Integral Launch and Reentry Vehicle: Triamese (1968-1969)

Triamese target: a large Earth-orbital "Space Base" assembled from modules launched atop two-stage Saturn V rockets. The Space Base, expected to be operational by about 1980, would be staffed by up to 100 people. Image credit: NASA.

The Triamese concept originated in 1967 in a reusable launch and reentry vehicle study General Dynamics Convair (GDC) performed on contract to the U.S. Air Force (USAF). Triamese owed its peculiar name to its peculiar launch configuration. At liftoff it would comprise one orbiter element and two booster elements. The boosters would together serve as the first stage; they would also provide propellants to the orbiter's engines during first-stage boost. One booster would attach to the orbiter's flat belly and the other to its rounded back. 

Space launch vehicle concepts with separate reusable booster and orbiter elements were not exactly new in 1967. What was different about Triamese was its strict reliance on a common booster and orbiter design. The Triamese orbiter and booster elements were intended to be virtually identical. GDC explained that

[i]n order to achieve the economy predicted for the Triamese system, the orbital and boost elements must have a high degree of commonality and must represent essentially a single development program. . .This commonality has been obtained by "overdesigning" the boost elements. . .[which] creates performance penalties that are accepted.

GDC called Triamese "a new mixture of aircraft, spacecraft, and launch vehicle." The Initial Point Design (IPD) Triamese launch stack (A, above) would have comprised two booster elements and one orbiter element, all virtually identical. It would have measured 149.5 feet (45.6 meters) tall from the trailing tips of its six rudder fins (two per element) to its three noses. B, a tail-on view of one element, shows the V-shaped, 46.1-foot (14-meter) spread of the rudder fins, 21-foot-wide (6.4-meter-wide) flat belly, and twin XLR-129 rocket engines arranged one above the other. Turning view B 45° horizontally yields view C. The IPD Triamese element would measure 31.4 feet (9.6 meters) from its belly to the tops of its rudder fins. View D displays "switchblade" wings deployed for stable subsonic flight. Wingspan is 107.5 feet (32.8 meters). Image credit: General Dynamics Convair/DSFPortree

The Triamese concept helped to shape NASA's May 1968 Integral Launch and Reentry Vehicle (ILRV) study Statement of Work and the ILRV Request for Proposal the space agency released to U.S. industry in October 1968. When time came for NASA to select four industry proposals for ILRV study contracts in January 1969, it was a foregone conclusion that Triamese would be counted among them.

NASA Marshall Space Flight Center (MSFC) in Huntsville, Alabama, was tasked with managing the GDC ILRV study contract. NASA MSFC was home of the three-stage Apollo Saturn V rocket. At the time of the ILRV study, Apollo Saturn V development, manufacture, and testing were drawing to a close. Managers at the Huntsville center hoped, however, that a two-stage Saturn V variant designated INT-21 might launch a series of increasingly complex space stations in the 1970s.

INT-21 consisted of the first two stages of the Saturn V — the S-IC first stage and S-II second stage — both of which measured 33 feet (10 meters) in diameter. An Earth-orbital payload measuring up to that diameter — for example, a large space station module — would replace the 21.7-foot-diameter (6.6-meter-diameter) S-IVB third stage of the Apollo Saturn V. 

One station program scenario, favored by NASA Administrator Thomas Paine, would see INT-21-launched Apollo Applications Program (AAP) Orbital Workshops — converted S-IVB stages — lead in 1975 to a large drum-shaped station with up to 12 crewmembers. Multiple INT-21-launched large station modules might then be joined together in orbit as early as 1980 to form a "Space Base" with up to 100 staff.

In that scenario, the ILRV shuttle would serve as a Saturn V supplement. The big rocket would do the heavy lifting all through the 1970s, leaving to the smaller reusable shuttle the specialized task of affordably launching astronauts, supplies, replacement parts, and scientific experiment apparatus to the space station and returning astronauts, experiment results, and data products to Earth. 

GDC began its ILRV Triamese study with an Initial Point Design (IPD) based on its USAF study results and inputs from NASA engineers. The IPD Triamese was designed to deliver up to 25,000 pounds (11,340 kilograms) of supplies and equipment to the space station and return up to 2500 pounds (1130 kilograms) to Earth during a single flight. The two boosters and the orbiter would each carry a flight crew of two astronauts, for a total of six. In addition, the orbiter would include a passenger compartment for transporting 10 astronauts to and from the space station. 

Orbiter and booster commonality was not the only cost-saving principle underpinning the IPD Triamese system. Another was use of off-the-shelf technology. GDC proposed, for example, that the design of the Triamese "switchblade" wings, which would enable stable flight at subsonic speeds, should be based on the variable-geometry wing system of the F-111 "Aardvark" aircraft the company manufactured for the USAF. 

The variable-geometry wings of the supersonic F-111 in action. In 1967, the F-111 became the first variable-geometry aircraft to enter active service. Image credit: U.S. Air Force.

GDC envisioned that the IPD Triamese elements would, like operational airplanes, fly repeatedly with minimal refurbishment between flights. The company acknowledged, however, that the elements would be subjected to greater stress during flight than would most aircraft, leading to greater potential for component failure.

GDC proposed to solve this problem by equipping IPD Triamese subsystems with sensors linked to on-board magnetic-tape flight recorders. After landing, data on subsystem performance would be carefully analyzed. Hardware that showed signs of actual or impending trouble would be subjected to detailed inspection and possible repair or replacement. 

The sensors would also enable a detailed on-board checkout capability that would slash costs by allowing NASA to get by with only a simple launch control center. KSC's Apollo Saturn launch control center was expansive and expensive, with many control consoles and an army of highly trained personnel; IPD Triamese launch control might more closely resemble an airport control tower. 

GDC expected that the IPD Triamese design, development, and test program would begin on 1 November 1971 and last until the first operational IPD Triamese flight on 1 January 1977, a period of 62 months. Engineering design would occur between 1 November 1971 and 1 July 1974. Development of the Pratt & Whitney XLR-129 rocket engine, which GDC called a "pacing item," would last from 1 November 1971 to 1 August 1974. Rocket engine tests using IPD Triamese vehicles that were captive  — that is, bolted down so that they could not take off — would take place between 1 March 1975 and 1 March 1976.

GDC proposed a "fatigue test vehicle" to help to ensure that the IPD Triamese elements would be as reusable as expected. This would take the form of a skeletal IPD Triamese element with all systems installed except for the metal plates and insulation blankets of its heat shield. 

Beginning on 1 November 1974, the fatigue test vehicle would undergo repeated propellant tank and cabin pressurizations, switchblade wing, turbofan jet engine, and landing gear deployments, computer starts, and other subsystem activations so that engineers could gain insight into malfunction characteristics and operational lifetimes. The tests would continue into the period of operational IPD Triamese flights.

The Initial Point Design (IPD) Triamese orbiter element differed from the booster element only in detail. Unless otherwise noted, all features called out in the side view drawing above are features of both the orbiter and the booster. A: cockpit for two astronauts seated side by side; B: passenger compartment for 10 Space Station crewmembers with seating arranged in three rows (orbiter only); C: forward landing gear (stowed). D: forward landing gear (down and locked); E: short liquid oxygen tank (orbiter); F: leeward forward pin connection (orbiter only); G: windward forward pin connection; H: cargo bay hatch (orbiter only); I: cargo bay (orbiter only); J: main landing gear (stowed); K: main landing gear (down and locked); L: short liquid hydrogen tank (orbiter only); M: switchblade wing compartment; N: leeward propellant feeds; O: windward propellant feeds; P: XLR-129 rocket engine (one of a pair); Q: body flap with elevons; R: rudder fin (one of a pair); S: rudder flap (one of a pair); T: extendible engine skirt (orbiter only — retracted); U: extendible engine skirt (orbiter only — extended). Image credit: General Dynamics Convair/DSFPortree.

Top view of IPD Triamese element. Unless otherwise noted, all features called out in the top view drawing above are features of both the orbiter and the booster. 1: cockpit windows; 2: cockpit crew hatch; 3: passenger compartment crew hatch/docking unit (orbiter only); 4: turbofan jet engine (stowed); 5: turbofan jet engine (deployed and locked); 6: long liquid oxygen tank (booster only); 7: reinforcing ring for attachment of forward pin connection (booster only) or connections (orbiter only), landing gear, and switchblade wing pivot; 8: switchblade wing pivot (one of a pair); 9: switchblade wing (deployed — one of a pair); 10: switchblade wing flap (one of a pair); 11: switchblade wing (stowed — one of a pair); 12: main landing gear (stowed); 13: cargo bay hatch/docking unit (orbiter only); 14: long liquid hydrogen tank (booster only); 15: aft attachment pin actuator (booster only); 16: leeward propellant feeds (one of a pair); 17: rudder fin (one of a pair); 18: rudder flap (one of a pair); 19: non-extendible XLR-129 rocket engine skirt (booster only); 20: body flap with elevons. Image credit: General Dynamics Convair/DSFPortree.

IPD Triamese flight testing would use "an aircraft approach." All flights would carry two test pilots per element — there would be no unpiloted IPD Triamese test flights. GDC allotted three booster elements and three orbiter elements for the IPD Triamese test program. Of these, two boosters and one orbiter would be carried over to operational flights. 

GDC scheduled 50 horizontal test flights at Edwards Air Force Base, California, between 1 October 1974 and 1 March 1976. During these tests, individual IPD Triamese elements would use their twin TF-34 turbofan jet engines to take off from a runway with their switchblade wings extended to verify subsonic flight and landing characteristics. 

