She was not returning from an ambitious, two-week science mission, trailing double sonic booms in her wake as she swept into desolate Edwards Air Force Base or the marshy expanse of the Kennedy Space Center. Nor was she blasting into orbit under the combined thrust of two Solid Rocket Boosters and three main engines to deploy an important satellite, upgrade a world-class observatory or haul hardware aloft for the International Space Station.
Rather, she was hurtling several thousand feet over Birmingham, right in the heart of England, securely affixed to the top of a Boeing 747 Shuttle Carrier Aircraft (SCA). Her name was Enterprise.
It was June 1983. It was sports day at my infant school in Tyseley. I was six years old. As I stumbled clumsily along, desperately trying to keep the egg affixed, equally securely, to the spoon in my hand, Enterprise flew overhead, midway through her tour of Germany, Italy, England and ultimately Canada and the United States.
She had already stolen that year’s Paris Air Show. Today, she sits in the McDonnell Space Hangar at the Smithsonian’s Steven G. Udvar-Hazy Center, near Dulles Airport in Washington, DC. The museum, into which she was placed in November 2003, now hosts a myriad of other rockets and spacecraft, according to curator Valerie Neal. Even three decades after she was built and flight-tested over the Californian desert, she remains in good shape, albeit somewhat dusty and with parts still missing.
Enterprise’s career has been distinctly overshadowed by those of her spacegoing sister ships Columbia, Challenger, Discovery, Atlantis and Endeavour. Yet in some ways she is the most important of them all, for she cleared much of the uncertainty in the 1970s over the handling, approach and landing characteristics of the delta-winged Shuttle as it plunges its way, brick-like, through the atmosphere. Removing this uncertainty led directly to the triumphant homecoming of Columbia on April 14th 1981: the first manned, orbital spacecraft to land like an airliner.
Enterprise, sadly, was never given the opportunity to travel into space. Instead, since her completion in March 1976 and rollout six months later, she has served as a ‘hangar queen’ – a Second World War reference to an aircraft which, though non-flying itself, provides parts for those which do fly – and proven instrumental in testing safety modifications across the Shuttle fleet. Most recently, in the summer of 2003, only months before being transferred to the Udvar-Hazy Center, NASA ‘borrowed’ fibreglass panels from her wings to assist the Columbia Accident Investigation Board in its inquiry.
During the course of that inquiry, which included firing chunks of insulating foam at Shuttle wing panels, the sections of fibreglass from Enterprise – though not broken – suffered permanent damage to their seals. Since the Reinforced Carbon Carbon (RCC) panels on Columbia’s wings had been two-and-a-half times weaker than Enterprise’s fibreglass, the hangar queen’s contribution to the investigation helped to validate the theory that flyaway foam and a breached thermal-protection system were indeed responsible for the tragedy.
Her contribution to the Shuttle fleet, however, goes still deeper. Although best known for a series of captive and free flights performed with the Boeing SCA in 1977, she was brought out of retirement in the wake of the Challenger disaster to evaluate a new net-like ‘barricade’ system for capturing an orbiter whose brakes had failed during rollout. She also supported practice runs for new crew-escape techniques, ‘lent out’ her nose landing gear and elevon flipper doors for testing, evaluated hardware for the Shuttle Amateur Radio Experiment and even demonstrated procedures for transferring propellant between her forward thrusters and rear-mounted Orbital Manoeuvring System (OMS) pods.
“NASA, over the years, has had lots of different uses for Enterprise,†structural engineer Julie Kramer White has said. “As the Shuttle has evolved and as we considered a second-generation orbiter, we have used Enterprise as a testbed for subsystems developmentâ€. She added that recent inspections of the hangar queen’s composite payload bay doors have shown them to be in excellent condition, to such an extent that they may be pressed into service as spare parts should one of the spacegoing orbiters suffer damage to its own set.
However, those, together with boron-aluminium tubes in her mid-fuselage, are parts “that we covet,†said White, “that we can’t touch unless they’re really neededâ€.
Remarkably, therefore, three decades after her completion, Enterprise continues to play a key role in the Shuttle’s evolution. It is a pity that she was never upgraded to make her spaceworthy; an irony not lost on recent Star Trek moviemakers, whose own Enterprise carried a commemorative mural in its ready room showing Space Shuttle Enterprise docked onto the International Space Station. In fact, Gene Roddenberry’s fictional series is an appropriate place to begin exploring her history, for it was responsible for changing the name on her payload bay doors from NASA’s preferred ‘Constitution’ to that which she bears today.
