The Adapter They Built on Apollo 13

At 7:30 in the evening of 15 April 1970, in a back room of the Grumman Aerospace plant in Bethpage, Long Island, a 41-year-old engineer named Ed Smylie sat down with a roll of gray duct tape, a square canister, a clear plastic bag from a liquid-cooled garment, an EVA cue card from the Lunar Module checklist, and a length of hose. The canister was a Command Module lithium hydroxide cartridge, designed to remove carbon dioxide from the air of the Apollo Command Module. The hose was the Lunar Module’s red hose. The four items on Smylie’s table, none of which had been designed to work together, were the materials he had to work with. He had been working on this problem, on and off, for two days. The Apollo 13 spacecraft had suffered an oxygen-tank explosion on 13 April, and the three astronauts, James Lovell, John Swigert, and Fred Haise, were now in the Lunar Module Aquarius, which was never designed to support three people for the four days the mission now required. The Lunar Module’s lithium hydroxide canisters, the filters that scrub carbon dioxide from the cabin air, were designed for two people for 30 hours. With three people, they would be saturated within a day. The Command Module had spare canisters, but the Command Module canisters were square and the Lunar Module’s filter housings were round. The square canisters did not fit. The carbon dioxide was climbing toward dangerous levels. The carbon dioxide, in the way that carbon dioxide does, would make the crew drowsy, then unconscious, and then dead. The adapter that Ed Smylie was about to design, and the materials he had to do it with, were the only thing between three astronauts and the rising carbon dioxide.

The problem

The carbon dioxide problem on Apollo 13 had been understood within hours of the explosion. The Lunar Module’s Environmental Control System had two lithium hydroxide canisters, designed to handle the carbon dioxide output of two people for the 30 hours of the planned lunar landing mission. With three people in the Lunar Module, the same two canisters would have to handle three people for the 90 hours of the trip home, and would saturate well before they reached the Earth’s atmosphere. The Command Module’s lithium hydroxide canisters, of which the Odyssey had several spares, were designed to handle three people for the duration of the mission, but the Command Module canisters were square and the Lunar Module’s filter housing was round. The square canisters did not physically fit into the round filter housing. The problem was a physical mismatch in a piece of hardware that had been built by two different contractors, in two different buildings, in two different programmes, a generation apart. The two systems had been designed to be used only in their own spacecraft, and the design assumption that each would never have to work with the other had never been challenged in the decade of the Apollo programme. It had been challenged now.

Ed Smylie was the right person for the problem. He was the chief of the Crew Systems Division at the Manned Spacecraft Center in Houston, and the man who had been responsible for the design of the Lunar Module’s Environmental Control System. He had been at Mission Control on 13 April when the explosion happened, and he had been working on the carbon dioxide problem since the evening of that day, with a small team of engineers who had stripped the spare Command Module canisters out of the Crew Systems Division’s stock and laid them out on a series of tables to work out how to fit a square canister into a round filter housing using only the materials known to be available on the spacecraft. The materials on the spacecraft were known to be, in approximate inventory, duct tape, plastic bags, the EVA cue cards, the liquid-cooled garment (LCG) plastic bags, the hose from the Lunar Module’s Environmental Control System, and the spacesuit’s filters. Smylie and his team had, by the evening of 15 April, drawn an adapter on paper. They had tested it on the ground using spare parts from a Lunar Module test article. They had not yet built it in flight.

The adapter

The adapter, in its final form, was a piece of equipment that looked like nothing so much as the inside of a school locker after a particularly bad lesson. The crew were instructed to take the duct tape and cut two pieces about three feet long, then to make two belts around the sides of the canister, one near the top and one near the bottom, with the sticky side out, as tight as possible. They were then to take a one-and-a-half-inch strip of EVA cue card, which was a piece of stiff cardboard about a quarter-inch wide, and form an arch over the top of the canister, tucking one end of the cardboard under one ridge on the top of the canister and the other end against the ridge on the other side, so that the cardboard formed a rounded arch. They were then to cut a diagonal hole in one corner of the LCG plastic bag, slip the red Lunar Module hose through the hole, and tape the bag to the hose where it entered. The crew were instructed to make two of these units, each with a single canister, and to install them in the Lunar Module’s Environmental Control System one at a time. The cardboard arch, the duct tape belts, and the plastic bag together held the square canister against the round filter housing’s intake, and the air flowed through the canister, through the plastic bag, and into the Lunar Module’s cabin.