The General Electric-built TF-34 engine generated 12,600 pounds (5715 kilograms) of thrust. GDC was familiar with the engine because it used it in its proposal for the U.S. Navy's S-3 Viking aircraft. The engine produced a characteristic low rumble, a sound that would no doubt have become associated with piloted spaceflight had NASA given GDC the nod to build the IPD Triamese.

A U.S. Navy S-3 Viking aircraft descends to a carrier landing. Visible is one of its two General Electric-built TF-34 jet engines. The IPD Triamese shuttle orbiter and booster elements would each have included two such engines. In the unlikely event that a returning IPD Triamese element missed its first attempt at a landing on the runway at NASA Kennedy Space Center, the jet engines would have permitted a second try. Image credit: U.S. Navy.

The company scheduled 15 single-element rocket-propelled vertical flights at NASA Kennedy Space Center (KSC) on Florida's east coast between 1 September 1975 and 1 November 1976. The tests would, among other things, enable verification of IPD Triamese flight characteristics at transonic and supersonic speeds. 

The IPD Triamese element under test would lift off from one of two launch pads built at KSC specifically for IPD Triamese launches, climb to a specified altitude, and shut down its twin rocket engines. It would then pitch over to horizontal attitude, deploy its wings and jet engines, and fly to a runway at KSC built specifically for IPD Triamese landings. 

In December 1975, the flight test program would shift into high gear as preparations began for suborbital two-element test flights, the first IPD Triamese flights to launch astronauts into space. A pair of joined booster elements would lift off vertically from a KSC IPD Triamese pad on 15 February 1976, separate, and undergo a reentry virtually identical to that they would experience during operational Triamese flights. They would then land on the KSC IPD Triamese runway. NASA would repeat this test on 1 April 1976. 

About two weeks later, on 15 April 1976, the first booster-orbiter suborbital flight test would take place. It would closely resemble the booster-booster tests. The second booster-orbiter test would occur on 1 June 1976. 

The IPD Triamese flight test series would end with a pair of three-element orbital flight tests on 1 August and 1 November 1976. The missions would see the first IPD Triamese dockings with an Earth-orbiting space station. 

The boosters and orbiter flown during the second orbital test flight would be used for "refurbishment verification" — a rehearsal of the normal IPD Triamese post-flight checkout and maintenance "turnaround" process — then the orbiter and one booster would be held in reserve as "standby elements" for the first operational flight of the IPD Triamese program on 1 January 1977.

Availability of standby elements — a backup orbiter and a backup booster — would be a standard part of preparation for every operational IPD Triamese mission. If an active orbiter or active booster suffered damage or malfunctioned and required time-consuming repairs, a standby element would fill in for it so that launch could go ahead as scheduled. This approach recognized the critical role reliable space transportation would play in NASA's space station program. 

GDC proposed that, in addition to the two standby elements, NASA's IPD Triamese fleet should include four active orbiters and six active boosters. The orbiters would each fly once per month, for a total of 48 orbiter flights per year. The boosters would each fly 16 times per year, for a total of 96 booster flights. 

Diagram of IPD Triamese orbiter and booster turnaround flow. In one month, four active orbiters would lift off from Kennedy Space Center, Florida. In the same period, four active boosters would fly once and two would fly twice. A fifth orbiter and a seventh booster would serve as "standby elements" ready to enter the turnaround flow if an active orbiter or booster should be grounded for repairs. Image credit: General Dynamics Convair/DSFPortree.

At the start of every operational IPD Triamese mission, turnaround technicians would load the 17.5-foot-diameter (5.3-meter-diameter), 12.4-foot-long (3.8-meter-long) payload bay located between the orbiter's liquid oxygen tank and its liquid hydrogen tank with 25,000 pounds (11,340 kilograms) of supplies and equipment bound for the Space Station. The orbiter propellant tanks would be made shorter than the booster tanks to make room for the 3000-cubic-foot (85-cubic-meter) bay.

Turnaround technicians would next pump consumables into the IPD Triamese elements. These would include 4660 pounds (2110 kilograms) of jet fuel for each booster and 1610 pounds (730 kilograms) for the orbiter, along with 3820 pounds (1730 kilograms) of attitude control propellants for the orbiter and 1420 pounds (645 kilograms) for each booster. 

The three elements would then be towed to the launch pad on their extended tricycle landing gear, hoisted vertical, and, after their landing gear was retracted, mounted on the pad on three support struts each. After the vehicles were joined to each other by three "pin connections," one forward and two aft, five support struts (the three supporting the orbiter and one each supporting the boosters) would be removed, leaving in place two per booster. 

Launch pad technicians would connect propellant feed lines linking the orbiter and the booster propulsion systems and attach umbilical hoses for propellant tank loading. After a leak check using on-board checkout equipment, they would fill the orbiter's tanks with 362,800 pounds (164,560 kilograms) of liquid oxygen and 51,830 pounds (23,510 kilograms) of liquid hydrogen. Each booster would be loaded with 424,500 pounds (192,550 kilograms) of liquid oxygen and 62,890 pounds (28,525 kilograms) of liquid hydrogen. Before vacating the vehicles, the pad technicians would conduct a final check of the propulsion system using on-board checkout equipment. 

The three flight crews and passengers would board, then the flight crews would perform a final check of all on-board systems save propulsion. Finally, at a time selected to enable a quick rendezvous with the Space Station, the six XLR-129 engines would ignite and power up to 20% of maximum sea-level thrust. There they would briefly hold to allow the flight crews to check engine performance. If all six engines were found to be operating normally, they would power up to 100%, hold-down attachments on the four support struts would disconnect, and the IPD Triamese stack would lift off.

IPD Triamese launch and ascent: the IPD Triamese launch stack (A) would stage at an altitude of 160,000 feet (48,770 meters) (B). The twin boosters would undergo a low-stress suborbital reentry (C), then would level off at 15,000 feet (4570 meters). Their flight crews would extend their jet engines and wings, then fly back in tandem to their NASA KSC base (D), a distance of 185 nautical miles (340 kilometers). The orbiter, meanwhile, would continue its journey (E) to the Space Station in 270 nautical-mile (500-kilometer) low-Earth orbit. Image credit: General Dynamics Convair/DSFPortree.

At liftoff, the four booster engines would each generate 394,500 pounds (178,715 kilograms) of thrust; the two orbiter engines, 380,000 pounds (172,365 kilograms) each. GDC calculated that the IPD Triamese stack would weigh 1,751,000 pounds (794,240 kilograms) at liftoff. Of this, the boosters would each account for 596,450 pounds (270,545 kilograms) and the orbiter, 558,100 pounds (253,150 kilograms).

During the first stage of ascent, the twin booster elements would supply all propellants to their own engines and the two orbiter engines. GDC did not specify how long first-stage flight would last. The company calculated, however, that the entire journey from launch pad to orbit would last only 6.2 minutes. Acceleration during ascent would top out at four times the pull of Earth's gravity.

GDC assumed that NASA's space station destination would circle the Earth in an orbit inclined 55° relative to Earth's equator. IPD Triamese launch azimuth would, however, be set at 35° to avoid overflight of the U.S. east coast early in the ascent phase. This meant that the orbiter would have to perform a westward yaw ("dogleg") maneuver to reach 55° orbit.

GDC estimated that flight conditions during ascent were 500 times more likely to cause a system failure than were conditions in space. As might be expected, engines, propellant feeds, and avionics were the systems most likely to malfunction. The company cited possible failure modes virtually certain to lead to structural failure and loss of life in as little as one second — for example, a hydraulic system failure that would cause the engines of one of the three elements to gimbal (pivot) and lock suddenly. 

To avoid such catastrophic failures, GDC proposed automatic malfunction detection and switchover to backup systems. This approach would, the company estimated, reduce the IPD Triamese catastrophic failure rate to one in 2000 flights.

Switching to backups might allow an IPD Triamese mission to proceed as normal. Even if an abort were necessary, under most circumstances the boosters would return to the KSC runway as normal. The orbiter, on the other hand, might seek to return directly to KSC, reach a low orbit and return to KSC after circling the Earth once (the generally preferred option), bank eastward and land downrange on the North Atlantic island of Bermuda, or, in the worst-case scenario, ditch at sea or crash-land on the Arctic ice cap. 

Booster thrust per engine would increase to 433,300 pounds (196,540 kilograms) just before burnout. The orbiter engines, meanwhile, would each extend an expendable skirt just before staging, allowing an increase in thrust per engine to 460,500 pounds (208,880 kilograms). 

The boosters would expend their propellants as the IPD Triamese stack reached a speed of 6800 feet per second (2070 meters per second). After booster separation, thrust per orbiter engine would steadily decrease until it reached 310,000 pounds (140,620 kilograms) just before shutdown. 

After they separated from the orbiter, the boosters would perform a suborbital reentry and turn toward KSC. They would deploy their switchblade wings and jet engines and fly back to base at a speed of 225 miles (365 kilometers) per hour. 

Staging during ascent to orbit: the operations illustrated above would last no longer than nine seconds. The orbiter (A) is shown with twin XLR-129 engines firing and engine skirts extended. Pyrotechnic bolts would fire in the booster (B) forward pin connections, allowing aerodynamic drag and inertia to cause the boosters to tip away from the orbiter. C: aft pin connection actuators on the boosters simultaneously extend to ensure adequate clearance between the booster body flaps and the orbiter engine bells. D: when the boosters tipped back to an angle of 20° relative to the orbiter center line, pyrotechnic bolts sever the two aft pin connections. E: the aft pin connection actuators on the boosters retract. The boosters would then roll to turn their windward sides toward their direction of flight and begin descent and return to NASA Kennedy Space Center. Image credit: General Dynamics Convair/DSFPortree.

