Space Shuttle abort modes

Space Shuttle abort modes were procedures by which the nominal launch of the NASA Space Shuttle could be terminated. A pad abort occurred after ignition of the shuttle's main engines but prior to liftoff. An abort during ascent that would result in the orbiter returning to a runway or to a lower than planned orbit was called an "intact abort", while an abort in which the orbiter would be unable to reach a runway, or any abort involving the failure of more than one main engine, was called a "contingency abort". Crew bailout was still possible in some situations where the orbiter could not land on a runway.

Redundant set launch sequencer (RSLS) abort
The three Space Shuttle main engines were ignited roughly 6.6 seconds before liftoff, and computers monitored their performance as they built up thrust. If an anomaly was detected, the engines would be shut down automatically and the countdown terminated before ignition of the solid rocket boosters (SRBs) at T - 0 seconds. This was called a "redundant set launch sequencer (RSLS) abort", and happened five times: STS-41-D, STS-51-F, STS-51, STS-55, and STS-68.

Ascent abort modes
Once the shuttle's SRBs were ignited, the vehicle was committed to liftoff. If an event requiring an abort happened after SRB ignition, it was not possible to begin the abort until after SRB burnout and separation about two minutes after launch. There were five abort modes available during ascent, divided into the categories of intact aborts and contingency aborts. The choice of abort mode depended on how urgent the situation was, and what emergency landing site could be reached.

The abort modes covered a wide range of potential problems, but the most commonly expected problem was a Space Shuttle main engine (SSME) failure, causing the vehicle to have insufficient thrust to achieve its planned orbit. Other possible non-engine failures necessitating an abort included a multiple auxiliary power unit (APU) failure, a cabin leak, and an external tank leak.

Intact abort modes
There were four intact abort modes for the Space Shuttle. Intact aborts were designed to provide a safe return of the orbiter to a planned landing site or to a lower orbit than planned for the mission.

Return to launch site (RTLS)
Return to launch site (RTLS) was the first abort mode available, and could be selected just after SRB jettison. The shuttle would continue downrange to burn excess propellant, as well as pitch up to maintain vertical speed in aborts with an SSME failure. After burning sufficient propellant, the vehicle would pitched all the way around and begin thrusting back towards the launch site. This maneuver was called the "powered pitcharound" (PPA), and was timed to ensure less than 2% propellant remained in the external tank by the time the shuttle's trajectory brought it back to the Kennedy Space Center. Additionally, the shuttle's OMS and reaction control system (RCS) motors would continuously thrust to burn off excess OMS propellant to reduce landing weight and adjust the orbiter's center of gravity.

Just before main engine cutoff, the orbiter would be commanded to pitch nose-down to ensure proper orientation for external tank jettison, since aerodynamic forces would otherwise cause the tank to collide with the orbiter. The SSMEs would cut off, and the tank would be jettisoned as the orbiter used its RCS to increase separation. Once the orbiter cleared the tank, it would make a normal gliding landing about 25 minutes after lift-off.

Should a second SSME have failed at any point during PPA, the shuttle would not have been able to make it back to the runway at KSC, but the crew would be able to bail out. A failure of a second engine during the PPA maneuver would have led to loss of control and subsequent loss of crew and vehicle (LOCV). Failure of all three engines as horizontal velocity approached zero or just before external tank jettison would have also resulted in LOCV.

The capsule communicator would call out the point in the ascent at which an RTLS was no longer possible as "negative return", approximately four minutes after lift-off, at which the vehicle would be unable to safely bleed off the velocity it had gained in the distance between its position downrange, and the launch site. This abort mode was never needed in the history of the shuttle program. Astronaut Mike Mullane referred to the RTLS abort as an "unnatural act of physics," and many pilot astronauts hoped that they would not have to perform such an abort due to its difficulty.

Transoceanic abort landing
A transoceanic abort landing (TAL) involved landing at a predetermined location in Africa or western Europe about 25 to 30 minutes after lift-off. It was to be used when velocity, altitude, and distance downrange did not allow return to the launch point via RTLS. It was also to be used when a less time-critical failure did not require the faster but more dangerous RTLS abort.