The crew built the two units, following the radioed instructions from Mission Control, in the afternoon and evening of 16 April, while the spacecraft was on its way back to Earth. The first unit, installed at 6:30 in the evening Houston time, immediately began to draw carbon dioxide out of the cabin air. The carbon dioxide level, which had been climbing toward 15 millimetres of mercury, a level that would have produced unconsciousness within hours, began to fall. By 9:00 in the evening, with both units installed, the carbon dioxide was back to safe levels. The adapter had worked, and the three astronauts were, in the standard phrase of the flight directors, “in good shape.”

How the mission went wrong

The Apollo 13 mission launched from the Kennedy Space Center on 11 April 1970 at 13:13 Central Standard Time, on a Saturn V rocket. The crew, James Lovell, John Swigert, and Fred Haise, were to be the seventh crewed Apollo mission and the third crewed lunar landing. The mission was to land in the Fra Mauro highlands, a region of the Moon’s surface chosen for its geological interest, on 16 April. The mission was, until the evening of the third day, a normal Apollo mission. The crew had completed the translunar injection burn, the manoeuvre that put the spacecraft on its trajectory to the Moon, on 13 April at 1:53 PM EST. The spacecraft was, by the evening of 13 April, about 200,000 miles from Earth, on its way to a lunar landing.

The accident happened at 9:08 PM EST on 13 April. The crew, in the Command Module Odyssey, had just finished a television broadcast and was preparing to shut down for the night. Jack Swigert, the Command Module Pilot, was going through the routine of stirring the cryogenic oxygen in Service Module tank number 2. The stir procedure required the heater on the tank to be turned on; the heater, on Apollo 13, was being supplied by a 65-volt bus at the time of the stir, rather than the 28-volt bus the procedure assumed. The heater, when energised at the higher voltage for a long enough time, generated enough heat to damage the Teflon insulation of the fan motor inside the tank. The fan motor’s Teflon, when it melted, ignited. The fire damaged the wiring of the fan and the wiring of the heater. The damaged wiring shorted the heater to the tank shell. The short, in a damaged tank with exposed wiring, generated a spark. The spark, in a pure-oxygen atmosphere at high pressure, ignited the insulation of the fan motor and the wiring. The fire spread through the Service Module bay that held oxygen tank number 2. The heat raised the pressure in the tank. The pressure, when it stopped, it blew off the panel of the Service Module that contained the tank, and damaged the adjacent oxygen tank number 1. Both Service Module oxygen tanks, the spacecraft’s supply of breathing oxygen and the supply of oxygen for the fuel cells, were now venting into space. The fuel cells, which were supplied by oxygen tank number 1, began to fail. The Command Module, which was supplied by the fuel cells, began to lose power. The spacecraft, in the standard phrase of the mission report, had suffered a near-catastrophic failure of the Service Module’s electrical and life-support systems.

The crew’s first indication of the problem was a sharp bang and a vibration through the spacecraft, accompanied by a master alarm and a cascade of caution and warning lights on the Command Module’s instrument panel. Jack Swigert, the Command Module Pilot, radioed Mission Control: “OK, Houston, we’ve had a problem.” The phrase, which was to become famous in the retelling, was not, in the moment, as dramatic as it sounded. The phrase was, in the moment, a routine pilot report. The problem, as it developed over the next few minutes, was to become the most serious in-flight emergency in the history of crewed spaceflight. The spacecraft, at the time of the call, was about 200,000 miles from Earth, on its way to the Moon, with a damaged Service Module and a Lunar Module that was not designed to support the mission it was now being asked to fly. The spacecraft’s abort modes, the procedures the crew were to follow in the event of a mission failure, called for the crew to abort the lunar landing and return to Earth. The abort modes had been designed for failures of the Lunar Module’s propulsion, the Command Module’s propulsion, or the spacecraft’s electrical system, but not for failures of the Service Module’s oxygen tanks. The crew, on the evening of 13 April, were in a position that the abort modes had not anticipated.

