Space Shuttle design process

• 1. Introduction

  ›

  The Space Shuttle design process began before Neil Armstrong had even set foot on the Moon. As early as October 1968, NASA engineers were already sketching out what might come next. The agency that had just achieved something no civilization had done before was already asking a harder question: how do you make space travel routine?

  What followed was not a single inspired decision but a decade of competing visions, political maneuvering, budget battles, and technical trade-offs. A fully reusable spacecraft was imagined, then scaled back. The Air Force reshaped the cargo bay. The Office of Management and Budget weighed in on the rockets. A president nearly canceled it twice. And in the spring of 1972, four of America's biggest aerospace firms submitted rival proposals to build the thing.

  How did a program born of ambition become something so different from its original form? And who actually decided what the Space Shuttle would be?

• 2. Post-Apollo Crossroads

  ›

  In 1969, United States Vice President Spiro Agnew chaired the National Aeronautics and Space Council, the body tasked with mapping out where America would go after the Moon. Four paths were on the table: a crewed Mars mission, a follow-on lunar program, a low Earth orbital infrastructure program, and discontinuing human spaceflight altogether.

  President Nixon ultimately chose the orbital infrastructure option, which meant building a space station and developing a shuttle to service it. But funding restrictions made doing both simultaneously impossible. NASA decided the shuttle had to come first, with the station to follow.

  At the same time, proposals for a wide variety of post-Apollo missions were circulating inside the agency, some of which would have cost as much as or more than Apollo itself. Each competed for a NASA budget that was being severely constrained. Three major proposals were presented to Agnew in 1969, and the shuttle rose to the top largely through tireless advocacy by its supporters. By 1970 it had been selected as the one major project for the near-term post-Apollo period.

• 3. The Reusability Debate

  ›

  The earliest vision for the shuttle was fully reusable. A large winged crewed booster would carry a smaller winged crewed orbiter to altitude, then separate. The booster would return and land horizontally on a runway; the orbiter would complete its mission in orbit and glide home the same way. The logic was straightforward: if both vehicles could fly again, operating costs would fall sharply.

  Studies quickly revealed a problem. To lift an orbiter with the desired payload capacity, the booster had to be enormous. In aviation and space systems, cost tracks closely with mass, so a bigger booster meant a vastly more expensive program. Both vehicles would also need rocket engines for space flight and jet engines for atmospheric maneuvering, along with separate fuel and control systems for each mode. The complexity compounded the cost.

  Shuttle advocates countered that a high launch rate would spread development costs thin enough to undercut expendable rockets. Theoretical studies mentioned 55 launches per year. But the final design never came close to supporting that number. The maximum production rate for the external tank was limited to 24 tanks per year at NASA's Michoud Assembly Facility in Louisiana.

  A competing school of thought proposed keeping the Saturn V production line running, using its enormous lift capacity to launch a space station in a few large payloads rather than many smaller shuttle ones. Another idea was using the Air Force Titan III-M to launch an enlarged Gemini capsule called Big Gemini, or a stripped-down glider shuttle with no main engines and a 15 by 30 ft payload bay. These alternatives kept the debate open well into the early 1970s.

• 4. Air Force Leverage

  ›

  During the mid-1960s, the Air Force had watched both of its major piloted space projects canceled: the X-20 Dyna-Soar and the Manned Orbiting Laboratory. That record made one thing clear to military planners. If they wanted to put astronauts and payloads in orbit, they needed to work with NASA. Between 1959 and 1970 the Air Force had launched more than 200 satellite reconnaissance missions, and the sheer volume of military payloads was exactly the kind of launch traffic that could make the shuttle economically viable.

  After the Six-Day War and the Soviet invasion of Czechoslovakia exposed gaps in the American satellite reconnaissance network, Air Force interest sharpened around one specific requirement: the ability to launch spy satellites southward into polar orbit from Vandenberg Air Force Base in California. Reaching polar orbit from Vandenberg demanded more energy than lower inclination orbits. Returning to Earth after a single orbit also presented a geometric challenge: the Earth rotates roughly a thousand miles beneath the orbital track during that time, so the vehicle needed a large delta wing to maneuver back to a runway. That single requirement pushed the shuttle toward a configuration more complex than Maxime Faget's preferred straight-wing design, which would have required the vehicle to fly in a near-stall throughout reentry.