GDC proposed an IPD Triamese Reaction Control System (RCS) with 24 nitrogen tetroxide/hydrazine thrusters, most of which would cluster near the nose and tail. Of the 24, half would generate 1420 pounds (644 kilograms) of thrust and half 1160 pounds. 

Eight of the former would serve as orbital maneuvering thrusters, with four facing forward and four aft. These would permit the orbiter flight crew to circularize their orbit at space station altitude and perform rendezvous and station-keeping with the station. The company noted that the eight orbital maneuvering thrusters could be omitted from the boosters if doing so would save money.

The IPD Triamese orbiter mission would last 25 hours. Of this, the orbiter would spend 17.3 hours attached to the space station, during which time it would rely on station electricity, attitude control, life support, and communications. 

Precisely how the orbiter would link up with the space station was not explained. The liquid oxygen tank would be located between the cargo bay and the passenger compartment, preventing movement between them; for this reason, each would require an exterior hatch. This implies the existence of two docking units, one for each hatch, or a station hangar surrounding both hatches that could be pressurized. Though drawings show the cargo bay hatch as round, GDC described it as square and five feet (1.7 meters) wide. 

The company also did not describe the method of cargo transfer. No doubt the transfer of 25,000 pounds (11,340 kilograms) of supplies and equipment to the space station would need to be carefully orchestrated if it was to be completed in 17.3 hours. In addition, 2500 pounds (1130 kilograms) of cargo would be loaded into the cargo bay and 10 passengers at the end of their space station tour-of-duty would board the orbiter for return to Earth.

Shortly after departing the space station, the flight crew would use the orbital maneuvering thrusters to perform a deorbit burn, then carefully orient the orbiter for reentry. It would enter the atmosphere moving at 25,912 feet (7900 meters) per second at an altitude of 400,000 feet (122,000 meters) and would slow to 20,000 feet (6100 meters) per second at an altitude of 200,000 feet (61,000 meters). At these speeds, the orbiter would compress the thin air in its path, causing severe aerodynamic heating.

GDC described the IPD Triamese Thermal Protection System (TPS) heat shield in greater detail than any other system. Mostly it would comprise overlapping metal "cover panels" backed by thermal insulation blankets. The company divided the TPS into windward (nose, belly, and leading edge) and leeward (everywhere else) sections.

The composition of the TPS cover panels and the composition and thickness of the insulation behind them would depend on many factors. These would include orbiter reentry angle, banking angle, potential for air cooling, location on the orbiter, and the existence of new development programs aimed at perfecting existing TPS materials or producing new ones. 

The majority of the panels would be mounted on posts attached to the propellant tanks, which were meant to serve as "primary structure." GDC modeled its tank design on that of the Saturn V S-II second stage, which it said was made up of "cylindrical integrated pressure tanks." These could carry structural loads while unpressurized except during launch and ascent. In areas where no propellant tanks were available — mainly over the cockpit and passenger compartment, the cargo bay, and the engine compartment — the panels would be mounted on posts attached to a "trapezoidal framework." 

For its IPD Triamese TPS calculations, the company assumed an entry angle no greater than 1°. This would yield skin temperatures ranging from 3950° Fahrenheit (F) (2180° Celsius — C) on the windward side of the orbiter nose to 700° F (370° C) on the leeward side of the fuselage 90 feet (27 meters) aft of the nose. 

Most of the IPD Triamese would be covered by TD Nickel-Chromium (TD Ni-Cr) panels capable of withstanding a reentry temperature of up to 2400° F (1315° C). TD Ni-Cr is a thorium oxide-coated alloy. The panels would measure just 0.01 inches (0.254 millimeters) thick. At that thickness, they would weigh 1.75 pounds (0.8 kilograms) per square foot (0.09 square meters). GDC estimated that the typical TD Ni-Cr panel could withstand 50 reentries before it would need to be replaced. 

The nose and rudder fin leading edges would create special TPS problems. GDC called a thorium oxide-coated tungsten nose cap a "representative" state-of-the-art system. This would, however, need to be replaced after every third flight, so the company called for accelerated development of new TPS materials. The rudder fin leading edges, which would be made of costly coated tantalum, would need to be replaced after every 10th flight. 

The insulation blankets behind the panels would comprise layers of Microquartz and Dynaflex, products of the Johns Manville Corporation. Microquartz, which would make up one-third of the thickness of the blanket when used with Dynaflex, would be made of silica microfibers. It could withstand temperatures up to 1600° F (870° C). Dynaflex, an aluminum oxide, silica, and chromium oxide microfiber material that could withstand temperatures up to 2800° F (1540° C), would make up the remaining two-thirds of the blanket thickness.

Insulation blanket thickness and composition would depend on location on the vehicle. It would, for example, consist of Microquartz and Dynaflex and measure 3.7 inches (9.4 centimeters) thick on the windward side of the cockpit/passenger compartment area. A layer of Microquartz alone just 0.8 inches (2 centimeters) thick would suffice on the leeward side beginning about 60 feet (18.3 meters) aft of the nose.

The orbiter would maneuver during hypersonic reentry using its rudder fin-mounted flaps and body flap-mounted elevons. Initial calculations showed that a 20° bank initiated at 400,000 feet (122,000 meters) would permit a landing up to 450 nautical miles (830 kilometers) off the orbital track while causing an average increase in surface temperature of only 40° F (23° C). More detailed calculations suggested a different approach: a 45° bank gradually reduced to 10° at 200,000 feet (61,000 meters), then gradually increased again to 45°.

GDC proposed that vehicle primary structure temperature be controlled through "detailed air injection" during flight. Vents in the fuselage would be opened during descent to admit air, then ducts would channel it to hot areas to keep the temperature below 200° F (93° C). The company calculated that failure to air-cool the IPD Triamese orbiter would allow heat to "soak" into the vehicle, driving primary structure temperature to a punishing 330° F (166° C) 50 minutes after landing.

Like the boosters during their return to KSC, the orbiter would slow to subsonic speed at an altitude of 15,000 feet (4570 meters). It would, however, reach that altitude nearer the KSC landing strip than would the boosters. The orbiter would then deploy its TF-34 jet engines and switchblade wings. Subsonic flight under jet power would last no more than 10 minutes. 

About 400 feet (120 meters) above the ground, the flight crew would lower the landing gear and perform a flare maneuver, raising the orbiter's nose so that its main landing gear would touch the runway first. The flight crew and passengers would feel a deceleration equal to two times Earth's gravity at touchdown. Landing would occur at a speed of 180 miles (290 kilometers) per hour; rollout would measure less than 10,000 feet (3050 meters) with switchblade wing flaps down and less than 13,000 feet (3960 meters) with flaps up.  Maximum landing weight was 135,300 pounds (61,370 kilograms).

Desk model of Triamese launch (left) and landing flare configurations. The landing flare configuration model displays switchblade wings (colored orange), one of two deployed TF-34 jet engines (colored silver), and tricycle landing gear. Image credit: National Air and Space Museum.

Immediately after landing, the orbiter would again enter the turnaround flow, joining the boosters with which it had launched a little more than a day before. GDC determined that, under normal circumstances, an IPD Triamese orbiter would require 810 person-hours of turnaround servicing, while a booster would need 490 person-hours. A normal orbiter turnaround could be completed in a week by two teams of 23 technicians working two eight-hour shifts. Flight data recorder analysis, mission planning, and payload preparation would need additional time. 

Occasional additional tasks would add to turnaround time. GDC envisioned a special engine inspection every six months and an annual three-day "calendar inspection," which would see technicians visually inspect the interior of the liquid oxygen and liquid hydrogen tanks along with all wiring and plumbing. Every two years, technicians would spend three weeks performing "progressive rework" maintenance, during which they would remove the entire TPS to allow a detailed inspection of all vehicle systems and system replacement and updating as necessary.

As the ILRV study continued into the Spring of 1969, NASA, often acting at the request of the USAF, imposed new requirements on its contractors. Most new requirements reflected an ongoing shift in reusable vehicle purpose away from low-cost space station resupply and crew rotation and toward general spaceflight cost savings. 

In April 1969, NASA asked the ILRV contractors to add a 15-foot-wide-by-60-foot-long (4.6-meter-wide-by-18.4-meter-long) payload bay to the orbiter component of their designs. The contractors were also directed to study designs that could place 50,000 pounds (22,680 kilograms) or 100,000 pounds (45,360 kilograms) of payload into low-Earth orbit. 

At about the same time, the space agency requested that they study orbiter missions independent of a station lasting up to 30 days. Such missions would, in effect, see the orbiter function as a short-term space station. This was an ill omen for NASA's ambitious space station aspirations. 

Adding a large payload bay and long-duration missions to the IPD Triamese orbiter undermined the cost-saving principle of boost element and orbiter element commonality. GDC sought to accommodate the new requirements within its Triamese proposal; for example, the company proposed clustering more than two booster elements around an expendable second stage attached to a large payload. By October 1969, however, it was clear that the Triamese concept's days were numbered. 

On 13 January 1970, NASA Administrator Paine announced that the Saturn V assembly line would be shut down permanently. AAP would, however, continue under the new name Skylab. The Apollo 20 Moon mission would be canceled so that its Saturn V could be stripped of its S-IVB third stage and put to work launching Skylab into Earth orbit. 

