BACK TO MAIN SITE

CHALLENGER'S FINAL VOYAGE

By William Harwood

 


 

"We're not going to launch this thing and take any kind of risk because we have that schedule pressure. We're going to continue to abide by the flight rules that we've established in this program and we'll sit on the ground until we all believe it's safe to fly." - Jesse Moore, associate administrator for space flight, Jan. 26, 1986

In a sense, the shuttle Challenger's fate was sealed by a sequence of events that conspired to push the launch date, originally set for July 1985, into January 1986. Like most shuttle missions, flight 51-L evolved through a series of changes that inevitably affected its launch date, including 10 major payload revisions. Blastoff ultimately was set for Jan. 22, but problems launching the shuttle Columbia in December 1985 forced Challenger's date to slip even further behind. This was a major concern to NASA because on board Challenger were two time-critical satellites. One was needed to complete an around-the-world space communications system, and the other, a small automated science platform, was set to study Halley's comet, which was racing toward its close encounter with the sun on Feb. 9. With each launch delay, the time available to study the fabled comet before it disappeared into the sun's glare was running out.

The first sign of last-minute trouble came on Dec. 19, 1985, when Columbia's countdown reached 14 seconds to launch at 7:54 a.m. EST only to be interrupted when the shuttle's four flight computers detected high turbine speeds in a hydraulic assembly used to steer the right-side solid-fuel booster rocket's nozzle. Columbia's countdown was recycled to the T-minus 20 minute point but it was hopeless: the launch would have to be delayed. NASA subsequently decided to replace the hydraulic assembly and Columbia's launch was rescheduled for Jan. 4. Challenger's flight remained scheduled for Jan. 22, but Columbia's problems were not yet over. Launch was rescheduled for Jan. 6 to give launch crews the Christmas holidays off and to allow Columbia's crew two additional days of practice in flight simulators at the Johnson Space Center in Houston.

At the same time, Challenger's launch was rescheduled for Jan. 23 and then to Jan. 24 to give its crew additional training time. On Jan. 6, Columbia was grounded for the third time because of a sluggish oxygen propellant supply valve in one of its three liquid-fueled main engines. Launch on Tuesday, Jan. 7, was postponed a fourth time because of bad weather. Launch was rescheduled for Jan. 9, but the day before, engineers discovered a temperature sensor had broken off inside a fuel line, requiring another delay for repairs. The same day, Challenger's crew practiced launch procedures in the cockpit on pad 39B, just a mile and a half away from Columbia on pad 39A. Incredibly, Columbia was grounded on Jan. 10 by driving rain, but finally, at 6:55 a.m. on Jan. 12, Columbia took off.

Three days later, NASA rescheduled Challenger's launch for Saturday, Jan. 25.

Challenger's crew planned to fly to the Kennedy Space Center on Wednesday, Jan. 22, to prepare for blastoff. But dust kicked up by desert storms reduced visibility at an emergency landing field at Dakar, Senegal, where the crew would have to attempt an emergency landing in the event of an engine failure during the latter phases of the ascent. While mission managers debated their options, including use of the municipal runway at the Casablanca airport on Africa's east coast, Challenger's crew was told to delay the trip to Florida for 24 hours. Later that day, launch was rescheduled for Sunday at 9:36 a.m. and Challenger's countdown began on time Thursday at 10 a.m. The astronauts arrived in Florida later that day to begin final preparations. All seven shuttle fliers appeared at ease dressed in light blue NASA flight suits. It was an especially joyous moment for school teacher Christa McAuliffe, whose dream of one day flying in space was about to pay off. During Challenger's six-day mission, she planned to teach two lessons from orbit to classrooms across the nation as the first teacher in space, indeed, the first truly private citizen to win a seat on a space shuttle.

"I'm so excited to be here!" she said, standing on the tarmac at the Kennedy Space Center's 3-mile-long shuttle runway. "I don't think any teacher has ever been more ready to have two lessons in my life. I've been preparing these since September and I just hope everybody tunes in to watch the teacher teaching from space."

But Challenger's planned Sunday blastoff quickly turned into a cliff-hanger. A cold front from Texas was barreling toward the Cape bringing menacing cloud cover and weather conditions in Africa showed little sign of improvement. "We may not know until the last minute Sunday whether we can launch," said NASA spokesman Charles Redmond.

