57: Chapter 57 Let the Mechanical Structure Speak for Itself

The next day, Sunday, early morning.

"I'm heading out for a bit and will be back in the afternoon."

"Where are you going?"

"Jiangcheng University. I have an appointment with a teacher to ask about a few problems."

"A teacher at the university? Then you must be polite. Don't go empty-handed. Should I get you something to take?"

"No need, Mom. We're just talking about study matters, and bringing things would actually be weird. I'm leaving."

"Ride your bike slowly on the way and come back early for dinner."

Jiang Lin responded and walked out the door.

Passing two traffic light intersections, a wide inland river lay across the path ahead, spanned by an aging two-way, four-lane concrete bridge.

Jiang Lin rode his bicycle onto the bridge deck.

There was fog over the river, like a translucent veil suspended above the slowly flowing water, blurring the silhouettes of the scattered modern buildings on the opposite bank.

He had never seen fog in the Wasteland.

The air in the Wasteland was dry with a low moisture content. The early morning wilderness had only wind, without this kind of dampness that softened everything.

He took a deep breath of the moisture-filled air, exerted strength in his legs, and his bicycle accelerated on the bridge deck, plunging headfirst into the thin mist ahead.

Forty minutes later, he arrived at Building B of the School of Physics at Jiangcheng University.

Parking his bicycle downstairs, he took out his phone, opened the WeChat account he had just added yesterday, and sent a message to Teacher Lu.

"Teacher Lu, I've arrived. I'm at the main entrance of Building B."

Two minutes later, Professor Lu Zhixing replied with a message.

"Wait a moment, I'll come down to get you."

Jiang Lin put his phone back into his pocket, took two steps back, and looked up at the building before him.

Building B of the School of Physics looked like a Soviet-style building built in the 1980s.

Its gray pebble-dash exterior walls, narrow high windows, and heavy concrete load-bearing pillars exuded the heavy atmosphere unique to classical Physics.

Jiang Lin waited less than five minutes before Teacher Lu, who had talked eloquently in the popular science video Quantum Mechanics Isn't That Mystical, hurried out of the dim stairwell.

In person, he looked leaner than in the video.

He was wearing a gray turtleneck sweater with a somewhat loose collar, carelessly layered with a long black down jacket over it. His hair was messy, and there were obvious traces of staying up late under his eyes, clearly showing he hadn't bothered to groom himself in the past few days.

When Jiang Lin went up to greet him proactively, Professor Lu Zhixing's footsteps paused abruptly.

A look of astonishment lingered on Jiang Lin's young face for a full three seconds, his brain briefly recalibrating a certain expectation in an instant.

Because the person who had come was much younger than he had imagined.

In his subconscious, the person who could write an email that hit the pain points of fundamental mechanics directly, with detailed and restrained data, should at least be a senior PhD student who had spent five or six years in the Institute of Engineering Mechanics, or perhaps a young senior engineer who had spent years rolling about in military-industrial enterprises.

Yet the person before him wore a dark gray jacket, a black hooded sweatshirt, old sneakers, and carried a backpack full of student vibes.

At most, he looked like an undergraduate student.

However, after a brief surprise, Professor Lu Zhixing quickly put away his exposed emotions.

In a School of Physics at a university like Jiangcheng, there was never a shortage of overly young monsters.

Geniuses pulled in by the Qiangji Plan, scientific research madmen who entered research groups in their sophomore year of undergraduate studies to derive formulas in advance, problem-solving machines soaked in Olympic competitions since childhood, and those monsters in the adjacent juvenile class who started calculating Feynman diagrams at sixteen or seventeen.

All kinds of unreasonable growth paths existed here.

In the classes taught by Professor Lu Zhixing himself, there were a few sophomore students whose questioning angles were so tricky that they could make him break out in cold sweat right on the podium.

A few high-level undergraduate deities could even put second-year PhD students on the spot during group meetings.

Professor Lu Zhixing looked Jiang Lin up and down once more.

There was an engineering student vibe about him that hadn't been completely eroded by laboratory dogma.

Very steady.

"You're Jiang Lin, right?" Professor Lu Zhixing reached out and shook hands with Jiang Lin.

His palms were ice-cold, but very strong.

"It's me, Teacher Lu."

"Let's go, it's windy outside. Let's go upstairs to talk." Professor Lu Zhixing didn't have any superfluous pleasantries and directly turned to lead the way.

The two of them went one after the other, walking up the stairs.

They stopped in front of a wooden door on the third-floor corridor with a sign reading "Room 302 Conference Room".

Professor Lu Zhixing pushed open the door.

This was a typical university research group activity space.

The room was quite large, with several long tables pieced together in the center, and a dozen folding chairs scattered haphazardly around.

On a huge whiteboard against the wall, it was densely written with the derivation steps of Hamiltonians and the Schrödinger equation, as well as some tensor abbreviations that outsiders couldn't understand at all. The handwriting was in red and black, and some places were haphazardly wiped by an eraser, leaving patches of turbid shadows.

In a corner of the room were stacked several Amazon delivery cartons, filled with various models of operational amplifiers, capacitors, and jumper wires.

An old Epson projector was hanging from the ceiling, projecting a complex spectrum analysis chart onto the screen directly ahead.

Three people were already sitting in the room, but the air pressure was so low that it made one breathless.

Sitting by the window at the long table was a boy with a bird's nest-like hair. He was staring dead at the screen of the Alienware laptop in front of him, his fingers tapping frantically on the keyboard.

On the screen was a Python development environment with colorful text on a black background, and a piece of extremely complex Kalman Filtering function code was running in the window.

Jiang Lin later learned that his name was Meng Che, a third-year PhD student in the group responsible for software algorithms and control theory, currently under high pressure on the verge of graduation.

Standing against the experimental bench on the other side was a slightly chubby boy wearing black-framed glasses.