That same month, the ILRV study was redesignated Space Shuttle Phase A. On 28 January 1970, GDC teamed up with North American Rockwell (NAR) to compete jointly for a Space Shuttle Phase B contract, which they subsequently won. GDC applied its ILRV study experience to the design of a reusable Booster for an NAR reusable Orbiter.

Sources

"Togetherness," M. Getler, Aerospace Technology, 17 July 1967, p. 70.

"MOL Switch Forthcoming," Aerospace Technology, 1 January 1968, p. 3.

Memorandum, Douglas Lord, Deputy Director, Advanced Manned Missions Program, NASA Headquarters, to Maxime Faget, Manned Spacecraft Center, "Manned Spacecraft Center Revised FY 1967 Advanced Study Program," 10 April 1968.

"Pace of Post-Apollo Planning Rises," W. Normyle, Aviation Week & Space Technology, 3 February 1969, pp. 16.

"NASA Aims at 100-Man Station," W. Normyle, Aviation Week & Space Technology, 24 February 1969, pp. 16-17.

"Large Station May Emerge as 'Unwritten' U.S. Goal," W. Normyle, Aviation Week & Space Technology, 10 March 1969, pp. 103, 105, 109.

Triamese Reusable Launch Vehicle/Spacecraft Status Report II, Report No. GDC-DCB69-014, General Dynamics - Convair Division, 7 May 1969.

A Shuttle Chronology 1964-1973: Abstract Concepts to Letter Contracts, Volume I: Abstract Concepts to Engineering Data; Defining the Operational Potential of the Shuttle, Management Analysis Office, Administration Directorate, NASA Johnson Space Center, December 1988, pp. I-10 - I-15, I-81 - I-83, I-85, I-87 - I-95, I-101 - I-102, II-108 - II-110, II-138 - II-140, II-156, II-158 - II-159, II-166 - II-167, II-182 - II-184.

More Information

"Without Hiatus": The Apollo Applications Program in June 1966

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

"A True Gateway": Robert Gilruth's June 1968 Space Station Presentation

Think Big: A 1970 Flight Schedule for NASA's 1969 Integrated Program Plan

McDonnell Douglas Phase B Space Station (1970)

Star-Raker (1978)

Star-Raker (right), a single-stage-to-orbit space plane, parks next to a 747 at a conventional airport. Image credit: M. Alvarez/Rockwell International.

Elsewhere in this blog, I have described the 1970s joint NASA/Department of Energy Solar Power Satellite (SPS) studies (see "More Information" below). Had even a single SPS been assembled, it would have been by far the largest human construction project in space; it would have weighed more than 100 times as much as the 420-metric-ton (460-U.S.-ton) International Space Station. The SPS studies envisioned assembly of two such satellites per year between 2000 and 2030, bringing the total number in the SPS constellation to sixty. 

NASA envisioned boosting SPS components to low-Earth orbit (LEO) in the payload bays of massive reusable launch vehicles. One such launcher, Boeing's winged, two-stage Space Freighter, would have weighed about 11,000 metric tons (12,125 U.S. tons) at liftoff and delivered about 420 metric tons (463 U.S. tons) to LEO. For comparison, the two-stage Saturn V rocket used to place 77-metric-ton (85-U.S.-ton) Skylab into LEO weighed about 2800 metric tons (3086 U.S. tons) at liftoff.

The Space Freighter would have risen vertically from a launch pad and pointed itself generally toward the east. As its first stage, the Booster, expended its propellants, it would have separated. The second stage, the Orbiter, would then have ignited its engines to complete its climb to LEO. In orbit, it would have maneuvered to rendezvous and dock with a large space station designed specifically for handling SPS cargo modules.

The Space Freighter Booster would have been a fully reusable winged vehicle closely resembling the Space Freighter Orbiter. After Space Freighter Orbiter separation, the Space Freighter Booster would have turned, deployed jet engines, and flown to a long, wide runway at its launch site. 

To begin return to Earth, the Space Freighter Orbiter in LEO would have separated from the cargo-handling space station, then would have turned its tail forward and ignited rocket motors to slow down, lowering its orbit so that it intersected Earth's atmosphere. Following a fiery reentry, it would have landed on the runway near its launch pad. 

After launch pad, Orbiter, and Booster refurbishment, the two Space Freighter stages would have been hoisted vertical. After the Orbiter was placed atop the Booster's nose, a cargo module would have been loaded into its payload bay. The Space Freighter would then have been moved to a launch pad to begin another flight. Launching parts for two SPS into LEO in a year would have required about 240 Space Freighter launches, or about one launch every 36 hours.

In October 1977, a team of 14 Rockwell International engineers studied a Space Freighter alternative. The Star-Raker space plane, 103 meters (310 feet) long with a wing span of about 93 meters (280 feet), would have carried a maximum of 89.2 metric tons (98.3 U.S. tons) of cargo into LEO. More than 1100 flights would have been required each year to support the SPS program, or about one launch every eight hours.

In its fully developed form, however, Star-Raker would have had important advantages over Space Freighter which might have made its required high flight rate feasible. For example, it would have begun its flights to LEO by taking off horizontally from a conventional 2670-to-4670-meter-long (8000-to-14,000-foot-long) runway at virtually any civilian or military airport capable of supporting 747 or C-5A Galaxy cargo planes. No specialized launch and landing site would have been required.

Every bit as important, Star-Raker would have been capable of flying routinely between such airports. The Rockwell team explained that this would "reduce the number of operations required to transport material and equipment from their place of manufacture on Earth to [LEO]." For example, rolls of solar cell blankets would not need to be shipped by train, barge, or plane to a specialized launch and landing site; they would, potentially, need only be transported to a local airport for Star-Raker pickup.

Though the 1977-1978 Star-Raker study focused on its possible use in the Department of Energy/NASA Solar Power Satellite program, Star-Raker would have had potential as a general-purpose space cargo plane. In the image above, three Star-Rakers, their nose sections hinged back to expose their cargo bays, take on payloads bound for destinations ranging from low-Earth orbit to deep space. Image credit: M. Alvarez/Rockwell International.

David Reed, an engineer at North American Rockwell (NAR), as the company was then known, originated the Star-Raker concept in 1968, as NASA began earnest efforts to develop a reusable Space Shuttle. Key elements of the concept had been proposed — and rejected — earlier in the 1960s decade. These included wings packed with lightweight structurally integral tanks holding liquid hydrogen fuel and liquid oxygen oxidizer and a complex jet engine/rocket engine propulsion system.

The 1968-1969 study determined that, as it burned the propellants in its wings and maneuvered through ascent from subsonic speed to Mach 6 (six times the speed of sound), aerodynamic pressure on its structure would become excessive. This led NAR to examine wing designs developed in 1970 for the proposed (and subsequently abandoned) U.S. Supersonic Transport program. 

A "tridelta flying wing" design appeared to solve the pressure problem; by then, however, NASA had narrowed its Shuttle design requirements, excluding Star-Raker from consideration. NAR continued Shuttle studies and became Shuttle prime contractor in July 1972. 

Rockwell revived study of the tridelta flying wing Star-Raker as SPS studies ramped up in 1976. The Star-Raker study that began in October 1977, led by Reed and performed for NASA Marshall Space Flight Center (MSFC) in Huntsville, Alabama, continued into late 1978, yielding the design described in this post.  

The 1977-1978 study benefited from computer modeling that enabled Rockwell to further refine Star-Raker wing shape and flight profile. It also allowed Reed's team to take more fully into account the benefits of propellant-saving "lifting ascent." 

Star-Raker's propellants, liquid hydrogen and liquid oxygen, were not typically found at airports in 1977-1978; this remains true in 2020. The Star-Raker study team might have assumed that airports would evolve to provide them by the time SPS cargo flights began in 2000. This would, perhaps, not have been an unreasonable assumption, given that the 30-year SPS program was expected to create a lucrative new industry spanning the continental United States. 

One Star-Raker takes off as another undergoes airport servicing. With its landing gear extended, Star-Raker ground clearance would have been 1.52 meters (five feet). Image credit: M. Alvarez/Rockwell International.

For the 1977-1978 study, however, they hedged their bets by assuming that liquid hydrogen fuel would be available at airports only in sufficient quantities for airport-to-airport subsonic air-breathing jet engine Star-Raker flights. Liquid oxygen would, of course, not have been required. Flights to LEO, which would have needed both propellants in large quantities, would have begun on a runway at NASA's Kennedy Space Center (KSC) in Florida, at Vandenberg Air Force Base, California, or at any other launch sites the U.S. might have deigned to establish. 

The propellant tanks in Star-Raker's wings would have been approximately conical in shape. They would have extended from the space plane's body to its wing tips and been designed to strengthen the wings with minimal weight penalty. They would have been reinforced with regularly spaced "cell web" walls. Foam-filled glass-fiber honeycomb would have surrounded the tanks, defining Star-Raker's shape.

The Rockwell team described in detail a Star-Raker flight from KSC to 556-kilometer-high (345-mile-high) LEO and back to a U.S. airport. It would have begun with arrival at KSC of a Star-Raker space plane loaded with cargo bound for LEO at the end of a subsonic flight from a conventional airport. 

Following a limited airplane-type checkout, crews would have installed three sets of jettisonable orbital-takeoff main landing gear, each with eight wheels, and pumped liquid hydrogen and liquid oxygen propellants into Star-Raker's tanks. Fully loaded with propellants and cargo and with its orbital-takeoff gear attached, Star-Raker would have weighed about 1935 metric tons (2130 U.S. tons). 

