The Douglas ASTRO: An Air Force Launcher


The ASTRO, as pictured in the September 3, 1962 issue of Missiles and Rockets. Image artist unknown and copyright status uncertain, but believed to be in the public domain. Via the Internet Archive.

What it was: A lifting body craft proposed to the USAF by Douglas Aircraft. It would initially be used as a suborbital trainer then, after up-scaling and being paired with a second lifting body in an unusual nose-to-tail arrangement, evolve into a fully reusable vehicle with a nine-tonne payload capacity to LEO.

Details: In late 1962, the USAF was on the cusp of deciding how it would go forward with its plans to put military men in space. The X-15 had made its first flight mid-year, and the X-20 program was ramping up. Doubts about the latter were getting stronger, though, and would ultimately result in the Air Force deciding to work on the Manned Orbiting Laboratory instead.

It was at this point that an article was published in the now-defunct Missiles and Rockets magazine outlining a proposal from Douglas Aircraft that was supposedly being evaluated by the USAF. What it outlined was a two-part development program that would check the usual laundry list of military applications for space as perceived in the early 1960s.

The core of the ASTRO (Advanced Spacecraft Truck/Trainer/Transport Reusable Orbiter) was the answer to a question the USAF had proposed to North American Aviation and Douglas, as well as Boeing, Vought, and Republic: how to train pilots for the X-20 on actual flights prior to the X-20 being built. North American had come back with what they called the STX-15, which was a way of reconfiguring an X-15 to have the projected flight characteristics of an X-20 (except for, of course, the highest speed and orbital parts). The Phase I of Douglas’ ASTRO was their significantly more ambitious counter to the NAA proposal.


A schematic of the ASTRO’s A2 vehicle, which would be both independent for suborbital hops, or be boosted to the point that it could be lifted into orbit by a derivative of the same vehicle. Note the booster nose’s ghostly presence at the far right of the image. Same source as previous. Click for a larger view.

Unfettered by the previously existing X-15, Douglas wanted to build a completely new craft dubbed A2, which would be capable of suborbital hops of about 5000 miles (8000 kilometers) after taking off from a runway under the impetus of a J-2 engine, the same rocket engine used by the Saturn V’s second and third stages. Pilots would get their space training, the USAF would have themselves a reusable vehicle with intercontinental range which could carry ten people, or a similar amount of payload. Two RL-10s, as used on the Centaur, would provide a little extra oomph.

Phase II was where Douglas diverged from the question being asked. Take the A2, modify it so that it only carried one crew and two extra J-2 engines, then stick it nose to bumper on the end of another A2 built to the Phase I spec. Turn it 90 degrees and launch it vertically, with the two separating from each other at altitude and speed (both unspecified). The sole crew member aboard the booster would glide back to Earth, while the uppermost A2 would ignite its engines, hopefully after allowing a bit of distance to build from the booster, and carry on into orbit. Douglas projected two crew and about a tonne of cargo to LEO in this configuration.

Phase III scaled up the booster, now dubbed B, and equipped it with two J-2s and one M-1, a never-built LH2/LOX engine that dwarfed even the F-1 engines used on the Saturn V’s main stage. Also launched vertically, this would be the ultimate version of the craft.

The full, two-stage Phase III vehicle was to have been 159 feet long (48.5 meters) and while mass was not mentioned the propellant capacity of the stages (165,000 pounds for the A2 and 594,000 pounds for the B) are—this suggests a total loaded vehicle mass at launch of about 380 to 400 tonnes. Total payload, as mentioned previously, was about nine tonnes, including crew, and there’s a sign that Douglas was nervous about this: the article specifically mentions wanting to launch due east from the Equator, which is an odd thing to be suggesting in 1962, well after the US had committed to launching from the continental USA.

If built, the program was expected to run from 1964 to 1970, with the first flight of the Phase III craft at the end of that period.