At a management meeting that began at 10 p.m. Saturday, launch engineers were told by Air Force weather officers that fog, low clouds, rain and thunderstorms could be expected in the launch area Sunday morning. At 10:16 p.m., the decision was made to recycle the countdown, which was at T-minus nine hours, to the 11-hour mark targeted for a launch at 9:37 a.m. Monday in a bid to thread a narrow one-day "window" and get Challenger off the ground between two threatening weather systems. As it turned out, the front slowed dramatically overnight and Challenger could have taken off safely on Sunday, Jan. 26. But it was too late - the countdown already had been recycled for a Monday launch try. The seven Challenger crew members took the delays in stride and relaxed in their motel-like quarters at the Operations and Checkout building near Kennedy Space Center headquarters, boning up on their flight plans. McAuliffe and Hughes satellite engineer Gregory Jarvis went for a bicycle ride to unwind, only to be intercepted by a television crew. McAuliffe, appearing ill at ease but smiling, said she was looking forward to launch on Monday. It was the last time she ever spoke in public.

"We're not going to launch this thing and take any kind of risk because we have that schedule pressure," said Jesse Moore, associate administrator for space flight. "We're going to continue to abide by the flight rules that we've established in this program and we'll sit on the ground until we all believe it's safe to fly."

 

 

The Jan. 27 Launch Attempt

At 1:26 a.m. Monday, engineers began loading a half-million gallons of supercold liquid oxygen and liquid hydrogen fuel into Challenger's giant external fuel tank. The crew was awakened at 5:07 a.m. for a traditional breakfast of steak and eggs before donning their light blue flight suits. Smiling and waving to NASA employees and news photographers, the astronauts left their quarters about 6:50 a.m., climbed aboard NASA's silver "Astrovan," a modified Airstream mobile home, and headed for the launch pad. Launch director Gene Thomas ordered extra weather balloons to be launched as blastoff time neared to get additional data. High winds were a concern because of the shuttle's record payload weight - 48,361 pounds - and its stability during the thunderous climb to orbit.

The shuttle fliers assembled in the "white room" at the pad. One launch pad technician greeted McAuliffe wearing an academic mortarboard, to the teacher's obvious delight. Each astronaut then donned an airtight helmet and a sort of high-tech Mae West life vest before climbing aboard, assisted by technicians who dutifully wiped the bottom of each crew member's shoes to make sure no one tracked dirt into the pristine environment of Challenger's cockpit. Once inside, the astronauts connected hoses from their flight helmets to the cabin's oxygen supply system and to "personal egress air packs, called "PEAPs," which provided a six-minute supply of unpressurized air for use during an emergency exit when dangerous fumes might be in the area. With all seven aboard and strapped securely to their seats by 7:56 a.m. and with the external fuel tank topped off with its load of propellant, the technicians closed Challenger's hatch. Incredibly, they were unable to get indications from microswitches that the hatch was locked shut.

"If we have a preference, we'd like to have one of the guys who's very familiar with the latches to come inside, close the door, check them to make sure they're all OK and send him back out again," commander Dick Scobee radioed launch control.

Meanwhile, cloud cover began to roll over the Cape and Challenger's three-hour "launch window" was quickly running out. The countdown was delayed at 9:10 a.m. to give technicians time to fix the hatch. The microswitch problem was quickly corrected but engineers had run into trouble removing a special tool from the hatch used to dog it shut. They called for a drill to bore out a stripped bolt. The drill came but its batteries were dead. The technicians sent out for fresh batteries. At 10:30 a.m., launch control was forced to recalibrate the shuttle's navigation equipment so the ship's computers would know where they were after the delay, an hour-long procedure. Meanwhile, the astronauts remained strapped in their seats. Launch was rescheduled for 12:06 p.m., but high winds were kicking up at the spaceport and finally, at 12:36 p.m., at the end of Challenger's launch window, blastoff was postponed for the day because crosswinds at the nearby shuttle landing strip were beyond acceptable limits in the event Challenger's crew was forced to attempt an emergency landing because of an engine failure or other problem during the first four minutes of flight. The countdown was recycled to the T-minus six-hour mark.

