09 November 2018

Another Look at Staged Reentry: Janus (1962-1966)

The M2-F1 lifting-body glider (left) and its successor, the M2-F2. Of the experimental lifting bodies NASA built and flew, the Janus spacecraft would have most resembled these pioneering aircraft. Image credit: NASA
In 2013, while spending a gleeful Sunday afternoon searching through old patent applications (don't judge me), I stumbled upon an intriguing design for a piloted spacecraft using "staged reentry." I wrote about it on my old Beyond Apollo blog on the WIRED website.

In 2017, I expanded that post with more context details on the history of lifting body research and better illustrations and posted it on this blog (see the link at the end of this post). At the time, the patent application, filed in January 1964 by TRW engineers C. Cohen, J. Schetzer, and J. Sellars and granted in December 1966, remained my only source of information on the staged reentry concept.

No longer. One benefit of working at a university is that journal articles formerly locked up behind paywalls, out of reach of independent scholars on a budget, are now readily accessible. Last month, while spending a gleeful Sunday afternoon searching through the 1965 volume of The Journal of Spacecraft & Rockets, I stumbled upon a staged reentry design named for Janus, the two-faced Roman god of endings and beginnings. Closer examination confirmed that the Janus spacecraft was indeed the unnamed spacecraft of the 1966 patent.

Janus is an apt name for the proposed spacecraft design, because its most unique features are related to launch and (especially) landing - that is, the beginning and ending of its mission. The name was first used in a confidential May 1962 TRW Space Technology Labs report by I. Spielberg and C. Cohen.

Spielberg, whose name does not appear on the patent application, presented the staged reentry concept at the first conference of the American Institute of Aeronautics and Astronautics in Washington, DC (29 June-2 July 1964) along with Cohen, whose name was the only one to appear on the 1962 report, the 1964 presentation, the 1965 Journal of Spacecraft & Rockets paper based on the presentation, and the 1966 patent. It seems likely, given his continuous involvement, that Cohen originated and championed the Janus staged reentry concept.

Patent applications are not engineering papers; or, perhaps, one may say that lousy is the engineering paper that reads like a patent application. In addition to being more readable, the 1965 Spielberg and Cohen paper provides considerably more detail than the patent application.

The TRW engineers explained the rationale behind the staged reentry concept:
A manned system should provide precision and flexibility in its landing characteristics. It should be capable of routine launch and routine return without a large recovery task force. Moreover, these criteria must be satisfied without curtailing payload volume or weight or reducing the reliability of reentry protection. In general, these requirements conflict, since efficient entry vehicles (e.g., blunt lifting bodies) have poor landing characteristics, whereas vehicles that land well (winged configurations) tend to have low volumetric efficiency and serious reentry design problems. The staged reentry concept. . . circumvents the difficult design compromises that otherwise must be made to ensure good landing qualities, high volumetric efficiency, and desirable reentry characteristics.
The Janus spacecraft comprised two parts that would separate in flight. The largest part was a 26.8-foot-long, 16-foot-wide, 10-foot-deep "pod." Designed to carry three astronauts, it was an 11,660-pound half-cone lifting body with flat aft and top surfaces and a curved, blunt nose.

The TRW engineers described the pod's double-walled structure. Its inner hull, the pressure vessel, would be manufactured from aluminum sheet. The outer hull would be made of aluminum honeycomb with aluminum alloy plates for added strength. Aluminum frames with "I" and "Z" cross-sections would link the two hulls. An ablative heat shield (that is, one that chars and erodes to carry away heat) would cover the aluminum honeycomb, and low-density insulation would fill the space between the inner and outer hulls.

Cutaway view of the Janus spacecraft. Image credit: U.S.Patent Office
The other part of the Janus spacecraft was a 4000-pound delta-wing jet aircraft measuring 21 feet long, 13.3 feet across its wings, and 5.33 feet tall. It would include twin downward facing rudder fins and a belly-mounted air intake feeding a Continental 356-23 turbojet engine. The engine could be started at 18,000 feet of altitude using ambient air or at up to 45,000 feet with supplemental oxygen. Cruise speed at 30,000 feet was about Mach 0.6 (370 knots) and range with a full load of 77 gallons (500 pounds) of jet fuel was 200 nautical miles.