What happened to make it fail: It’s difficult to fit the ASTRO into the chronology of the X-20. Phase I appears to have been an attempt to come up with a “Gemini” for the X-20’s “Apollo”, giving the USAF the capability of sending pilots on long suborbital jaunts to train them for the environment they’d encounter when aboard the fully orbital X-20. Phase III would then have been a follow-up to the X-20, increasing crew capacity and payload over that craft.

If this is the case, then, it explains why the ASTRO never went anywhere. The craft made its sole notable public appearance in September of 1962, and American Secretary of Defense Robert McNamara was definitely thinking about cancelling the X-20 no later than March 1963—and possibly earlier. When the X-20 was stopped, then ASTRO would go with it. This is particularly true if one assumes, as seems likely, that the USAF was never very warm about the idea at all, and that it primarily existed as a pitch from Douglas leaked through Missiles and Rockets magazine to drum up support. There’s essentially no reports or discussion of ASTRO post-dating the magazine’s unveil.

What was necessary for it to succeed: It’s not easy to see a way forward for this one. X-20 was dead in the water less than six months later (eventually being formally cancelled in December 1963), and the payload capacity of even the Phase III ASTRO was marginal for what would have been an expensive program. There’s also the issue of Douglas vastly exceeding the question posed by the USAF—it’s unclear that there was any interest on the part of the Air Force in anything other than Phase I. This in turn defeated the purpose of building a fully operational craft for pilot training.


“Air Force Studies Space Trainer”, Missile and Rockets. September 3, 1962.

STAR: The USAF’s “Everything” Spacecraft


STAR, the Space Technology and Research Vehicle. Based on a Poseidon missile MIRV (though upscaled), by the 1980s it was a candidate to be a research spaceplane in the mold of the X-15 as well as a cheap, re-usable operational craft for the USAF. Public domain image from the DARPA document Spaceplane Technology and Research (STAR). Click for a larger view.

What it was: An early 1980s proposal to build a research spaceplane along the lines of the X-15 program of the 1960s. To defray costs it would also have been an operational system, designed to do as many things as possible as a supplement to the relatively limited Space Shuttle. It was a slim, small spaceship capable of taking one crew and would have been taken to orbit in the cargo bay of the Shuttle, on top of a modified Peacekeeper missile, or eventually part of an air-launched stack fired from underneath a heavy-lift version of a 747.

Details: In the early 1970s the US Navy looked at a submarine-launched manned spacecraft intended to attack Soviet spy satellites in the event of a war—the idea being that it could be launched from an undetected submarine, perform its mission in less than one orbit, and then return to Earth before being picked up by Soviet radar.

The basic concept foundered on fitting the so-called “Space Cruiser” into a Poseidon missile tube aboard a sub: even a much stripped-down design was hard pressed to fit into one. Anything that could fit wasn’t even going to get to orbit on its own. There’s not a lot of detail available about this early phase of STAR, but one presumes that a submarine looking to send one into space would have to surface, and then have the sub crew remove it from one tube and place it on top of a warhead-less missile in another.

Ultimately the Navy lost interest, but one part of the initial design lived on. The Space Cruiser was a very long, thin cone, taking advantage of work that had been done on the hypersonic characteristics of the MIRV warheads for the Poseidon missile—though much larger, the Space Cruiser would have encountered the same conditions during re-entry.

The designer of the alpha version of the craft, Fred Redding, was a civilian contractor and so by the late 70s he had managed to extract his work from the Navy and revive it under the auspices of DARPA and the US Air Force. Now the Space Cruiser—redubbed Spaceplane Technology and Research (STAR)—would be launched more conventionally, but otherwise was very similar: long and pointed, with only a minor change from a circle-based cone to one with an elliptical base. This had the twin advantages of increasing its internal volume (as the STAR was always starved for propellants) and turning the craft into a lifting body: the original Cruiser needed small aerodynamic strakes, which were difficult to make in a way that could withstand re-entry, but STAR would have stability and cross-range capability solely as a consequence of the shape of its fuselage.