The crew moved into the Lunar Module Aquarius, the spacecraft’s lifeboat. The Lunar Module was designed for two people for 30 hours of lunar surface operations, not for three people for 90 hours of coasting back to Earth. The Lunar Module’s environmental control system, its propulsion system, its electrical system, and its water system were all designed for the mission the Lunar Module was meant to fly, not for the mission the Lunar Module was now being asked to fly. The Lunar Module, on 13 April, was now being asked to be a four-day lifeboat for three people, with a damaged Service Module providing only the spacecraft’s heat shield and the consumables that the Service Module still had. The Lunar Module, in the standard phrase of the mission report, was being asked to be a lifeboat, and the lifeboat was not designed for the mission.

How the answer got back to the crew

The instructions for the adapter had been worked out on the ground, in a back room of the Grumman Aerospace plant in Bethpage, on the evening of 15 April, and tested on a Lunar Module test article in the same room. The test article was an actual Lunar Module ECS test rig, with the same round filter housing the Lunar Module had in flight. The adapter, when Smylie and his team had taped it together, drew carbon dioxide out of the test air at the rate the trajectory predicted. The next problem was to get the instructions from Bethpage to the spacecraft. The Apollo 13 spacecraft, at the time, was about 200,000 miles from Earth, on a free-return trajectory that was to swing around the Moon and back to a splashdown in the Pacific. The spacecraft had lost its high-gain S-band antenna in the explosion, and was communicating with Mission Control in Houston only through a low-gain S-band antenna and the VHF link. The voice loop, for most of the mission, was functional, and the telemetry that told Mission Control what the carbon dioxide was doing was also functional, though reduced.

The first call from Mission Control to the Apollo 13 crew, on the evening of 16 April Houston time, was a set of instructions read up by Ed Smylie himself. The crew were instructed to gather the materials: a roll of duct tape from the Command Module’s stowage, a plastic bag from a liquid-cooled garment, an EVA cue card from the Lunar Module’s checklist, and a length of Lunar Module hose. The crew, at the time, were running the spacecraft on the Lunar Module’s systems, in the cold and the dark, with the Command Module powered down and most of its systems off. The Lunar Module’s cabin temperature was about 4 degrees Celsius; the Command Module’s, where the duct tape was stowed, was colder still. The crew had to go into the Command Module, find the duct tape, and come back to the Lunar Module. Jack Swigert, the Command Module Pilot, made the trip. The duct tape was in a stowage compartment behind the Command Module Pilot’s seat.

The crew then built the two adapters, in the Lunar Module, one at a time. The first adapter, with the duct tape belts, the cardboard arch, and the plastic bag, was installed in the Lunar Module’s ECS at 6:32 PM Houston time on 16 April. The carbon dioxide level, which had been climbing toward 15 millimetres of mercury, immediately began to fall. By 8:00 PM, with both units installed, the carbon dioxide was back to about 4 millimetres of mercury, a safe level for sustained exposure. The Lunar Module’s ECS was now scrubbing carbon dioxide from three people’s air, with the improvised adapter, on the spacecraft’s way back to Earth. The crew were, in the standard phrase, in good shape.