A TAL abort would have been declared between roughly T+2:30 minutes (2 minutes and 30 seconds after liftoff) and main engine cutoff (MECO), about T+8:30 minutes. The shuttle would then have landed at a predesignated airstrip across the Atlantic. The last four TAL sites were Istres Air Base in France, Zaragoza and Morón air bases in Spain, and RAF Fairford in England. Prior to a shuttle launch, two sites would be selected based on the flight plan, and were staffed with standby personnel in case they were used. The list of TAL sites changed over time and depended on orbital inclination.

Preparations of TAL sites took four to five days and began one week before launch, with the majority of personnel from NASA, the department of defense and contractors arriving 48 hours before launch. Additionally, two C-130 aircraft from the manned space flight support office from the adjacent Patrick Air Force Base, delivering eight crew members, nine pararescuers, two flight surgeons, a nurse and medical technician, and 2500 lb of medical equipment to either Zaragoza, Istres, or both. One or more C-21s or C-12s aircraft would also be deployed to provide weather reconnaissance in the event of an abort with a TALCOM, or astronaut flight controller aboard for communications with the shuttle pilot and commander.

This abort mode was never needed during the entire history of the space shuttle program.

Abort once around
An abort once around (AOA) was available if the shuttle was unable to reach a stable orbit but had sufficient velocity to circle the Earth once and land, about 90 minutes after lift-off. The time window for using the AOA abort was very short: just a few seconds between the TAL and ATO abort opportunities. Therefore, taking this option was very unlikely.

This abort mode was never needed during the entire history of the space shuttle program.

Abort to orbit
An abort to orbit (ATO) was available when the intended orbit could not be reached but a lower stable orbit was possible. This occurred on mission STS-51-F, which continued despite the abort to a lower orbit. The Mission Control Center in Houston, Texas (located at Lyndon B. Johnson Space Center), observed an SSME failure and called "Challenger-Houston, abort ATO".

The moment at which an ATO became possible was referred to as the "press to ATO" moment. In an ATO situation, the spacecraft commander rotated the cockpit abort mode switch to the ATO position and depressed the abort push button. This initiated the flight control software routines which handled the abort. In the event of loss communications, the spacecraft commander could have made the abort decision and taken action independently.

A hydrogen fuel leak in one of the SSMEs on STS-93 resulted in a slight underspeed at MECO, but was not an ATO and the shuttle achieved its planned orbit; if the leak had been more severe, it might have necessitated an ATO, RTLS, or TAL abort.

Preferences
There was an order of preference for abort modes:
 * 1) ATO was the preferred abort option whenever possible.
 * 2) TAL was the preferred abort option if the vehicle had not yet reached a speed permitting the ATO option.
 * 3) AOA would have been only used in the brief window between TAL and ATO options.
 * 4) RTLS resulted in the quickest landing of all abort options, but was considered the riskiest abort. Therefore, it would have been selected only in cases where the developing emergency was so time-critical the other aborts were not feasible, or in cases where the vehicle had insufficient energy to reach the other aborts.

Unlike all previous United States crew vehicles, the shuttle was never flown without astronauts aboard. To provide an incremental non-orbital test, NASA considered making the first mission an RTLS abort. However, STS-1 commander John Young declined, saying, "let's not practice Russian roulette."

Contingency aborts
Contingency aborts involved failure of more than one SSME and would generally have left the orbiter unable to reach a runway. These aborts were intended to ensure the survival of the orbiter long enough for the crew to bail out. Loss of two engines would have generally been survivable by using the remaining engine to optimize the orbiter's trajectory so as to not exceed structural limits during reentry. Loss of three engines could have been survivable outside of certain "black zones" where the orbiter would have failed before bailout was possible. These contingency aborts were added after the destruction of Challenger.

Post-Challenger abort enhancements
Before the Challenger disaster during STS-51-L, ascent abort options involving failure of more than one SSME were very limited. While failure of a single SSME was survivable throughout ascent, failure of a second SSME prior to about 350 seconds would mean an LOCV, since no bailout option existed. Studies showed that an ocean ditching was not survivable. Furthermore, the loss of a second or third SSME at almost any time during an RTLS abort would have caused an LOCV.