In his hand, he was leafing through a long printed parameter table, wearing a pair of white anti-static gloves that were meant for dust prevention but were already pilling.

His name was Yin Hang, a second-year PhD student and the hardware and circuit person-in-charge of the group.

In the corner, there was a separate small desk.

A girl was sitting there, organizing massive amounts of data waveforms facing dual-screen monitors.

She wore a baggy plaid shirt, her hair was casually tied behind her head with a shark clip, noise-canceling headphones were on her ears, and her brows were tightly locked.

Her name was Yao Siyu, a third-year master's student.

Hearing the door open, the two PhDs and one master's student raised their heads simultaneously and glanced at Jiang Lin.

Their eyes carried a trace of annoyance at being interrupted and obvious puzzlement: why would Professor Lu bring a kid who looked like an undergraduate in at this critical juncture?

Professor Lu Zhixing didn't introduce Jiang Lin's true origin, just saying indifferently: "This classmate is sitting in today; report as usual."

Hearing this, the three of them withdrew their gazes one after another.

At the group meeting, the person responsible for reporting was Meng Che.

He sighed, walked to the projection screen, picked up a laser page-turner, and pressed it.

The spectrum chart on the screen disappeared, replaced by an extremely complex Allan variance chart, accompanied side-by-side by a power spectral density chart.

"Teacher Lu, fellow students, this is our latest data analysis on the low-frequency drift of that interferometer over the past 72 hours."

He drew a circle on the screen with the dot of the laser pointer.

"Everyone can clearly see that the red curve represents the raw interference signal collected by our detector. In the ultra-low frequency range of 10⁻³ to 10⁻⁴ in the frequency domain, there is a distinctly raised 1/f noise envelope. This bulge corresponds to that forty-minute cycle of slow drift that the reviewer is biting down on and refusing to let go."

Meng Che pressed the page-turn button, switching to the next slide, which was densely packed with state-space matrix equations.

"I've repeatedly checked the control loop these past few days, and I strongly suspect that our original PID control loop has insufficient integral gain in the low-frequency band, causing the integrator to experience a winding effect when facing this slow, cumulative thermal disturbance, failing to generate strong enough low-frequency feedback to effectively suppress this drift."

"Therefore, I tried a bold scheme. In the feedback loop, I abandoned the traditional PID and forcibly added an adaptive Kalman Filtering with a forgetting factor. Everyone look at the state update equation and covariance matrix prediction here..."

Meng Che's mathematical foundation was indeed quite solid. He spoke and wrote in front of the whiteboard, and the derivation process was seamless and impeccable.

As he put this new filter algorithm into the computer software next to him, a green processed curve was immediately generated on the right side of the screen.

That green curve was extremely gorgeous. The annoying low-frequency bulge was forcibly ironed flat by an invisible iron, and the residual data instantly returned to within the horizontal line that PRL reviewers could barely accept.

"Everyone look here." Meng Che's tone carried a hint of urgency of grasping at a straw, and even some excitement, "As long as we modify the underlying FPGA code of the lower computer this afternoon, burn this new filtering algorithm into it, and replace the old PID, our data can be used, and this paper can be saved."

"No."

A voice directly interrupted Meng Che's fanaticism.

The speaker was Yin Hang sitting by the window. He had already taken off his glasses at some point, vigorously rubbing his bridge of the nose with two fingers.

"You are simply deceiving yourself, Meng Che."

Yin Hang unceremoniously pointed out the problem, his tone full of irritability after pulling an all-nighter.

He stood up abruptly, strode to the computer, snatched the mouse, and brought up another monitoring chart hidden in the background.

"Everyone look at this."

Yin Hang pointed to a red polyline on the screen that was almost about to pierce through the ceiling.

"You used Kalman Filtering at the software level to forcibly suppress that low-frequency data. It looks very beautiful. But have you ever thought that that drift has not disappeared at the real physical level."

He turned his head and stared at Meng Che.

"Look at the compensation voltage you output to the PZT. In order to maintain the locked state of the interference fringes, your algorithm commands the piezoelectric ceramic to stretch frantically. The compensation voltage currently applied to the PZT has approached 80% of its full scale."

"Do you know what this means? This means our core optical path has actually undergone huge physical displacements of tens or even hundreds of nanometers. You are just using the deformation of the piezoelectric ceramic, like a paper-hanger, to forcibly pull it back to the origin. If the magnitude of this drift increases by even another twenty percent, the PZT voltage will completely saturate. By then, the entire interference system will lose lock directly within a second, all fringes will completely disappear, and the data we measure inside will be a pile of garbage."

Yin Hang's voice echoed in the conference room, carrying a kind of despair unique to engineering guys.

Meng Che's face turned alternatingly green and pale, and he retorted: "Then what do you say we do, huh? You're so awesome at hardware, why don't you find the cause? We tested the dark noise of the photodetector three times, and the thermal drift is at the picoampere level, no problem. The frequency stabilization loop of the external cavity semiconductor laser is locked tightly onto the rubidium atom absorption line, no problem. That damn constant temperature room of ours has an air conditioning control precision of plus or minus 0.1 degrees, no problem either."

Meng Che slapped the table with a palm: "Besides relying on algorithms for software compensation, can you, Dr. Yin, find any other noise sources in this million-level system?"

"It must be that operational amplifier chip at the front end of the photoelectric amplifier that has a hard-to-monitor thermal drift effect!"

Yin Hang insisted on his hardware perspective, refusing to yield an inch.

"With current semiconductor technology, even if the ambient temperature only changes by 0.1 degrees, the junction temperature change inside the chip caused by the current thermal effect may cause the input bias current to experience non-linear drift. I suggested the day before yesterday to scrap the entire front-end amplification circuit board, redraw the board, and replace it with a military-grade operational amplifier chip with an internal constant temperature bath."