Original plans, dating back to July 1972, when NASA awarded a contract to North American Rockwell for her construction, called for just two spacefaring Shuttles: Constitution (Orbiter Vehicle 101), which would perform a number of Approach and Landing Tests (ALT) in the low atmosphere before being upgraded for operational missions, and Columbia (Orbiter Vehicle 102). However, in the first of many changes that would be applied to Constitution, following a mass influx of more than 100,000 letters to then-President Gerald Ford from ‘Trekkies’, NASA bowed to White House pressure that she should be renamed ‘Enterprise’.
The agency, apparently, did not approve of the name, preferring ‘Constitution’ as it honoured the upcoming 1977 bicentennial of the United States’ declaration of independence and its aftermath. Nonetheless, when Enterprise finally rolled out of Rockwell’s Air Force Plant 42 in Palmdale, California, on September 17th 1976, the Trekkies’ wish was very visibly granted. She had been structurally ‘complete’ since March, but had spent several months undergoing a series of Horizontal Ground Vibration Tests to validate her integrity and ability to withstand the stresses of several simulated ‘launch’ and ‘landing’ phases.
Although Enterprise was not of an identical configuration to sister ship Columbia, Rockwell engineers understood their differences well enough to incorporate them into mathematical testing models. Unlike Columbia, Enterprise was not equipped with provisions for attaching ‘real’ OMS pods in her aft fuselage (she carried only ‘boilerplate’ replicas); nor did she have the same integrally-machined vertical stabiliser tailfin as her spacegoing sibling. Four months after her very public rollout, on January 31st 1977 she was towed 36 miles overland to Edwards Air Force Base in readiness for her atmospheric flight tests.
Following her arrival in the gigantic Mate-Demate Device, Enterprise was hoisted atop the Boeing SCA on February 7th in readiness for three ‘taxiing tests’ along the runway on the 15th. Even after three decades, the sight of the heavily-modified jumbo with a Shuttle riding piggyback is among the most iconic and awe-inspiring images to emerge from the space programme, as the August 2005 return of Discovery to Florida after her STS-114 mission amply demonstrated. The original airliner was purchased by NASA from American Airlines in June 1974, partly because it was the largest available to accommodate the DC-3-sized orbiter.
Even so, it required the removal of virtually all of its interior equipment – including passenger seating and galley – and extensive modifications to its air-conditioning ducts, electrical wiring and plumbing, together with the installation of higher-thrust engines. Its upper fuselage was beefed-up with internal support structures and, following wind-tunnel tests, endplate-style vertical stabilisers were fitted to its horizontal tail. Like Enterprise, it was also provided with an emergency ejection capability for its four-man crew.
All three taxiing tests were performed in the early morning to reduce heating on the 747’s tyres and brakes – which, in addition to carrying the 400,000 lb load of the airliner itself, also had the 150,000 lb Enterprise on top – and were performed without incident. The combo’s speed along the runway was steadily increased from 89 to 140 and finally 157 miles per hour, after which procedures for an aborted takeoff at high speed were simulated, together with tests of full braking, thrust reversers and speedbrakes. The success of the taxiing runs cleared the way for Enterprise’s first captive flights.
During a two-week period from February 18th, the Boeing’s crew of pilots Fitz Fulton and Tom McMurtry and flight engineers Vic Horton and Louis ‘Skip’ Guidry performed five airborne runs with the still-unmanned Enterprise to assess the combined vehicles’ structural integrity and handling characteristics under flight conditions. So successful were these so-called ‘captive-inert’ tests that a planned sixth run was dropped. In fact, both Fulton and McMurtry found that the presence of the Shuttle above them had little adverse effect on their ability to control the aircraft, partly due to a large aerodynamic tailcone installed onto Enterprise’s aft fuselage.
Moreover, the presence of the orbiter’s delta-shaped wings actually turned out to generate more ‘lift’ than expected, leading some journalists after the flight to comment that they had just seen the debut of the world’s largest biplane. During the last two captive-inert tests, Fulton and McMurtry perfectly executed a so-called ‘short field’ landing in readiness for an anticipated touchdown on the 7,500-feet-long runway at NASA’s Marshall Space Flight Center in Huntsville, Alabama. The combo demonstrated that it could slow to a complete halt in less than 6,000 feet.