The Mission Control room, at the time the first readings came back, erupted. The flight controllers, who had been working on the problem for three days, who had tested the adapter on the ground and watched the carbon dioxide drop on the rig, who had then sat and listened while the crew built the same adapter in flight, who had watched the carbon dioxide on the spacecraft’s telemetry and seen it stop climbing and begin to fall, the flight controllers got up and clapped. It was the only time, in the Apollo programme, that the controllers in Mission Control applauded a specific engineering solution. The spacecraft’s carbon dioxide would stay at safe levels for the rest of the mission, and the spacecraft splashed down, on 17 April, with all three astronauts alive.

Why the two systems were never meant to fit

The square Command Module canister and the round Lunar Module filter housing were, in the design history of the Apollo programme, the result of two independent design efforts that had never been required to work together. The Command Module’s ECS had been designed by Hamilton Standard, the Environmental Control Division of the Hamilton Standard Corporation, under a 1962 contract with NASA’s Manned Spacecraft Center. The Lunar Module’s ECS had also been designed by Hamilton Standard, under a 1963 contract for the Lunar Module’s environmental systems. The two systems were designed by the same contractor, in the same division, by teams that worked in the same building in Windsor Locks, Connecticut, but they were designed for different spacecraft, with different crew sizes, different mission durations, and different assumptions about what could be available on board. The design assumption that each canister would only ever be used in its own spacecraft was, in 1963, a reasonable assumption. The spacecraft had been designed, throughout, for the missions they were to fly, and the missions had been designed, throughout, for the spacecraft they were to fly on. The two programmes, the Command Module and the Lunar Module, were joined at the interface and never joined at the carbon dioxide scrubber.

The two systems were reviewed, in the course of the Apollo programme, by a number of boards. The Apollo spacecraft’s environmental systems were reviewed by the Manned Spacecraft Center’s Crew Systems Division, of which Ed Smylie was the chief, by the prime contractors’ systems engineering teams (North American Aviation for the Command Module; Grumman Aerospace for the Lunar Module), and by NASA’s Headquarters Office of Manned Space Flight. The reviews looked at each system in isolation, and at the interfaces between systems. The interfaces between the Command Module ECS and the Lunar Module ECS were, in the design, at the cabin air and the cabin pressure, not at the carbon dioxide scrubber. The two systems were designed to share an atmosphere, not a scrubber. The lithium hydroxide canister was, in both systems, an internal sub-component that was not reviewed at the interface. The square Command Module canister and the round Lunar Module filter housing were, in this sense, an oversight of design, not an oversight of engineering. The two design teams had both done what they had been asked to do, and the oversight of design had been one of the predictable costs of designing two spacecraft in parallel for two different missions.

The oversight was, in the design community, a known risk. The design philosophy of the Apollo programme had been, from the start, to design for the mission and to design with a margin of safety, and the cost of doing so was that the two systems, having been designed for two different missions, were not interchangeable. The cost of designing for interchangeability, in 1963, would have been a more expensive and heavier Lunar Module ECS, a more expensive and heavier Command Module ECS, and a delay in both spacecraft. The cost-benefit calculation that the engineers of the Crew Systems Division and Hamilton Standard had run, in 1963, had been that the missions would not require it. The Apollo 13 mission had required it. NASA, in the aftermath of the mission, modified the design of the Lunar Module’s filter housing so that it would accept either a round or a square canister, and added a third lithium hydroxide canister to the Lunar Module’s stowage on the later Apollo missions. The redesign was, in the standard account of the programme, the predictable consequence of having flown a spacecraft in conditions the designers had not anticipated, and the predictable lesson from the redesign was that the next time the spacecraft was asked to fly a mission the designers had not anticipated, the mission was to be flown with a smaller amount of improvisation, and a larger amount of design.

