It was a reckoning, and it had only just begun. The drive to Huntsville took 47 minutes. Malcolm spent most of it looking out the window. Not at the road, not at the passing landscape, but at something that wasn't there anymore. Rebecca Sloan sat across from him in the second row seat, giving him the silence the way someone gives a man space when they can tell he's working something out internally that has nothing to do with the data on her tablet.

She was good at reading rooms. Malcolm noticed that about her in the first 10 minutes. She didn't fill silence with noise. The third man in the vehicle, the one who hadn't introduced himself, sat up front and said nothing the entire drive. At the 30 minute mark, Malcolm spoke first.

How long have you been tracking the instability? 6 weeks. Rebecca said it appeared during the third full duration ground simulation. We thought it was instrumentation error initially. By the fifth simulation, we knew it wasn't.

Who were the two contractor teams? A pause, short, but present. Vertex Dynamics was the first. They spent 8 days on it and concluded the oscillation was thermal in origin and within manageable parameters. We ran two more simulations after their report.

Both ended in automatic shutdown. She looked at the tablet in her lap. The second team was from Apex Meridian Systems. The name landed in the vehicle like something dropped on a hard floor. Malcolm didn't move.

Who led the Apex team? Rebecca looked at him directly. I need you to know that I found out about your history with that organization after we identified your paper. Not before. I understand that.

Who led the team? Richard Vale. Malcolm turned back to the window. 7 seconds of silence. Vale told us the resonance was a software artifact.

Rebecca continued her voice, measured and precise the tone of someone delivering a briefing they've rehearsed. He recommended we update the ignition control firmware and recalibrate the turbo pump governor settings. We did both. The next simulation 2 days ago produced the worst oscillation we've recorded. The automatic shutdown triggered at 9 seconds.

She paused. The simulation after that is what I showed you this morning. Malcolm said, "You followed Vale's recommendations and the problem accelerated." Yes. And then someone found my paper.

One of my junior engineers, Daniel Cho, he's been pulling every publication on cryogenic restart instability he can find. He came to me 3 days ago with your paper flagged and highlighted and asked me if I'd considered the possibility that the problem was resonant rather than discreet. She glanced at her notes. I had your file pulled within the hour. Malcolm was quiet again.

What's Vale's current role on the program? He asked. Senior contractor adviser. He doesn't have operational authority, but he has access to all program meetings and review boards. She hesitated.

He has relationships with people at the senior level. I know what he has, Malcolm said. The vehicle went quiet again. And this time, Rebecca let it stay quiet because she understood she had just told Malcolm Reed something important. Not about the engineering problem, but about the environment surrounding it.

And the environment was its own kind of problem. She watched him stare at the window. She did not know the full shape of what had happened 8 years ago. She knew the official record. Malcolm Reed, formerly of Apex Meridian, separated from the program following an internal compliance dispute.

She knew the paper trail was thin and one-sided in the way that internal compliance disputes always were when the person being disputed had less institutional power than the person doing the disputing. She had seen that pattern before in her 22 years in the industry. She knew what it looked like. What she did not know was what it had cost him. she was beginning to. They arrived at Marshall Space Flight Center at 11:43 in the morning.

Malcolm had been inside facilities like this before. Enough times that the size of the buildings and the weight of the security protocols didn't register as impressive anymore, only as familiar the way a soldier might feel walking back into a base after years away. Familiar and loaded with memory at the same time. He passed through badging and clearance with Rebecca beside him, received a visitor credential, and was escorted through two corridors before they stopped at a set of double doors. Rebecca pushed them open.

The propulsion analysis lab was large, climate controlled to a degree that was immediately noticeable and occupied by 12 people who all looked up simultaneously when Malcolm walked in. Several of them looked at his visitor badge. A few looked at his hands, still faintly marked with the ghost of grease that three rounds of industrial soap hadn't completely removed. One woman in the back looked at him and then looked at Rebecca with an expression that said she expected an introduction. A young man near the central workstation stood up quickly.

Early to mid20s Korean, the specific energy of someone who had been waiting for something to arrive and was trying not to look too eager now that it had. Daniel Cho he said extending a hand. I found your paper. Malcolm shook it. I know.

I was told. I flagged the resonance model on page 11, the frequency coupling diagram. Daniel was already reaching for a printed copy of the paper on his desk. It was marked up heavily in three colors of pen. I've read it probably 15 times in the last 3 days.

The way you modeled the interaction between chamber oscillation and feed line thermal varants. Nobody had done that before. I mean people had considered them as separate phenomena but the coupling model was Daniel. Rebecca's voice was gentle but firm. He stopped.

Sorry. Sorry. He exhaled. Welcome to Marshall. We're glad you're here.

Tell me about the three stable runs. Malcolm said. He hadn't sat down yet. Out of 14 simulations, three were stable. What was different?

Daniel blinked at the directness, then turned immediately to his workstation. I can pull the comparison data right now. Do that. While Daniel typed, Malcolm moved along the wall of monitors displaying the telemetry from the most recent simulation. He looked at each one methodically, left to right, top to bottom.

The way he read any data set, full picture first, then detail. The room watched him do it. A man in the far corner had not stood up when Malcolm entered. He was older, gray-haired, with the deliberately relaxed posture of someone who was accustomed to being the most important person in any room and was managing the fact that this particular moment was not organized around him. Malcolm recognized him without looking directly at him.

Richard Vale. He was on his phone talking quietly, but his eyes tracked Malcolm across the room. Malcolm did not acknowledge him. Dr. Reed Daniel turned from his workstation.

The three stable runs I found two consistent differences. Fuel temperature at turbo pump inlet was 1.7° warmer in all three cases. And the restart timing interval, the gap between engine cutff and restart initiation was between 0.3 and 0.5 seconds longer than in the unstable runs. Malcolm stopped walking. He turned around.

Show me the timing intervals for all 14 runs. Exact values not rounded. Daniel pulled them up. Malcolm leaned over the workstation and looked at the numbers. 11 unstable runs.