He handed the tablet back without commenting. He moved to the engine compartment. One of the five engineers, young guy, tired eyes, the specific exhaustion of someone who'd been staring at screens for 8 hours, stepped aside to give him room. He looked like he wanted to say something, but wasn't sure if he was allowed to. " What did you check?"
Ethan asked him quietly. The engineer glanced toward Vera's group, then back at Ethan. Everything on the diagnostic tree. We ran the full software suite twice, checked the fuel system, the FADEC hydraulic pressure rotor head. What did the FADEC report?
That's the thing. The engineer lowered his voice further. It's not reporting anything. No fault codes. Systems reading nominal across the board, which is why we can't.
He stopped, shook his head. It shouldn't be possible. An S-76D has 17 redundant monitoring systems. If there's a failure mode, something should be flagging. Unless the failure isn't in a monitored system, Ethan said.
The engineer stared at him. Everything on this aircraft is monitored. Almost everything, Ethan said. He leaned into the engine compartment. Here was the thing about the S-76D that most mechanics who hadn't worked on them extensively didn't fully appreciate.
The aircraft was as of its D series redesign almost entirely fly by wire and computerized which meant that when something went wrong every trained instinct told you to go to the computers, go to the data, go to the software because that was where the problems lived in a modern aircraft. And that was exactly why a sufficiently obscure mechanical failure could hide in plain sight. Ethan wasn't looking at the computers. He was looking at the fuel control unit. The physical mechanical fuel control unit, the component that sat upstream of all the electronic monitoring, the piece that governed fuel delivery to the engine before the FADEC ever saw a number.
He checked the line connections. He checked the mounting bracket. He checked. He stopped. He got closer.
He pulled a small flashlight from his toolbox and shone it at a specific bracket on the fuel control system where a particular safety component, a physical lock that prevented inadvertent changes to the fuel mixture ratio under high vibration conditions was seated. It was installed, but it was installed at 180° from where it was supposed to be. It was such a small thing, such a simple thing. The component itself was functional. It wasn't broken.
It wasn't defective. It hadn't failed. It had simply been put in upside down during what the maintenance log recorded as a routine fuel system inspection 6 weeks ago. And because it was functional, just installed wrong. Every digital diagnostic system on the aircraft read it as nominal because to the computers it was.
The computers didn't know the orientation of a physical component. The computers knew what the sensors reported, and the sensors reported normal values right up until the moment the helicopter was at a certain RPM under a certain load condition, at which point the incorrectly installed lock allowed a micro flutter in fuel delivery that compounded into a cascade failure of the engine timing sequence. It wouldn't show up on any screen. It wouldn't show up unless you were looking at it, unless you put your hands on the machine and looked. Ethan straightened up and turned around.