Chapter 31: The Engineering Problem
Howard arrived at the shared workspace at 9:07 AM with the specific urgency of a man whose project timeline was three days from catastrophic slippage.
"I need eyes," he said.
Leonard looked up from his laptop. Raj turned from the whiteboard where he had been sketching orbital calculations. Adam set down the documentation he had been reviewing.
"What's happening?" Leonard asked.
"Thermal coupling." Howard pulled up a chair and dropped into it. "Sensor array. NASA validation test in 72 hours. The coupling component is behaving unexpectedly at temperature and I cannot figure out why."
"Define unexpectedly," Adam said.
Howard ran his hands through his hair — the gesture of someone who had been up late, probably multiple nights in a row. "Micro-expansion is throwing calibration off by 0.003 millimeters. That doesn't sound like much, but for the tolerance we need, it's catastrophic. I've run the standard diagnostics. Nothing. The thermal models say it should be stable. It is not stable."
Adam's Molecular Conductor was already registering the problem through Howard's description alone. The material composition implied by "thermal coupling" in a NASA-grade sensor array. The expansion coefficient of the standard alloys used in that application. The way micro-expansion behaved differently under gradient versus magnitude stress.
He did not say any of this.
"What material is the coupling housing?" he asked instead.
"Titanium alloy — Ti-6Al-4V. Standard aerospace grade."
Two more questions. Both sounding like an engineer thinking aloud. Both confirming what the Molecular Conductor had already mapped.
"And the mount orientation?"
"Perpendicular to the thermal source. Standard configuration."
"What's the temperature differential across the component during operation?"
Howard checked his phone — notes from last night's testing, probably. "Fourteen degrees Celsius gradient from the source side to the ambient side."
Gradient, not magnitude. The expansion is directional.
Adam could feel it now — the way the Ti-6Al-4V would behave under asymmetric thermal load. The crystalline structure responding to the temperature difference not radially, but along the gradient axis. The micro-expansion accumulating in a direction the standard models did not account for.
"The problem is directional expansion," Adam said. "The coupling is responding to the thermal gradient, not the thermal magnitude. Your models are calculating radial expansion because that's the standard assumption for uniform heating. But with a 14-degree differential, the expansion is predominantly along the gradient axis."
Howard stared at him.
"Try reorienting the component mount by 11 degrees," Adam continued. "Rotate it so the gradient axis aligns with the lowest-tolerance direction. The expansion will still happen, but it won't affect your calibration."
"How did you get that," Howard said, "without running the thermal model?"
"The material composition suggests it. You can verify with the model."
Howard was already pulling up his laptop. His fingers moved across the keyboard with the particular intensity of someone who wanted to either confirm or disprove something immediately.
The simulation ran.
Ninety seconds.
The result populated.
"Eleven point three degrees," Howard said. "You got it within a third of a degree from a material description."
"It is a pattern the material makes."
"That is not how that calculation works."
"The pattern is consistent. The material tells you if you know what to listen for."
Howard looked at him for a long moment. Then he closed his laptop and stood up.
"You just saved my NASA timeline."
"The component told you. I just translated."
"It really did not."
Howard paused at the door. "You are weirdly good at materials."
"It is in the documentation focus."
Howard accepted this — or at least, he accepted it enough to stop asking questions. He left, walking faster than he had arrived, already pulling out his phone to coordinate the reorientation with his team.
---
Leonard had been watching the entire exchange.
He did not say anything until Howard was gone. Then he turned back to his laptop, typed something brief, and closed the window.
"That was impressive," he said.
"It was pattern recognition."
"From a material description. Without a model."
Adam shrugged.
Leonard did not push. He returned to his work. But Adam could see it in the way his attention had shifted — a new data point filed, added to whatever informal log Leonard was keeping.
Four times now. The elevator stairs comment during the first week. The Raj synthesis question at the coffee shop. The boundary condition redirect with Sheldon. And now this.
Leonard did not have a formal investigation. He did not have a calibration log or a Folder A or a spy hypothesis. He had something potentially more dangerous: the quiet accumulation of observations by someone who liked Adam and was starting to notice that things did not quite add up.
---
[ADAM'S APARTMENT — EVENING]
The coffee had been terrible.
Howard had bought it from the engineering building vending machine as a thank-you — C-4, the button for what the machine optimistically labeled "Premium Blend." It had tasted like burnt plastic filtered through disappointment.
Adam had drunk it anyway.
Now, alone in his apartment, he opened his notebook and wrote:
"Molecular Conductor: social use ceiling approaching. The solution method is becoming recognizable if I use it too often."
He paused. Read the sentence back.
"Vary the approach."
He underlined "vary."
The problem was frequency. Once, the "material tells you" framing was interesting. Twice, it was a quirk. Three times, it became a pattern. Patterns invited investigation.
He needed to space out the demonstrations. Mix in conventional reasoning. Appear to struggle before arriving at solutions, even when the Molecular Conductor had already provided the answer.
The cover requires imperfection. Perfect accuracy is suspicious. Reasonable accuracy is background noise.
He wrote: "Frequency: no more than once per five chapters."
Then he crossed out "chapters" and wrote "major interactions." The structural bleed was subtle, but it mattered.
---
Howard's sensor array would pass the NASA validation test in 72 hours.
Four months later, the mission would be mentioned in a press release. The thermal coupling component would work exactly as designed, oriented at 11.3 degrees from its original position.
Adam's contribution would not appear in any documentation.
This was exactly correct.
Get Early Access to New Chapters
Thank you for reading. For those who want to skip the wait, my Patreon is currently 21 chapters ahead of the public sites.
Schedule: 7 new chapters released every 10 days.
Benefit: Gain a significant lead of 7 to 21 chapters depending on your tier.
Support the project and start reading the next arc now: Patreon.com/IsekaiStories
more chapters + every language you asked about — it's all on unwrittenrealm.com
Arabic, Korean, Spanish, Hindi, Russian, and 9 more. completely free.