After the loss of Challenger in STS-51-L, numerous abort enhancements were added. With those enhancements, the loss of two SSMEs was now survivable for the crew throughout the entire ascent, and the vehicle could survive and land for large portions of the ascent. The struts attaching the orbiter to the external tank were strengthened to better endure a multiple SSME failure during SRB flight. Loss of three SSMEs was survivable for the crew for most of the ascent, although survival in the event of three failed SSMEs before T+90 seconds was unlikely due to design loads being exceeded on the forward orbiter/ET and SRB/ET attach points, and still problematic at any time during SRB flight due to controllability during staging.

A particular significant enhancement was bailout capability. Unlike the ejection seat in a fighter plane, the shuttle had an inflight crew escape system (ICES). The vehicle was put in a stable glide on autopilot, the hatch was blown, and the crew slid out a pole to clear the orbiter's left wing. They would then parachute to earth or the sea. While this at first appeared only usable under rare conditions, there were many failure modes where reaching an emergency landing site was not possible yet the vehicle was still intact and under control. Before the Challenger disaster, this almost happened on STS-51-F, when a single SSME failed at about T+345 seconds. The orbiter in that case was also Challenger. A second SSME almost failed due to a spurious temperature reading; however the engine shutdown was inhibited by a quick-thinking flight controller. If the second SSME had failed within about 69 seconds of the first, there would have been insufficient energy to cross the Atlantic. Without bailout capability, the entire crew would have been killed. After the loss of Challenger, those types of failures were made survivable. To facilitate high-altitude bailouts, the crew began wearing the Launch Entry Suit and later the Advanced Crew Escape Suit during ascent and descent. Before the Challenger disaster, crews for operational missions wore only fabric flight suits.

Another post-Challenger enhancement was the addition of East Coast abort landings (ECALs). High-inclination launches (including all ISS missions) were now able to reach an emergency runway on the East Coast of the United States under certain conditions.

An ECAL abort was similar to RTLS, but instead of landing at the Kennedy Space Center, the orbiter would attempt to land at another site along the east coast of North America. Various emergency landing sites extended from South Carolina and Bermuda up into Newfoundland, Canada. ECAL was a contingency abort that was less desirable than an intact abort, primarily because there was so little time to choose the landing site and prepare for the orbiter's arrival. The ECAL emergency sites were not as well equipped to accommodate an orbiter landing as those prepared for an RTLS abort.

Numerous other abort refinements were added, mainly involving improved software for managing vehicle energy in various abort scenarios. These enabled a greater chance of reaching an emergency runway for various SSME failure scenarios.

Ejection escape systems
An ejection escape system, sometimes called a "launch escape system", had been discussed many times for the shuttle. After the Challenger and Columbia losses, great interest was expressed in this. All previous US manned space vehicles had launch escape systems, although none were ever used.

Ejection seat
The first two shuttles, Enterprise and Columbia, were built with ejection seats. It was only these two that were planned to be flown with a crew of two. Subsequent shuttles were built only for missions with a crew of more than two, including seats in the lower deck, and ejection seat options were deemed to be infeasible, so Challenger, Discovery, Atlantis, and Endeavour were built with no ejection seats. The type used on the first two shuttles were modified versions of the seats used in the Lockheed SR-71. The approach and landing tests flown by Enterprise had these as an escape option, and the first four flights of Columbia had this option as well. But STS-5 was the first mission to have a crew of more than two, and the commander made the decision that the ethical thing to do was to fly with the ejection seats disabled. Columbia's next flight (STS-9) was likewise flown with the seats disabled. By the time Columbia flew again (STS-61-C, launched on January 12, 1986), it had been through a full maintenance overhaul at Palmdale and the ejection seats (along with the explosive hatches) had been fully removed. Ejection seats were not further developed for the shuttle for several reasons:


 * Very difficult to eject seven crew members when three or four were on the middeck (roughly the center of the forward fuselage), surrounded by substantial vehicle structure.
 * Limited ejection envelope. Ejection seats only work up to about 3400 mph and 130,000 feet (39,624 m).  That constituted a very limited portion of the shuttle's operating envelope, about the first 100 seconds of the 510 seconds powered ascent.
 * No help during a Columbia-type reentry accident. Ejecting during an atmospheric reentry accident would have been fatal due to the high temperatures and wind blast at high Mach speeds.
 * Astronauts were skeptical of the ejection seats' usefulness. STS-1 pilot Robert Crippen stated:

"[I]n truth, if you had to use them while the solids were there, I don’t believe you’d—if you popped out and then went down through the fire trail that’s behind the solids, that you would have ever survived, or if you did, you wouldn't have a parachute, because it would have been burned up in the process. But by the time the solids had burned out, you were up to too high an altitude to use it. ... So I personally didn't feel that the ejection seats were really going to help us out if we really ran into a contingency."

