60: Chapter 60 Controlled Environments

At 8:20 on Saturday morning, when Jiang Lin arrived at Building B of the Physics building, Yao Siyu was waiting for him at the entrance.

She was wearing a beige chunky knit sweater today, her hair casually twisted up at the back of her head with a claw clip, and she was holding a transparent document bag in her hand.

"Classmate Jiang Lin."

"Classmate Yao, morning."

Yao Siyu raised her wrist to glance at her sports watch and said, "Teacher Lu is estimated to arrive at nine. He asked me to take you to B304 first to sign the registration form for temporary off-campus visitors and the safety commitment letter."

"Has it been approved already?"

"It's approved. Not as a student status, but as a temporary visitor of the research group." Yao Siyu seemed to be suppressing the smile at the corners of her mouth.

Jiang Lin nodded.

Holding the document bag against her chest, Yao Siyu walked as she spoke, "Teacher Lu made several trips for this registration. The school administration was very strict at first, mainly because you are a minor and an off-campus student. Later, he submitted your troubleshooting report, the reply letter from No. 7 High School, the parental consent form, the risk control checklist, and the residual comparison data before and after the B104 reinstall."

She paused.

"Meng Che said Teacher Lu wrote a sentence in the application reason."

"What?"

"This student has already shown a rare error chain recognition ability in real scientific research troubleshooting, possessing further observation and cultivation value. Instead of letting him explore on his own outside the rules, it is better to put him into a controlled environment for observation, training, and protection."

"Thank you for telling me these, Classmate Yao." Jiang Lin nodded.

Yao Siyu reached out to tuck a strand of loose hair hanging by her cheek behind her ear, smiling a little embarrassedly.

The two went up to the third floor and pushed open the door of B304.

The room was not very large, with four rows of workstations by the window, and filing cabinets and a HP black-and-white printer against the wall.

In the innermost position against the wall stood a dark gray workstation, with a piece of white paper pasted next to the monitor.

Written on the white paper was——

[Temporary Literature Terminal]

[Restricted to citation retrieval, open data reading, and intra-group controlled literature database access]

[Full-text requests for commercial databases require filling out an application form]

[Full texts are provided after review by official members of the research group and must not be disseminated outward]

[Batch downloading prohibited]

[External personal storage devices prohibited]

[Using personal accounts to log into on-campus databases prohibited]

There were already two people in the room.

Meng Che was sitting at the second workstation by the window, with a cup of suspiciously dark Americano in front of him. His hair was still messy, but he had slightly less of a near-death vibe than last time.

Yin Hang was disassembling a small circuit box, with an anti-static mat and a multimeter placed beside him.

Hearing the door open, both of them looked up.

The atmosphere between Meng Che and Yin Hang was unnatural for a split second.

When Jiang Lin came last time, they still regarded him as an auditing kid of unknown origin.

As a result, in less than half a day, that auditing kid used a piece of waste silicon wafer, a laser pointer, and a bit of red lead oil to pin down the source of the low-frequency drift they couldn't find for three weeks onto a burr of an aluminum plate.

For any PhD student, this kind of experience was not very easy to digest naturally.

"Classmate Jiang Lin, your report is extremely rigorous, especially on the error chain. The part where you broke down the four factors—cable parasitic stress, different metals thermal expansion difference, uneven preload, and local burr fulcrum—is almost a standard Physics troubleshooting teaching case. I have already compiled it into a memorandum. There is just one point: it reads a bit like an engineering accident review rather than a supplementary material for a paper."

Jiang Lin put down his schoolbag: "Where is it inappropriate?"

"You used too many words like catastrophic failure; you can't write like that in the reviewer response." Yao Siyu opened the file. "What we should write is that a low-frequency drift source with insufficiently constrained mechanical boundary conditions was discovered, and the residuals decreased significantly after reassembly. We cannot write it as us previously acting like fools letting a burred aluminum plate drag the whole system around for three weeks."

Hearing this, Meng Che couldn't help coughing once.

Yin Hang's face also darkened slightly.

Amidst this interaction, Jiang Lin finished signing the registration form, and the wall clock pointed to 8:57.

Right at this moment, Professor Lu Zhixing walked in with large strides.

Seeing the few people, he waved his hand. "Let's not talk about the paper first; let's go through the procedures today."

He took out a stack of materials from the folder in his hand.

The first copy was the laboratory safety training record sheet.

The second was the basic laser safety test questions.

The third was the data and literature usage specifications.

The fourth was the research group public area usage instructions.

Meng Che moved his coffee aside, stood up, and said in a somewhat lazy tone: "I am responsible for literature and data specifications."

