by examining its stillborn cousin, the SST. In 1971 Congress killed the U.S. SST program, citing its dubious commercial prospects, probable airport noise and sonic-boom problems, possible negative effects on ozone and climate, and the impropriety of spending large sums of public money on a project to benefit a select few. These same questions will inevitably dog an HST.
Take, for example, Congress’s fears about profitability. They are borne out by the Anglo-French Concorde, which, though a technological achievement, is a financial disaster. A viable HST program would need convincing, favorable economic projections supported by both government and the aerospace industry.
However, many industry officials see the HST as strictly a military program. “This whole idea of a hypersonic airplane is good from a military standpoint, but is being way overplayed as an Orient Express,” asserts John Steiner, who was a vice-president of Boeing for 22 years. He recently chaired the White House Aeronautical Policy Review Committee. Steiner favors a second-generation supersonic transport because airlines would be willing to invest in it. He says that at a time when many airlines are struggling just to stay out of the red, talk about developing an entirely new system based on hydrogen or methane is “baloney.”
Evidently there are those on the other side of the Atlantic who agree with Steiner. France’s Aerospatiale, Ltd., is busy designing a Mach 3 successor to the Mach 2 Concorde. Speaking of the HST, one Aerospatiale official has remarked, “Does this mean airlines are going to qualify passengers for space flight in order to carry them in these technological marvels?” He adds, “We French are often accused of going off with our heads in the clouds on new ideas, but I think in this case it is we that have our feet on the ground while our American friends are doing the daydreaming.”
The HST would almost certainly be too expensive and too risky to develop in the face of a competing second-generation SST from France. The French already have an extensive technology base in supersonic-transport design, so the project lends itself to international cooperation, a stated goal of U.S. civilian aerospace policy. The French have expressed interest in such cooperation. However, because the HST and the highly classified aerospace plane share technology, cooperation appears a long way off.
No matter what U.S. aerospace officials think of the HST, few dare criticize the plan publicly, in case the program goes forward. This has produced some creative arguments in favor of the HST. Lockheed justified the plane to the House Committee on Science and Technology on the basis of the “nominal comfort limit of passengers,” which is three hours. “To extend beyond the three-hour limit, services such as meals and entertainment must be provided to entertain passengers,” Lockheed’s representatives said. “These services cost fixed weight, add crew members, and are expensive in themselves.” It strains the imagination to think that the cost and weight of sandwiches, soda, and videotaped movies justify a multi-billion-dollar development project.
The Aerospace Plane as a Military Vehicle
The military is extremely tight-lipped about the aerospace plane. Contractors have been ordered not to discuss the project and NASA officials have been chastised for providing overly realistic concept drawings. Because of the secrecy, it is unclear what applications the military is considering.
Yet as recently as 1983, Robert S. Cooper, former DARPA director, told Congress that “currently defense has no mission for a hypersonic aircraft.” At the same hearing, retired Brigadier General Charles E. Yeager, a staunch advocate of the Orient Express, admitted that while the military is interested in researching hypersonic aircraft, it is not quite sure what it would do with one. Hans Mark, deputy administrator of NASA, stated that an aerospace plane might be useful for the Strategic Defense Initiative (SDI), but qualified his remark: “We are not at the stage in the hypersonic area where there are any missions that one can hang a program on … I thoroughly agree with General Yeager … whether there were missions for hypersonic vehicles. The answer right now is no.”
Because the Department of Defense is reluctant—or unable—to specify missions for an aerospace plane, analyzing military applications is difficult. Some possibilities include tactical air defense, strategic reconnaissance, strategic offense, strategic defense, and space combat.
Tactical air defense of civilian and military installations requires vehicles that can take off quickly. Such vehicles would use long-range missiles or cannon fire to locate and destroy enemy cruise missiles and aircraft. The high speed of the aerospace plane would not be much of an advantage, since F-15s can be forward-based in Great Britain, West Germany, Greenland, Spain, and the Azores. Moreover, as Cooper said, “Aircraft ultimately end up in their final throes of engagement with one another in subsonic regimes. And so supersonic aircraft beyond about Mach 2 are probably of little or no utility for the military.” He added that missiles can be used above Mach 2.
Nor does the aerospace plane appear to be the most cost-effective air defense system. My calculations indicate that the aerospace plane configured as a launch vehicle would cost about $1 billion per vehicle. Configured as a tactically equipped, hypersonic fighter plane, it would cost less, but the F-15s cost only about $27 million. The Air Force has said it would not build another plane, the Advanced Tactical Fighter (ATF), if it ends up costing more than $40 million a copy.
Designing an aerospace plane for reconnaissance would be a special challenge since integrating a camera system into the vehicle’s fuselage would further complicate already formidable thermal-protection and aerodynamic problems. Military reconnaissance currently employs both sophisticated satellite systems (KH-11 and Big Bird) and high-altitude, high-speed aircraft (primarily the SR-71). Satellites are said to provide the best resolution, but the SR-71, able to cruise at Mach 3 at 86,000 feet, can map over 100,000 square miles in less than an hour.
An aerospace plane configured as a reconnaissance aircraft would cover an area between those of the SR-71 aircraft and the KH-11 satellites. Thus it appears that any reconnaissance the aerospace plane might provide could also be obtained using the KH-11 or the SR-71.
Another possible military use is strategic offense—that is, delivering nuclear weapons. The aerospace plane’s potentially high speed and wide operating envelope could increase the probability of penetrating enemy air defenses. However, to ensure that a Soviet first strike could not destroy aerospace planes on the ground, they would have to be able to take off 15 minutes after being alerted. Strategic bombers can take off this fast, but they do not face the problems of storing liquid hydrogen.