"Are you crazy?" Yao Siyu, who hadn't spoken in the corner all along, suddenly took off her headphones and couldn't help shouting out, "Redrawing the board, finding a manufacturer for prototyping, mailing it over, then manual surface-mount patching, debugging and troubleshooting—going through this entire process will take at least two weeks even at the fastest."

Yao Siyu's eyes were even a bit red as she pointed at the calendar on the desk: "The revision deadline given to us by the PRL editorial department is only twenty days away. In the middle of this, we still need to set aside time to re-run dozens of hours of stability data and revise the main text. Boss Professor Wang Heng just sent a WeChat message yesterday asking about the progress, and now you say you want to overthrow the hardware and start over? Do you want to drag everyone in the entire laboratory down with you?"

The meeting room instantly plunged into a fierce argument.

Meng Che from the software faction believed they could only rely on algorithms as a safety net, Yin Hang from the hardware faction stubbornly believed that there were flaws in the circuit board chips, while Yao Siyu, who was responsible for the data, was on the verge of a breakdown at the edge of the deadline cliff.

The air was filled with anxiety and frustration, as well as that panic of impending rejection.

Professor Lu Zhixing had been sitting in the main seat, his hands crossed against his chin, looking at the table with a deep gaze without saying a word.

He knew that Meng Che's algorithm was actually a form of self-deception, using software to cover up a physical ulcer.

He also knew that Yin Hang's suggestion to replace the chip not only could not quench an immediate thirst with distant water, but based on his many years of experience, the noise spectrum characteristics of the chip's thermal drift did not look like this at all. Replacing it with a top-tier chip might not truly solve this slow drift with a cycle of up to forty minutes.

But what could they do?

They were all trying to use the methods they were most skilled at and most reliant on to patch an unknown loophole that was completely in a cognitive blind spot.

This was like someone who only had a hammer in their hand, looking at every troubled place in the world as a nail that needed to be hammered.

The argument in the meeting room gradually exhausted its energy and entered an insoluble dead end.

Meng Che dejectedly pulled out a chair and sat down, holding his head in agony with both hands and messing up his hair.

Yin Hang also irritably closed the lid of his laptop, turned around to look at the gray sky outside the window, and panted heavily in big gasps.

No one spoke again.

However, Professor Lu Zhixing had no intention of ending the meeting at all; he turned his head to look at Jiang Lin.

From entering the door until now, this young man had been listening quietly.

He had not taken out a notebook to take notes, had not shown blank confusion upon hearing those complex technical terms, and had not appeared cramped because of this fierce argument.

"Classmate Jiang Lin, is there anything you would like to say?" Professor Lu Zhixing spoke, breaking the stiff atmosphere of the meeting room.

The gazes of the two doctoral students and one master's student turned to Jiang Lin in unison; they did not understand why the boss would ask someone just sitting in as an auditor at this time.

Jiang Lin thought for a moment, looked calmly at Yin Hang, and said: "Could you please pull up the most original sampling file of that drift data from just now and let me take a look?"

Yin Hang frowned and waved his hand somewhat impatiently: "When we do frequency domain analysis, of course we have already looked at the raw data. The low-frequency envelopes are all there; looking directly at the time-domain graph cannot reveal any physical essence."

"I am talking about that most primitive layer, without any Kalman Filtering, without feedforward compensation, without normalization processing, and without eliminating the outliers you think exist. The hexadecimal sampling sequence spat directly out from the registers of the ADC chip," Jiang Lin explained.

Yin Hang turned to look at Professor Lu Zhixing, and seeing that his advisor had no intention of objecting, he muttered somewhat stiffly: "Scientific research data processing has a standard noise reduction workflow; what is the point of looking at raw data full of glitches?"

"I know," Jiang Lin looked at him, "but I just want to see it before processing."

Yin Hang had no choice but to pull out a massive raw .csv file from the underlying database and spread it across the main screen using drawing software.

The data chart came out.

This was a line chart covered in messy glitches, looking like a Wasteland overgrown with weeds.

Jiang Lin walked closer, stood in front of the screen, and like a sculpture, stared quietly at the chart for about three minutes.

The main component of the low-frequency drift was indeed, as Meng Che said, lurking beneath massive amounts of high-frequency noise.

That slow undulation with a cycle of about forty minutes had its amplitude forcibly suppressed to a few tenths of a percent of the measurement range after filtering, but in the unpolished raw picture before his eyes, its massive outline was still clearly distinguishable.

However, he noticed one thing.

The waveform of this drift was by no means a symmetrical sine wave or cosine wave.

Its rising edge was significantly steeper than its falling edge.

More crucially, after filtering out those random electronic white noises in his mind, he keenly captured that within the steep rising segments of several specific time periods, there was a small step with an extremely short duration.

That step was very abrupt, with a width on the time axis of only about three to five minutes.

It abruptly interrupted the originally smoothly rising slope. During this period, the drift amount remained constant, and five minutes later, the slope continued to climb upwards at its original slope.

This feature was too ugly, so in Yao Siyu's standard data cleaning workflow, it was treated as an occasional outlier of some mechanical vibration or was directly swallowed up by the smoothing algorithm.

But in Jiang Lin's eyes, this was by no means any kind of outlier.

In his mind, he instantly overlaid this abrupt step onto the mechanical blueprint of the aluminum-steel composite adapter plate containing locating pinholes described in Professor Lu Zhixing's email early yesterday morning.

Contact stiffness mutation... The difference in linear expansion coefficients between the two metals reached a critical point... The microscopic gap between the locating pin and the hole wall was completely eaten up and underwent hard extrusion... The conduction constant of thermal equilibrium time... The relaxation width of the system transition state...

Everything matched up.

Seeing that he had seen almost enough, Professor Lu Zhixing opened his mouth and asked: "Can you show me that set of data you mentioned in the email?"