When the last captive-inert flight ended on March 2nd 1977, it was at last time to modify Enterprise to accept her first crew. Three ‘captive-active’ tests were planned for June and July, in which she would remain attached to the Boeing, followed by a series of up to eight ‘free flights’ beginning in August in which she would separate and glide to an unpowered, ‘deadstick’ touchdown at Edwards.
Aboard Enterprise for her first captive-active flight were astronauts Fred Haise – a veteran of the ill-fated Apollo 13 lunar mission – and rookie Gordon Fullerton. Far from simply being ‘operators’ of the vehicle, however, both astronauts were heavily involved in the design and development of its software and procedures both for the ALT series and for operational missions. “If I’d never flown the Enterprise,†Fullerton would say later, “doing the training was challenging and intriguing in its own right.
“People say ‘How do you train?’ thinking, ‘Well, you go to a school and somebody tells you how to do it’. It’s not like that at all. Somebody’s got to write the checklist, so you end up writing the checklist, working with each subsystems person and trying to come up with a pre-launch checklist for the Approach and Landing Tests. So you’re doing the work [and] the learning comes from doing jobs that needed to be done.
“We worried about doing this deadstick landing, so we had to train for that. I built a gadget to work on the T-38 [training jets] that would allow you with any given weight to set the power with the speedbrakes down to simulate what the data said the orbiter would fly it at, so that we could go fly the pattern we intended to fly in T-38s, making steep descents, flaring and touching down.
“The Shuttle training airplane, a Gulfstream 2, was built as an airborne trainer, and so the four of us [himself, Haise and fellow astronauts Joe Engle and Dick Truly] assigned to ALT served as the Shuttle pilots along with a Gulfstream pilot to do many dives at the ground to get the aircraft built and working rightâ€.
Haise and Fullerton’s first captive-active flight was a day late in coming when one of Enterprise’s four general-purpose computers exhibited a fault and had to be replaced on June 17th. All was ready by the following morning, however, and the Boeing lumbered off the runway with its oversized cargo at 8:06 am local time. The peculiar sensation of sitting aboard the Shuttle at such a great height came as a surprise to both astronauts. “When we first rode on top,†Haise recalled, “you couldn’t see the 747, no matter how [much] you’d try to lean over and try to look out the side windows.
“Not even a wingtip! It was kind of like a magic carpet ride. You’re just moving along the ground and you take off; and something below you – you knew it was there, but you couldn’t see what was taking you aloft. It was also deceptive sitting up that high. Things always looked like it was going slower than it was, for your taxiing and particularly the first takeoff. I really thought Fitz had rotated too early. It didn’t look like we were going fast enoughâ€.
The feeling of nervous excitement and anticipation was equally intense for the crew of the Boeing. “Fitz was a great leader,†Tom McMurtry recalled years later, citing the quiet moment, seconds before taxiing for takeoff on June 18th, when Fulton had turned and shaken hands with him, Horton and Guidry. “I thought that was a nice gesture. I think he just wanted to do that as a friendâ€.
During their time in flight, Haise and Fullerton were able to briefly test Enterprise’s aerosurfaces, rudder and speedbrake. Ten days later, their colleagues Engle and Truly set off for a second run, conducting low-speed tests of her control system and simulating the separation manoeuvre from the Boeing that they would follow on the free flights. To accomplish these tasks, Fulton climbed to 22,030 feet, before pushing-over and descending at around 3,000 feet per minute, allowing the astronauts to position Enterprise’s elevons in a ready-to-separate orientation.
Haise and Fullerton completed the captive-active roster on July 26th to finalise avionics and control surface checks before the green light was given for the free flights to commence in August. The only minor problem was a faulty sensor in one of Enterprise’s auxiliary power units, which triggered a caution-and-warning alarm in the cockpit and obliged Fullerton to shut it down. After landing on concrete Runway 22 at Edwards, whilst still attached to the top of the Boeing, Haise was able to deploy the Shuttle’s landing gear in readiness for the free flights.
Sixty-five thousand people, including 900 accredited members of the press and some 2,000 special guests, had assembled at Edwards in the pre-dawn hours of August 12th to see Enterprise perform her independent landing. With Haise and Fullerton once again at the Shuttle’s controls, Fulton duly ran the Boeing’s engines up to full power at precisely 8:00 am and within seconds the combo swept – remarkably quietly, onlookers would later comment – off Runway 22 and into the steadily-brightening desert sky, accompanied by five T-38 jet trainers.