The bigger change in STAR was its goal. Reading the project’s final report from 1984, one gets the sense that Redding felt burned by the Navy withdrawing funding. Accordingly this time he spread STAR’s purpose as far and as wide as possible. For DARPA he was proposing a research craft, specifically modelled on the X-15, that would provide insight into flight into hypersonic travel in the atmosphere, in low Earth orbit, and even as high as geocentric orbit. Paired with this were suggestions from a large number of defense contractors for research questions, with the goal of demonstrating that private industry might pony up some or all of the necessary money for flights that investigated them.

The Air Force got a research vehicle too (Redding specifically mentions a mandate from the Air Force’s Aerospace Medical Division to gather biometric data on humans in microgravity), but for them and the Department of Defense STAR was more an operational vehicle. While the idea was that at first it would be primarily for research it would be extended in a variety of ways as more was learned about flying it. STAR was also specifically tuned to address a number of failings in the Space Shuttle both from the standpoint of the American military and the Shuttle’s overall capabilities, such as lack of maneuverability in space, inflexible launch schedules, and the vulnerability of its launch facilities to military attack.

The basic STAR was intended to be as small and cheap as possible. It would take only one man to space, and do it in Spartan style. The crew compartment would be unpressurized and was only big enough to sit in: the astronaut would have to stay seated in a spacesuit for the duration of his mission. The craft would have no hydraulics, or an ejection seat, or even landing gear. Instead it would finish its ride home under a parawing, like the one originally planned for Gemini. As it would usually end up on land in the US and was relatively simple in design, refurbishing costs and turnaround time would be kept at a minimum. As the lack of an ejection seat suggests, Redding was also a bit contemptuous of the safety culture that had evolved in NASA since the mid-60s, and spec’d his proposal to the test-as-you-fly/fly-as-you-test standards of years earlier.

STAR itself wasn’t intended to get into space under its own power, and is best thought of as an orbital runabout. It would have been eight meters long and only a meter and a half tall at its aft end, tapering down to a fine point at its nose. The nose itself was designed to fold back at a hinge four meters down from the tip of the STAR, producing a compact package just four meters long. While it wasn’t the primary reason for the folding nose, Redding points out that the cost of shipping something in the Space Shuttle’s cargo bay was the greater of two numbers based on length and mass (a mere 4500kg in the case of STAR, folded or unfolded), and so the compacted version would save quite a bit of money at a time when NASA’s carriage fees were rising dramatically.


The more-capable STAR/Centaur-SP combination, which would have been capable of getting to geosynchronous orbit. The Centaur tank could be left behind for use as a small space station module. Public domain image from Spaceplane Technology and Research (STAR).

At first the Shuttle would be used to lift a STAR into orbit, and potentially even two or even three at a time, and deploy them from its cargo bay. Once there it could tool off on its missions and either return to the Shuttle when done or head back to Earth on its own. In situations where the STAR needed to go higher than its on-board propellant would allow (a figure of about 1650 kilometers is quoted), the Shuttle could instead lift a STAR mated to a truncated Centaur stage with a single RD-10 engine, dubbed the Centaur-SP, an arrangement which would just fit into the NASA craft’s cargo bay lengthwise. On top of one of these, a STAR could travel as far as geosynchronous orbit and return, plus the plan was to keep the emptied Centaur tank in orbit to use as the base of a small space station if so desired.

Being able to divorce themselves entirely from NASA was apparently one of the Air Force’s goals, because the plan was to eventually move STAR on top of its own launch vehicle. Mid-term, the idea was to put it on top of a man-rated three-stage MX Peacekeeper missile, then the newest and hottest rocket in military hands (the first of them had been test-fired the year before the STAR proposal was published). One wouldn’t be able get a STAR to orbit by itself, but the STAR could do the rest of the work and reach LEO with some internal propellant remaining.