Jiang Lin turned around, took out that transparent plastic folder from his schoolbag, pulled out the stack of thirty pages of well-organized manuscripts and data descriptions, walked to the long table, and handed them to Professor Lu Zhixing.

Professor Lu Zhixing took the materials.

He first glanced at the data description sheet on top.

When he saw the ridiculously large system error estimates listed above and the appallingly crude descriptions of the measurement environment, the corners of his mouth twitched slightly and uncontrollably.

He didn't speak, directly turning the explanation paper to the back, and began to look at the data line chart recorded by Jiang Lin in the Wasteland.

Professor Lu Zhixing's way of looking at data was somewhat different from what Jiang Lin expected.

Instead of looking at the labels of the horizontal and vertical axes first, he first looked at the distribution of the data points, watched for about thirty seconds, then tapped a certain region with his fingertip and asked: "The error bars you marked for the three data points in this temperature drop interval are nearly three times larger than those of the data points in the normal area nearby. What caused this? Is it an accidental reading error, or a systematic bias?"

The two doctoral students and one master's student were all stunned; the three of them did not understand why their advisor would take a data chart that looked like it was hand-drawn so seriously.

"Those few measurements were made at two o'clock in the morning when the ambient temperature dropped sharply. The industrial thermocouple probe I used for temperature measurement at that time was aged. In this specific low-temperature range, the thermoelectric potential output had a clear non-linearity. Due to the lack of a standard source, my later mathematical corrections were not accurate enough. For the sake of rigor, I could only forcibly amplify the uncertainty of that interval," Jiang Lin answered without hesitation.

Professor Lu Zhixing nodded, a flash of approval crossing his eyes.

In the field of experimental Physics, honestly exposing the limits of instruments was ten thousand times more precious than forcibly drawing a perfect false curve using algorithms.

He continued turning down, flipping to Jiang Lin's handwritten model derivation part.

That was the multi-material contact stiffness evolution model derived by Jiang Lin with a pencil, lighting a camping lamp on countless nights when sandstorms howled.

Professor Lu Zhixing read the boundary condition assumptions on the first page.

He stopped, closed his eyes to think for a moment, and then reread the first page line by line from the beginning.

After confirming that there were no errors, he continued turning backward.

He turned the pages very slowly, as if savoring an ancient document.

Occasionally he would stop, take a mechanical pencil from his chest pocket, and make a tiny mark in the margin of the manuscript.

That mark was so small that even Yin Hang, who was sitting next to him, couldn't see what he had written.

When reading the section on the dynamic evolution of plastic relaxation of preload under thermal cycling in Chapter 3, Professor Lu Zhixing stopped and raised his head.

"Your relaxation function uses an exponential decay model here. In the exponential model, the time constant τ is the core that determines the speed of the entire process. But in complex mechanical assemblies, this constant is extremely difficult to determine. How did you determine it?"

"Estimated using the thermal diffusion time. The characteristic time is roughly taken on the order of L ² / α, with L taking the equivalent thickness in the main thermal gradient direction in the assembly, and α taking the equivalent thermal diffusion rate of the aluminum-steel contact pair."

"How is the equivalent thermal diffusion rate taken, series thermal resistance, or weighted by contact area?" Professor Lu Zhixing pursued, his tone carrying an air of testing.

"I made a weighted average based on the respective thermal diffusion coefficients of aluminum and steel on both sides of the contact surface. The weighting coefficient is the area proportion of their actual microscopic contact under the constraint of roughness."

Jiang Lin's tone was calm, but everyone present could hear the massive amount of calculation contained within it.

Professor Lu Zhixing looked at that page of the manuscript again, his eyes growing brighter and brighter.

He did not continue to pursue this question, but accelerated his speed, finishing the entire stack of more than thirty pages of materials in one breath.

Finally, he closed it, gently patted it on the experimental table, and sighed deeply: "The precision of your data is really poor. So poor that if this were a paper, any reviewer would reject it in the first second. I have read your data description, and I know under what extremely restricted conditions you produced it."

The topic shifted.

"However, there is one thing that interests me very much. In your data description, you wrote four harsh limiting conditions for yourself, which shows that you are very clear about where the fatal flaws of your data set lie, and furthermore, you used an academic term—proxy indicator."

"What you measured in this model is not contact stiffness itself at all, because that kind of microscopic stiffness cannot be measured without top-tier equipment. What you measured is a certain proxy quantity of the change in contact stiffness."

"Yes," Jiang Lin admitted frankly, "under that environment, I had no way to directly measure the normal contact stiffness K ₙ. So I settled for second best, and what I measured was the distribution of indentation area left by micro-asperities after being squeezed on the contact surface. I took this indentation as a proxy variable and used plastic mechanics to deduce the relative change rate of contact stiffness."

"Have you done a rigorous Monte Carlo analysis on the error propagation chain of this proxy quantity?"

"Yes, in the last few pages of the model derivation appendix, I listed the transfer error limits introduced due to the imperfect Gaussian distribution assumption of surface roughness."

Professor Lu Zhixing didn't need to flip through it anymore; he had just seen that part of the error derivation that could be called obsessive-compulsive.

"In your derived model, that low-frequency angular disturbance predicted at the end. If your parameter values are accurate, then this angular disturbance caused by mechanical deformation must lag behind the actual change of the ambient temperature in phase. And the magnitude of this lag depends entirely on the thermal conduction time constant of the entire metal assembly."

"Correct," Jiang Lin nodded.

"Jiang Lin," Professor Lu Zhixing called his name, "this feature, this phase lag feature, we have seen it repeatedly in that drift data over the past three weeks."

As Professor Lu Zhixing spoke, he pulled up all the raw residual comparison charts of the past three weeks from the data acquisition computer and displayed them side by side on the big screen.

The low-frequency drift was clearly visible among these raw data like ghost-like giant snakes.

The cycles were all around forty minutes.