"We have just had an announcement from launch director Gene Thomas to the crew and the launch team that we are going to scrub for today," said NASA spokesman Hugh Harris.

It was a frustrating delay for the crew and for thousands of spectators, including McAuliffe's husband and two children, and scores of educators and teachers who stood by in chilly weather awaiting the space teacher's pyrotechnic launching. The astronauts, appearing none the worse for the wear but clearly disappointed, climbed out of Challenger at 1:06 p.m. after spending more than five hours on their backs. Scobee inspected the ship's balky hatch. Blastoff was rescheduled for 9:38 a.m. Tuesday, Jan. 28, and while the weather was expected to be clear, temperatures were predicted to dip below freezing overnight, prompting implementation of a special weather plan to prevent key launch pad water systems from freezing. The astronauts left the shuttle and returned to crew quarters for what turned out to be their last day on Earth.

 

 

The Launch-Eve Debate

While the crew slept, engineers with NASA's Marshall Space Flight Center in Huntsville, Ala., and with Morton Thiokol Inc., builder of the shuttle's giant solid-fuel booster rockets, held an unscheduled teleconference to discuss the effects of the record cold weather on sensitive rubber O-ring seals in Challenger's booster rockets. The boosters, the largest solid-fuel rockets ever flown, are assembled in sections, with four fuel segments making up each motor. The O-ring seals are used in the joints between fuel segments to maintain internal pressure and prevent hot gas or flame from escaping in a catastrophic "burn through." Morton Thiokol engineers unanimously recommended a launch postponement because they feared the O-rings might not seat properly because of the cold weather.

But NASA managers at Marshall, where the solid rocket program is managed, strongly objected to the Morton Thiokol recommendation, claiming the engineers did not have enough data to support their concerns, despite a known history of past O-ring erosion during flight. This philosophy of "prove it's not safe to fly" was a direct reversal of a long-standing NASA tradition of "prove it's safe or we won't launch." Bowing to pressure from NASA, Morton Thiokol managers who originally voted with the engineers to delay launch, reversed the decision and sent a telefax to the Kennedy Space Center approving blastoff. Word of the debate was never passed on to higher level NASA managers because the issue had been resolved.

 

 

Launch Preparations

Other problems also surfaced during the night. The Liberty Star and the Freedom Star, ships used to recover spent booster casings after they fall away from the shuttle, were unable to stay on station because of high seas and had to point into the wind and head toward shore for safety. Arnold Aldrich, manager of the shuttle program at the Johnson Space Center, was told the recovery crew probably would not be able to salvage the booster parachutes and forward nose cones because of the wind and the time it would take the ships to return to station. Aldrich decided to proceed with the countdown anyway, judging it was more important to get Challenger aloft given the upcoming flight schedule than to worry about recovery of relatively minor hardware. And throughout the night, engineers worried about the buildup of ice on the launch pad.

And so, engineers nursed Challenger through another final countdown, hampered by howling wind and ice and battling wind chills of 10 degrees below zero. Keeping a lookout for ice buildups in critical launch pad systems, workers evacuated the pad and began pumping another load of liquid oxygen and liquid hydrogen into Challenger's external tank about 1:25 a.m. But less than 20 minutes later, the carefully timed procedure was aborted, before any fuel had reached the tank, because of trouble with a computer interface system that routed data from hydrogen fire detectors and other safety equipment to launch controllers during the hazardous fueling operation. Two hours later, the fueling procedure was still on hold but mission managers took advantage of the delay to send an ice inspection team back to the launch pad, which was closed during the fueling operation, to inspect water systems for ice damage. After using up all of the countdown's available "hold" time, technicians were able to fix the trouble with the fire detector system and they were cleared to resume loading fuel at 3:55 a.m.. Blastoff was tentatively rescheduled for 10:38 a.m.

So in the early morning darkness at the Kennedy Space Center, Challenger stood bathed in million-candlepower spotlights, clearly visible to travelers on Interstate 95 more than 10 miles away. The crew was allowed to sleep an extra hour because of the fueling delay but they got up on time anyway. After another "last" breakfast, the astronauts made their way to the Astrovan. McAuliffe was wearing gloves to keep out the cold. The white room technicians at the pad who greeted McAuliffe with mortarboards the day before, presented her with a shiny red apple this time around and the crew appeared relaxed and calm as they went through the now familiar ritual of climbing aboard beginning at 8:23 a.m. It was 61 degrees inside the shuttle's unheated crew cabin. Infrared scans of the boosters showed temperatures at the base of the right-side rocket were well below freezing.