Star-Raker would have lifted off from the runway at a speed of 415 kilometers per hour (260 miles per hour) under "supercharged afterburner" power from its 10 multicycle jet engines. The Rockwell team explained that it had consulted with leading jet engine manufacturers to arrive at its jet engine design; these included General Electric, Pratt & Whitney, Aerojet, Marquardt, and Rocketdyne. The resulting engine was more a wish list than a firm design, though it was an informed wish list. 

The Rockwell team envisioned four operational cycles for its jet engine ranging from conventional turbofan to ramjet. Liquid hydrogen would have been used to cool the engine and then burned as fuel. Large, slot-shaped inlets on the underside of Star-Raker's wings, arranged in two groups of five on either side of the space plane's body, would have funneled air to the engines, which would have been mounted at the wing trailing edge. The inlets would have been equipped with "ramp" doors that could close partially or fully to moderate or halt airflow.

Shortly after leaving the ground, the space plane's crew would have dropped the three sets of orbital-takeoff landing gear (they would have lowered to the ground on parachutes for recovery and reuse), then would have retracted its nose and main landing gear. The space plane would then have switched its jet engines to turbofan power, climbed to 6100-meter (20,000-foot) cruise altitude, and increased its speed to Mach 0.85. It would have turned due south and, over the next hour and fifty minutes, flown directly to Earth's equator.

Star-Raker would have flown to the equator and turned east so that it could get a boost from Earth's rotational velocity, which at our planet's midriff can, in theory, add about 1600 kilometers (1000 miles) per hour to the orbital velocity of ascending launch vehicles. 

In addition, and more importantly, the turbofan flight to the equator would have amounted to a plane-change maneuver; that is, it would have enabled Star-Raker to reach equatorial LEO without performing the rocket-propelled plane-change maneuver in LEO required if Star-Raker flew directly to orbit from a non-equatorial launch site, such as KSC. The Rockwell team hoped that this would save propellants, enabling an increase in cargo weight.

Following the eastward turn, the space plane would have climbed to 13,710 meters (45,000 feet) under supercharged afterburner power, then would have begun a shallow dive to 11,280 meters (37,000 feet). During the powered dive, a propellant-saving maneuver, Earth's gravity would have helped it to break the sound barrier and accelerate to Mach 1.2. 

Go for orbit: the Star-Raker space plane design included 10 multicycle air-breathing jet engines, three high-pressure rocket engines akin to the Space Shuttle Main Engine, and two advanced Orbital Maneuvering System rocket engines. In the image above, the 10 jet engines are throttling up to begin the transition to supersonic flight. Image credit: M. Alvarez/Rockwell International.

Star-Raker would then have begun ascent to orbit in earnest, with a supersonic climb to 29 kilometers (18 miles). During this phase, the space plane's jet engines would have throttled up to "full ramjet" power, accelerating it to Mach 6.2. Throughout its climb to orbit, Star-Raker would have maneuvered to put to good use lift provided by its wings. 

Upon reaching Mach 6.2, the three rocket motors in Star-Raker's tail would have ignited, adding rocket power to ramjet power. The three engines, with a combined thrust of 1.45 million kilograms (3.2 million pounds), would have drawn liquid hydrogen from a sturdy tank located at the aft end of the long, narrow Star-Raker cargo bay. The tank, to which the engines would have been mounted, would have served as the load path that would have distributed their thrust to the space plane's structure.

At Mach 7.2, Star-Raker would have switched to full rocket power. As it throttled up the rocket motors to full thrust, it would have shut down the jet engines and closed completely their air inlet doors. 

When Star-Raker reached a 51-kilometer-by-556-kilometer (32-mile-by-345-mile) equatorial orbit, the main rocket motors would have shut down. At apogee, the high point in its orbit, the crew would have ignited the twin advanced Orbital Maneuvering System (OMS) engines at the base of its tail to raise its perigee (orbit low point) and circularize its orbit. Upon attainment of circular equatorial orbit, Star-Raker would have used the OMS to maneuver to a rendezvous with the SPS cargo-handling space station.

Star-Raker in low-Earth orbit. Image credit: M. Alvarez/Rockwell International.

The weight of cargo Star-Raker could carry would depend on its mission profile. For the profile described here, cargo weight delivered to orbit would have amounted to only about 48.6 metric tons (53.6 U.S. tons). The aerodynamic flight to the equator under jet power, meant to steal some of the Earth's rotational energy and avoid a plane change maneuver in LEO, had under close examination turned out to be expensive. 

The Rockwell team proposed improving the equatorial profile's payload performance by loading liquid oxygen at the equator, either during flight using a new-design tanker aircraft, or after a landing at an equatorial facility with an adequate runway, orbital-takeoff gear attachment and recovery capability, and ability to provide liquid oxygen. Either approach would, however, have complicated Star-Raker operations.

To unload cargo, Star-Raker would have swung its nose, which would have contained its crew compartment, sideways out of the way, exposing one end of its six-meter-high-by-six-meter-wide-by-43-meter-long (20-foot-high-by-20-foot-wide-by-141.5-foot-long) cargo bay. The bay's arched ceiling would have made it a point of structural strength, not weakness, in the Star-Raker design.

The crew would have moved to the rear of the crew compartment to assist with cargo transfer. Windows at the rear of the two-deck crew compartment would have provided a 121° field of visibility. 

The Rockwell team did not describe its cargo transfer system in any detail, though it is clear that Star-Raker would not have docked in the conventional sense. Brief mention was made of a transfer rail system in the cargo bay that would have linked to equivalent rails on the space station.

Return to Earth would have begun with cargo bay closure. After moving away from the space station, the crew would have turned Star-Raker so that its tail faced in its direction of orbital motion, then would have fired its OMS engines to slow down. 

Maximum deceleration during the unhurried shallow-angle reentry would have reached no more than 2.3 gravities. Star-Raker would, in general, have experienced reentry temperatures lower than the Space Shuttle Orbiter, though nose and wing leading-edge temperatures were expected be somewhat higher. The higher leading-edge temperature was attributable to its relatively blunt shape. 

The Rockwell team proposed two types of reusable Thermal Protection System (TPS) for Star-Raker. Both would have been mounted on an outer facing sheet covering a honeycomb layer. The honeycomb layer would in turn have been attached to an inner facing sheet covering the honeycomb core that surrounded the propellant tanks.

The first TPS design closely resembled that baselined for the Space Shuttle Orbiter. Ceramic tiles individually molded and milled to match Star-Raker's curves would have been glued to fabric strain-isolator pads affixed to the outer facing sheet. 

The second TPS design, similar to one developed for the B-1 Bomber, was more complex. Metal panels — titanium-aluminum for low-temperature areas and "superalloy" for high-temperature areas — would have been attached to the outer facing sheet using flexible standoffs. The standoffs would have permitted the overlapping panel edges to slide over each other as they grew hot and expanded or cooled and contracted. Foil-wrapped thermal insulation blankets affixed to the outer facing sheet would have provided additional thermal protection.

Both TPS designs would have included a system for detecting breaches in the TPS. The Rockwell team provided no details of its design and did not describe what the crew might do if a breach were detected.

Star-Raker on approach. Image credit: M. Alvarez/Rockwell International.

When Star-Raker slowed to Mach 6, it would have begun cross-range maneuvers designed to shed energy and slow it to Mach 0.85. The crew would then have opened the inlet ramps and started "some" of its jet engines. 

The Rockwell team provided the space plane with enough liquid hydrogen to permit a 556-kilometer (345-mile) subsonic cruise and two powered landing attempts. Landing velocity would have been about 215 kilometers per hour (135 miles per hour). At wheels stop at an airport capable of supporting a cargo 747 or a C-5A Galaxy, Star-Raker would have weighed about 281 metric tons (310 U.S. tons).

Star-Raker weights given in this flight description are based on data the Rockwell team generated in the period spanning December 1977-January 1978. In February-March 1978, NASA MSFC and NASA Langley Research Center (LaRC) in Hampton, Virginia, reviewed the Rockwell team's Star-Raker weight numbers. 

The NASA centers found that Rockwell's estimates were low if "normal" technology were assumed and high if "acceleration" (advanced) technology were assumed. Whereas Rockwell had placed Star-Raker's "dry" weight with orbital-takeoff gear at 293.5 metric tons (323.5 U.S. tons), MSFC/LaRC determined that, with normal technology and a 10% cushion for weight growth during development, Star-Raker would weigh 407.6 metric tons (449.3 U.S. tons) without propellants; with advanced technology and the cushion, it would weigh only 257.6 metric tons (284 U.S. tons). 

The Rockwell team and NASA MSFC engineers met in May 1978 to try to reconcile the weight estimates. They made one important change in Star-Raker's flight profile: they abandoned the subsonic flight to the equator in favor of a KSC launch and direct climb to a 556-kilometer (345-mile) LEO inclined 28.5° relative to Earth's equator (that is, the latitude of KSC). 

The NASA and Rockwell teams settled on a Star-Raker weight without propellants (but with orbital-takeoff gear and 10% cushion) of 330.4 metric tons (364.2 U.S. tons). As it began ascent to orbit on a KSC runway, the space plane would have weighed 2280.5 metric tons (2514 U.S. tons). Of this, Star-Raker's maximum weight, 89.2 metric tons (98.3 U.S. tons) would have comprised cargo for the SPS project.

Sources

Independent Research and Development Data Sheet, Project Title: Earth-to-LEO Transportation System for SPS, Rockwell International, 15 December 1978.