The flat top of the small jet would form the largest part of the top of the lifting body. The jet's underside would form the "ceiling" of the lifting body's 860-cubic-foot pressurized internal volume; that is, the plane's belly, including its air intake, would protrude into the main crew living and working space. Ceiling height, though variable, would measure no less than seven feet.

The jet would ride on three rod-like "pneumatic/explosive actuators" attached to the pod. Latches would link the actuators to holes in the plane's nose and on the underside of its wings. Other latches would anchor the jet's wing leading edges.

Spielberg and Cohen recognized that creating an air-tight seal between jet and pod would pose significant design challenges. They proposed an inflatable or "fluted" (grooved) gasket, presumably made of a rubberized fabric. They admitted that their seal system, though "feasible," was not yet "optimized."

Atop a booster on the launch pad, jet and lifting body would point their noses at the sky. Spielberg and Cohen envisioned that the flat aft surface of the pod would sit atop a launch vehicle adapter that would measure 10 feet in diameter where it linked to the pod. The bottom of the adapter would match the larger diameter of the launch vehicle upper stage.

Just before launch, the astronauts would pass through a hatch in the side of the adapter. Overhead they would see the flat aft surface of the pod, which would include a round hatchway. The hatchway would lead into a cylindrical airlock just large enough to hold one space-suited astronaut. A round hatch in the airlock would in turn lead into the pod. In the near-vacuum of low-Earth orbit, the airlock would permit astronauts to spacewalk without depressurizing the pod.

Forward-facing crew couches would be arranged single-file, one behind the other, in a line beneath the jet fuselage. This would place the astronauts one above the other on the launch pad.

The pod would contain the Janus spacecraft main control console. Intended for use in orbit, it would be mounted on the pod's aft interior wall next to the inner airlock hatch. This would place it out of reach of the reclining astronauts. Critically important controls would be mounted on couch arms.

The patent application said nothing about possible launch vehicles, but in their paper Spielberg and Cohen specified two candidates: Titan III (probably the Titan IIIC variant) and Saturn C-1 (otherwise known as Saturn I). The former could boost 28,000 pounds into the 140-nautical-mile orbit required to forestall orbital decay long enough to carry out a two-week Janus mission; the latter, 20,000 pounds. The total weight of the Janus spacecraft (crew, pod, and jet) was 15,660 pounds, so in theory it could transport 12,340 pounds of unspecified payload if launched on a Titan III and 4340 pounds if launched on a Saturn C-1.

It is worth noting that Janus included no docking mechanism, and that was it not designed to perform significant maneuvers in space (apart from a deorbit burn). This ran against the grain of NASA requirements in the first half of the 1960s, when both Gemini and Apollo were under development. Though it could carry a hefty payload, it could not deliver it anywhere. Presumably, this meant that its payload would always take the form of equipment that would remain inside the pod. It is conceivable, however, that small payloads could be tossed out its airlock and larger ones assembled outside by spacewalkers - Spielberg and Cohen did not, however, suggest these possibilities.

A successful mission would begin with launch from Cape Kennedy on Florida's east coast. The launch vehicle would ascend vertically, then roll toward the southeast on a course that would avoid Caribbean islands and South America. About 10 minutes after liftoff, Janus would reach its operational orbit and separate from the upper stage of its launch vehicle. The crew would then unstrap from their couches and begin work in the pod's large pressurized volume.

They would also work in the jet cockpit. The jet's glass canopy, which would stand higher than the rest of the Janus spacecraft's mostly flat top, would make the cockpit the prime spot for conducting Earth and astronomy observations.

Spielberg and Cohen proposed a novel method for entering and leaving the cockpit. The crew couches would each be mounted on a pair of rails, and the underside of the jet's fuselage would include automatic doors. Operating controls on the couch arms would cause the doors to open and the couch to ride the rails from pod to cockpit and vice versa. The TRW engineers explained that a single set of couches shared between the pod and the jet would save weight, though with the large Janus payload capability this would probably have been a minor concern.