"Let's hope we go today," orbiter test conductor Roberta Wyrick radioed the crew from launch control some four miles away.

"We'd like to do that," Scobee replied.

This time, the hatch worked properly - astronaut Ronald McNair took a moment to inspect the now famous handle - but launch was delayed until 11:08 a.m. to give inspection teams more time to assess ice buildups on the launch pad. It was the first indication launch controllers were seriously concerned about ice.

"One of the concerns is these icicles, some of which are several feet long, could possibly break off during liftoff and damage the orbiter and its thermal protection system," launch commentator Harris said.

By 8:44 a.m., the ice team had completed its second inspection of pad facilities. Aldrich decided to delay the countdown to allow more time for ice to melt. At 11:15 a.m., a second ice inspection had been completed and finally, Moore gave the final "go" to pick up the countdown, shooting for liftoff at 11:38 a.m. Computers in launch control were reconfigured for the terminal countdown phase and the shuttle's guidance and navigation systems were updated. A final "poll" of the launch control console operators was taken and the engineers voted to proceed with blastoff. The countdown clock began ticking once more at 11:29 a.m. at T-minus nine minutes to launch.

As commander, Scobee was strapped into the cockpit's left forward seat with pilot Michael Smith sitting to his right, his white flight helmet clearly visible in photographs taken by launch pad cameras. In front of them were banks of instruments and controls along with three computer screens to help them monitor the shuttle's trajectory and the performance of the ship's three high-tech liquid-fueled main engines. Seated between the two pilots directly behind them was Judy Resnik, working as "flight engineer" during launch to help monitor critical instruments. To her right, up against the wall of the cockpit, was Ellison Onizuka and below in the cabin's middeck area, McNair was seated just inside the shuttle's hatch. Ahead and to his right was McAuliffe and seated to her right, Jarvis. Of the three crew members in the middeck, only McNair would have a view during the climb to space through a small window in the hatch to his left. There was no way out barring a safe landing.

 

Shuttle Abort Options Reviewed

When the space shuttle was designed in the early 1970s, engineers faced a crucial decision: what escape systems were possible and which could be economically implemented? For the Mercury and Apollo programs, NASA opted to use solid-fuel rockets to pull the crew capsule away from a malfunctioning booster. For the two-man Gemini program, ejection seats were employed and for the first four shuttle flights, Columbia was equipped with rocket-powered ejection seats to allow the two-man crews to catapult away from a major malfunction. But with the completion of the fourth shuttle mission, the program was declared "operational" and crew sizes were increased. The flight deck of the crew module could only accommodate two ejection seats and the two aboard Columbia later were removed. The other three shuttles in NASA's fleet were never so equipped. Instead, NASA designed the shuttle for "intact" aborts, meaning no matter what the emergency, barring the most extreme catastrophe, the shuttle would be able to make an intact landing somewhere.

The shuttle launch phase is broken down into four basic abort regimes, based on the weight of the spaceship and the thrust level of its main engines. From liftoff through about four minutes, the shuttle could attempt a "return to launch site" abort, or RTLS, in the event one or more of the ship's main engines failed, which would leave the shuttle with too little power to reach orbit. After four minutes, the shuttle is too far from Florida to make it back to the Kennedy Space Center and so the next abort regime is called the "transAtlantic Landing," or TAL, contingency. In this scenario, if one or more main engines fail the crew would attempt to land at an emergency runway in Europe or Africa. But this option is only good between slightly before the end of the RTLS "window" and about seven minutes after blastoff. Depending on the weight of the shuttle's payload, two other abort options may be available, both preferable to the first two: abort once around, in which the crew loops around the planet for an emergency landing in California, or an abort to orbit, in which the shuttle is high enough and going fast enough to achieve a lower-than-planned orbit. In practice, the abort scenarios come in many combinations depending on the weight of the shuttle and the main engine thrust level.