Yin Hang raised the screwdriver in his hand: "I am responsible for the hardware area boundary, especially what things you cannot touch."

Yao Siyu pointed at her computer: "I am responsible for the troubleshooting report and subsequent residual data annotation."

This division of labor was simple and crude.

And also very clear.

The safety training had no drama whatsoever.

Professor Lu Zhixing personally went through laser levels, goggles wavelengths, reflected light hazards, items that cannot be brought into the clean area, equipment reservation systems, and raw data saving rules one by one.

Jiang Lin listened very attentively.

When Yin Hang talked about the boundaries of the hardware area, his attitude was still a bit stiff.

"You cannot enter B104 alone right now. Even if you can enter in the future, it can only be at specified times, wearing a cleanroom suit and corresponding wavelength goggles, accompanied by me or Meng Che. Do not casually touch the core optical path, detector, high-voltage PZT control line, and laser power supply."

Jiang Lin nodded.

"Also, for your optical lever scheme last time, although the result was correct, future similar operations need to write a risk checklist in advance. Stray light, silicon wafer fixation, reflection path, and detector protection must all be listed out and cannot rely on on-site judgment."

"Understood."

Seeing him answer so readily, Yin Hang instead didn't know how to respond.

Meng Che chuckled from the side: "He is more rule-abiding than you."

Yin Hang glared at him: "You haven't written your reflection on that Kalman Filtering yet."

The smile on Meng Che's face instantly disappeared.

Yao Siyu calmly dealt a critical blow: "Teacher Lu asked you to write it, and I've even thought of the title for you: 'How Beautiful Residual Curves Conceal True Physics Instability'."

Meng Che snapped the coffee cup lid shut: "Yao Siyu, you are talking more and more like Reviewer Number Two now."

The atmosphere in the room finally relaxed a bit.

When Jiang Lin sat in front of the temporary literature terminal, it was already 10:40 in the morning.

Meng Che pulled over a chair and sat next to him, slamming a piece of A4 paper onto the desk.

[Literature Application Form]

There were five columns on it.

Problem number, keywords, intended literature to check, usage description, whether full text is needed.

Meng Che tapped on the usage description column with his pen.

"The focus is here. You can look up literature, but you can't download whatever you search for like an undergraduate writing a course paper. Every time you want a full text, you must tell me which problem this piece of work serves."

Jiang Lin looked at the form.

"Can I look up more citations?"

"Citations are fine. Abstracts, DOIs, citation chains, authors, years—jot them down however you like, but full texts won't do. Full texts involve database authorization and our own management. If you want a PDF, write down the reason first."

"What about the intra-group shared library?"

"Teacher Lu opened read-only permissions for you." Meng Che clicked open an interface. "But it's only the parts we've organized. Don't expect to go in and see some peerless secret manual; most of them are papers, supplementary materials, failure records, and group meeting notes."

Jiang Lin looked at the tab bar on the screen.

[Low-frequency drift]

[Thermal noise]

[PZT nonlinearity]

[Mechanical hysteresis]

[Contact stiffness]

[Assembly and adjustment failure]

[Reviewer doubts]

[Irreproducible]

[Excluded]

[Suspected but unconfirmed]

His gaze lingered on [Irreproducible].

Meng Che noticed it.

"That tag is the best to look at," he said. "Successful papers are all written brightly and glamorously. If you really want to avoid detours, look at irreproducible ones."

Jiang Lin clicked it open.

There weren't many inside, only twenty-something entries.

But each entry was more valuable than an ordinary paper abstract.

[Low-frequency noise index of a certain literature cannot be reproduced: suspected different sample holder material]

[Supplementary material does not specify fixture preload method; repeated experiment drift is significant]

[Temperature control record missing; not recommended to cite long-term stability data]

[Author admits in reply that raw data had a line-switching operation, not written in the main text]

[This route has been tested twice in this group, both showing baseline drift after thermal cycling]

Jiang Lin slowly read down.

He suddenly understood that what Professor Lu Zhixing truly gave him was not a database, but a failure map in the scientific research path.

Failure is also a public asset.

Seeing him not moving for a long time, Meng Che asked: "What do you want to look up for your first batch?"

Jiang Lin took out his own notebook from his schoolbag.

Written on the first page: [Literature Pool Rules]

1. Downloading does not count as entering the literature pool.

2. Reading the abstract does not count as entering the literature pool.

3. Only literature that has participated in a specific problem, experimental scheme, error troubleshooting, or failure review is given an official number.