Again, cost-effectiveness must be considered. An MX missile, carrying 10 warheads, costs about $50 million to $100 million—5 to 10 percent of the cost of the aerospace plane as a launch vehicle. The projected aerospace plane payload could accommodate approximately 30 warheads—the equivalent of only three MX missiles.
Hans Mark once suggested a strategic defense role for the plane in intercepting ballistic missiles in flight. There are two main problems with this idea. First, the plane would need to react almost instantaneously and accelerate at rates a human could not tolerate. Second, even if an aerospace plane with suitable weapons were in orbit at the time of attack, the missiles would be so dispersed that very few could be intercepted during the 20-minute coast phase.
Finally, some people have suggested a space-combat function for the aerospace plane—either offensively, to destroy enemy space assets, or defensively, to protect our space systems.
The defensive role does not appear feasible for several reasons. Our aerospace plane would be manned, while its target, a Soviet anti-satellite weapon (ASAT), would be unmanned and thus able to travel faster. The current Soviet ASAT takes three hours to hit a target, but if the United States proceeds with a military aerospace plane, the Soviets could develop a faster ASAT. The U.S. ASAT takes only a few minutes to destroy its target.
In an offensive role, the aerospace plane would be just another ASAT. Paul Czysz, McDonnell Douglas’s aerospace plane program manager, thinks the vehicle could neutralize selected enemy space assets without being detected. He offers two examples: an aerospace plane might fire needle-like projectiles into an enemy tracking and fire-control radar to overwhelm the antennae, or it might use low-intensity lasers to “blind” satellite sensors.
However, disabling a satellite with an aerospace plane does not appear to offer any advantages over destroying it with existing ASATs. In addition, satellite sabotage, as described by Czysz, would not be covert since Soviet space-tracking radar would easily detect an aerospace plane.
Why is the Aerospace Plane Being Pushed?
Only three years ago civilian and military experts testified before Congress that they had no missions for hypersonic aircraft. Now hypersonics are the rage in NASA and DOD, and the president has endorsed the aerospace plane.
Technological progress and the economic incentives of reduced launch costs are ostensibly the reasons behind the decision to proceed with the program. But political factors may be overshadowing the technical and economic arguments.
Aerospace plane support in the executive branch comes from the top. Reagan’s backing may result partly from his keen interest in the space program. He has twice before endorsed sizable space plans in State of the Union addresses. In 1983 it was SDI and in 1984, the space station, but both proposals met substantial criticism. Some speculate that the president wishes to be remembered for a landmark contribution to the space program, much as Kennedy is remembered for challenging NASA to put a man on the moon before 1970.
Reagan may also be motivated by the importance he attaches to SDI. Officials in SDIO readily admit that they are leaders in the push for an aerospace plane. The projected development and production costs of SDI’s infrastructure alone are alarmingly high. With launch costs at today’s rates added in, the figure becomes colossal, so SDIO officials stress the need to reduce those costs by “an order of magnitude.” The current baseline SDI architecture calls for 50 million pounds to be orbited at a total launch cost of $130 billion. While the aerospace plane may greatly benefit the SDI program, it can be sold as a civilian aircraft. This would appease a public generally supportive of the space program but increasingly concerned about the “militarization of space.”
Regardless of military applications, the Air Force has its own reasons to promote the aerospace plane. The vertically launched rocket described by Bekey and Eldred might be more economical, but it lacks “political sexiness.” On the other hand, the aerospace plane arouses interest in almost everyone who hears about it and may be an easier concept to sell.
A Plane Without a Mission
In short, the aerospace plane is being oversold. It is expecting too much to believe that the plane would combine the vantage point of a reconnaissance satellite with the maneuverability of an SR-71, deliver the ordnance of a bomber with the speed of an ICBM, launch payloads into orbit with the ease of a DC-9, and whisk civilian passengers across oceans in a few hours. The futuristic vehicle’s feasibility is critically dependent on scramjet engines performing over the most optimistic operating range. Moreover, aerospace plane costs are all likely to be significantly higher than advertised.
Also, the aerospace plane is most definitely a military program: 80 percent of the funding comes from DOD and 20 percent from NASA. This creates two main problems. First, while much of today’s civilian aviation technology can trace its roots back to military programs, the research orientation of NASA and the procurement orientation of DOD conflict. Military requirements may hasten prototype development at the expense of basic research that would make the vehicle cheaper or better.
Second, the organization that contributes the most money to a project usually has the biggest say in the direction it takes. The factors that drive technology for military applications are very different from those which drive technology for commercial applications. Military specifications stress operational objectives. Commercial developments tend to emphasize operating efficiency, safety, reduced production costs, and high availability with low maintenance.
This country needs a sensible alternative to the space shuttle, but we must be wary of repeating the mistakes of the past. Perhaps the most important lesson of the space shuttle is that a space vehicle designed to perform many functions is optimal for none. Some form of an aerospace plane could drastically reduce the cost of a launch, but this objective must be explicitly made number one. This will not happen by itself and it will not happen without giving precedence to civilian applications over military ones.
Preeminence in aeronautics depends on technical excellence and clear-cut cost advantages. A low-cost, low-noise, fuel-efficient vertical take-off and landing vehicle for use in urban areas might be an even greater boon to civilian air travel than a hypersonic aircraft. Likewise, the military’s desire for a hypersonic interceptor must be balanced against the need for maneuverable and maintainable aircraft. We must avoid the trap of thinking that higher and faster is necessarily better.