Moreover, just as Jiang Lin had pointed out earlier, their rising edges were, without exception, steeper than their falling edges.

That qualitative physical feature arising from the asymmetry of thermal expansion and contraction and mechanical friction hysteresis achieved an astonishing consistency deep in the soul with that ugly chart measured by Jiang Lin in the Wasteland using a thermocouple and vernier caliper.

Professor Lu Zhixing looked at Jiang Lin, who was silent in front of the screen.

After a brief contemplation, he proposed a suggestion that had almost never existed in this extremely conservative key laboratory.

"Just looking at data is useless. Why don't we go down to the laboratory downstairs to take a look on-site and apply your model in this real world?"

On the second basement floor of the Jiangcheng University Physics building, B104 was the Ultra-Clean Precision Measurement Laboratory.

Security here was strict.

Professor Lu Zhixing swiped his access card three times and passed through an air shower room for dust removal before that heavy soundproof fire door was heavily pushed open, accompanied by a hiss of releasing air pressure.

Before entering, Jiang Lin changed into a white cleanroom suit, and Professor Lu Zhixing took several pairs of laser safety glasses from the cabinet by the door, handing them over according to wavelength while snapping the main optical path's dust cover back into place.

The moment he stepped into the laboratory following Professor Lu Zhixing, Jiang Lin's skin and pores instantly and acutely sensed anomalies in the air flow.

There was no wind.

Or rather, the wind speed was extremely precisely controlled at an exceptionally low value.

The ceiling was densely covered with micro-perforated diffusers, which forced down the filtered cold air in a uniform downward flow.

The numbers displayed by the temperature and humidity sensors on the wall were: temperature 22.00 ± 0.05 ℃, relative humidity 45%.

This was a physical fortress built to resist the random thermodynamic fluctuations of the real universe.

In the center of the laboratory coiled the core of this fortress: an optical table three meters long and two meters wide.

The stainless steel honeycomb tabletop, thirty centimeters thick, was held high by four robust air-spring damping legs.

These four damping legs with active servo feedback functions completely isolated this massive metal plate from the fundamental vibrations of the entire building, and even from the tiny low-frequency seismic waves caused by the subway passing two kilometers away.

On the tabletop, complex interference optical paths composed of various black anodized aluminum brackets, high-reflectivity dielectric membrane reflectors, polarization beam splitters, and collimating lenses were densely spread out.

Laser beams shuttled and refracted between these lenses, resembling a miniature modern jungle of steel and photons.

Yin Hang and Meng Che, who followed them inside, looked at Jiang Lin standing before the table, their faces carrying a hint of resentment and tension.

They were deathly afraid that this kid would break something.

But Jiang Lin didn't even look at those extremely expensive core optical components.

He knew that lenses manufactured by top-tier manufacturers like Zeiss or Nikon using ion-beam polishing typically achieved a surface accuracy of λ/10 or even higher; they were extremely stable in physical properties and rarely made mistakes on their own.

What truly easily bred errors and hid disasters were forever those mechanical joints where humans performed secondary assembly.

Those bolts, flanges, and adapter plates.

Jiang Lin's gaze was like a high-resolution industrial X-ray machine, carrying an extremely strong sense of purpose as it swept one by one across the metal bases supporting the optical components and among the rows of matrix-arranged threaded holes.

He was searching for breakpoints of force, the vulnerable links of stress accumulation.

On the Wasteland, in order to build the wind turbine that could save the shelter, he had fought against gravity, material fatigue, and the tearing force of sandstorms for many years using the most primitive files, hand scrapers, and rusted steel bars picked up from the ruins.

He knew how force was conducted within metal lattices, and he also knew how stress accumulated within unreasonable structures and ruthlessly tore open cracks in the most vulnerable places.

In less than a minute.

Jiang Lin's line of sight lingered on an inconspicuous corner at the edge of the optical table.

There was an optical adjustment mount used to fix a half-wave plate in order to adjust the polarization state of the beam.

The problem was not the adjustment mount itself, but beneath it.

Unlike other precision components, it was not directly fixed onto the optical table's standard threaded hole positions spaced 25 millimeters apart.

Due to an error in the optical path height design, an additional rectangular adapter plate was padded underneath it.

This adapter plate was grayish-white in color, with clear machining tool marks on the surface, and had undergone no black anodizing treatment whatsoever.

This was a spacer temporarily milled from ordinary 6061 aluminum alloy on a machine tool in the workshop outside the laboratory just to save trouble.

This aluminum adapter plate was tightly pressed onto the stainless steel table below via counterbored holes at its four corners using four M6 hexagon socket head cap screws.

Furthermore, Jiang Lin keenly noticed that on the side of this adjustment mount, due to messy wiring, a black coaxial radio frequency cable the thickness of a thumb was bound to it.

The other end of this cable hung heavily outside the edge of the optical table, tracing an arc in the air before extending all the way down to the humming instrument control cabinet on the ground.

After a brief thought, Jiang Lin asked Yao Siyu to overlay the PZT compensation voltage and the platform temperature differential curve when the drift occurred.

Only then did he stop, pointing his hand at that forgotten corner, and said, "That drift curve of yours is most likely not an electrical noise problem from some front-end operational amplifier chip, nor can it be solved by software filtering algorithms; it might purely be a mechanical and physical problem."

"Why?" Professor Lu Zhixing looked at him.

"I used to assemble similar things before." Jiang Lin's gaze seemed to pierce through the equipment before him, returning to the Wasteland World, "Dissimilar metal combination—aluminum alloy on top, carbon steel or stainless steel below—with four bolts unevenly pre-tightened during manual assembly. On this plate, the bolt in the right-rear corner has the lightest locking force. Under long-term thermal cycling, microscopic stress redistribution occurred inside the contact surface. The machine I built back then started showing fatal periodic jitter in its main shaft after experiencing similar working conditions for three months."