4. All literature must be annotated: what it solved, what it didn't solve, why it cannot be copied blindly, and whether I can reproduce it now.

5. Wasteland experience only serves as a source of problems, and real conclusions must be verified by real data.

Meng Che originally just glanced casually, but when he saw the fifth rule, his brow furrowed.

"Wasteland experience?"

"My own code name, meaning experience under extreme low-resource conditions."

Jiang Lin very naturally turned the page.

Meng Che did not continue asking.

Every researcher has their own quirks.

Some people name failure data graveyards, and some people call equipment that can never be fixed ancestors.

Jiang Lin called his low-resource experience Wasteland. Although strange, it was not unacceptable.

"Fine," Meng Che said. "Then what is your first question?"

Jiang Lin wrote down on the first line of the application form: [P-001: Contact hysteresis and angular drift of dissimilar metal adapter plates under minor temperature cycling]

Keyword column——

thermal cycling

contact stiffness

bolted joint

micro-slip

hysteresis

optical mount drift

bimetallic effect

preload loss

Usage description——

[Used to explain the asymmetric rising edge, residual deformation, and PZT compensation voltage anomaly observed in the B104 low-frequency drift troubleshooting; focus on finding bolted connections, micro-slip, thermal cycle preload decay, and optical mount mechanical drift.]

After reading it, Meng Che said: "Your keywords look more decent than mine when I first entered the group as a first-year PhD student."

Yin Hang snorted coldly on the other side: "At least he didn't write Kalman Filtering right off the bat in his keywords."

"If you bring this up again, I'm really going to get mad at you," Meng Che said.

Jiang Lin continued to write the second question.

[P-002: Boundaries of low-cost non-invasive optical lever verification in precision optical platform troubleshooting]

Keywords:

optical lever

angular displacement measurement

non-invasive diagnostics

laser safety

stray reflection

precision alignment

Usage description:

[Rewrite the temporary verification method from last time into an auditable, low-risk troubleshooting procedure, clarifying applicable conditions, error ranges, safety boundaries, and unsupported conclusions.]

Yao Siyu leaned her head over from the side.

"I want this one. The troubleshooting memorandum is just lacking this section. Don't just write that you can see the light spot moving; also write what it cannot prove."

"What can it not prove?" Yin Hang asked.

Yao Siyu said, "We cannot prove that the drift source is one hundred percent exclusively this single board, nor can we prove that all long-term stability issues have been resolved, nor can we prove that the data for over twelve hours after reassembly will definitely meet standards. We can only prove that in that specific operation, the adapter board had an observable angular response to the cable load."

Professor Lu Zhixing had walked up behind them at some point, and upon hearing this, he nodded.

"This is the language of scientific research."

He looked at Jiang Lin.

"Your problem in the past was that your engineering language was too strong. In engineering, once you find the source of a fault, fix it, and get the system running, you can say it's resolved. But papers won't do; papers require boundaries."

For the next two hours, Jiang Lin did not download a single paper.

He checked the bibliographic records, authors, citation chains, and whether the supplementary materials were publicly available.

He checked which titles in the references of a certain review actually concerned bolted joint micro-slip.

Meng Che originally thought he would be overwhelmed by the search results like an undergraduate entering a database for the first time.

As it turned out, Jiang Lin was much more restrained than he had imagined.

He would click open the abstract of a paper, look at it for thirty seconds, and then write beside the bibliographic record: [Only look at the experimental setup diagram, ignore the conclusions]

Or: [Model relies on AFM roughness input, currently non-reproducible]

Or: [Temperature control precision is too high, unsuitable for explaining air conditioning periodic disturbances like B104's]

Or: [Keywords are similar, problems are different, exclude]

...

By the end of the morning, his literature requisition form had only nine full-text requests filled out.

Yet there were seventy-three bibliographic records.

Meng Che picked up that requisition form and looked at it very slowly.

"Why don't you want this paper?" he pointed at a highly-cited review among them.

"It talks about contact adhesion in MEMS devices, the scale is too small, and both the material and loading conditions are different. It can serve as background, but it's not suitable as a method."

"What about this one? The citation count is very high."

"It assumes uniform bolt preload, and the experimental fixture rigidity is much higher than the sample. B104's problem happens to stem from uneven assembly preload and support constraint defects, so it cannot be directly applied."

Meng Che pointed to another one. "Why did you mark this one as needing the full text?"

"Because the abstract mentions that they retained the original loading-unload hysteresis loops, and the supplementary material has the fixture structure diagram. I want to see the hysteresis loop, not the conclusion."

Meng Che put down the requisition form.

