Isaac had spent six weeks preparing it.
The concern began with a component almost nobody outside the propulsion group ever discussed: the secondary pressure-regulation valve inside the third-stage drive assembly.
Under conventional qualification testing, the valve performed correctly.
Static pressure tolerance fell within specification.
Thermal testing passed.
Vibration testing passed.
Cycle life passed.
Separately, every requirement passed.
The problem appeared only when Isaac stopped treating those qualification categories as separate.
A polar-orbit vehicle did not experience temperature, vibration, and pressure in isolated laboratory departments.
It experienced all of them together.
Every ninety-three-minute orbit moved the spacecraft from direct solar exposure through shadow and back into sunlight.
Small dimensional changes accumulated.
Materials expanded.
Contracted.
Recovered.
Not quite completely.
Then did it again.
Isaac’s simulation suggested the secondary valve sat inside a tolerance window narrow enough that repeated thermal cycling could gradually alter its behaviour.
The initial effect would be insignificant.
A pressure deviation so small a program manager might call it sensor drift.
By approximately eighteen to twenty-four months, depending on orbit and duty cycle, the deviation would become cyclic.
By roughly month twenty-eight, control authority could deteriorate beyond the propulsion system’s certified operating margin.
Not an explosion.
Not cinematic destruction.
Something aerospace engineers feared more.
A system continuing to operate while quietly becoming less controllable.
Isaac asked for four months to redesign the valve housing and modify the tolerance scheme.
The government contract required hardware delivery in eleven months.
Four months was unacceptable.
At the far end of the boardroom sat Cameron Ashford.
Vice president of technical programs.
Cameron was brilliant in the way corporations rewarded most reliably.
He understood engineering well enough to speak its language and management well enough to translate uncertainty into schedules.
Beside him sat Giselle Hartmann, Nexora’s chief executive.
Giselle had spent her career acquiring smaller aerospace contractors and turning them into divisions that met deadlines investors appreciated.
Isaac respected her intelligence.
He distrusted what she did when intelligence and schedule came into conflict.
A corporate attorney sat near the wall.
Isaac had not known legal would attend.
He remembered noticing the attorney’s yellow notepad before starting the presentation.
That should have warned him.
At slide thirty-one, Cameron interrupted.
“Your model parameters exceed the validated heritage range.”
“They exceed the test profile,” Isaac said.
“That’s the validated range.”
“No. That’s the tested range.”
Cameron folded his hands.
“The customer accepted those parameters.”
“Because they weren’t given the combined model.”
“The customer was provided the approved qualification package.”
“Which treats thermal cycling and mechanical vibration separately.”
“They are separate certification regimes.”
“The spacecraft doesn’t know that.”
A program manager looked down.
Isaac advanced to the next slide.
The graph showed predicted pressure deviation over repeated orbital transitions.
“This is what happens when the two transfer functions interact.”
Cameron leaned back.
“That interaction has not been demonstrated experimentally.”
“We didn’t test for it.”
“Then it remains hypothetical.”
“Every failure is hypothetical until the hardware proves you wrong.”
Giselle finally spoke.
“Isaac.”
He looked toward her.
“We understand the concern.”
“No, you understand the schedule impact.”
Silence.
The attorney began writing.
Isaac knew he had crossed some invisible line.
He continued anyway.
“I am not signing design approval.”
Cameron’s expression changed.
Very slightly.
“Then someone else will.”
“That doesn’t make the valve safe.”
“It makes the program compliant.”
“No. It makes the paperwork complete.”
The meeting ended nine minutes later.
Two weeks afterward, Isaac received a termination letter.
Violation of internal security protocol.
Improper retention of restricted analysis.
Failure to follow established review procedures.
The charges were specific enough to sound serious and vague enough to be almost impossible to challenge without litigating against one of the best-funded aerospace corporations in the country.
Isaac’s access badge stopped working at 3:12 p.m.
Security met him outside his office.
He was allowed one cardboard box.
A colleague named Martin avoided eye contact.
Another whispered, “I’m sorry,” while pretending to search a cabinet.
Cameron never appeared.
Giselle did not call.
Human resources gave Isaac a settlement agreement and reminded him of confidentiality obligations.