Jiang Lin took a step closer, his index finger pointing empty air towards that grayish-white adapter plate.

"The root cause might be right here."

Meng Che had been checking the parameter table nearby just a moment ago. Hearing this sentence, he abruptly turned his head, glanced at the place Jiang Lin pointed to, and then looked at Yin Hang, who was equally shocked.

"Do you even understand the structure of an interferometer? That's just a temporary half-wave plate adapter plate. The cable next to it is the high-voltage modulation control line connected to the piezoelectric ceramics; what possible problem could this have? This part of the optical path is not on the main interference arm for measuring length at all; it's merely a preparation module for the early-stage beam polarization state. How could its slight sway produce such a large low-frequency displacement residual?" Yin Hang asked back somewhat annoyed.

"No matter which position of the optical path it is in, it and your extremely sensitive main optical path are both on this single, whole physical platform."

Jiang Lin was not overwhelmed by Yin Hang's momentum; his voice remained calm.

"At the nanometer measurement scale, even the reaction torque caused by a fly taking off at the edge of the platform would propagate through the metal continuum to the interference arm. Let alone this is a metal plate in an extremely terrible state of internal stress distortion."

Professor Lu Zhixing didn't speak immediately. His gaze shuttled back and forth between the adapter plate and Jiang Lin's face for a while before saying, "How do you plan to verify this deduction?"

"In the past when I didn't have precision instruments, I used an extremely ancient method." Jiang Lin turned his head to look at the console, "Do not move any optical paths in your core interference arms. Just use this board as a reflecting surface, along with that white wall opposite. Do you have any coated, smooth silicon wafers in your laboratory, even discarded ones are fine."

Looking at the young man in old clothes before him, Yin Hang's eyes intertwined with an extremely intense sense of absurdity and offended annoyance.

An unknown young man who hadn't read their tens of thousands of lines of FPGA control code, hadn't touched the precision instruments they spent half a year tuning, and hadn't even systematically glanced at the optical path diagram of the entire interferometer set.

dared to boldly stand in this constant-temperature room worth millions and claim that this group of elite university Ph.D.s had calculated the most basic physical boundary conditions wrong.

"Not touching the core optical paths, using a piece of waste reflector and a laser pointer used for lecturing?" Yin Hang finally couldn't help but sneer, saying in a sharp tone, "Do you think this is doing light reflection experiments from junior high school Physics textbooks? What we are measuring now is nanometer-level displacement with eight zeros after the decimal point. What can your caveman-like toy scheme possibly see on this wall, molecular thermal motion?"

Meng Che also furrowed his brows and looked at Teacher Lu: "Teacher Lu, this is too absurd. Letting him shine a laser pointer randomly on the optical table—what if stray light hits our avalanche photodiode and burns out the detector?"

Professor Lu Zhixing didn't immediately call a halt, seeming to want to hear what Jiang Lin had to say further.

Jiang Lin quietly stood in place, without refuting or showing any impatience, grievance, or anger after being looked down upon by elite university Ph.D.s, simply calmly explaining his own scheme.

Undeniably, this scheme was almost flawless in terms of experimental protocol.

Non-invasive, extremely low-cost, with zero risk of damaging core equipment.

Professor Lu Zhixing quickly evaluated the risk-reward ratio of this scheme in his mind.

One minute later, he made a decision.

"Yin Hang, go to the sundries box by the door and find a cut-and-discarded silicon wafer. Meng Che, go up and get a collimated laser pointer."

Yin Hang froze in place, his mouth slightly open: "Professor Lu, are you really going to accompany him playing house here?"

"You've been stuck for three days." Professor Lu Zhixing interrupted him, his voice as stern as if he were scolding undergraduates, "Since the most cutting-edge software algorithms and top-tier circuit analyses can't find the cause, then return to the starting point and let the most basic mechanical structure speak for itself."

Moments later, Meng Che ran back holding a laser pointer.

Yin Hang was also somewhat reluctant, handing over a coated silicon wafer the size of a fingernail that had been discarded due to a cutting error at the edge.

Jiang Lin took these two things and asked for a small piece of ubiquitous 3M double-sided tape.

Everyone quietly watched his movements.

Without blinking, Jiang Lin carefully glued the waste silicon wafer onto the side of that controversial aluminum alloy adapter plate, on a flat edge perpendicular to the ground.

Throughout the entire process, his hands were so steady that there wasn't a trace of tremor, completely avoiding the adjustment mount and main optical path above the adapter plate, and not touching anything that shouldn't be touched.

Then, he looked around and brought over a mobile instrument cart loaded with an oscilloscope, completely independent of the optical table.

Pushing the cart to a place about one meter away from the platform, he locked the universal wheels.

Then, using thick cloth-backed tape, he tied the black laser pointer to the metal pillar of the cart, ensuring it wouldn't experience any relative shaking.

With a soft "click,"

Jiang Lin pressed the switch of the laser pointer.

An extremely thin and bright green laser beam instantly shot out. Under the scattering of tiny dust particles in the air, the beam struck like a sharp sword precisely in the dead center of the waste silicon wafer glued to the side of the adapter plate.

Because the coating on the surface of the silicon wafer had extremely high reflectivity, the beam was reflected immediately.

This green reflected light crossed the long distance of about ten meters across the entire constant-temperature laboratory, finally projecting onto the wall opposite painted with matte white paint.

On the white wall, a bright green circular light spot appeared.

Jiang Lin pulled out a black fine-tipped water-based pen from his hoodie pocket, walked over to that wall, and steadily drew an extremely small and fine cross mark right at the exact center of that green light spot on the wall.

"Optical lever." Professor Lu Zhixing stood beside the platform, watching Jiang Lin finish setting all this up, and softly spat out these three words.

As an associate professor of the Department of Physics, he certainly knew this principle.

Not only did he know it, but he had also countless times derived its geometric equations for students on the podium.