The Soviet shuttle Buran was planned to be fitted with the crew emergency escape system, which would have included K-36RB (K-36M-11F35) seats and the Strizh full-pressure suit, qualified for altitudes up to 30,000 m and speeds up to Mach three. Buran flew only once in fully automated mode without a crew, thus the seats were never installed and were never tested in real human space flight.

Ejection capsule
An alternative to ejection seats was an escape crew capsule or cabin escape system where the crew ejected in protective capsules, or the entire cabin is ejected. Such systems have been used on several military aircraft. The B-58 Hustler, XB-70 Valkyrie, General Dynamics F-111 and early prototypes of the Rockwell B-1 Lancer used cabin ejection.

Like ejection seats, capsule ejection for the shuttle would have been difficult because no easy way existed to exit the vehicle. Several crewmembers sat in the middeck, surrounded by substantial vehicle structure.

Cabin ejection would work for a much larger portion of the flight envelope than ejection seats, as the crew would be protected from temperature, wind blast, and lack of oxygen or vacuum. In theory an ejection cabin could have been designed to withstand reentry, although that would entail additional cost, weight and complexity. Cabin ejection was not pursued for several reasons:


 * Major modifications required to shuttle, likely taking several years. During much of the period the vehicle would be unavailable.
 * Cabin ejection systems are heavy, thus incurring a significant payload penalty.
 * Cabin ejection systems are much more complex than ejection seats. They require devices to cut cables and conduits connecting the cabin and fuselage. The cabin must have aerodynamic stabilization devices to avoid tumbling after ejection. The large cabin weight mandates a very large parachute, with a more complex extraction sequence. Air bags must deploy beneath the cabin to cushion impact or provide flotation. To make on-the-pad ejections feasible, the separation rockets would have to be quite large. In short, many complex things must happen in a specific timed sequence for cabin ejection to be successful, and in a situation where the vehicle might be disintegrating. If the airframe twisted or warped, thus preventing cabin separation, or debris damaged the landing airbags, stabilization, or any other cabin system, the occupants would likely not survive.
 * Added risk due to many large pyrotechnic devices. Even if not needed, the many explosive devices needed to separate the cabin entail some risk of premature or uncommanded detonation.
 * Cabin ejection is much more difficult, expensive and risky to retrofit on a vehicle not initially designed for it. If the shuttle was initially designed with a cabin escape system, that might have been more feasible.
 * Cabin/capsule ejection systems have a patchy success record, likely because of the complexity.

Space Shuttle abort history
Source:

Emergency landing sites
Pre-determined emergency landing sites for the orbiter were determined on a mission-by-mission basis according to the mission profile, weather and regional political situations. Emergency landing sites during the shuttle program included: Sites in which an orbiter has landed are listed in bold, but none is an emergency landing.

Algeria
 * Aguenar – Hadj Bey Akhamok Airport, Tamanrasset

Australia
 * Kingsford-Smith International Airport, Sydney, New South Wales (until 1986)
 * RAAF Base Amberley, Ipswich, Queensland
 * RAAF Base Darwin, Darwin, Northern Territory
 * RAAF Base Pearce, Perth, Western Australia

Bahamas
 * Lynden Pindling International Airport, Nassau

Bermuda
 * NAS Bermuda, St. David's Island

Canada 
 * CFB Goose Bay, Goose Bay, Labrador
 * CFB Namao, Edmonton, Alberta (until 1994)
 * Gander International Airport, Gander, Newfoundland
 * Stephenville International Airport, Stephenville, Newfoundland
 * St. John's International Airport, St. John's, Newfoundland
 * Halifax Stanfield International Airport, Halifax, Nova Scotia

