Read Chapter 7: FRANK’S NOTEBOOK

Delaney’s widow called three nights later.

Luc had never given her his number.

“My husband mentioned you,” she said. Her voice was tired but steady. “The French engineer who argued about the power box.”

Luc glanced at Élodie across the room.

“Yes.”

“Can you come by?”

Mary Delaney lived in a modest house in Pasadena. Toys were stacked beneath the living-room window. A photograph of Frank in shirtsleeves sat on the mantel.

She handed Luc a spiral notebook stained with grease.

“I found this in the garage. I don’t know what any of it means. Your name is in it.”

Luc opened the book.

Most pages contained calibration notes, connector pinouts and reminders. Then he found the DCP entries.

Z AXIS DROP REAL. NOT FIXTURE.

P. WANTS RETEST LOW.

S.B. SAYS WAIVER ALREADY SIGNED.

Later:

UNIT 9 DROP ARTICLE — POWER LOSS SAME PATTERN.

And finally, one week before Delaney’s death:

FOUND ORIGINAL LOT TEST. HOT/VIB FAIL. COPY IN FILE DOESN’T MATCH.

ASK L.M.? NEED SECOND SET OF EYES.

Luc looked up.

Mary Delaney watched him. “Is it bad?”

“Yes.”

“Bad for Frank?”

He understood the question. Did this mean her husband had caused his own death? Had he made a mistake? Had he been involved in something improper?

“No,” Luc said. “It suggests he found a problem and was trying to document it.”

She sat down slowly.

“The company told me the accident was his fault.”

Luc said nothing.

“They said he opened the wrong valve.”

“I cannot tell you what caused the accident.”

“But you don’t believe them.”

“I do not know enough.”

Mary looked at the notebook. “Frank never opened anything without checking twice. He checked the oven before bed three times.”

On the drive home, Élodie asked, “Do you think he was killed?”

“I think Pike said something no innocent manager should say.”

“That is not an answer.”

“No. I do not know.”

But the notebook changed the scale of the problem. Delaney had found an original lot test—evidence that the relay vendor data had also been altered. If true, the falsification was not a single improvised decision after Galveston. It was systematic.

Luc brought the notebook to Reed the next morning.

Reed read the entries, then sat very still.

“We need to go outside Asterion.”

“To NASA?”

“Maybe higher.”

“Without the original vendor data, they will call this a dead man’s suspicion.”

“We have the strip charts.”

“They will say fixture error.”

“We have the thermal-vacuum dropout.”

“They will say non-repeatable anomaly.”

Reed closed the notebook. “You sound like Pike.”

“I sound like the men who will defend him.”

They needed the original lot test.

MidSouth Electromechanical was located near Huntsville, Alabama, and supplied relays to several aerospace contractors. The supposed records-room fire had destroyed environmental test logs, but Delaney’s note implied that a copy had survived long enough for him to see it.

Luc requested travel on the pretext of supplier engineering review.

Pike denied it within an hour.

“Why?” Luc asked.

“Budget.”

“I can combine it with the Huntsville integration meeting.”

“No.”

“Why?”

Pike looked at him. “Because I said no.”

That evening Reed bought two airline tickets with cash.

“You are paying,” he said. “I have children.”

Luc stared at him.

“You do not have children.”

“I plan ahead.”

For the first time in days, Luc laughed.

Then Reed’s face became serious.

“We go Friday night. Back Sunday. Nobody at Asterion knows.”

Luc nodded.

In his notebook he wrote one new line:

Find what Frank found.

Mary Delaney had not known that Frank kept engineering records at home.

When Luc called to ask permission to examine anything connected with the relay work, she led him to a workbench in the garage. A coffee tin held loose screws. A child’s bicycle leaned against a cabinet. The ordinariness of the room made the technical pages feel intrusive.

Frank had drawn the relay contact mechanism three times, each sketch more detailed than the last. Beside one diagram he had written HEAT → SPRING LOSS? and circled the question twice.

Luc understood the proposed mechanism. The contact spring depended on retained temper. At elevated temperature, a marginal heat treatment could reduce force. Under vibration, the contact might separate long enough to open the circuit.

“He was trying to prove the metallurgy,” Luc said.

Mary stood in the doorway. “Was he qualified to do that?”

Luc looked at her.

“He was qualified to notice.”

The answer seemed to matter to her.

At Asterion, titles drew thick boundaries around knowledge. Engineers designed. Technicians executed. Quality signed. Managers decided. But machines did not respect job descriptions. A technician who watched hardware every day could see patterns a graduate engineer missed.

Frank’s notes also contained a phone number in Huntsville and the name W. KESSLER.

Reed called the number from a pay phone rather than an office line.

An elderly man answered.

When Reed mentioned lot 62-B, the silence on the other end lasted five seconds.

“Who gave you this number?”

“Frank Delaney.”

Another silence.

“Frank dead?” Kessler asked.

“Yes.”

The man exhaled.

“Then come Saturday. Not the plant. I’ll tell you where.”

That was how Huntsville entered the case: not through NASA channels, but through a dead technician’s pencil note.

The meeting with Kessler lasted far longer than Luc expected because the old supervisor wanted to tell the whole history of MidSouth's relay troubles.

The spring supplier had changed furnace settings in 1962 to improve production yield. The change passed dimensional inspection but altered retained hardness after prolonged heat exposure. Most relays still worked perfectly in ordinary service. Only under combined heat and vibration did the reduced contact force matter.

“That is why management hated the finding,” Kessler said. “It was not a clean defect. Ninety-nine parts looked fine until you tested the one condition the customer cared about.”

“Did MidSouth correct the process?” Luc asked.

“Eventually. New lots were good. But Asterion already owned 62-B and did not want to scrap them.”

“How much money?”

Kessler named a figure that seemed small compared with Apollo budgets.

Reed whistled. “They risked all this over that?”

Kessler shook his head. “Not over the relay cost. Over what admitting the bad lot would trigger. Requalification. Schedule review. Maybe redesign. In big programs, the expensive part of a bad component is admitting it is bad.”

Luc wrote the sentence down.

Kessler noticed.

“Don't quote me.”

“I am only remembering.”

“That is worse.”

Before they left, the old man showed them one more document: an internal MidSouth telegram from Asterion asking whether the relay could be described as “acceptable with application controls.”

Kessler had handwritten his answer at the time: ONLY IF APPLICATION EXCLUDES SIMULTANEOUS HIGH TEMP / VIBRATION.

The DCP application did not exclude either.

Luc photographed the telegram.

Kessler watched him carefully.

“You know what happens when this gets out?”

“Yes.”

“No, you know what you hope happens. Investigations don't obey engineers.”

Luc lowered the camera.

“What do you suggest?”

“Fix the machine first. Then decide how much of the institution you need to break.”

That advice, unlike Kessler's refusal to sign a statement, Luc respected immediately.