How the spacecraft got home

The Apollo 13 spacecraft, after the explosion, was on a trajectory that would have taken it past the Moon and back to Earth, but the trajectory was a free-return trajectory that would have skipped off the Earth’s atmosphere and back into space if the spacecraft did not perform an engine burn to drop it into the atmosphere. The free-return trajectory, in other words, was a safe trajectory for a short time, but it was not a safe trajectory for getting home. The spacecraft had to perform an engine burn to put it on a trajectory that would re-enter the Earth’s atmosphere. The engine burn, on Apollo 13, was to be performed with the Lunar Module’s descent engine, the engine that was designed to land the Lunar Module on the Moon’s surface. The Lunar Module’s descent engine, in the modified flight plan for Apollo 13, was to be used to perform two burns: one to drop the spacecraft out of the free-return trajectory, and one to fine-tune the trajectory before re-entry. The burns were performed by the Lunar Module Pilot Fred Haise and the Commander James Lovell, on the morning of 14 April and again on 15 April. The burns worked, and the spacecraft was put on a trajectory to a Pacific splashdown on 17 April.

The Service Module, which had been damaged in the explosion, was jettisoned before re-entry. The jettison of the Service Module, which was performed by the crew on the morning of 17 April, exposed the Command Module’s heat shield, the heat shield that was to protect the crew during the re-entry. The heat shield, when the Service Module was jettisoned, was intact, though the inspection of the heat shield from the Lunar Module window showed that the heat shield had been damaged by the explosion. The damage, in the moment, was a worry, but the heat shield was, on inspection, intact enough to perform the re-entry. The Command Module, with the three astronauts inside, separated from the Lunar Module, oriented itself for re-entry, and entered the Earth’s atmosphere at about 12:53 PM EST on 17 April. The re-entry, which was to take the spacecraft from a speed of about 25,000 miles per hour to a splashdown speed of about 20 miles per hour, took about four minutes. The Command Module’s heat shield, which had been subjected to temperatures of about 3,000 degrees Fahrenheit, held. The Command Module splashed down in the Pacific Ocean, south of Samoa, at 1:07 PM EST on 17 April 1970. The three astronauts were recovered by the aircraft carrier USS Iwo Jima, which had been standing by in the Pacific for the splashdown. The Apollo 13 mission, which had been the third crewed lunar landing mission and was to become the third crewed mission to abort a lunar landing, was over.

The post-mission investigation, which was carried out by the Manned Spacecraft Center’s Apollo 13 Review Board, found that the cause of the explosion was the damage to the Teflon insulation of the fan motor in Service Module oxygen tank number 2. The damage had been caused by the heater being run at a higher voltage than the procedure called for, which had caused the Teflon to melt and ignite. The heater had been run at the higher voltage because the heater had been designed to operate at the higher voltage, but the ground support equipment at the Kennedy Space Center had been modified to operate the heater at the higher voltage for the Apollo 13 mission, and the modification had not been documented in the spacecraft’s procedures. The investigation also found that the damage to the Teflon had been a known risk before the mission, and that the risk had been accepted by the engineers of the Crew Systems Division on the grounds that the modification to the ground support equipment was not significant. The investigation recommended that the modification be reversed, and that the design of the Service Module’s oxygen tanks be reviewed. The recommendations were implemented before the next Apollo mission, and the redesign of the spacecraft’s environmental systems was, in the standard account of the mission, the predictable consequence of having flown a spacecraft in conditions the designers had not anticipated.

What was on the spacecraft

The improvised adapter was, in retrospect, a triumph of design within constraint. The materials that Ed Smylie had to work with, the duct tape, the cardboard, the plastic bag, the hose, were not the materials an engineer would have chosen; they were the materials that happened to be on the spacecraft. The duct tape was standard military-issue gray tape, with a polyethylene backing and a rubber-based adhesive, which had been chosen for the Apollo spacecraft for its wide temperature range and its ability to stick to a range of surfaces including metal, plastic, and fabric. The EVA cue card was a piece of cardboard that the astronauts used to remind themselves of the steps in a procedure; the cue cards had been chosen for the spacecraft as low-cost, low-mass, easily-replaceable items, and they had a stiffness that turned out to be ideal for the adapter’s arch. The LCG plastic bag was a transparent polyethylene bag that was the water container of the astronauts’ liquid-cooled garment, the undergarment that kept them cool during spacewalks; the bag had been chosen for its heat-sealable seams and its ability to hold about a litre of water. The red hose was the Lunar Module’s cabin air intake hose, a flexible plastic hose that had been chosen for its compatibility with the Lunar Module’s air system. None of the four items had been designed to work with the others, and the four together, in the right configuration, were enough to keep three people alive for the four days the spacecraft needed to get back to Earth.