"Fine, I'll pull these nine for you."

"Thank you."

"Don't thank me too early," Meng Che said, "After reading, you need to hand in a card detailing what it solved, what it didn't solve, whether it can provide evidence for our current problem, and making sure no single sentence is over-quoted. Each must not exceed one page."

Jiang Lin nodded; this suited his intentions perfectly.

Lunch was eaten on the second floor of the cafeteria.

Professor Lu Zhixing had something temporary come up and didn't come; at the dining table there were only two doctoral students, one master's student, and Jiang Lin.

The four of them carried their meal trays and sat at a window-side spot.

The atmosphere was more natural than last time, but still somewhat subtle.

Meng Che picked up a chopstickful of shredded pork with green peppers and suddenly asked, "Don't you usually need to grind through practice problems during your third year of high school?"

"I do."

"Then when you come here on weekends, what about your homework?"

"I write it at night."

"What about the Gaokao?"

"It won't affect it."

Looking at that calm face of his, Meng Che suddenly felt that he had asked a redundant question.

Yin Hang took a sip of soup, put down his bowl, and asked, "Have you really done machining before?"

"I've done some."

"Lathes?"

"Haven't touched them."

"Milling machines?"

"Haven't touched them."

"Scraping?"

"I've done that a bit more."

Yin Hang fell silent.

After a moment, he said, "Over at the Engineering Training Center, there are basic practical training equipment on the weekends. Once the paperwork comes through, I'll take you to see them. Let's make it clear first: you can only look and do basic operations, you can't jump straight into making complex parts."

"Okay."

"Also," Yin Hang added, seemingly a bit unnaturally, "that adapter board from last time—I've redrawn a version of it. The material has been changed to stainless steel, the bottom face ground, the hole-edge chamfers written into the drawing, and the cable fixing point made independent. Could I trouble you to take a look at it for me when you have time this afternoon?"

Meng Che immediately raised his head.

"Oh."

Yin Hang said expressionlessly, "Shut up."

Yao Siyu lowered her head to eat, but the corners of her mouth curled up slightly.

Jiang Lin said, "Sure, but I can only look at the mechanical boundaries; you yourselves need to confirm the optical height and installation space."

Yin Hang nodded, "I know, thank you."

After these words, a certain tension at the dining table finally dissipated.

In the afternoon, at the temporary workstation in B304, Jiang Lin received the first batch of official full texts.

The third chapter of a doctoral dissertation, two sets of supplementary materials, and an optical mount manufacturer's manual.

A bolted connection assembly specification, and an excerpt of old troubleshooting records from inside Professor Lu Zhixing's research group.

When Meng Che opened the folder permissions for him, he said, "These can only be viewed on this machine and cannot be copied away. You can extract bibliographic records and your own understanding, but do not copy large paragraphs."

"Understood."

Jiang Lin opened the first paper.

Regarding micro-slip and stiffness degradation of bolted connection structures under thermal cycling.

He read very slowly.

The first pass only looked at the diagrams.

Experimental setup diagrams, loading path diagrams, temperature cycling curves, and hysteresis loops.

The second pass looked at the methodology.

Sample dimensions, material grades, bolt specifications, preload application methods, temperature control rates, and sensor models.

The third pass looked at errors.

Number of repetitions, error bars, sources of uncertainty, and whether the authors wrote about the thermal drift of the fixture itself.

When reading a certain section, he stopped.

There was a figure in the paper showing that after the twelfth temperature cycle, the joint interface stiffness did not continue to decrease linearly, but instead appeared a plateau period.

The authors interpreted this as the contact interface re-stabilizing.

Jiang Lin stared at that figure for a very long time.

Then he wrote on the literature card: [Note: A plateau period does not necessarily represent stability; it may represent that the current measurement proxy variable is insensitive to continued degradation. It needs to be judged in combination with residual displacement and loading-unload hysteresis loops.]

When Meng Che passed by and saw this line of text, he stopped in his tracks.

"Your suspicion makes sense," he said, "Many people want to write stabilization as soon as they see a plateau period, but reviewers will ask you whether it's just that you can no longer measure it."

"Therefore, it cannot be directly cited."

"Correct."

Meng Che pulled up a chair and sat down.

"The way you read literature is strange."

"What's strange about it?"

"You aren't in a hurry to find sentences that support your own views."

Jiang Lin thought for a moment.

"I'm more afraid that it's deceiving me."

Meng Che was stunned for a moment.

Undergraduates reading literature often do so to find a sentence that can be put into an assignment.

Master's students reading literature do so to find a precedent that can support their experimental design.