  Despite this influence, the Air Force was not desperate for the shuttle. Its expendable boosters were working fine. That gave the Defense Department and the National Reconnaissance Office leverage. NASA had originally planned a 40 by 15 ft cargo bay; the NRO specified 60 by 15 ft, anticipating that future intelligence satellites would grow larger. When Faget proposed narrowing the bay to 12 ft, the military almost immediately insisted on keeping the 15 ft width. In exchange for these concessions, the Air Force testified before the Senate Space Committee on the shuttle's behalf in March 1971.

  Congress reportedly told the Defense Department it would not fund satellites designed to fly on anything other than the shuttle. The potential for using the shuttle to verify Soviet compliance with the SALT II treaty was likely a factor when President Jimmy Carter chose not to cancel the program in 1979 and 1980, even as it ran years behind schedule and hundreds of millions of dollars over budget.

• 5. Design Choices Locked In

  ›

  With the fully reusable booster ruled out on grounds of cost, technical complexity, and development risk, NASA settled on a partially reusable architecture: a winged orbiter with three liquid-fueled main engines, a large external tank holding liquid propellant that would be discarded after each launch, and solid rocket boosters that could be recovered and refurbished.

  The external tank was a consequential trade. Early designs had the orbiter carry its own propellant internally, but studies showed that moving the fuel to an external tank allowed a much larger payload bay inside a smaller vehicle overall. Losing the tank after each flight added to operating costs, but it was judged a relatively small fraction of total expenses.

  The jet engines question was settled by precedent. Earlier plans had the orbiter use jet engines to maneuver through the atmosphere after reentry, but NASA drew on experience with the X-15 and various lifting-body research aircraft and chose a gliding orbiter instead. Eliminating the jet engines and their fuel reduced weight and complexity while increasing payload capacity.

  The booster choice came down to four options examined by engineers at NASA's Marshall Space Flight Center, the center that had managed Saturn V development: an upgraded Saturn lower stage, new pressure-fed liquid-fuel engines, a single large solid rocket, or two or more smaller solids. Marshall engineers were specifically concerned about solid rocket reliability for crewed missions. The Office of Management and Budget ultimately forced the decision toward solid boosters because they carried lower projected development costs, even though liquid boosters offered better performance, lower per-flight costs, and less environmental impact.

  On crew size, some argued the shuttle should carry no more than four, the maximum who could be accommodated with ejection seats. A commander, pilot, mission specialist, and payload specialist would cover any mission profile. NASA anticipated flying additional participants as payload specialists, however, and designed the vehicle to carry more.

  The shuttle's avionics ran on the IBM AP-101 computer, using the HAL/S programming language. The first microprocessor used in the program was the 8088, later upgraded to the 80386.

• 6. The Contract Competition

  ›

  In the spring of 1972, four of America's largest aerospace firms submitted proposals to build the shuttle: Lockheed Aircraft, McDonnell Douglas, Grumman, and North American Rockwell. The NASA selection group worked through each bid carefully.

  Lockheed's proposal was judged too complex and too expensive, and the company lacked experience building crewed spacecraft. McDonnell Douglas's entry had both cost and technical problems. Grumman offered an excellent design that nonetheless came in over budget. North American Rockwell's proposal stood apart on two counts: it had the lowest cost and the most realistic cost projections, and it was rated the easiest to maintain over the long term.

  North American also brought a specific track record to the table. The Apollo 13 accident, which involved a failure in North American's command and service module, had demonstrated how the company managed electrical system failures under pressure. NASA announced its selection of North American Rockwell on the 26th of July, 1972.

  The program was eventually completed at a cost measured in billions of 1971 dollars against an original estimate of $5.15 billion. Operational costs, flight rates, payload capacity, and reliability all diverged from initial projections in ways that observers continue to debate. The gap between the shuttle that was imagined in 1968 and the one that first flew is a record of every compromise that accumulated along the way.