Here are the actual abort regime times for Challenger's flight:

 

2:44 - Two engine TAL at Dakar in the event of a single engine failure
3:24 - Two engine TAL at Casablanca
3:51 - End of RTLS regime
5:20 - Single engine TAL at Dakar in the event of two engine failures
5:20 - Shuttle can achieve lower-than-planned orbit with single engine failure
6:01 - Last chance for two-engine TAL at Casablanca
7:07 - Shuttle can reach lower-than-planned orbit with a single engine at 104%
7:14 - Last chance for a three-engine or two-engine TAL at Dakar
7:26 - Last chance for TAL at Dakar with single engine running
8:26 - MECO - normal "main engine cutoff" in orbit

All four of these contingency aborts were designed with two problems in mind: main engine failures or major on-board system malfunctions, such as loss of cabin air pressure or electrical power. In fact, on July 29, 1985, Challenger's No. 3 main engine shut down prematurely when two temperature sensors failed, misleading the shuttle's flight computers into believing the powerplant was overheating. The failure happened five minutes, 45 seconds after blastoff and commander Gordon Fullerton was able to complete a safe abort to orbit.

All abort modes are inherently risky affairs and the prospect of multiple engine failures is especially disconcerting. Here's how former shuttle commander Robert Overmyer once described the consequences of multiple engine failures during an RTLS.

"If you lost that second prior to starting the RTLS, you're in (an ocean) ditch capability," he said in an interview. "If you lose three engines, forget it, you're probably dead. We advertise all kinds of neat things about tricks we might pull, but the chances of pulling that off are slim to none. We do practice, and that's one of the hardest flying tasks that I've ever had to try and fly, and we practice all the time, killing that second engine and flying the rest of the RTLS on one engine, and because you're getting some aerodynamics, you're at about 225,000 feet but you're starting to get some aerodynamics, the vehicle wants to yaw around on you a little bit and if the only engine you've got running is one of the side engines, it's problematical and nobody has a whole lot of success in completing those in the simulator.

"Things happen so fast that before you can even spit you're out of control. You might recover but by the time you recover, you've lost so much energy you're not going to make it back to the runway, anyway. I can't say we don't pull it off occasionally but losing two engines on an RTLS is a very bad day."

The intact abort modes do not begin until about 30 seconds after SRB separation. During the first two minutes of flight when the solid-fuel boosters are firing a shuttle crew has no survivable abort options.

Early on during the design of the shuttle, engineers considered whether to attempt implementation of systems that would allow survival in the event of a solid rocket booster - SRB - failure. Milton Silveira, a former chief engineer of NASA, said the agency considered the addition of small solid rockets attached to the rear of the shuttle that would help blow the craft away from the solid rockets and external fuel tank in the event of an impending catastrophe. Such rockets would be needed because simply detaching from the fuel tank during SRB firing would result in the rear of the shuttle crashing into the base of the external tank because of aerodynamic forces. Silveira said such systems were prohibitively costly in terms of money and extra weight but perhaps more important, because of additional complexity and risk. For example, such systems would require an extensive network of sensors to detect the problems in the first place and with each new level of complexity the odds go up for potential malfunctions.

The shuttle builders were, of course, aware of this two-minute period of vulnerability. In any case, the decision was made to design the shuttle's booster rockets so they would be fail proof. Because as everyone in the program knew, a major booster problem would doom a shuttle and its crew.

 

 

T-Minus 9 and Counting

At nine minutes to launch, a final 10-minute hold in Challenger's countdown ended and an automatic computer sequence began to initiate the terminal countdown. McAuliffe and Jarvis were strapped tightly into seats on the shuttle's lower deck and had nothing to look at during the climb to space but rows of lockers just a few feet in front of them bathed in the soft glow of fluorescent lights. As the countdown finally picked up at the T-minus nine-minute point at 11:29 a.m., they must have looked over at each other and smiled, elated that their grand adventure was about to begin. McNair, sitting behind and to their left by the lower hatch window, probably enjoyed the view, not having the responsibilities of his colleagues on the flight deck. Then again, he might have been thinking ahead to his role in the deployment of the TDRS satellite later in the day.