The adapter that Ed Smylie and his team had designed, and that the Apollo 13 crew had built, was, in the standard account, an example of NASA’s ability to improvise in the face of an engineering problem. The improvised adapter was, however, also an example of the design philosophy that had produced the Apollo spacecraft in the first place. The Apollo spacecraft had been designed, throughout, with a margin of safety that included the assumption that things would go wrong, and with a redundancy that included a number of single-point failures. The Lunar Module’s Environmental Control System, the most critical single system in the spacecraft after the main propulsion, had been designed to be repairable in flight, and the materials that Ed Smylie had to work with were the materials that the Lunar Module’s designers had decided to put on the spacecraft. The improvised adapter was, in this sense, the predictable consequence of a design philosophy that expected its hardware to fail and tried to ensure that, when it did, the failure was recoverable.

What Ed Smylie did later

Ed Smylie, who had been chief of the Crew Systems Division at the Manned Spacecraft Center since 1964, continued in the role through the rest of the Apollo programme and into the early Shuttle programme. He retired from NASA in 1981, having seen five more Apollo missions flown and the first three Space Shuttle missions planned. He died in 2009, at the age of 80, having outlived the spacecraft programme he had been part of and the spacecraft itself. The story of the Apollo 13 adapter, which had been an internal NASA story for decades, became public in 1994 with the publication of the astronaut Jim Lovell’s memoir Lost Moon (reissued as Apollo 13), in which Lovell told the story of the carbon dioxide problem in the way that the mission was remembered in the Mission Control community. The story of the adapter has, since then, become a fixture of the engineering and management literature, and the question “What would Ed Smylie have done?” has become a small tradition in the operations community. The answer, on the evidence, would have been to sit down with what was on the table and figure out how to make it work.

References

1. NASA Manned Spacecraft Center (1970). Apollo 13 Mission Report, MSC-02680, September 1970. The official post-mission report, including the adapter description and the air-quality data. https://www.nasa.gov/wp-content/uploads/static/history/alsj/a13/A13_MissionReport.pdf
2. “Building an Apollo 13 LiOH Canister Adapter.” Apollo Lunar Surface Journal, NASA History Office. The detailed technical description of the adapter, with the original radio transcripts and the 15 April 1970 photograph. https://www.nasa.gov/history/alsj/a13/a13_LIOH_Adapter.html
3. Lovell, Jim, and Jeffrey Kluger (1994). Lost Moon: The Perilous Voyage of Apollo 13. Houghton Mifflin. Reissued in 2000 as Apollo 13. The astronaut’s memoir, the source of the public story of the carbon dioxide problem.
4. Mindell, David A. (2008). Digital Apollo: Human and Machine in Spaceflight. MIT Press. The scholarly account of the Apollo spacecraft’s man-machine systems, including the Environmental Control System and the adapter design philosophy.
5. Apollo 13 Review Board (1970). Report of the Apollo 13 Review Board, NASA Headquarters, June 1970. The official accident investigation, including the Service Module oxygen tank fan motor and the heater modification history. https://www.nasa.gov/history/alsj/a13/a13review.html
6. Chaikin, Andrew (1994). A Man on the Moon: The Voyages of the Apollo Astronauts. Viking. Reissued by Penguin. The narrative history of the Apollo programme, including the Apollo 13 mission.
7. “Apollo 13: The People and the Procedures.” NASA History Office. The oral-history account of the Apollo 13 mission, including interviews with Ed Smylie and the Mission Control team. https://www.nasa.gov/history/alsj/a13/a13.html