"Star-Raker: An Airbreather/Rocket-Powered, Horizontal Takeoff Tridelta Flying Wing, Single-Stage-to-Orbit Transportation System," SSD 79-0082, D. Reed, H. Ikawa, and J. Sadunas, North American Rockwell Space Systems Division; paper presented at the American Institute of Aeronautics & Astronautics Conference on Advanced Technology for Future Space Systems in Hampton, Virginia, 8-11 May 1979.

More Information

Electricity from Space: The 1970s DOE/NASA Solar Power Satellite Studies

NASA Johnson Space Center's Shuttle II (1988)

Flying Brickyard Postponed: A 1972-1973 Study of an Interim Ablative Space Shuttle Heat Shield

Space Shuttle Orbiter during atmosphere reentry as conceived by North American Rockwell (NAR) in July 1972. The following month, NASA would make NAR the Space Shuttle prime contractor. Image credit: NASA.

Launch, ascent to orbit, and Earth atmosphere reentry are the most risk-fraught phases of most piloted space missions to date. They are also the mission phases that most tax the ingenuity of engineers who design reusable spacecraft.

Aerodynamic heating creates challenges during reentry and, to a lesser degree, during ascent to orbit. Before the Space Shuttle, almost all piloted spacecraft designed to operate for some portion of their mission in an atmosphere withstood such heating by employing single-use ablative heat shields. (The only exception was the X-15A-2 rocket plane, which, for part of its career, included a replaceable ablative heat shield — please see "More Information" at the end of this post.) During reentry, ablative heat shields char and break away, carrying away heat.

The Space Shuttle, approved for development by President Richard Nixon on 5 January 1972, marked a dramatic departure in heat shield technology. Originally conceived as a fully reusable, economical Space Station resupply and crew rotation vehicle, Nixon's partially reusable Shuttle had as its only approved goal a dramatic reduction in the cost of launching things into space. A reusable heat shield was believed to be essential for achieving that objective.

Over the decades, engineers have considered many reusable heat shield concepts, typically in combination. High on the list was a layer of overlapping "shingles" made of exotic metal alloys. Other approaches included liquid or solid heat sinks, thick metal or composite adjoining plates, or even an "active" system with cooling fluid circulating through a network of tubes behind a metal-alloy hull.

Unfortunately, all of these concepts would be heavy. To compensate for a heavy heat shield, engineers could design a more powerful booster system or could cut back on payload capacity (or both). Both approaches would boost development and operations costs. The Nixon White House had made clear that the Shuttle development budget of $5.15 billion was carved in stone, leaving NASA with little choice but to find new approaches — including some that accepted a significant increase in eventual operations cost.

Partial cutaway of the NAR ("NR") Shuttle Orbiter showing optional jet engine module installed in the aft portion of the Payload Bay, main engines, fore and aft reaction control system thruster clusters, aluminum wing structure, twin robot arms, crew module, and ring-shaped, nose-mounted docking unit forward of the crew module. Image credit: NASA.

For the Shuttle Orbiter, NASA and contractor engineers chose a lightweight combination of fabrics and brittle silica ceramic tiles, which they dubbed Reusable Surface Insulation (RSI). The tiles could withstand temperatures of up to 2300° Fahrenheit. Reinforced Carbon-Carbon composite panels would protect the Orbiter's wing leading edges, nose, and other areas subject to the highest reentry temperatures (as high as 3000° Fahrenheit).

Though RSI was meant to block almost all heat, enough would get through that, combined with aerodynamic buffeting, the Orbiter's mostly aluminum skin would tend to warp and flex ("flutter"). This meant that large ceramic panels affixed to the skin would crack, leaving it vulnerable to reentry heating.

Shuttle engineers sought to avoid damage by gluing RSI ceramics to a flexible fabric "strain isolator" layer glued to the Orbiter's skin and by making individual ceramic elements small in size. By resorting to many small "tiles" in place of a relatively few large panels, engineers designed an RSI heat shield that was in effect "pre-cracked."

The tiles, each milled to conform to its place on the Orbiter's complexly curvaceous hull, would number in the tens of thousands. By late in the 1970s decade, when their number hovered around 31,000, the tiles earned the Orbiter the nickname "The Flying Brickyard."

Some engineers harbored doubts about RSI; enough that NASA Langley Research Center in Hampton, Virginia, paid the Denver Division of Martin Marietta Corporation (MMC) to examine an alternative. Between May 1972 and August 1973, MMC engineers sought to determine whether Space Shuttle Orbiters could employ an ablative heat shield.

The ablative shield was seen as a stand-in system meant to provide NASA with more time for RSI development should problems arise. In his October 1975 report on the ablative heat shield study, Rolf Seiferth, who managed the MMC study between 5 September 1972 and its conclusion on 31 August 1973, envisioned that the ablative shield might fill in for RSI for five years. Based on a November 1972 NASA-generated Space Shuttle traffic model, this meant that 151 flights between 1979 and the end of 1983 would rely on the stand-in ablative system.

Seiferth noted that, in past programs, ablative heat shield materials had been glued directly to the spacecraft hull. This was, he explained, a cost-saving, weight-saving approach; scraping away a used directly applied ablative shield would, however, add time to Orbiter refurbishment between flights and generate considerable debris, including invasive dust.

In addition to the directly applied heat shield, MMC examined three types of "mechanically attached" ablative panels. These had ablative material glued to panels made of aluminum, magnesium, graphite composite, or beryllium/aluminum "Lockalloy" sheet or honeycomb.

The panels would be joined to oversized holes in the Orbiter's skin using nut-and-bolt fasteners, enabling entire panels to be replaced as necessary. The oversized holes would allow for thermal expansion of the heat shield components.

The simplest mechanically attached ablative panel would see ablative material glued to a metal or composite sheet. Adhesive and sheet would together measure only about 0.06 inches thick. Attachment points for the sheet panel design would typically occur five inches apart over much of the Orbiter, though larger spacings (up to 20 inches) were also possible.

The two more complex mechanically attached ablative panels substituted metal or composite "honeycomb" for the metal or composite sheet. One had ablative material glued to the honeycomb, which was then bolted to oversized holes in the Orbiter's skin.

The other — to which MMC gave considerably less attention — added rib-like standoffs to the Orbiter's skin. The honeycomb was then mechanically attached to oversized holes in the standoffs, leaving a gap between the underside of the honeycomb and the Orbiter skin.

Honeycomb panel attachment points would typically occur 10 inches apart over much of the Orbiter. Larger (up to 20 inches) and smaller (down to five inches) spacings were possible.

Seiferth's team used computer models to determine required ablator thickness, which would vary depending on its location on the Orbiter. All models assumed a maximum reentry deceleration equal to 2.5 times Earth's surface gravity (that is, 2.5 G) and a maximum allowable Orbiter aluminum skin temperature of 350° Fahrenheit, variables which indicated a relatively benign reentry environment (as compared to an Apollo lunar-return reentry, for example).

MMC used for its calculations properties of several types of ablative material it had developed for other missile and space projects (notably, the Titan missile family and the Viking Mars lander). It found that, for most locations on the Orbiter, its least robust ablator would be sufficient.

The ablative layer for most locations could be surprisingly thin. For the simplest mechanically attached panel design, for example, the MMC computer models indicated that a point on the Orbiter's underside on the fuselage centerline 50 feet aft of its nose would need a layer of ablative material only 1.7 inches thick.

Assessing the cost of the ablative designs relative to RSI was difficult in part because Space Shuttle Program cost estimation was, for want of a better term, eccentric. Seiferth supplied no development or operations cost estimate for RSI in his report, though he did provide estimates for several of MMC's ablative designs.

A system with an ablator glued directly to the Orbiter's aluminum skin would, Seiferth estimated, cost a total of $164.8 million for 151 flights over five years. Of this, installation and removal would account for $27.9 million.

A mechanically attached system comprising an aluminum sheet, adhesive, and an ablator (that is, the simplest mechanically attached ablative system) with attachment points five inches apart would cost $168.3 million with an installation and removal cost of $21.9 million. The aluminum honeycomb system with no standoffs and attachment points five inches apart came in at $187.1 million with $25.7 million for installation and removal.

NASA provided MMC with an RSI weight estimate of 30,240 pounds, enabling an RSI/ablative system weight comparison. The MMC study determined that an ablator directly attached to the Orbiter's skin would weigh 27,199 pounds, while the sheet and honeycomb (no standoffs) mechanically attached systems would weigh 32,577 pounds and 32,158 pounds, respectively.

Seiferth noted that modifications to the Orbiter's aluminum skin design would need to be put in place soon if mechanically attached ablative panels were used. Delaying until after the Orbiter's skin was in place would make prohibitive the cost and difficulty of adopting the ablative Space Shuttle heat shield. By the time Seiferth's report saw print in October 1975 — more than two years after the MMC study concluded — a stand-in ablative heat shield, never high on NASA's list of Space Shuttle priorities, was in fact no longer an option.

Late in the 1970s decade, problems with the Space Shuttle Main Engine, RSI, computers, and other systems contributed to delays in STS-1, the Space Shuttle's orbital maiden flight. RSI problems in particular became very public in March 1979, when the Space Shuttle Orbiter Columbia was flown from California to NASA Kennedy Space Center (KSC), Florida, atop its 747 carrier aircraft. It was the first Orbiter's first visit to its home base. At the time, Columbia was scheduled to carry out STS-1 in November 1979.

Columbia rolls into the Orbiter Processing Facility at Kennedy Space Center, Florida, on 25 March 1979. Though the image displays only the area around the front of the fuselage, many RSI gaps are evident. Image credit: NASA.