Higher-than-normal air temperatures at altitude delayed the original 8:30 am release time, but at 8:48 am, flying at 310 miles per hour, Fulton nosed the aircraft into a seven-degree dive. “The Enterprise is set,†Haise radioed. “Thanks for the liftâ€. Without further ado, he pushed the separation button on his instrument panel and seven explosive bolts fired, ‘popping’ the Shuttle away from the Boeing. To increase the separation distance between them, Fulton immediately put his aircraft into a descending left turn, while Haise placed Enterprise into a right-hand turn and pitched upwards.
Fullerton, meanwhile, was busily scrabbling to remove circuit breakers and reset switches within seconds of leaving the top of the Boeing. “The instant we pushed the button to blow the bolts and hop off the 747,†he said years later, “the shock of that actually dislodged a little solder ball and a transistor on one of the computers and we had the caution tone go off and the red light. We had three [cathode-ray tube monitors] and one of those essentially went to halt. This is pretty fundamental. All your control of the airplane is through fly-by-wire and these computers.
“I had a cue card with a procedure if that happened, that we’d practiced in the simulator, and I had to turn around and pull some circuit breakers and throw a couple of switches to reduce your susceptibility to the next failure. I did that and, by the time I looked around, I realised, hey, this is flying pretty good, because I was really distracted from the fundamental evaluation of the airplane at firstâ€.
As Fullerton worked, Haise held two degrees of pitch for three seconds, before banking 20 degrees to the right and heading for dry lakebed Runway 17. Maintaining a nine-degree, nose-down attitude, Enterprise executed a pair of 90-degree turns and Haise aimed her for the runway centreline and opened the speedbrake. Mission Control in Houston erroneously radioed that he had a lower lift-to-drag ratio than predicted in wind-tunnel tests; from the flight deck, Haise responded by flying his final approach at higher speed, conserving energy to extend the glide. In fact, Enterprise’s lift-to-drag ratio was exactly as expected.
Realising that the Shuttle was actually “high and hotâ€, and that he would land ‘long’, Haise opened the speedbrake from 30 to 50 percent to slow down and began the landing flare at an altitude of 900 feet. As Enterprise levelled out, he deployed her landing gear and touched down 3,000 feet ‘long’ at some 213 miles per hour. As expected, in view of her low lift-to-drag ratio, the Shuttle had remained independently airborne for almost five-and-a-half minutes.
One unusual aspect of the touchdown was provided by the shortness of the nose landing gear, which caused the wings to tip downwards in a ‘negative’ angle-of-attack as soon as all six wheels were on the ground. “It’s almost funny,†Haise recalled, “the first time you de-rotate, or try to the put the nose down: for a little bit, you almost think you don’t have a nose gear, because it goes down so far! It does present a problem today, where the vehicle’s heavier, with actually [having to follow] a ritual on de-rotating to get the nose gear on.
“I’ve never been on an airplane that you had to actually worry about a sequence to do that effectively, because if you do de-rotate too fast, too early, while you’re still at high speed, the effect of that negative lift – putting pressure down on the tyres – can conceivably blow [them]. You have to go down to a point in pitch to hold and wait till you get below a certain speed to then continue the de-rotation to effectively get the nose gear on the ground.
“At the same time, you can’t hold it off too long, while it’s still too high, or you’ll lose the ability to arrest the fall [through], and so, if you kept it up too long, it would fall through and damage the nose gear from the standpoint of hitting down too hard. You’ve kind of got to work in between [with] a scheme of getting the nose gear on the runwayâ€.
Engle and Truly took Enterprise aloft for her second free flight on September 13th for a more extensive series of manoeuvres and a perfect landing. In fact, the only problem of note was a radar failure at Edwards, which could have led to an abort had it not been promptly brought back online. Ten days later, Haise and Fullerton evaluated the ‘autoland’ system that NASA hoped would ultimately be capable of landing the vehicle without crew interaction. The astronauts allowed the computers to guide them down to the 900-feet flare point, before taking over and hard-braking Enterprise to a perfect halt.