The crew would breathe a 47% oxygen/53% nitrogen air mix at a pressure of 7.5 pounds per square inch. Water for crew needs would come from fuel cells, the primary task of which would be to generate 2.5 kilowatts of continuous electricity by combining liquid hydrogen and liquid oxygen. Fluid circulating in pipes in the pod walls would gather and carry waste heat from the pressurized volume and the fuel cells to a radiator mounted on the pod's aft surface.

For return to Earth, the astronauts would sit in their couches in the pod, turn the Janus spacecraft using small thrusters so that its aft end pointed in its direction of motion, and ignite its 1100-pound solid-propellant retrorocket. After burnout, the retrorocket casing would be cast off and Janus reoriented with its nose aimed forward. Descent toward 400,000-foot reentry altitude would last 14 minutes. At start of reentry, the Janus spacecraft would be moving at about 250 feet per second (fps).

Reentry would be a balancing act. The lifting-body pod would need trim flaps for stability and steering; however, four trim flaps attached in pairs to the bottom edge of its flat aft surface would tend to tip its nose down (that is, give it a negative angle of attack). This would permit hot reentry plasma to course over the pod's top surface, destroying the jet canopy. At the same time, the pod would be tail-heavy, raising its nose and making it aerodynamically unstable.

Spielberg and Cohen proposed a two-part solution: cautiously reshaping the pod's nose and packing its triangular nose volume with heavy subsystems (for example, the fuel cells and their reactants). The former would tend to level its angle of attack and the latter, they calculated, would shift its center of gravity forward to a point 54% of its length (about 11 feet) aft of the pod's nose, yielding a slightly "nose up" angle of attack. The pod's nose would thus bear the brunt of reentry heating, and no reentry plasma would reach the jet canopy.

The Janus spacecraft would reenter at a very shallow angle (just 2°). It would thus shed speed gradually in a low-density atmosphere, preventing maximum deceleration from exceeding 1.9 gravities. An automated attitude control system would operate the trim flaps and small thrusters to maintain stability as the pod descended.

During reentry, the outer hull, safe behind its heat shield, would maintain a temperature below 600° Fahrenheit (F). The inner hull would remain at 70° F throughout the mission. The hot outer hull would tend to expand. If the aluminum frames linking the inner and outer hulls were rigidly attached at both ends, differential expansion would tear them apart. To avoid this, Spielberg and Cohen proposed that the frames be attached to the outer hull by flexible connections and to the inner hull by rigid ones.

A little less than 12 minutes after reentry start, at an altitude of about 120,000 feet, the Janus spacecraft would slow to a velocity of about 50 fps. Deprived of lift, its angle of descent would increase in a little over a minute to about 55°.

At 50,000 feet of altitude, the Janus spacecraft would slow to subsonic speed and begin to lose stability. The mission commander would activate the motors that would raise the three couches into the jet cockpit. Beneath the astronauts' feet, the fuselage doors would close and seal. At 45,000 feet, the spacecraft would slow to Mach 0.9, and jet separation from the pod could occur.

Separation would begin with a command to fire explosive bolts. This would release the latches linking the jet to the pod so that the three rod-like pneumatic actuators could extend, pushing the jet away from the pod with a jolt. The pressure seal would be breached, exposing the pod's interior to the outside environment.

The commander would ignite the jet's engine and fly at a cruise altitude of 30,000 feet to a waiting airfield up to 200 nautical miles away. The jet would land on a nose wheel and skids on the ends of its rudder fins. The pod, meanwhile, would deploy parachutes from its aft surface and descend to a landing on its nose.

In the event of an abort on the launch pad or during first-stage operation, a pair of solid-propellant abort rocket motors mounted on the pod's aft surface outside the adapter linking it to the launch vehicle would ignite to boost the Janus spacecraft up and away. The motors would propel it to an altitude of 6600 feet in 19 seconds. If no first-stage abort took place, the abort motors would eject after second-stage ignition so that the launch vehicle would not need to carry their weight to orbit.