At seven minutes, 30 seconds to launch, McNair was able to watch the white room slowly pull away from the side of the orbiter. At five minutes to launch, co-pilot Smith fired up the shuttle's three auxiliary power units, the devices that provide the power to move the shuttle's engine nozzles for steering and operate the wing flaps and rudder. The ground computer ordered Challenger's engine nozzles, wing flaps and rudder to move through a programmed pattern to make sure they would be ready to steer the shuttle after blastoff. At three minutes to launch, the "beany cap" on the nose of the shuttle's external tank, which had been carrying oxygen vapor away to prevent ice buildups during fueling, lifted up and rotated out of the way as the two propellant tanks that made up the external fuel tank were brought up to flight pressure.

"OK, there goes the LOX (liquid oxygen) arm," Smith said over the shuttle's intercom system.

"Doesn't it go the other way?" Onizuka joked. With his crewmates laughing, he continued, "Now I see it; I see it."

"God, I hope not, Ellison," Smith replied.

At 31 seconds to launch, Challenger's four flight computers took over the countdown. It was a beautiful day for a shuttle launch and Scobee took time to update the astronauts below on the middeck of the countdown status: "Thirty seconds down there."

At T-minus 30 seconds, launch commentator Harris read through the final paragraphs of his prepared script: "And we've had a go for auto sequence start. The SRB hydraulic power units have started. T-minus 21 seconds and the solid rocket booster engine gimbal now underway. T-minus 15 seconds...."

The shuttle's flight computers moved the solid rocket booster engine nozzles through a programmed pattern to ensure they would be ready to steer the giant spaceship during the first two minutes of flight. At 13 seconds to launch, huge valves at the base of a nearby water tower opened to direct torrents of water to the launch pad to muffle the acoustic shock of booster ignition. The countdown now carried a sense of urgency with events happening with computer-driven rapidity. At 11 seconds to launch, the self destruct systems of each 14-story-tall solid rocket booster were armed. In the event of a catastrophe, shaped explosives would blow the rocket casings open to neutralize thrust, preventing an out of control rocket from reaching any populated areas. Unlike the liquid-fueled main engines, once an SRB is fired it cannot be shut down until its fuel is exhausted. At 10 seconds to launch, the ground computers gave a final "go" for main engine start and two big sparklers under Challenger's three main engines ignited with a shower to burn away any excess hydrogen that may have accumulated prior to engine start.

"T-minus 10 seconds, GLS (ground launch sequencer) go for main engine start," a technician said over the NASA audio circuit. Harris picked up the final countdown: "...10, 9, 8, 7, 6..."

At 9.5 seconds to launch, the engine start sequence began and prevalves were commanded open to allow liquid hydrogen and liquid oxygen to flow to the engine turbopumps. At eight seconds to launch, Challenger switched to internal cooling and recirculation of liquid hydrogen through key engine systems was terminated. Finally, the shuttle thundered to life. Main engine No. 3 shuddered and belched fire 6.6 seconds before zero, its dirty orange flame quickly changing to a nearly transparent bluish white as the fuel mixture stabilized. Engine No. 2 followed suit 120 milliseconds later and at 6.3 seconds to go, engine No. 1 flashed into action.

"There they go, guys," Scobee told his crewmates over the intercom.

"All right!" Resnik exclaimed.

During the next three seconds, each $35 million Rocketdyne powerplant throttled up to 100 percent power and Challenger's flight computers dutifully checked to make sure the engines were operating properly. They were, and because the nozzles are offset from Challenger's vertical axis, the entire shuttle "stack" leaned toward the external tank with the tip of the tank actually moving some 25 inches before the vehicle rebounded in a phenomenon the astronauts call the "twang."

At three seconds to launch, mechanical systems engineer K.A. Reiley in mission control at the Johnson Space Center in Houston told flight director Jay Greene vents in Challenger's fuselage had opened as planned to maintain proper pressurization in the 60-foot-long payload bay during the ascent. Even at this late stage in the countdown, the ground computers monitoring the shuttle could have called a halt and shut the engines down if a malfunction had been detected. But all systems were "go."

 

 

Challenger Takes Off

Harris: "...3, 2, 1 and liftoff..." An electronic command was sent to each solid rocket booster and eight giant exploding bolts at the base of each SRB detonated, freeing Challenger from the launch pad. A scant quarter of a second later, the first continuous vertical motion was recorded and Challenger was triumphantly on its way, committed to flight.

"Aaall Riight," Resnik drawled over the intercom.