Original plans called for four flights with the tailcone in place, followed by two final attempts to land with the dummy main engines and OMS pods exposed, to assess the ‘real’ aerodynamic obstacles a Shuttle might encounter whilst returning from space. However, managers felt that the first three flights were so successful that Enterprise’s fourth landing could be done without her tailcone. It had been expected that the Boeing would encounter increased aerodynamic ‘buffeting’ as the vehicle separated, although Fulton and his crew reported conditions as moderate, but acceptable.
Enterprise did, however, descend far more steeply and rapidly than on her three previous flights and, despite some problems with the tactical air navigation system, Engle and Truly landed perfectly in just over two minutes. The final test, on October 26th, with Haise and Fullerton at the helm for the last time, involved a precision touchdown on concrete Runway 22. Although the flight itself was successful, with both pilots reporting that the Shuttle performed better than predicted, trouble cropped up during their final approach to the landing site.
As he emerged from the pre-flare manoeuvre, Haise found that Enterprise was dropping at 334 miles per hour, considerably faster than planned. In an effort to slow the orbiter, he opened her speedbrake early, but instead of slowing down, her speed increased. In response, he deployed the landing gear and pitched her nose down to achieve the desired touchdown point on the runway. The wings dipped and Haise struggled to correct the problem as Enterprise’s main landing gear hit the concrete hard, before suddenly taking off again and ‘bouncing’ 20 feet into the air.
Eventually, after several long seconds, the vehicle stabilised herself for the remainder of the rollout and her nose finally dropped down onto the runway surface. The problem was ultimately traced to a 270-millisecond ‘time lag’ problem in the flight control software: the delay between Haise’s inputs on the stick and Enterprise’s response had led him to overcorrect, resulting in the ‘pilot-induced oscillation’, a porpoising motion and a bouncy landing. Although very unwelcome for Haise himself, the touchdown and rollout provided necessary data for engineers monitoring the landing gear.
“The landing gear folks were quite chagrined through most of that test programme,†he said later, “because we were not landing hard enough to get them good data for the instrumentation they had on the landing gear struts; although I solved their problem on the fifth landing flight, where I landed and bounced the vehicle! That gave them the data [they needed] and they were very happy with that, although I wasn’t!â€
Despite the somewhat problematic landing, NASA decided that the ALT series had achieved its objectives: to guide the approach and landing phases of a Shuttle mission, test the orbiter’s autoland capability and demonstrate her subsonic airworthiness and the performance of her systems. Haise would also comment that verifying the performance of Enterprise’s brakes and nosewheel steering had to be accomplished in reverse, compared to normal flight-test projects. “In an airplane,†he explained, “you have a jet engine and normally approach flight test by first of all doing some taxi tests around the ramp and then some runs down the runway, progressively getting faster and faster.
“We had no way of doing taxi tests because Enterprise had no engines, so we were going to have to face taxi tests from the upper end of the speed spectrum, backwards! In other words, after we landed at 190 knots or so, somewhere down that rollout we were going to do taxi tests and we did it by each flight: [the] first flight started it at very low speed [and we] didn’t touch anything till we got down slow. Then each flight stepped backwards up the speed spectrum to check out braking and nosewheel at progressively higher speeds. That’s the way taxi tests were done with Enterpriseâ€.
Overall, the astronauts were happy with the performance of the Shuttle. “It handled better, in a piloting sense, than we had seen in any simulation,†said Haise, “either our mission simulators or the Shuttle Training Aircraft. The term I use is: it was tighter. [It was] crisper in terms of control inputs and selecting a new attitude in any axis and being able to hold that attitude; it was just a better-handling vehicle than we’d seen in the simulationsâ€. The cause of Haise’s bouncy landing – the time lag in the flight software – was subsequently corrected in time for the Shuttle’s first orbital mission in April 1981.
Although Enterprise had completed her approach and landing test series with flying colours, she still had several more tests to complete before, it was hoped, she would be extensively upgraded to make her capable of travelling into space. In March 1978, she was moved to NASA’s Marshall Space Flight Center and spent the following year undergoing vertical tests, attached to an empty External Tank and two inert Solid Rocket Boosters. The results of this work ultimately led to several design changes to the system, particularly the boosters. It was after the completion of these tests that Enterprise was scheduled to be modified for her first orbital voyage.
As history has shown us, it never happened.