The deorbit rocket motor would play two possible abort roles: in an abort off the launch pad, it could be ignited after the twin abort rocket motors burned out to boost the Janus spacecraft higher and farther downrange, providing more time for successful jet separation; it would also become the primary abort rocket motor after the twin abort motors ejected.

An abort within 200 nautical miles of Cape Kennedy would see the commander separate the jet from the pod as during a normal descent, then fly back to the launch site. The jet could also remain attached to the pod throughout the abort, in which case the entire Janus spacecraft would descend nose down on parachutes to a landing or splashdown at 25 feet per second. Spielberg and Cohen included 1030 pounds of recovery gear in the Janus spacecraft mass budget.

Down-range aborts - for example, during second stage flight - would occur over open ocean, placing land - never mind suitable airports - outside the jet's 200-nautical-mile range. Spielberg and Cohen noted that the lifting body would during second-stage flight be high enough to use its trim flaps and steering thrusters to maneuver closer to land. This would, they judged, permit jet separation within 200 miles of airfields on Caribbean islands or in northeastern South America.

Here is the link to my staged reentry post based only on the Cohen, Schetzer, and Sellars patent of December 1966. In addition to a summary history of lifting body development in the United States, the post contains detailed labeled drawings from the patent application.


"Janus: A Manned Orbital Spacecraft with Staged Re-Entry," I. N. Spielberg and C. B. Cohen, The Journal of Spacecraft & Rockets, Volume 2, Number 4, July-August 1965, pp. 531-536

Patent No. 3,289,974, "Manned Spacecraft With Staged Re-Entry," C. Cohen, J. Schetzer, and J. Sellars, TRW, 6 December 1966

Related Links

X-15: Lessons for Reusable Winged Spaceflight (1966)

Where to Launch and Land the Space Shuttle? (1971-1972)

What if a Shuttle Orbiter Struck a Bird? (1988)

NASA Johnson Space Center's Shuttle II (1988)


  1. David, thanks for updating and refreshing this article. It reminds us tha sometimes in the early days of complex endeavors, untenable ideas are not quickly recognized. The complex sequence of transition from the habitation compartment into the jet cockpit after reentry and while dropping through 50,000 feet must surely be a good example of that. I don’t see why they couldn’t be buttoned up safely in the cockpit before deorbiting. Failure of the seat repositioning systems or of the fuselage panels closing mechanisms at the last minute would have had tragic effects, not to mention the complications of doing all that with one or more injured or incapacitated pilots. Remember the June 1966 XB-70 ejection scenario in which one man was injured by the closing clamshell doors and the other failed to eject at all.

  2. Hi, John:

    All that you say is true. Perhaps more fundamental, the jet seems very small - maybe too small for the task assigned to it. It's hard for me to see how three astronauts and their couches could fit into the fuselage ahead of the air intake. It's telling that their patent drawing shows only two couches.

    Still, I wonder whether this could have some application - maybe off Earth. Perhaps it would be a way to deploy automated gliders at Venus or Titan?


  3. David;

    Thanks for updating the staged-reentry concept it's an interesting one to say the least. Which I'm sure (saying the least :) ) is what you'd prefer I'd do I'm going to take some shots at 'justification' for the concept. Yes it's complex and unwieldy but despite the patent and now expanded "Janus" concept it is still a very rough concept which obviously could have been refined over time if anyone had been interested.

    As you note the use of 'separate seats' would have obviously been something that could be done in the design but in "reality" as the design went along the major bias of the authors would have been shown that they considered the "lifting body" the main vehicle and the "jet" simply as a means to get the astronauts to a runway. Which is arguably the opposite of how it should have been done. the Jet was and should have been the main 'residence' of the astronauts from launch to landing. Moving the seats was a too complex and risky system to assume would function fully at and under any given situation or scenario. The suggested length of an air tight 'seal' is also something that would never pass a second glance. Worse the idea of having your 'aircraft' intrude into your pressurized volume is, well as a professional wrench turner I'd tend to use one for 'attitude correction' of the engineers who would suggest such. Aircraft leak, they out-gas, and they are very clear paths for your 'pressure' to leak out along the pathways they normally use in the air to have required flow. In other words even if your ungainly long 'seal' works as advertised your going to leak all your air out THROUGH the jet unless it won't work as a jet. No what you do is put a 'top' on your pressure vessel at the 7 foot mark (noted the 'least' ceiling height) and put a hatch in the bottom of the jet and top of the pressure vessel. Bit more mass but well worth the cost.

