Report: Flight control system problem caused F/A-22 crash

Air Force Link ^ | 6/8/2005 | AFPN Staff

[Excuse_My_Bellicosity]

6/8/2005 - LANGLEY AIR FORCE BASE, Va. (AFPN) -- A flight control system problem caused an F/A-22 Raptor to crash on the runway at Nellis Air Force Base, Nev., on Dec. 20, according to an Air Force report released June 8.

The pilot ejected and sustained minor injuries. The $133.3-million aircraft, assigned to the 422nd Test and Evaluation Squadron at Nellis, was destroyed when it crashed. Additional damage was limited to an arresting cable, runway guide sign, runway light and the runway itself.

The flight control system malfunction was caused by a brief power interruption to the aircraft’s three rate sensor assemblies, which caused them to fail. The assemblies measure angular acceleration in all three axes: pitch, roll and yaw. With three failed assemblies, the F/A-22 is not able to fly, investigators said.

When the pilot shut down engines for maintenance servicing, he left the auxiliary power unit running. Based on technical order guidance, he believed the power unit would supply continuous power to the flight control system. However, there was a less-than-one second power interruption to the assemblies during engine shutdown.

There is no automatic warning of this condition. To discover it, the pilot would have had performed a diagnostic test. The pilot accomplished a successful test before engine shutdown, and because the power unit was on, he believed a second test was unnecessary. (Courtesy of Air Combat Command News Service)

[PAR35]

apparently due to a problem with the system handbook.

No, it was caused by a lack of electricity. If a battery (or as another poster pointed out, a capacitor) had been wired into the system, there would not have been a power loss in the brief period of time between engine shutdown and apu takeover.

The problem might be solvable by either a manual or computer driven cut-over to the APU before the engine is shut down.

[sgtbono2002]

Sensor , Computer , its all part and parcel of the system.

[expatpat]

Well, there are multiple problems here, but the pilot believed after reading the 'manual' that the APU would continue to power the sensors with the engine off. That is a perfectly normal thing to expect, and the system handbook should have highlighted a deviation from this.

The lack of electricity would have not been an issue if the pilot had known this. He would not have tried to take off until the gyros were fully up to speed, and no battery or capacitor would be needed.

[bvw]

I'd call it a software design drop-out. Such sequence after unsual circumstance contingencies can be caught by astute designers or testers. However many aren't. I caught some on an F/B eject sequence wherein a system failover occurred. They had been there, latent, for years and years.

But I am not representative of the average sw engineer. Sui generis.

[Excuse_My_Bellicosity]

I'm sure this mishap caused some tech order changes.

[expatpat]

There's an old saying in the software business: "You never find the last bug". You could call it a software problem, I suppose, but it seems more like a system-design problem. The power to flight-critical sensors should itself be regarded as flight-critical, and the lack of such power in the case of an engine failure/shut-down in flight is IMO a BIG mistake.

It's a good job it was caught during test and eval, and no one was hurt.

[expatpat]

Yeah, and maybe one or two engineering vacancies.

[Wiz]

Fly by light when it is about F/A-22...