## Chapter 4: The Price of Progress

  The approval arrived not as a meeting but as a document, slid under his dormitory door on a Thursday morning by Zhang's aide, who did not knock.

  ---

  > **GENERAL ARMAMENTS DEPARTMENT — PROGRAM AUTHORIZATION**

  > *Reference: PGM-152 Development Study*

  > *Date: 12 November 1987*

  >

  > Authorization is granted for a twenty-four month development program under Program Director General Fang Liming, Artillery Corps. Budget allocation: Category B, discretionary. Technical consultant: Lin Wei (civilian, research grade). Program classification: Internal restricted.

  >

  > *Approved: Zhang [signature]*

  ---

  Lin Wei read it once. Set it on the desk. Read it again.

  Twenty-four months. Category B budget — modest but real. His name on a program document for the first time. His name under General Fang's, which was correct; under Zhang's authorization, which was the thing that made it real.

  He did not celebrate. He made a list.

  ---

  **Beijing, January 1988. 2:17 AM.**

  The guidance section was not surviving the firing impulse.

  On the test rig — a hydraulic ram that simulated the shell's launch acceleration — the prototype seeker failed three consecutive times. Different failure modes each time. The first: the laser detector array cracked along a pre-existing stress line in the housing. The second: the signal processing circuit's solder joints fractured. The third: the detector array itself delaminated at the bonding interface.

  Lin Wei sat alone in the test cell at two in the morning with the failed module on the bench and a cup of cold tea.

  The silence was the kind that belongs to problems without obvious solutions.

  Wei Hua appeared in the doorway. The young optics engineer had been working late in the adjacent lab. He looked at the module. He looked at Lin Wei.

  "What killed it this time?"

  "Delamination at the bond interface. The acceleration stress is higher than the adhesive rating."

  Wei Hua came in and sat on a stool across the bench. He picked up the failed module and turned it over in his hands, studying the fracture surface.

  "The gimbal is the problem," he said. "Any mechanism that moves under those loads is going to fail. The adhesive, the bearings, the pivot — something breaks."

  "I know," Lin Wei said. "That's why I want to remove the gimbal entirely."

  Wei Hua looked at him.

  "Fixed cruciform detector array," Lin Wei said. "No moving parts. The spot location is computed from the differential signal between four quadrants. The mathematics is straightforward — it's basically a weighted centroid calculation."

  Wei Hua was quiet for a moment. Lin Wei could see him working through it — the physics, the signal processing architecture, the manufacturing implications.

  "The field of view narrows," Wei Hua said finally.

  "Fourteen degrees half-angle. I've run the numbers."

  "Is that enough?"

  "The shell is already on a trajectory close to the target. We're not searching for the designator spot — we're steering toward it. Fourteen degrees is enough if the observer keeps the spot on target through the terminal phase."

  "And if he doesn't?"

  "Then the round misses. Same as every other guided munition when the designation breaks." Lin Wei spread his hands. "Nothing flies itself."

  Wei Hua set down the failed module. He looked at the bench — the tools, the test documents, the cold tea, the two of them in a fluorescent-lit room at two in the morning over a problem that didn't exist yet anywhere else in the world.

  Support the creativity of authors by visiting Royal Road for this novel and more.

  "Show me the centroid calculation," Wei Hua said.

  Lin Wei pulled a notepad from under the bench.

  They worked until 5 AM. When they stopped, the mathematical framework was solid. The engineering path was clear.

  "Four months to a prototype guidance section that survives the rig," Wei Hua said.

  "Three," Lin Wei said.

  Wei Hua looked at him with the expression of an engineer being told an impossible timeline by someone who has never been wrong before.

  "Three," he agreed, slowly.

  He went home at 5:15 AM. His wife was already up — she worked early shift at the hospital — and she looked at him standing in the doorway in yesterday's clothes and said nothing for a moment.

  "Lin Wei's problem?" she asked.

  "My problem now," Wei Hua said.

  She handed him tea. He sat at the kitchen table and looked at the centroid calculation he had copied into his notebook and thought about the three months of nights it represented. He had been attached to Lin Wei's programs for two years now. He had published less than his peers. His institute supervisor had mentioned this, carefully, twice.

  He thought about the fixed-array solution. The elegance of it — removing the moving part, absorbing the constraint into the mathematics, making the failure mode disappear by refusing to need the thing that failed.