This was the oldest, most classic, and even most elegant pure mechanical amplification technology for tiny displacements in the history of human Physics.

Back then, in order to measure the extremely weak universal gravitational constant, and Coulomb to measure the torsional force between microscopic charges, they all used this set of methods that did not rely on any electronic chips, relying solely on pure geometric reflection principles to amplify microscopic deformations.

The optical path of geometry was extremely merciless.

If the adapter plate experienced an extremely minute angular deflection Δθ, then according to the law of reflection, the deflection angle of the reflected beam would be amplified twice, becoming 2Δθ.

After this deflected beam was amplified by the long arm of distance L (here, spanning the entire laboratory, a full ten meters away), the actual displacement Δy of the light spot projected on the wall strictly equaled L × tan(2Δθ).

At extremely small angles, it was approximated as 2L · Δθ.

[part:gemini-3.6-flash]

A ten-meter optical lever arm was sufficient to transform tiny angular variations into visible movements of the spot of light on the wall.

"What does this prove anyway?" Yin Hang stared at the spot of light and the cross mark on the wall, still filled with confusion and irritation. "Beneath us is a thirty-centimeter-thick stainless steel plate, and everything above is rigidly connected with high-strength bolts. Do you really think this half-inch-thick aviation aluminum plate is bending like a noodle on its own in a constant temperature room at twenty-two degrees Celsius?"

Jiang Lin paid no attention to Yin Hang's mockery.

He walked straight back to the optical table, fixed his eyes on Meng Che, and said in a non-negotiable tone, "Please hold up this dangling black cable beside the adjustment mount with your hand. Mind you, don't push upward with force, nor sway it side to side. Just gently support it enough to relieve the downward pull of its own weight on this adapter plate. Hold it for ten seconds."

Meng Che froze for a second, then glanced back at Professor Lu Zhixing.

Seeing no sign of opposition from his advisor, he took a deep breath, extended his right hand clad in an anti-static glove, and carefully supported the suspended coaxial cable.

The ten-second countdown began.

The laboratory was so quiet that only the extremely faint airflow from the air conditioner vents could be heard.

"Alright," Jiang Lin said, watching the second hand on his watch. "Now, release your hand extremely slowly. Let this cable dangle naturally again and return to its original state."

Meng Che swallowed hard, opened his fingers bit by bit, and slowly pulled back his hand.

Losing support at the bottom once more, the heavy coaxial cable's gravity turned back into a continuous pulling force, hanging off the edge of the adapter plate.

"Look at the wall," Jiang Lin's voice was completely devoid of emotion.

Yin Hang, Meng Che, Yao Siyu, and even Professor Lu Zhixing—everyone whipped their heads around, staring at the white wall ten meters away.

"Hss—"

Yin Hang gasped, his eyes instantly widening.

That green spot of light had drifted away from the perfect center of the cross Jiang Lin had drawn earlier with a black pen.

On a macroscopic scale, the deviation was not large—roughly zero point two to zero point three millimeters.

However, for these people who dealt with extreme precision year in and year out, it was glaring enough.

A ten-meter optical path with a two-fold optical lever amplification meant that the adapter plate had undergone an angular change on the order of 10⁻⁵ radians.

For everyday mechanical assembly, this was practically nonexistent, but for nanometer-scale interferometry, it was a disaster.

Δy = 0.2 mm

L = 10 m

Δθ ≈ 1×10⁻⁵ rad

This meant that the aluminum alloy adapter plate—which they had believed to be rock-solid, clamped down onto a thirty-centimeter-thick stainless steel base by four high-strength socket head cap screws—had actually undergone a spatial angular displacement of approximately five hundred-thousandths of a radian during that brief moment when it lost and regained the tiny two-hundred-gram weight of the cable.

Five hundred-thousandths of a radian was less than one percent of the diameter of a human hair in the everyday macroscopic world.

Even if measured with the most precise micrometer caliper, no change could be detected.

Yet for an interferometer that required optical path stability on the order of a few nanometers and measured using the phase of light waves, this was a massive catastrophe capable of distorting the entire complex optical path and causing the interference fringes to jitter frantically.

And an even more chilling physical phenomenon was yet to come.

When Meng Che released his hand completely, the spot of light that had drifted about zero point two to zero point three millimeters did not snap back to the center of the cross immediately as an ideal elastic body would upon force restoration.

It paused for a moment.

Then, over the next dozen seconds or so, that green spot of light was like an extremely slow snail crawling on glass, creeping back toward the center of the cross inch by inch, with a suffocating sense of hysteresis.

In the end, it stopped about 0.2 millimeters away from the center of the cross and remained completely still.

No matter how long they waited, it moved no further back.

A hysteresis loop.

Residual deformation.

This was the phantom of microscopic friction.

"How is this possible?"

Meng Che muttered to himself, looking down at his right hand in utter dismay before suddenly snapping his head up to stare at the black cable hanging quietly.

"It's just an ordinary cable! The total suspended weight isn't even two hundred grams. How could it... how could it possibly pull a metal plate locked down by four high-strength M6 bolts? This goes against force analysis!"

"Because metal is never an absolutely perfect rigid body, and moreover, your bolts didn't lock it down at all."

Jiang Lin walked back to the optical table, pointed at the four corners of the offending aluminum alloy adapter plate, and said,

"When you first assembled this system, to save time and effort, did you just use a regular Allen key and tighten these four bolts sequentially in a single direction? What's more, you didn't use a torque wrench to control the tightening torque of each bolt."

Yin Hang, who had been responsible for the initial optical path hardware assembly, turned pale in an instant, feeling as if he had been stripped naked in public.

He opened his mouth, his Adam's apple bobbing, but could not utter a single word of refutation.

Jiang Lin hit the nail right on the head.