"Here we go," Smith chimed in a second later.

The flight deck crew and McNair on the middeck could see the launch pad tower dropping away to their left, agonizingly close the the screaming spaceship. Smith, as a rookie, no doubt carefully watched his instruments, making sure all three main engines throttled up to 104 percent power as planned seconds before tower clear. At this point, the shuttle was in "attitude hold" with the flight computers steering the ship straight up. At four seconds after launch, the flight computers switched to "major mode 102," a computer program that allowed solid rocket steering once the shuttle had cleared the tower. The 36-story Apollo Saturn 5 moon rockets took 17 seconds to clear the top of the launch pad tower. Challenger did the job in just seven seconds - zero to 70 mph straight up - and immediately began executing its "roll program." The shuttle's computers moved the booster nozzles in opposite directions to roll Challenger about its vertical axis so as the ship climbed toward space, the shuttle would be positioned below the external tank with the crew upside down relative to Earth and the shuttle's wings level in relation to the horizon. This reduces aerodynamic loads on the spaceship and ensures alignment on the proper trajectory.

Scobee and Resnik were busy monitoring their own banks of instruments as the commander called Houston in a routine communications check to note the beginning of Challenger's roll program. Over the intercom, Smith said: "Go you mother" and Resnik reminded the pilot to check his attitude instrumentation.

For observers three miles away at the press site, Challenger's launch was a visual feast. As usual, several seconds went by after main engine ignition before the sound caught up and a low-pitched but smooth roar suddenly snapped into being with almost physical force. A huge cloud of steam billowed into the sky to the right of the launch pad as the flame from the main engines instantly vaporized water being poured on the launch stand to reduce acoustic energy. At the moment of SRB ignition, a gush of dirty brown smoke shot out to the left of the pad and Challenger instantly began moving, coming into view over the top of its launch tower, the flame from the SRBs almost too bright to believe, rivaling the very light of the sun. And then the sound reached the press site from those awesome boosters: a crackling, booming roar that shakes an observer with its raw power. The shock wave from the ignition hit like a minor earthquake and books fell off shelves in press trailers as the shuttle majestically climbed toward space.

Challenger's launch was unique in that it took off from launch pad 39B in the first use of that Apollo moon rocket facility since 1975 when three astronauts blasted off in a Saturn 1B rocket to link up with two Soviet cosmonauts in the Apollo-Soyuz Test Project. The previous 24 shuttle missions were launched from pad 39A and the new perspective made Challenger's take off all the more dramatic to observers.

Harris (continuing) "... Liftoff of the 25th space shuttle mission, and it has cleared the tower." TV tracking cameras show Challenger climbing smoothly and rolling over on its back.

"Houston, (this is) Challenger. Roll program," Scobee radioed eight seconds after blastoff. Astronaut Richard Covey in mission control, serving as "capcom," or capsule communicator for Challenger's flight, reassuringly replied: "Roger roll, Challenger."

Mission Control spokesman Steven Nesbitt in Houston then came up on the Public Affairs audio loop to report on the progress of the flight from inside mission control: "Good roll program confirmed. Challenger now heading downrange." The shuttle, trailing incandescent streamers of flame from its solid rockets arced up through a crystalline blue sky and out over the Atlantic Ocean. Twenty eight seconds after blastoff, Nesbitt reported Challenger's three main engines were decreasing power, as planned, to reduce flight loads as the shuttle approached the region of maximum aerodynamic pressure, called "Max Q."

Aboard Challenger, the crew felt a stronger than remembered buffeting as the shuttle knifed through upper air turbulence with Smith saying over the intercom: "Looks like we've got a lot of wind here today." Indeed, NASA said later wind shear was more extreme for flight 51-L than for any of the previous 24 shuttle missions. Nonetheless, the array of instruments in the cockpit showed all systems operating normally and within guidelines. Nesbitt came back up on the audio loop to report: "Engines beginning throttling down, now at 94 percent. Normal throttle (setting) for most of the flight is 104 percent. We'll throttle down to 65 percent shortly."

At 45 seconds into the flight, Nesbitt said: "Engines are at 65 percent. Three engines running normally, three good fuel cells. Three good APUs (auxiliary power units). Velocity 2,257 feet per second (1,539 mph), altitude 4.3 nautical miles, downrange distance 3 nautical miles...