Doctoral students, reading on further, gradually come to understand that literature does not necessarily lie, but literature often tells only the part it wants to say.

Yet Jiang Lin stood at this position right from the start.

At half-past four in the afternoon, Jiang Lin finished three literature cards.

Each not exceeding one page.

First card: [BLT-001: Micro-slip and Stiffness Degradation of Bolted Connections under Thermal Cycling]

What it solved: Proved that tiny temperature cycles are sufficient to induce joint interface hysteresis.

What it didn't solve: Did not handle parasitic angular drift on extremely high-rigidity bases like optical platforms.

Cannot be copied directly: Sample size, temperature control rate, and preload application conditions are inconsistent with B104.

Can be borrowed: Processing method for loading-unload hysteresis loops; using residual displacement as an auxiliary indicator.

Second card: [OPT-004: Troubleshooting Records for Low-Frequency Drift of Optical Mounts]

What it solved: Provided a case study of similar thermal drift in optical brackets.

What it didn't solve: Did not publish raw data, only processed curves.

Risk: May rely too heavily on software smoothing.

Can be borrowed: Monitoring PZT compensation voltage as a mechanical drift proxy variable.

Third card: [STD-002: Specification for Step-by-Step Cross Preloading of Bolted Connections]

What it solved: Provided the basis for formal assembly sequence and torque verification.

What it didn't solve: Did not discuss micron-level burr pivot points on clean optical platforms.

Can be borrowed: Institutional basis to be written into the B104 reassembly process.

After reading them, Meng Che handed the three cards to Yao Siyu.

Yao Siyu scanned through them and said, "These can go directly into the memorandum attachments."

Yin Hang also leaned over to take a look, pointing at the third card.

"Can you send me this specification? When I draw the new adapter board later, I'll write the assembly requirements into the drawing notes."

"Have Meng Che give it to you," Jiang Lin said, "I don't have download permissions."

Yin Hang was stunned for a moment, then smiled a bit sheepishly.

Close to five o'clock, Professor Lu Zhixing returned from outside.

He was holding a freshly printed document, and neither joy nor anger could be seen on his face.

"The twelve-hour data is out."

Everyone in the room stopped what they were doing.

Meng Che asked first, "How is it?"

"Much better than before, the low-frequency humped peak didn't return, and the PZT compensation voltage stayed below twenty percent," Professor Lu Zhixing said.

Yao Siyu visibly let out a sigh of relief.

But Professor Lu Zhixing did not immediately declare victory.

He put the document on the table and tapped the last page.

"However, after six hours, a new slow drift appeared. The amplitude is very small, nowhere near the previous catastrophic level, but it exists."

Meng Che frowned, "Still mechanical?"

"Not necessarily," Professor Lu Zhixing said, "It could be the thermal equilibrium process after system reassembly, or the redistribution of local stresses on the platform, or the long-cycle control of the air conditioner, or another boundary condition we haven't seen yet."

As he spoke, he looked at Jiang Lin.

Jiang Lin was looking at that twelve-hour residual curve.

Compared to the massive low-frequency fluctuations with a forty-minute period from last time, this drift was much smoother—like a very shallow slope hidden within the noise, inconspicuous yet stubborn.

He opened a new page in his notebook.

[P-003: Source of Long-Cycle Slow Drift After Reassembly]

Below, he wrote the following lines.

1. Thermal equilibrium not yet complete.

2. Local stress on the platform re-settled.

3. Slowly varying traction between the cables and the external cabinet.

4. Current monitoring variables are insufficient, unable to determine.

Seeing this sentence, Professor Lu Zhixing nodded with satisfaction.

"Very good," he said, "This is the most important progress of today."

Meng Che didn't quite react in time: "Isn't the most important thing that the twelve-hour data has improved?"

"No," Professor Lu Zhixing said, "The most important thing is that we didn't declare peace across the world just because of one successful troubleshooting."

He handed that data to Jiang Lin.

"Starting next week, your first formal task is to establish a literature and evidence index for the low-frequency drift problem. Stratify everything we currently know, suspect, have excluded, and cannot yet determine. Papers, supplementary materials, instrument manuals, old troubleshooting records, and assembly specifications must all be included."

"Didn't you say you needed a literature pool? Then start from this problem. Don't build a pool that holds everything; first build a pool that can help us make fewer mistakes."

Jiang Lin looked at the twelve-hour data in his hand, and then looked at the sheet of paper beside the temporary literature terminal that prohibited batch downloading.

This was even better than giving him an unlimited database account.

Infinite data would only drown a person.

Only a real problem can sift data into a path.