Lessons learned during her construction were subsequently incorporated into the design of Columbia and it was realised in the 1978 timeframe that Enterprise weighed too much to transport a full payload into orbit; she would need a new set of plans, quite different from those of her sister ship, to make her spaceworthy. Moreover, she contained no propulsion system, plumbing, fuel lines or tankage of any kind, her main engines were dummies, her payload bay had no mounting hardware for cargo, its doors had no opening-and-closing mechanisms or cooling radiators and her ‘thermal-protection system’ was nothing more than black-and-white polyurethane foam for ‘tiles’ and fibreglass for ‘RCC’.
Inside Enterprise’s crew cabin, her instrument suite was sparsely populated with switches and dials compared to later orbiters: she had no guidance equipment, such as star trackers or heads-up displays and no indicators of External Tank or Solid Rocket Booster systems. Elsewhere, she had no aft flight deck or overhead windows, no airlock, no middeck lockers, no galley, no shower, no other crew-related items and her simple fuel cells were fed by high-pressure tanks, rather than cryogenic dewars. Her landing gear operated by explosive bolts and gravity, with no hydraulic mechanisms or manual backup systems available.
She did, however, contain Lockheed-built ejection seats for her pilots, which would have fired them through two blow-out aluminium panels in the ceiling in the event of an emergency. Modifying her for space missions, therefore, was envisaged to be a long, complex and expensive task. Additionally, the new design specifications called for much stronger wings and mid-fuselage than were installed aboard Enterprise and several aluminium components had to be changed to titanium to save weight.
Transportation and modification costs to accomplish these changes were simply unavailable and, as 1978 wore on, NASA was already looking to a high-fidelity Shuttle structural test article known as STA-099 as a cheaper option to upgrade for orbital service. On January 29th 1979, it was official: STA-099 would become the second space-capable orbiter – later renamed ‘Challenger’ – and NASA contracted with Rockwell to build two more vehicles (Discovery and Atlantis).
Enterprise’s dreams of space, it seemed, were over.
A couple of months after the contract to modify STA-099 and build Discovery and Atlantis was signed, she was transported to the Kennedy Space Center, mated to an empty External Tank and two inert Solid Rocket Boosters and rolled out to Pad 39A. The disappointment at having had the opportunity of spaceworthiness taken from her was sweetened somewhat in that she became the first structurally ‘complete’ Shuttle stack ever to sit on the launch pad. Her task for the next 11 weeks was to test the crew-escape systems and ensure that all work platforms were in the right places to access different parts of the orbiter, tank and boosters.
Returned to California in the autumn of 1979, she was juggled between Edwards Air Force Base and Rockwell for two years, before undertaking her European tour in the early summer of 1983 – providing me with my first chance to see her, albeit from afar – and returning to the United States to Louisiana as an exhibit in New Orleans for the 1984 World Fair. More tests were afoot in the winter of that year, when she was moved to Vandenberg Air Force Base in California for six months of fit checks at the Department of Defense’s Shuttle launch complex.
Ironically, this also made her the only complete Shuttle stack to sit on the launch pad at ‘Slick Six’, for the site was abandoned – at least for manned missions – in the wake of the Challenger disaster. A brief spell sitting next to the Saturn V exhibit at the Kennedy Space Center was followed by installation at Dulles Airport in November 1985, becoming the property of the Smithsonian. Following the loss of Challenger, the option of modifying her for space missions to return the fleet to four operational vehicles was briefly considered, but it was deemed cheaper to use an already-fabricated set of structural spares as the basis for a new orbiter, dubbed ‘Endeavour’.
Her last, remotely viable, chance of being upgraded into a spacegoing vehicle came almost a decade ago, in February 1996, when a team from the Johnson Space Center in Houston assessed her structural condition for possible refurbishment as an ‘additional’ Space Shuttle. In general, even after two decades of operations, she was in remarkably good condition, with no significant corrosion around her aft body flap or elevons and her wings were found to be in acceptable condition. More serious corrosion, probably caused by the effects of standing water, was detected on the floor of her aft fuselage and around her nose landing gear wheel well.
Ultimately, NASA derived greater benefit for the remaining members of its Shuttle fleet by using Enterprise as a testbed for continuing improvements to the system and helping to pinpoint the cause of the STS-107 tragedy. Today, as she sits in pride of place in the Smithsonian’s Udvar-Hazy Center, it is fortunate – in a way – that she was never made spaceworthy. Unlike Columbia, whose loss, astronaut Jay Buckey once lamented, had robbed him of the chance to someday take his grandchildren to see the ship he once flew, the happy ending for Enterprise, at least, is that she is still with us.