"Engines are throttling up," Nesbitt said 67 seconds after blastoff. "Three engines now at 104 percent."

"Challenger, go at throttle up," astronaut Dick Covey radioed Scobee. This call told Scobee and Smith that flight engineers agreed with onboard instruments that showed Challenger's systems were operating normally. Scobee and Smith already knew that but the calls from mission control are more verification than anything else and both were looking forward to the "PC less than 50" readout on their computer screens indicating chamber pressure in the solid rockets was less than 50 pounds per square inch, a precursor to SRB separation.

"Roger, go at throttle up," Scobee calmly replied to mission control, still watching his instruments.

Then, suddenly and without any warning, Smith said, "Uh oh," the crew was thrown to the right and almost immediately jolted back to the left. Then, in the blink of an eye, the astronauts may have seen a rush of cotton-candy like smoke wash over the cockpit windows, possibly accompanied by a brief burst of brilliant orange flame. Almost simultaneously, the astronauts were crushed down in their seats by a force at least 12 times greater than gravity as Challenger's fuselage torqued to the left, ripping away from the crew module.

As the nose continued its supersonic climb, at least three astronauts had the presence of mind to activate their "personal egress air packs," emergency air supplies available to each crew member. Of four "PEAPs" recovered after the incident, one, which belonged to Scobee, had not been activated. Smith's had, and because it would have been extremely difficult if not impossible for him to reach the activation switch given its location, NASA officials believe Resnik or Onizuka reached up and turned it on. Smith's PEAP air gauge indicated more than three quarters of the supply was used.

Milliseconds after the nose ripped away from the fuselage, the crew cabin instruments went dead. Scobee may have tried to radio mission control out of reflex, but Challenger's fuel cells, which generate the ship's electricity, were left behind in the fuselage along with the crew's oxygen supply.

Viewed from the press site, Challenger had disappeared behind the exhaust plume of its solid rockets as the shuttle arced out over the Atlantic Ocean. Just as Scobee acknowledged the throttle-up call from mission control, NASA tracking cameras switched to a closeup view of the shuttle. A strange, pinkish glow trailed the rear of the external fuel tank, but for those watching the launch on television, events happened too fast for conscious thought.

Suddenly, the image of the shuttle was swallowed in a mushrooming fireball, a terribly beautiful maelstrom of flaming debris. A sharp crackle of static, unnoticed at the time, came over the air-to-ground audio circuit. From the ground, observers at the press site did not realize what had happened. They saw the exhaust plume suddenly expand dramatically and there was an impression of debris flying through the air. First one and then two solid rocket boosters streaked away from the fireball, apparently intact, certainly still firing at full power. The shuttle was nowhere to be seen and a sense of numbed unreality settled over the Kennedy Space Center. Some observers at the press site thought Scobee was attempting a return to launch site abort. No one seemed able to translate the terrible images in the sky above, framed by the scorpion-like pincers of the booster contrails, into the reality they represented. Thick contrails of white smoke arced away from the explosion, falling toward the Atlantic Ocean. Nesbitt, reading instruments on his console, did not see Challenger's demise and he inadvertently heightened the sense of unreality by continuing, momentarily, to read off figures for a normal ascent.

"One minute 15 seconds," Nesbitt said. "Velocity 2,900 feet per second (1,977 mph). Altitude nine nautical miles. Downrange distance seven nautical miles."

Nesbitt's voice trailed away. On television screens in mission control, the awful reality was clear. One camera zoomed in on a single solid rocket, firing wildly and gyrating through the sky, a strangely riveting sight for NASA engineers because no shuttle solid rocket booster had ever before been seen firing on its own. In the same view, contrails from what must have been remains of the shuttle or its external tank arced toward the sea. Some 10 seconds went by in silence before Nesbitt spoke again: "Flight controllers here are looking very carefully at the situation. Obviously a major malfunction... we have no downlink..."

Finally, Nesbitt confirmed the worst fears of thousands of onlookers at the cape and millions glued to television sets across the nation.

"We have a report from the flight dynamics officer that the vehicle has exploded," he said, voice cracking. "The flight director confirms that. We are looking at checking with the recovery forces to see what can be done at this point."

The flight of mission 51-L was over.