    I'm pretty sure they were assuming the "jet" design was variable since as you note the patent shows only two seats while the article mentions three. Both the T-37 and T-38 were "available" for modification and I'll point out the T-38 was in fact considered as the basis of a "Spaceplane" trainer:

    (And I'll note there's a nice pic of a proposed four-seat+ "light airliner" using the same plane as a basis so up to four astronauts might be possible)

    And I'll also note this was far from the first time "jet" and "spacecraft" came together. I am unable to find an available link anymore, (thankfully I printed it out way back when, I just have to figure out where that printout is :) ) but perhaps you can:

    NASA Tech Memorandum (TM)X-656, dated 1963 for a “Boost-Glide Reentry Vehicle suitable for Operation as a Supersonic Aircraft” which details a sub-orbital or orbital delivered Mach-2 jet powered fighter OR a 5 or 6 person folding wing jet powered "shuttle" craft, take your pick.

    I agree that the design would have needed more work, the need for orbital maneuver and probably docking systems is I think a given and shouldn't have been that hard to incorporate in the design. Overall it address the "issues" that the authors put forward, but the main question is are those 'issues' as concerning as suggested?

    1. I think that this concept was probably left over from early (pre-lunar) Apollo planning. Lots of companies threw ideas around in the 1960-1961 timeframe, some of which were downright weird, and this concept seems to belong with those.

      What you say about the leaky jet is interesting. I suspect that preventing leaks would have caused the jet to be heavier than it would have been otherwise. The jet already had some problems, I think - it was too small to carry three people any distance.

      I don't take this concept very seriously, but I wonder whether it might have application in the robotic program. I could see a lifting body plummeting through the atmosphere of Mars or Titan, releasing gliders as it fell toward destruction.

      Thanks for your comments - they are astute.


  4. To put it bluntly the Lifting Body could easily be brought down in a general area required and then landed full by parachute and retro-rockets without resorting to a separate jet aircraft. That was the original idea behind the M1 Lifting Body itself. The M1 grew into the M2 due to a desire to retain control down through trans-sonic and subsonic speed first by adding an inflating aft section and then using a 'solid' aft section to create the M2F1 (and later M2F2/3) design. The eventual limitation of parachute landing due to parachute size made the idea of transitioning to a more fully capable flight vehicle tempting.

    But in a similar quandary the Lifting Bodies were always going to be 'hot' to land due to the mass-versus-lift. The flew "well" as long as they were lightly loaded but once you started putting actual flight system in them the landing speed grew accordingly.

    On the converse side a 'lifting' winged reentry vehicle below a certain size has increased thermal and L/D issue as well which has become clear over the course of such programs and Hermes and OSP. The X-37B gets away with a rather robust TPS because it's so light due to being unmanned and with minimum systems. The conceptualized manned X-37C requires either a very heavy metallic TPS or something like the Shuttle Tiles to survive reentry. Ablating heat shields don't work well with larger 'lifting' designs due to the uneven way they ablate during reentry.

    This concept tends to off-load the different and often contradictory requirements into a more 'proper' relationship. The lifting body has a larger surface over which to spread the heating and uneven ablating will tend to even out due to the lower LD factor of the entry profile. Meanwhile once below 'stable' flying speed the aircraft can take over and utilize the century old art of flying without having to compromise the design significantly to accommodate reentry and hypersonic flight. As a bonus you get an integrated mission module with extra pressurized volume to boot!

    In the end though the main question is at what point does this concept do the job better than others? And here the call is a lot closer due to assumed or imagined requirements. Arguably the concept fails to have the utility and simplicity of the capsule, offers some advantages over either the pure lifting body or lifting winged vehicle at a trade off in complexity, but probably not enough to be a clear 'winner' in any category. But it really depends on what your requirements are when boiled down to their essence.