  *That's it,* he thought. *That's the whole approach.*

  He finished his tea and went to sleep for three hours. When he woke up, he called the institute and said he would be in by noon, and spent the morning working through the full signal-processing derivation.

  He did not think about the publication gap again for quite a while.

  On a Tuesday in April 1988, the prototype guidance section survived the test rig firing impulse on the first attempt. Wei Hua ran the test three more times to be sure. The module survived all four.

  Lin Wei felt nothing dramatic. He had expected this result.

  He went home, slept four hours, came back, and started working on the base-bleed propellant grain.

  ---

  **The No. 5 Ordnance Factory. October 1988.**

  Five months of propellant grain testing and nothing was working.

  The base-bleed unit concept was simple: a small pyrotechnic grain burned at the shell's base during flight, filling the low-pressure region with combustion gas, reducing drag, extending range. The physics were right. The results were not.

  Too fast: the gas jet destabilized the shell's rotation. The round tumbled at 8,000 meters and impacted 400 meters short of the aiming point.

  Too slow: negligible range benefit. The test round landed 200 meters short of the model prediction.

  The burn rate problem was a function of three variables — grain geometry, propellant composition, ambient pressure — and every time the propellant team solved one, another drifted. Old Ma's best man, an engineer named Deng who had been in propellant chemistry for fifteen years, had run forty-seven grain configurations. Forty-seven failures.

  Lin Wei was at the factory on a Monday when configuration forty-eight went into the test cannon.

  Deng loaded the test round himself. He had the face of a man who has stopped expecting anything.

  The round fired. Downrange, on the tracking instruments: nominal rotation rate through burnout, predicted trajectory, terminal phase—

  Short. Sixty meters short. Burn rate still high.

  Deng sat down on the ammunition crate beside the test cannon and said nothing. He just sat.

  Lin Wei walked over and sat beside him. After a moment: "The star-perforation geometry. The arms."

  "We've tried three arm lengths."

  "Not the length. The thickness at the root. If the root is thicker, the surface area regression is slower at the start of burn and the burn rate flattens." Lin Wei had read this in a Swedish technical exchange paper — not classified, just carefully remembered. He described it as if he was working it out now. "The Swedish 155mm round uses a star perforation with a root thickness 40% of the arm length. We're at 25%."

  Deng looked at him. "That's specific."

  "I saw a paper in the foreign technical literature exchange. It mentioned the geometry indirectly. The implication was their result."

  Deng was quiet. He was a careful man. He did not accept someone else's solution without thinking it through. Lin Wei waited.

  "The initial surface area is lower with the thicker root," Deng said slowly. "So the early burn rate drops. But the late burn rate might spike when the perforations merge—"

  "Modified arm taper to maintain surface area through the merge," Lin Wei said. "Longer arms, tapered, thicker at root."

  Deng looked at the test cannon. He looked at the forty-eight failed configurations' data sheets on the clipboard in his hand.

  "Configuration forty-nine," he said. "Give me a week."

  Configuration forty-nine flew 24.3 kilometers and burned through its full run without rate excursion.

  ---

  **Zhurihe Range, March 1989.**

  The first live guidance test.

  General Fang Liming, standing at the observation post, had the expression of a man who had approved many programs and been disappointed by most of them.

  Downrange: a tank hull on a berm at 24 kilometers. Beside it, a forward observer with a laser designator, maintaining spot continuity. Waiting.

  The 155mm gun fired. The shell departed at 827 meters per second. Its base-bleed unit ignited at 0.3 seconds. Its trajectory was a long, high arc over the Inner Mongolian plain.

  At 23.8 kilometers, entering the terminal phase: the guidance section activated. The detector array found the laser spot. The differential signal computed the correction. Four steering fins deflected — two degrees, fractional adjustment, the shell already nearly there.

  The round impacted three meters from the laser spot.

  Silence.

  General Fang said: "Again."

  Rounds two through twelve. Average error: 2.8 meters. Range: 23–25 kilometers depending on charge selection.

  Eleven of twelve within five meters.

  "This changes everything," Fang said, quietly.

  Lin Wei was watching the data feed. He was already running the next problem in his head: seeker reliability through smoke and dust. Countermeasures. Supply chain for the guidance components. The export version's modified performance parameters.

  He was also thinking about something else.

  He was thinking about how to get this technology into a conflict where it would be seen.