I started with a ship full of max-level veterans, and I'm just a complete noob.

"That's the problem."

Tu Shou'e pinned the first spectrum graph to the drawing board, and then used a red pencil to trace along the three peaks in sequence, finally stopping at the installation position in the middle section of the instrument compartment.

After the projectile completes structural weight reduction, the longitudinal vibration and local torsional modes of the shell shift, and the three resonance peaks pass through the working areas of the gyroscope sampling, servo feedback and intermediate frequency amplification circuit, respectively.

The second overlap area is the widest. When the engine enters the thrust fluctuation range, the mounting bracket will directly send the casing vibration to the gyroscope.

Tu Shou'e turned to the guidance group.

"Continuing to thicken the structure would encroach on the space for propellant and cables. The current feasible solution is to reinforce the gyroscope mounting frame and add damping layers in certain areas to raise the instrument's own frequency."

Vasily, who was in the audience, turned over the translation and waited for the translator to finish speaking before tapping the table.

"The Soviet model had already solved the problem of basic reinforcement, while the Chinese side could not even complete the instrument mounting bracket independently, yet they were in a hurry to change the overall parameters."

The translator relayed the message, and the engineers in the room looked at the blueprints, not wanting to waste time on verbal exchanges.

Ren Xinmin opened the gyroscope parameter booklet and placed the sensitivity curve and drift error table side by side in front of Tu Shou'e.

"With increased mounting rigidity, local high-frequency vibrations will be transmitted to the rotor more quickly. Furthermore, with a thicker damping layer, both the startup response and the measurement of minute angular velocities will decrease."

He pressed down the sensitivity curve with a slide rule.

"The guidance system can only tolerate this range of response loss. If it exceeds the boundary, the initial attitude error will enter the integrator, and no amount of subsequent control input can completely compensate for it."

The structural engineers took the parameter sheet and rechecked the relationship between the mass and stiffness of the mounting frame. The calculated results were still within the same range.

The structural weight reduction plan has been finalized, and the instrument compartment space has been cut off by engine pipes and control cables. Any local modification will affect adjacent interfaces.

The guidance group also had no way to retreat. The sensitivity of the gyroscope was related to the lateral correction model. The Soviet parameter package had just been abolished, and China could not introduce a new error that could not be quantified.

Tu Shou'e stacked the two pictures on the lightbox, revealing three overlapping areas through the light.

"The structure can relinquish some stiffness, and the second peak will still remain within the working zone."

Ren Xinmin picked up a pencil and wrote down the allowable floating range next to the guidance sampling frequency.

"The sampling law can only be changed here. If it is moved outwards, the existing stable model will need to be recalculated entirely."

The discussions in the room repeatedly collided around several sets of boundaries, and the values ​​that could be moved were gradually crossed out, until the remaining blank space was not enough to accommodate the second resonance peak.

Jiang Ming flipped through the gyroscope sampling table, his attention not on the mounting bracket material, but on the time relationship between the sampling pulse, signal hold, and servo output.

In the systems engineering he had encountered in his previous life, early models often treated each subsystem as a fixed module, with the structure responsible for bearing the load, the control system responsible for stability, and the electronic equipment responsible for execution.

This division of labor works when the boundaries are loose, but once frequency, quality, and response time are squeezed against each other, any single optimization will push the cost to the adjacent system.

The current deadlock is not due to insufficient material strength, nor is it solely due to the sensitivity of the gyroscope. What is truly locked in is that everyone treats the operating frequency as a continuous and fixed benchmark.

Jiang Ming pulled out a blank sheet of calculation paper, converted the positions of the three resonance peaks into the time domain, and marked the guidance sampling period next to each item.

Fixed sampling ensures that each reading falls on a similar vibration phase, and the error signal is continuously fed into the integrator, so the vibration leaves a stable false deviation in the control system.

If the sampling interval can be varied according to a predetermined sequence, continuous reading will avoid the same vibration phase, and structural resonance will be difficult to accumulate continuously within the control chain.

Jiang Ming got up and walked to the podium, picked up the chalk, and drew a timeline on the blank space of the blackboard.

He first marked the engine thrust fluctuation zone, then divided the gyroscope sampling, holding, comparison and rudder output into four segments, and then projected the three resonance peaks onto the coordinate system.

Tu Shou'e read a few lines and then raised his hand to interrupt him.

"Comrade Jiang Ming, have you received complete training in projectile structural mechanics?"

"no."

Jiang Ming put down the half-used piece of chalk and looked calmly at the drawing board.

"Regarding the stress distribution and modal calculations of the structure, I can only rely on your conclusions. However, once the guidance signal enters the electronic circuit, the frequency overlap will interpret the structural vibration as an attitude change."