In the day-to-day setup of research experiments, in order to quickly verify optical principles, very few graduate students would act like senior technicians on an aerospace assembly line, strictly adhering to mechanical protocols by applying standard torque in stages and in a cross-diagonal pattern for every single bolt.

Everyone simply relied on feel, tightening them as long as they felt secure.

Jiang Lin's speaking speed remained steady, as if reading aloud a pre-written verdict of Physics.

"This was the start of the disaster. The preload was extremely uneven, with the two bolts on the left bearing immense force and the two on the right bearing very little. At a microscopic level, this seemingly flat aluminum plate was forcefully twisted into an extraordinarily bizarre stress state, filled with accumulated elastic stress inside."

"And this two-hundred-gram cable, though light, hangs off the edge, geometrically forming a long lever arm. At every moment, it exerts a tiny, continuous bias moment on this unevenly stressed aluminum plate. When you supported the cable, this moment suddenly vanished, allowing the accumulated elastic stress inside the aluminum plate to be instantly released, causing tiny relative slips between the microscopic asperities."

"When you released your hand, the gravitational moment was reloaded. However, because of the extremely complex static friction between the mating surfaces of the two metals, it couldn't immediately overcome that friction to return to its original equilibrium position. That is what caused the maddening mechanical creep and physical hysteresis you just saw on the wall!"

Jiang Lin reached out and pointed toward the central air conditioning vent on the ceiling of the constant temperature room.

"But even this isn't the root cause behind the forty-minute periodic drift in your interferometer. That cable was merely an amplifier, a trigger. The real phantom is right here."

He pointed back at the bottom surface of the adapter plate once more.

"The stainless steel platform underneath has a coefficient of linear thermal expansion of approximately 1.2×10⁻⁵ /K. The aluminum alloy adapter plate above has a coefficient of thermal expansion of 2.3×10⁻⁵ /K—nearly a two-fold difference."

Jiang Lin turned to look at Professor Lu Zhixing, revealing the ultimate answer to the puzzle that had baffled the entire top-tier laboratory for three full weeks.

"Teacher Lu, your constant temperature room is indeed very advanced, controlling room temperature within plus or minus 0.1 degrees Celsius. But this high-end central AC compressor has a fixed operating cycle. When the air conditioner cools and blows cold air, the aluminum plate contracts faster than the steel plate. When the cooling stops and the room temperature rises slightly, the aluminum plate expands faster than the steel plate."

"On the tightly mating contact surface between these two different metals, this creates a naturally sensitive bimetallic strip thermometer."

"The difference in their deformation rates generates a massive internal additional shear stress. With nowhere to go, this stress can only be conducted through the bolt contact surfaces. Because your bolt preload is extremely uneven, this additional stress inevitably releases first at the corner with the weakest preload—that is, the back-right corner, near the bolt that was tightened the least—thus creating a tiny, periodic angular deflection."

"To facilitate repeated assembly and disassembly, no dowel pins were installed underneath this temporary adapter plate. Without dowel pins, it lacks the absolute in-plane constraint expected in the design, effectively gaining an extra degree of freedom for minor rotation."

"Meanwhile, the dangling cable exerts a drag force. With every minuscule variation caused by the faint airflow from the air conditioner and subtle temperature fluctuations in the room, this force uses that redundant rotational degree of freedom to etch a tiny angular shift onto the benchmark mirror mount of your main optical path."

"Any one of these mechanical flaws, taken individually, is small enough to be negligible in the macroscopic world. But when they are all stacked onto a single optical path requiring nanometer-scale precision, in the eyes of the detector, they translate into that low-frequency drift curve you couldn't get rid of no matter what."

These words fell with absolute weight and force.

Silence reigned over the entire underground laboratory.

Yao Siyu snapped back to reality, turned around, and rapidly typed on the keyboard of the secondary data monitoring screen. She retrieved the raw residual data collected over the past two days and zoomed in to the limit on that strange little step feature on the rising edge that Jiang Lin had mentioned earlier.

The step appeared strikingly right in the center of the screen.

Clear and conspicuous.

Its width along the time axis was exactly around four minutes.

Its shape and slope matched Jiang Lin's description—made purely by eye back in the conference room—perfectly.

Meng Che stared blankly at the green cross on the wall and the spot of light that could no longer return to the center, his brain feeling as if it had completely crashed.

A intense sense of absurdity washed over him.

To flatten this residual, he had stayed up for three consecutive nights, thumbed through modern control theory books until they fell apart, and written hundreds of lines of exquisitely sophisticated adaptive Kalman Filtering algorithms.

And now, the algorithm he took such pride in was turned into an utterly laughable ornament in the face of a slightly warped aluminum plate due to thermal expansion and contraction, in the face of this extremely primitive mechanical creep.

His advanced software was calculating and compensating tirelessly with all its might, while the cold physical world was undergoing ruthless slippage at its lowest architectural layer.

Yin Hang stared at the ordinary M6 bolts on the optical table, his back soaked in cold sweat.

It was as if he were encountering Physics for the very first time, realizing for the first time that such basic mechanical fasteners—costing mere pocket change apiece at a hardware store—could harbor such terrifying destructive power capable of ruining millions in equipment.

Professor Lu Zhixing remained silent the entire time.

He stood less than half a meter away from Jiang Lin.

Looking at the young man's profile, a sense of absurdity so intense it sent a slight shiver through him welled up in his heart.

The logical deduction Jiang Lin had just demonstrated—that insight which allowed him to immediately discern preload distribution, thermodynamic deformation differences, and degree-of-freedom coupling processes from extremely chaotic phenomena—could never have been cultivated by sitting in a bright, spacious university classroom listening to an old professor read off PPT slides.

How immense a volume of training must one undergo to forge a pair of eyes as sharp and incisive as an electron microscope?

Professor Lu Zhixing took a deep breath, forcibly repressing the churning shock and inquisitive urge in his heart.

As the head of the research group, the most important thing right now was to solve the problem.