For the cross-country flight, about 26,000 permanent RSI tiles were installed on Columbia, along with about 5000 foam "dummy" tiles. By the time the Orbiter/747 combination set down on the Shuttle Landing Facility strip at KSC on 25 March 1979, Columbia had lost more than 200 RSI tiles. Many were lost as more than 4800 of the dummy tiles tore loose, a condition which would not occur during space flight.

Some permanent RSI tiles had, however, fallen off Columbia for other reasons. Close examination revealed tile manufacturing flaws, installation errors, and an overall unexpected degree of fragility. Even as Columbia entered the processing flow for STS-1, NASA conceded that the flight might be delayed until 1980.

Much was made of the "zipper effect," a hypothetical catastrophic failure mode that would begin with the loss of a single tile during reentry. The Orbiter was believed likely to survive loss of a single tile unless it occurred in an especially critical area. Loss of a single tile anywhere would, however, weaken surrounding tiles, potentially leading to a cascading loss of thermal protection. In fact, few tiles fell off Orbiters during the series of 135 Shuttle missions that began with Columbia's first launch on 12 April 1981.

The RSI system did, however, prove prone to impact damage during processing, launch, landing, and transport. The most extreme example before January 2003 occurred during STS-27 (2-6 December 1988), a classified Department of Defense mission. Eighty-five seconds after liftoff, debris broke free from the right Solid Rocket Booster, battering the right wing of Orbiter Atlantis. More than 700 RSI tiles were damaged and one was lost. Because the mission was classified, the near-disaster was not widely known for nearly 20 years.

This closeup of the right wing of the Orbiter Discovery was taken from the International Space Station (ISS) during STS-114 (26 July-9 August 2005), the first post-Columbia "Return-to-Flight" Mission. After the Columbia accident, NASA modified the External Tank design to eliminate the possibility of debris separation; nevertheless, two pieces of icy foam insulation broke free during STS-114, with one striking Discovery. In addition to a tile repair kit, which the STS-114 crew tested during a scheduled spacewalk, Discovery carried a Shuttle Remote Manipulator ("robot arm") extension that enabled its crew to inspect its RSI surfaces; it also performed a slow flip near the ISS so that astronauts on the station could inspect and photograph it. Though no damage was found, NASA prudently grounded the Shuttle fleet for another year after STS-114 returned to Earth so that it could continue its efforts to solve the External Tank debris problem. Image credit: NASA.

The Space Shuttle Orbiter Columbia lifted off on 16 January 2003 at the beginning of mission STS-107, its 28th flight and one of the few remaining non-ISS missions NASA had scheduled for the Shuttle fleet. During ascent, a piece of water ice-impregnated insulating foam weighing almost two pounds broke free from the External Tank to which Columbia was mounted. It struck the Reinforced Carbon-Carbon leading edge of the Orbiter's left wing, punching a hole at least 10 inches wide.

The debris strike was captured on video and immediately became the subject of urgent debate within the Shuttle Program. Knowledge of the strike was not shared widely. The viewing angle meant that the strike area was not visible in launch video recorded from the ground and its location meant that the STS-107 crew could not see it. Managers decided that Columbia's wing leading edge was probably intact.

The hole admitted hot gas as Columbia reentered on 1 February 2003. Its internal structure compromised, NASA's oldest Orbiter broke up over east Texas and western Louisiana, killing its seven-person crew and grounding the Space Shuttle fleet for 30 months.

The following January, President George W. Bush declared that the Space Shuttle would be retired after it performed its last International Space Station (ISS) assembly mission. The final Shuttle flight, STS-135 (8-21 July 2011), saw Atlantis, veteran of the STS-27 near miss, deliver supplies to ISS ahead of an anticipated gap in U.S. piloted space flights of indefinite duration.

Sources

"Space Shuttle Orbiter and Subsystems," D. Whitman, Rockwell International Corporation; paper presented at the 11th Space Congress in Cocoa Beach, Florida, 17-19 April 1974.

Ablative Heat Shield Design for Space Shuttle, NASA CR-2579, R. Seiferth, Denver Division, Martin Marietta Corporation, October 1975.

"Thermal Tile Production Ready to Roll," R. O'Lone, Aviation Week & Space Technology, 8 November 1976, pp. 51, 53-54.

"First Orbiter Ready for Florida Transfer," B. Smith, Aviation Week & Space Technology, 5 March 1979, pp. 22-23.

"Thermal Tile Application Accelerated," C. Covault, Aviation Week & Space Technology, 21 May 1979, pp. 59, 61-63.

"Space Shuttle Orbiter Status April 1980," S. Jones, NASA Johnson Space Center; paper presented at the 17th Space Congress in Cocoa Beach, Florida, 30 April-2 May 1980.

STS-27R OV-104 Orbiter TPS Damage Review Team, Volume I, Summary Report, NASA TM-100355, February 1989.

More Information

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

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

What If a Space Shuttle Orbiter Had to Ditch? (1975)

What If a Space Shuttle Orbiter Struck a Bird? (1988)

Chronology: Apollo-to-Shuttle Transition 2.0

Three years ago I published on this blog the first of my "Chronology" compilations of links to posts with a common theme. That first chronological compilation brought together links to posts on the transition from Apollo to the Space Shuttle. The aim was to impose chronology on posts that do not occur in chronological order in this blog as an aid to reader understanding.

This, my fourth "Chronology" compilation, updates that first compilation. I've added links to three posts dating from 1968, 1970, and 1972; that is, near the start, at the middle, and near the end of the planning phase of the Apollo-Shuttle transition.

"A True Gateway": Robert Gilruth's June 1968 Space Station Presentation

Series Development: A 1969 Plan to Merge Shuttle and Saturn V to Spread Out Space Program Cost (December 1969)

Think Big: A 1970 Flight Schedule for NASA's 1969 Integrated Program Plan (June 1970)

McDonnell Douglas Phase B Space Station (June 1970)

From Monolithic to Modular: NASA Establishes a Baseline Configuration for a Shuttle-Launched Space Station (July 1970)

An Alternate Station/Shuttle Evolution: The Spirit of '76 (August 1970)

Apollo's End: NASA Cancels Apollo 15 & Apollo 19 to Save Station/Shuttle (August-September 1970)

The Last Days of the Nuclear Shuttle (February 1971)

A Bridge From Skylab to Station/Shuttle: Interim Space Station Program (April 1971)

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

"Still Under Active Consideration": Five Proposed Earth-Orbital Apollo Missions for the 1970s (August 1972)

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

Final approach: the Shuttle Orbiter Discovery lands on the Shuttle Landing Facility at Kennedy Space Center, Florida, at the end of its longest mission (STS-131, 5-20 April 2010). Image credit: NASA.

The first NASA astronaut to die in the line of duty was U. S. Air Force Captain Theodore Freeman. Little known today, Freeman was a member of the third astronaut selection group, which NASA introduced to the world on 18 October 1963. The group included 10 astronauts who would become famous — Michael Collins, Edwin Aldrin, Alan Bean, David Scott, Russell Schweickart, William Anders, Eugene Cernan, Walter Cunningham, Donn Eisele, and Richard Gordon — and three besides Freeman who would perish before reaching orbit — Clifton Williams, Roger Chaffee, and Charles Bassett. Of the seven pre-Shuttle NASA astronaut groups, Group 3 experienced more pre-flight astronaut deaths than any other.

The astronauts had at their disposal Northrop T-38 Talon supersonic training aircraft. They used them in two basic ways: for training sorties to accumulate flight time so that they could keep their piloting skills well honed and retain their flight status, and as readily available, speedy transportation to NASA and contractor facilities and training sites across the United States. Transportation flights also contributed to the flight time requirement.

On 31 October 1964, 34-year-old Freeman took off alone in a T-38 from Ellington Air Force Base, located between downtown Houston, Texas, and NASA's Manned Spacecraft Center (MSC). He began his training sortie by flying over MSC, then out over Clear Lake and Galveston Bay.

NASA's Third Astronaut Group. Theodore Freeman is in the back row, fourth from left. Image credit: NASA.

As he returned to Ellington, a flock of Canadian geese took wing to one side of his flight path. As he made a turn, the flock rose up around his T-38, and one bird struck and shattered the plane's plexiglass forward canopy. Plexiglass shards entered the jet's twin engines through their air intakes. Moments later, the engines began to fail.

The eight-pound goose did not enter the T-38's intakes, though some sources report that it did. In fact, after striking the canopy, it struck the plane's rear seat, then spun away along the jet's upper fuselage.

Freeman tried to line up with an Ellington runway, but the engines flamed out and his plane began a steep dive at low altitude. He ejected, but before his parachute had time to open he struck the ground and was killed.

In October 1983, nearly 20 years after Freeman's untimely death, The Christian Science Monitor published a puff piece on NASA's efforts to keep wild pigs and alligators off the 15,000-foot-long, 300-foot-wide Shuttle Landing Facility (SLF) runway at Kennedy Space Center (KSC) in Florida. The story was timely because NASA aimed to achieve its first Orbiter landing at the SLF in January 1984. The space agency had planned to land Challenger at the SLF at the end of mission STS-7 on 24 June 1983, but had to divert it to Edwards Air Force Base (EAFB) in California after KSC became fogged in.

The north end of the SLF is about a mile from the Visitor Center for the Merritt Island National Wildlife Refuge (MINWR). MINWR and KSC both owe their origin to President John F. Kennedy's 25 May 1961 "Moon Speech." In 1962-1963, NASA acquired more than 140,000 acres of orange groves, swamp, and beaches to create a safety buffer around its Apollo Saturn V launch pads and other facilities. As landowners moved out, sometimes grudgingly, wildlife moved in.