Tu Shou'e didn't let him return to his seat, but instead pressed, "Are you planning to change the structural frequency, or the gyroscope?"

"First, change the timing of the dialogue between the two sides."

Jiang Ming picked up the chalk again and inserted different intervals between the originally equidistant sampling points.

"Fixed-period sampling repeatedly reads the resonance error in the same phase. The system accumulates it into an effective signal and changes it into a pre-set non-equal spacing sequence. The sampling points can bypass three danger zones."

Ren Xinmin stood up and walked to the blackboard to check the time coordinates.

"When the sampling interval changes, the delay term in the stability equation will also change. The existing integrator is designed with a fixed period, so compensation must be added."

"The compensation is placed after the hold loop, and the corresponding weight is used for each sampling period. The long-term average frequency remains unchanged, and the control model can still maintain the original time scale."

Jiang Ming wrote the widths of the four pulse sets below the coordinates and marked the cycle order.

Each interval avoids integer multiples of the resonance peak, and the average sampling rate returns to the original design value after the complete cycle ends, so the servo output timing does not need to be shifted back as a whole.

Several old-school control experts opened the stability derivation papers, their expressions gradually turning serious.

One of them pointed to the blackboard and said, "The sampling law is part of the core model. Changing it arbitrarily will break the linear assumption. Even a mechanical timer may not be able to achieve this allocation."

Jiang Ming did not skip this question.

"Therefore, we must first verify the electronic implementation and then perform stability boundary recalculation. Only after the scheme passes both verifications can it be incorporated into the overall system."

After listening to the translation, Vasily leaned back in his chair and chuckled a few times.

"To avoid a structural defect, the sampling law of the guidance core is modified, and then compensated with electronic devices that are not yet finalized. This kind of solution only exists on the blackboard."

His Soviet assistant opened a technical manual and pointed out that existing relay-type sequencers could not stably output multiple intervals, and that vacuum tube triggering circuits would also cause additional drift.

This questioning, grounded in the actual craftsmanship, is more difficult to evade than previous satire.

Jiang Ming looked at Ren Xinmin.

"If we use existing vacuum tube sequencers, we really can't control temperature drift and contact bounce."

Ren Xinmin caught the second half of his sentence and looked down at the silicon tube design report on the table.

"Can those 800MHz devices perform multi-phase triggering?"

"The cutoff frequency is sufficient. The real challenge lies in frequency division stability and pulse maintenance. We can first build a four-phase verification circuit with 502."

The venue fell silent again.

The structural team is awaiting the electronics conclusions, while the guidance team is estimating the stability risks brought about by non-equidistant sampling. Either party that signs first will bear the responsibility of going beyond the boundaries of their respective profession.

Tu Shou'e did not press the matter further.

He stood in front of the drawing board and looked at the sampling sequence several times. Then he picked up the slide rule and converted the three resonance peak periods into time intervals.

In five minutes, only a few sets of values ​​he wrote down appeared on the blackboard.

Tu Shou'e took down his own structural spectrum diagram, took a red and blue pencil, walked to the podium, and pasted the spectrum diagram next to Jiang Ming's time series coordinates.

The red pencil first circles the first overlapping point, and the blue pencil moves backward along the new sampling sequence, with the sampling pulse passing to the left of the peak.

The second resonance region is relatively wide. Tu Shou'e checked four sets of intervals in succession and found that after the pulse distribution was split, the error could not be continuously superimposed in the same phase.

The third overlap is located at the end of the servo feedback, and Jiang Ming's retention of the average sampling rate brings the overall timing back to the original interface range.

Tu Shou'e turned to look at the structural team.

"The rigidity of the mounting bracket is increased by only 30% compared to the original design, and the damping layer is halved to preserve the sensitivity of the gyroscope."

Ren Xinmin then asked, "Can the structural side control the three peak values ​​within the error band shown on the graph?"

"The overall team is responsible for retesting the differences in the trial production. If the measured peak value exceeds the error band, the structural team will re-weight it."

Tu Shou-e circled the permitted modification area with a red pencil, and then wrote temporary review comments in the lower right corner of the drawing.

He capped his pen and looked at the silicon tube report in front of Jiang Ming.

"As long as the electronics can implement this sampling sequence and the stability recalculation can pass, the overall projectile body will bear the responsibility for structural modifications."

Vasily's laughter had stopped.

Jiang Ming wrote the register, frequency division, and hold conditions required for the four-phase trigger circuit at the bottom of the blackboard, leaving the last cell blank for verification.

That one cell determines whether the plan can move from the timing diagram to the missile body.