Cape Verde
 * Amílcar Cabral International Airport, Sal Island

Chile
 * Mataveri International Airport, Easter Island

France
 * Istres-Le Tubé Air Base near Istres, France
 * Hao Airport, Hao, French Polynesia

The Gambia
 * Yundum International Airport, Banjul

Germany
 * Köln Bonn Airport, Cologne
 * Ingolstadt Manching Airport, Ingolstadt

Greece
 * Souda Air Base, Souda Bay, Crete

Iceland
 * Keflavík International Airport, Keflavík

Ireland
 * Shannon International Airport, Shannon, County Clare

Liberia
 * Roberts International Airport, Monrovia (until 1989)

Morocco
 * Ben Guerir Air Base, Morocco (1988-2002)
 * Mohammed V International Airport, Morocco (until 1986)

Portugal
 * Lajes Field, Lajes
 * Beja Airbase, Beja

Saudi Arabia
 * King Khalid International Airport, Riyadh

Spain
 * Zaragoza Air Base
 * Morón Air Base
 * Gran Canaria Airport, Gran Canaria

Somalia
 * Berbera Airport, Berbera (inactive since 1991)

South Africa
 * South African Air Force Base Hoedspruit

Sweden
 * Arlanda Airport, Stockholm

Turkey
 * Esenboğa International Airport, Ankara

United Kingdom
 * RAF Greenham Common, Berkshire, England (from 1981)
 * RAF Brize Norton, Oxfordshire, England
 * RAF Fairford, Gloucestershire, England
 * RAF Finningley, South Yorkshire, England (until 1996)
 * RAF Machrihanish, Campbeltown, Scotland
 * RAF Mildenhall, Suffolk, England
 * RAF Upper Heyford, Oxfordshire, England (until 1993)

British Overseas Territories
 * NAF Diego Garcia, British Indian Ocean Territory

United States
 * Andersen Air Force Base, Guam
 * Atlantic City International Airport, Pomona, New Jersey
 * Bangor International Airport, Bangor, Maine
 * MCAS Cherry Point, Havelock, North Carolina
 * Columbus Air Force Base, Columbus, Mississippi
 * Dover Air Force Base, Dover, Delaware
 * Dyess Air Force Base, Abilene, Texas
 * East Texas Regional Airport, Longview, Texas
 * Edwards Air Force Base, California
 * Ellsworth Air Force Base, Rapid City, South Dakota
 * Elmendorf Air Force Base, Anchorage, Alaska
 * Fort Huachuca, Arizona, Sierra Vista, Arizona
 * Francis S. Gabreski Airport, Long Island, New York
 * Grant County International Airport, Moses Lake, Washington
 * Grand Forks Air Force Base, Grand Forks, North Dakota
 * Griffiss International Airport, Rome, New York
 * Grissom Air Force Base, Kokomo, Indiana
 * Hickam Air Force Base, Honolulu, Hawaii
 * John F. Kennedy International Airport, New York, New York
 * Lehigh Valley International Airport, Allentown, Pennsylvania
 * Lincoln Airport, Lincoln, Nebraska
 * Mountain Home Air Force Base, Mountain Home, Idaho
 * Myrtle Beach International Airport, Myrtle Beach, South Carolina
 * Orlando International Airport, Orlando, Florida
 * Otis Air National Guard Base, Falmouth, Massachusetts
 * Pease Air Force Base, Portsmouth, New Hampshire
 * Plattsburgh Air Force Base, Plattsburgh, New York
 * Portsmouth International Airport, Portsmouth, New Hampshire
 * Stewart Air National Guard Base, Newburgh, New York
 * Westover Air Force Base, Chicopee, Massachusetts
 * White Sands Space Harbor, White Sands, New Mexico
 * Wilmington International Airport, Wilmington, North Carolina
 * Wright-Patterson Air Force Base, Dayton, Ohio

Democratic Republic of the Congo
 * N'djili Airport, Kinshasa (until 1997)

Other locations

In the event of an emergency deorbit that would bring the orbiter down in an area not within range of a designated emergency landing site, the orbiter was theoretically capable of landing on any paved runway that was at least 3 km long, which included the majority of large commercial airports. In practice, a US or allied military airfield would have been preferred for reasons of security arrangements and minimizing the disruption of commercial air traffic.