    So on the Gripping-hand I'll finish off with presenting the argument that the concept actually has utility if you think it through. Maybe not fully as an orbital vehicle but I'm reminded that Rocketplane Global and such concepts try and make an aircraft into a rocket when something like this as a basis gives you the ability to use a more 'off-the-shelf' aircraft by melding it with a rocket. In other words the "lifting body" becomes a rocket booster and reentry shield that is fully recoverable fitted to carry an rather conventional aircraft to 'space' and back. Sure it sounds a lot more complicated and in-efficient but more practically you are not seriously compromising the vehicles to turn them into less efficient (and more costly) versions by trying to make the hybrids which do neither job as well as the dedicated one would.

    A thought :)


    1. Randy:

      A host of interesting thoughts! Thanks for posting your comments.


    2. Anyway, you wrote:
      [quote]Still, I wonder whether this could have some application - maybe off Earth. Perhaps it would be a way to deploy automated gliders at Venus or Titan?[/quote]

      Too far afield really as your last two articles in fact point to the staged-entry being in rather common use and concept ;) After all isn't that how MOST atmospheric entry probes work in that they 'stage' their entry system prior to landing or operation? Ok, overthinking it I'm sure but really this is the 'logical' end point of such systems as applied to manned spaceflight. Separating the two 'stages' so as to maximize the utility of each but it begs the bigger question of economics since you now have two vehicles that have to be integrated balanced against not having to compromise a single vehicle in order to do two different jobs.

      One of the major 'crippling' factors is the Janus design itself if compromised in ways that don't themselves make a lot of sense.

      [quote]I think that this concept was probably left over from early (pre-lunar) Apollo planning. Lots of companies threw ideas around in the 1960-1961 timeframe, some of which were downright weird, and this concept seems to belong with those.[/quote]

      Makes sense especially as you note it had little utility for either Apollo application since it wasn't going to make a very good Space Station ferry either. Though the use of the 'pod' as a "mission module" clearly points to a pre-space station mission architecture. I suspect it got little official support from either NASA, (not a capsule) or TRW. I just don't think it was as well thought out as it could have been given the work to put together a presentation, article and patent.

      [quote]What you say about the leaky jet is interesting. I suspect that preventing leaks would have caused the jet to be heavier than it would have been otherwise. The jet already had some problems, I think - it was too small to carry three people any distance.[/quote]

      About a thousand pounds of caulk I'm sure :) As for the capacity that at least made some sense because you actually COULD get three people into a similar sized jet. If you look at the two nearest examples of working aircraft to the proposed Janus jet, you'll note that both the TT-1 Pinto (http://all-aero.com/index.php/55-planes-t-u/10990-temco-tt-1-pinto) and the Beech Jet Mentor, (http://all-aero.com/index.php/59-planes-b-c/14015-beech-73-jet-mentor) which are near the mass needed could easily fit three in-line once you remove the secondary instrument and control clusters of the trainer aircraft.

      I'd still think something like the T-37 would be better since I think it would be easier to expand but frankly the authors don't appear to have put as much effort into the design as I'd have thought. Both the speed and range is questionable on a single engine. (Note both above examples had issues don't 300 knots and require MUCH more than "77 gallons" to have a 200 mile range)

      In essence the Janus seemed to have initially been a suggestion to move beyond the simple Mercury program with a spacecraft that could 'double' as a mini-space station but lacked utility beyond that point. Lack of any support other than likely help getting the patent mean that the authors would not have official resources or help to expand and refine the concept leaving a lot of kludges and dead-ends in the concept. It's not as daft or out-there as a general concept but the lack of refinement really does not help 'sell' the idea.

      All in all not a really bad or 'weird' concept but it needs work to bring out the real advantages and identify and correct the built in issues. Thanks again for finding more information on this.


    3. Randy:

      You should write a blog - you're obviously very knowledgeable! I don't mean to say that you shouldn't share your knowledge here, of course (might be interpreted that way, I suppose).

      You're right about staged reentry being a normal thing - it's how InSight landed on Mars a few weeks ago. :-)



I like hearing from my readers. No rules except the obvious ones - please keep it civil and on topic.

Advertiser comments have led me to enable comment moderation.