He switched on the system.
Pumps came alive.
A low mechanical hum filled the garage.
Isaac adjusted several controls.
A pressure trace appeared on an old monitor.
Stable.
Then he activated the thermal cycle.
Another waveform appeared.
He introduced vibration.
Still stable.
Then altered command timing.
The pressure waveform developed a small oscillation.
Eleanor leaned toward the screen.
Isaac changed the thermal timing again.
The oscillation doubled.
A second harmonic appeared beneath it.
“Thirty-one minutes.”
Eleanor whispered.
Isaac nodded.
“The thermal transit period excites the valve housing.”
“But the real damage occurs because the pressure-control pulse intersects a structural resonance band every third cycle.”
“The material doesn’t simply fatigue from temperature.”
“The valve assembly slowly hammers itself apart.”
Eleanor watched the waveform.
“Why wasn’t this in your report?”
“Because I did not have enough data in 2017.”
“I suspected the coupling.”
“I couldn’t prove it.”
“And now?”
“You have twenty-one months of actual orbital telemetry.”
“That gives me what Nexora never did.”
“The real forcing function.”
Eleanor looked at him.
“What do you need?”
Isaac shut down the rig.
“Full telemetry.”
“Hardware drawings.”
“Control law.”
“Temperature history.”
“Propellant consumption.”
“Original qualification results.”
“Every change Nexora made after I left.”
“Can you fix it?”
Isaac looked toward the machinery.
“I can tell you whether it can be fixed.”
“That isn’t what I asked.”
“It’s the only honest answer you’re getting today.”
Six years earlier, honesty had cost Isaac almost everything.
In March 2017, Nexora Aerospace occupied twenty-two floors of a mirrored office complex outside Houston.
Isaac worked on twelve.
He had joined Nexora at twenty-five.
By thirty-eight, he was one of the most respected propulsion analysts in the company.
Not socially.
Isaac had never been especially good at offices.
He spoke too directly.
Did not network.
Forgot birthdays.
Avoided golf.
His shirts were often wrinkled.
He brought his own coffee because he thought the executive-floor machine tasted like “heated plastic regret.”
What he could do was see interactions.
Most engineers became extraordinarily deep in one discipline.
Isaac understood enough across several to notice when two correct answers created a wrong system.
Prometheus had bothered him from the first design review.
On paper, it was elegant.
A compact pressure-regulation assembly designed for long-duration spacecraft maneuvering.
Redundant valves.
Low mass.
High efficiency.
Beautiful qualification results.
The problem appeared only after Isaac compared several datasets nobody had combined.
Thermal cycling from orbit.
Valve actuation frequency.
Mechanical vibration modes.
Material fatigue behavior.
Each remained inside approved tolerance separately.
Together, the margins collapsed.
He spent six weeks checking himself.
Ran models again.
Asked colleagues to inspect fragments without revealing his full concern.
Changed assumptions.
Increased safety factors.
Reduced them.
The instability remained.
By month eighteen to twenty-four, the regulator response began drifting.
By later mission stages, pressure authority could become unreliable.
The schedule could absorb a redesign.
Barely.
The contract could not.
Nexora had promised hardware delivery in eleven months.
A four-month redesign threatened milestone payments worth tens of millions.
Isaac’s forty-seven-slide presentation lasted fifty-three minutes.
Cameron Ashford sat at the far end of the room.
Cameron had once been an excellent engineer.
That made everything worse.
He understood what Isaac was saying.
He simply understood the contract too.
“Your coupled model is theoretical.”
Cameron said.
“All models are theoretical until hardware confirms them.”
“Our qualification hardware passed.”
“Because the environmental cycles were tested independently.”
“They were tested according to approved procedure.”
“The procedure is wrong.”
That sentence changed the room.
Giselle Hartmann, Nexora’s chief executive, looked up.
“Isaac.”