ARLINGTON, Va. — The Feb. 23 crash of a B-2 bomber at Andersen Air Force Base, Guam, was caused by bad computer data sent to flight control computers from three tiny sensors on the bomber’s wings, according to the Air Force accident investigation report released Thursday.

The B-2 was on takeoff when the flight computer falsely told its pilots that it was moving along the runway at 140 knots, fast enough to become airborne, according to Mag. Gen. Floyd Carpenter. Carpenter, who is vice commander of the 8th Air Force at Barksdale Air Force Base, La., said the computer then told the aircraft it was going into a nosedive, when the pilots were actually in the process of lifting the craft off the ground.

The computer ordered the B-2’s nose to pitch up 30 degrees.

The pilots, who were on the last few days of a four-month Guam deployment from the 509th Bomb Wing from Whiteman Air Force Base, Mo., desperately tried to correct the movement.

"They tried their best to recover the aircraft," Carpenter said. "They didn’t make any mistakes."

But the bomber was moving too slow — only 130 knots — and the angle its nose was pitched was just too steep, Carpenter told reporters Thursday. The aircraft went into a fatal stall.

The B-2’s left wing hit the runway first, and the pilot and co-pilot ejected.

"By the grace of God, they were safe," Carpenter said.

The pilot sustained only local injuries, but the co-pilot’s spine was fractured. He was rushed to Tripler Army Medical Center in Hawaii for treatment and has since returned to Whiteman, with a full recovery expected, the report said.

The $1.4 billion plane was a total loss, according to the report.

Carpenter said the seven-week investigation by 14 Air Force officials and civilians revealed that the fault lay with the B-2’s sensors, a network of 24 tiny sniffing devices located throughout the skin of the aircraft that "weigh" air pressure across the body of the jet.

"This was not a failure by any pilot or technician," Carpenter said. "The aircraft [also] performed as it was designed. It truly was a procedure error."

The sensors have a calibration system — like scales that have to be reset to "0" before they weigh accurately — that the aircraft’s maintainers employ any time all 24 are not showing the same reading, Carpenter said.

But in Guam’s humid environment, three of these sensors had become moist and distorted the calibration data, telling the aircraft’s computers everything was well when it was not, the investigators said.

As a result, the B-2’s flight controls, as well as its airspeed and "angle of attack" data, were wrong, Carpenter said.

Moisture in the sensors is normal, and the bomber’s designers have planned for it by building "pitots," or heaters, that can dry out each device with a blast of 500-degree air, whether the B-2 is in the air or on the ground.

But until the 2008 accident, no one had ever known the sensors to get wet preflight, Carpenter said.

Although there was a hint of the possibility during a 2006 deployment to Guam, he said.

At that point, a few maintainers had noticed moisture in some sensors, he said. They began turning on the heaters as part of their own preflight checks.

"But it was applied sporadically," he said. "Some people know about it, and some didn’t."

The Air Force is not blaming the technicians or officers for failing to standardize the 2006 procedure because "it happened so infrequently, and even then, the data calibration typically fixed the problem," Carpenter said. "The jets continued to fly perfectly fine."

But as a result of the accident board’s findings, the Air Force has modified its orders for the B-2’s preflight procedure, Carpenter said: All sensor pitots must be turned to "on" before calibrations are performed.

Maintainers "are also doing a much more rigorous inspection of the entire [sensor] system," Carpenter said.

Sign Up for Daily Headlines

Sign up to receive a daily email of today's top military news stories from Stars and Stripes and top news outlets from around the world.

Sign Up Now