This was less than ideal, though, as the STAR’s peculiar shape didn’t allow for much fuel on-board and ideally you wanted its tanks full after leaving the atmosphere. This led to the long-term solution, developing an air-launched launch vehicle. Nose folded, the STAR would be put in an aerodynamic fairing on top of a two-stage rocket. The first stage would be a Titan III derivative and burn LOX and LH2, while the second would be one of the aforementioned Centaur-SPs. The Titan III stage would be assisted by two recoverable strap-on boosters again derived from a Titan III, but not as tall and so carrying less fuel.

This whole works would then be strapped to the underside of a 747-200F, the freighter version of the then-current intercontinental 747, with its landing gear increased in height by four feet and fixed into place to make room for its spacebound passenger. The jet would take off normally and lift the STAR stack to an unspecified height and then drop it, at which point the stack’s engines would fire and the astronaut aboard begin his climb to orbit.

However it got there, once the STAR was on its own in space it would burn N2O4 and a blended fuel based on UDMH located in two tanks immediately to fore of the pilot. In front of that, just past the nose hinge, was the payload bay. As can be imagined, the bay was not large: just eight cubic feet, or 0.2 cubic meters.

While it was up and moving around, STAR was intended to have a variety of operational missions. It was proposed to use it as a repair craft for satellites, a way of getting a man close up to a satellite just for inspection purposes (including potentially for objects not belonging to the US), and even as a weapons platform for shooting them down. It could be used as a rescue craft, and was cheap enough and small enough to engage in “buddy system” missions needing two STARs in orbit at the same time.

Once its mission was done, the STAR would return to Earth. As mentioned previously the craft would have an elliptical conical shape, which would give its pilot some control as it re-entered. This reveals one more interesting detail: the original Space Cruiser design, with its circular cone, was largely retained in the STAR vehicle and re-named the “substructure”. The external shape was maintained by a removable aeroshell, which had the advantage of greatly decreasing the turnaround time of a STAR: the internal “spacecraft” part could be extracted from the aeroshell and the latter replaced. While whatever necessary work was done on the bits that had actually been exposed to re-entry heat, the guts of the STAR could fly again in new clothes.

Once the STAR landed under its parafoil, it would be retrieved—and at 4500 kilograms, it wouldn’t be hard to retrieve from almost anywhere on land. If the mission was in the latter days of the program when the 747-based launcher was available, the jumbo jet could also serve as a carrying craft to get it back to base.

If the STAR program had gone ahead, three Shuttle payload opportunities in 1987, 1988, and 1989 were targeted for initial flights.

What happened to make it fail: The ground was shifting quite rapidly under STAR. When Redding made his final report to DARPA, the Air Force, and the Department of Defense in August 1984, the Soviet Union was seven months away from getting Mikhail Gorbachev as its new leader.  The US and Soviet Union would soon sign the INF Treaty, Eastern Europe would break free of Soviet domination, and SDI became a dead letter. The US military suddenly lost much of its interest in space.

After that STAR was left with only its worth as a research test bed. The USAF and the Department of Defense decided not to go ahead with it. DARPA apparently demurred too, though the reasons there are less obvious. One presumes that without military money the defense contractors which had expressed an interest in the program backed away too, and DARPA didn’t want to be entirely on the hook for funding the project.

What was necessary for it to succeed: Besides a change to the wider course of US/Soviet relations, you can also argue that STAR ran into trouble because it was at the tail end of a long-existing argument in space operations: “do we need a man on this, or can we get what we need with an automated system?” As a result, it’s difficult to get it to fly unless you can come up with some way to have it follow on to the X-15 more closely than it did, back into the era when a pilot was more necessary, or keep it in the 80s and get rid of the man on board.

It’s interesting to compare STAR to the current X-37B. So far as can be told from its classified flights the latter spacecraft covers much the same ground for the Air Force as STAR would have: testing spaceplane technology, apparently making dry runs of orbital rendezvouses, and landing horizontally on a regular landing strip. The major difference is that it does so unmanned, the state of the art having advanced even further than it had in the mid-80s—and it can do so for much longer periods of time (the two missions flown to date having been 224 and 469 days long). Something like STAR, with a one-man crew on board, was too extravagant for any time after the late 60s or early 70s.