On 28 August 1963, the space agency and the U.S. Fish and Wildlife Service agreed that the latter would manage the roughly 90% of KSC that NASA did not actively use. The interagency agreement assumed that KSC activities would increase over the course of the 1960s and 1970s and that its facilities would steadily expand. Apollo-era construction leveled off in 1966-1967, however.

Major facilities expansion did not begin again at KSC until April 1974, when the Morrison-Knudsen Company began work on the $22-million-dollar SLF. The facility, modeled on flight research runways at EAFB, was completed in 1976. It became KSC's airport, supporting astronaut T-38s, Gulfstream II Shuttle Training Aircraft, and other planes and helicopters. The first space-worthy Orbiter, Columbia, arrived at the SLF atop a 747 carrier aircraft in March 1979.

The Shuttle Landing Facility. Image credit: NASA.

A NASA spokesman told The Christian Science Monitor's reporter that KSC and MINWR played host to "all kinds of bald eagles, vultures, lots of brown pelicans, and ducks in winter." This was, however, not of great concern; the Shuttle Orbiter was a glider, he explained, so lacked air intakes that might ingest birds.

The Christian Science Monitor reporter wrote that the Orbiter had "triple-strength windows." This was a reference to the design of the six windows making up the flight deck windshield; each was three panes thick, with empty spaces between the panes. The outermost pane, the "thermal" pane, was attached to the fuselage structure; the innermost pane, the "pressure" pane, was attached to the crew cabin structure. Between these, also attached to the crew cabin structure, was a thick "redundant" pane.

The article affected an almost humorous tone as it described measures aimed at keeping alligators and wild pigs off the SLF. It seemed impossible that the Space Shuttle, a pinnacle of U.S. technological know-how, could ever be harmed by mere animals. Its author did suggest, however, that running over alligators basking in the Sun on the SLF runway might damage the Orbiter's "delicate landing gear."

On its second try, at the end of mission STS 41-B in February 1984, Challenger glided to a safe landing on the SLF runway. NASA hailed the landing, little more than five miles from the launch pad Challenger had left just eight days before, as a major step toward routine Shuttle flights and Shuttle launch rates of up to 25 per year.

A little less than two years later, on 28 January 1986, Challenger disintegrated 73 seconds after liftoff from KSC's Pad 39B, killing its seven-person crew. The disaster revealed that the Shuttle stack — twin reusable Solid Rocket Boosters, expendable External Tank, and reusable delta-winged Shuttle Orbiter — was much less robust than many had assumed.

Under intense scrutiny, NASA commenced a wide-ranging examination of Space Shuttle systems and operations. The U.S. civilian space agency soon found that many of its comfortable assumptions were incorrect.

Shuttle windshield: the Orbiter Endeavour during mission STS-123 (11-27 March 2008). Image credit: NASA.

Karen Edelstein, with NASA's Johnson Space Center, and Robert McCarty of the Wright Aeronautical Laboratories at Wright-Patterson Air Force Base in Ohio, reported on results of their study of bird impacts on the Orbiter windshield. They determined that, far from being triple-strength, it was "a poor barrier to bird impacts."

In fact, computer modeling using a refined version of the U. S. Air Force Material and Geometrically Nonlinear Analysis (MAGNA) program showed that, in every case, a four-pound bird — for example, a typical turkey vulture — would penetrate all three windshield panes in less than a second and enter the flight deck if the Orbiter were moving above an indeterminate speed between 150 knots (172 miles per hour) and 175 knots (201 miles per hour). They noted that the Orbiter traveled at up to 355 knots (408 miles per hour) as it fell past 10,000 feet and 195 knots (224 miles per hour) as its rear wheels touched the SLF runway.

This meant that at no time during descent through altitudes where birds fly did the Orbiter's windshield provide protection from bird strikes. In fact, the crew on the flight deck remained vulnerable until about the time the Orbiter's nose gear touched concrete.

Edelstein and McCarty did not examine in detail a bird impact leading to a partial window failure; for example, broken thermal and redundant panes and an intact pressure pane. This scenario was expected to occur at speeds as low as 150 knots. One may speculate that at the very least a partial failure would make the affected window essentially opaque; it might also create extra drag, altering the handling characteristics of the Orbiter.

A turkey vulture. Its wingspan is about six feet. Image credit: Wikipedia.

They noted that, short of a major redesign, there was little NASA could do to beef up the Orbiter windows. They urged designers of future space planes to seek materials more sturdy than glass when designing their windshields.

The Edelstein and McCarty paper did not lead to a major Orbiter redesign or new Orbiter window materials; NASA's allotted budget would not extend that far. Instead, the space agency redoubled its efforts to scare birds away from the SLF. Mostly it relied on loud noises.

For a time in the mid-1990s, however, KSC seriously considered putting falconers on its payroll. A June 1994 study noted that falcons had been used intermittently since the 1940s to kill or scare away birds at airfields in the U.K., the Netherlands, Spain, France, Canada, and the United States.

The study determined, however, that most of the more than 300 bird species that spent at least part of the year in MINWR had little experience with falcons, so were unlikely to be frightened by them. Falcons, for their part, were likely to be confused by wading birds such as herons and egrets.

The birds most threatening to Orbiters and other aircraft at the SLF, the 1994 study found, were various species of vulture. These were too large and numerous for falcons to tackle. It noted that groups of up to 30 individuals were frequently found around a single roadkill and that a "roost" of about 300 vultures had become established on the SLF runway's southern approach path.

The vultures, which weighed up to five pounds, took to the skies to ride thermals over KSC beginning in mid-morning. Mostly they glided lazily between 150 and 1800 feet above the ground. The air currents rising off the 526-foot-tall Vehicle Assembly Building were especially attractive to them. If the birds smelled a carrion buffet, however, they could fly rapidly, thwarting efforts to track and deter them. Loud noises, effective in driving away most other birds, were of little concern to vultures.

During the mid-morning launch of the Orbiter Discovery at the start of mission STS-114 on 26 July 2005, a vulture collided with the External Tank before the Shuttle stack cleared the Pad 39A launch tower. The bird probably weighed more than twice as much as the 1.7-pound chunk of External Tank foam insulation that had struck and breached Columbia's left wing leading edge on 16 January 2003, 82 seconds into mission STS-107. The foam chunk was estimated to have been moving at about 525 miles per hour when it hit the wing.

During Earth-atmosphere reentry on 1 February 2003, hot gases entered Columbia's left wing through the breach and rapidly destroyed its aluminum internal structure. NASA's oldest Orbiter broke up, killing the seven-member STS-107 crew.

Though the low-speed bird impact caused no obvious damage to the External Tank, NASA took notice because it occurred during launch of the first Shuttle mission since STS-107. The vulture might easily have struck a more vulnerable part of the Shuttle stack, or have struck it at a higher altitude, after the Shuttle had gained speed. KSC managers decided to apply SLF bird control techniques to the twin Shuttle launch pads. They also adopted a launch-day vulture "trap-and-release" policy.

By 2009, KSC's Bird Abatement Program relied on quick removal of roadkill to eliminate a major scavenger food source and pare down vulture numbers, bird detection radar and cameras, sirens, shotguns firing blanks and whistlers, and 25 liquid-propane-fueled "cannons." Installed along the SLF in 2007, the noise-producing cannons could be set off from the SLF runway control tower or by bird observers on the ground. They could also be set to fire automatically at random times and in random directions. Despite these measures, the risk to the Shuttle from bird strikes persisted until the Orbiter Atlantis rolled to a stop on the SLF runway at the end of STS-135, the final Shuttle mission, in July 2011.

Sources

"Space Shuttle Orbiter Windshield Bird Impact Analysis," ICAS-88-5.8.3, K. Edelstein and R. McCarty, Proceedings of the 16th International Council on Aeronautical Sciences Congress held in Jerusalem, Israel, 28 August-2 September 1988, Volume 2, pp. 1267-1274.

A Review of Falconry as a Bird Control Technique With Recommendations for Use at the Shuttle Landing Facility, John F. Kennedy Space Center, Florida, U.S.A., NASA Technical Memorandum 110142, V. Larson, S. Rowe, D. Breininger, and R. Yosef, June 1994.

"History of the Shuttle Landing Facility at Kennedy Space Center," E. Liston and D. Elliot; paper presented at The (40th) Space Congress in Cocoa Beach, Florida, 28 April-2 May 2003.

Fallen Astronauts: Heroes Who Died Reaching for the Moon, Revised Edition, C. Burgess and K. Doolan with B. Vis, University of Nebraska Press, 2016, pp. 1-45.

"NASA Tries To Keep The Hogs and 'Gators Off the Shuttle's Runway," G. Klein, The Christian Science Monitor, 12 October 1983 (https://www.csmonitor.com/1983/1012/101225.html - accessed 17 December 2017).

"It's a Jungle Out There!" L. Herridge, 26 June 2006 (https://www.nasa.gov/mission_pages/shuttle/behindscenes/roadkill.html - accessed 14 December 2017).

"Bye, Bye, Birdies," C. Mansfield, 30 June 2006 (https://www.nasa.gov/mission_pages/shuttle/behindscenes/avian_radar.html - accessed 16 December 2017).

"Bird Team Clears Path for Space Shuttles," L. Herridge, 12 August 2009 (https://www.nasa.gov/mission_pages/shuttle/behindscenes/clearbirds.html - accessed 14 December 2017).

More Information

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

What If a Shuttle Orbiter Had to Ditch? (1975)

What Shuttle Should Have Been: NASA's October 1977 Space Shuttle Flight Manifest