Why do helicopter parallel servos always break down? A maintenance worker’s experience of blood and tears

I have seen a large number of helicopters, and there is nothing abnormal when they are placed in a stationary state. However, once the collective pitch is pushed, the fuselage will experience uncontrollable shaking that is difficult to control.

It felt like someone was playing a drum in the cockpit.

When I took it apart, I found that the parallel servo was "emotional" again.

The scientific name of this kind of thing is "parallel telex steering gear". To put it simply, two motors share one output shaft, and what is used is the "dual-machine hot standby" method.。

There is a general aviation company that owns an R22. This R22 had a new steering gear. However, within 20 hours of flying, a "power dispute" occurred.

After taking it apart and measuring it, the current difference between the two motors was nearly 30%!

This is a typical "internal friction".

Why?

Because during installation, the lengths of the two tie rods were not adjusted to be absolutely equal.

Previously, one was pushing forward, while the other was pulling backward. Just like that, the two of them started fighting on their own. In this way, how could they care about the flying posture?

You see, this is the core weakness of the parallel servo:synchronicity。

What this thing pays attention to is "same frequency resonance", which is almost impossible.

I have seen an old machinist who never uses measuring tools when installing machines. He just relies on his feel.

The result? The plane rolled across the sky and almost didn't come back.

Since then, we have made a strict rule: when changing the steering gear, you must use a laser alignment tool.

This is the SOP gained through bloody lessons.

Did you know? Many times the servo "twitches", it is not the servo itself that is broken at all.

But the instructions given by its "brain", the flight control computer, were in conflict.

Once, a helicopter completed its regular inspection and everything was OK in the ground test.

As soon as it left the ground, the parallel servo began to make a "buzzing" sound.

Everyone thought it was the poor quality of the steering gear.

After checking for a whole day, I finally found that a parameter in the flight control had been changed by mistake.。

That parameter is called "cross-channel gain."

If the error is corrected, the phase of the commands sent by the flight control to the two motors will be 180 degrees different.

When one is extended and the other is retracted, can the steering gear rod not vibrate?

Therefore, when your parallel servo exhibits high-frequency oscillation, don’t rush to replace it.

Let’s take a look at the flight control log first.

I call this "soft first and then hard".

Sixty percent of "servo failures" are actually caused by software.

This is an open secret in the industry and the easiest place for newcomers to pay tuition.

So, how to judge whether it is mechanical internal friction or electrical draft?

I have a crude method, but it is extremely effective.

Listen to the sound.

The sound of strenuous struggle is a dull, low-frequency "dong, dong" sound, like a heart murmur.

The sound of an incorrect command is a sharp, high-frequency "squeak, squeak" sound, like a mouse.

Don't laugh, I've been using this trick for ten years, and it's faster than any spectrum analyzer.

Because sound is the most honest language of machinery.

Tips for article writing.

Okay, let's dig deeper.

I just talked about sound diagnosis, but this is just superficial.

The deeper problem lies in our superstitious belief in "redundancy".

We all think that dual-machine backup = security.

But in the case of parallel servos, dual-server backup often means "double trouble."

Why?

Because you have to introduce an extremely complex "force balancing" algorithm.

The algorithm itself has become a new point of failure.

It's like two people carrying water to drink. If the steps are inconsistent and the water is not drunk, the pole will break first.

I have dealt with one of the most classic cases.

A helicopter used for power line inspection had a parallel servo failure every month for three consecutive months.

Just replace it with a new one, but it will break after flying for dozens of hours.

Everyone suspected it was a batch quality problem.

Until I went to the site and took a look.

Every time that plane landed, the pilot would tap the collective pitch lever all the way to the bottom.

At that moment of hydraulic shock, the peak pressure was five times the normal value!

The force sensors equipped with the servos connected in parallel state, under such repeated violent impacts, those relatively fragile electronic components quickly entered a fatigue state.。

It's not that the steering gear is bad, it's that people's operating habits are too "violent".

This is philosophical speculation.

We always want to use more complex systems to defy the laws of physics.

But the laws of physics will always come back to bite you in more subtle ways.

The essence of failure in the game of uncertainty is every "non-command movement" of the parallel steering gear.

You want to trade 100% redundancy for 100% security.

As a result, redundancy itself creates a 1% failure rate.

In order to eliminate this 1%, you have to introduce more complex monitoring logic.

The more logic there is, the more holes there are.

Is this like a Möbius strip that can never be untied?

Back to the practical level.

Faced with these "squeamish" parallel servos, what should we do?

Create an "installation file".

It’s not just about recording the replacement time.

The initial "zero voltage" of both motors is to be recorded.

On the same aircraft, if parts are replaced twice before and after, the voltage value cannot differ by more than 50 millivolts.

Implement "static testing".

After installation, don't rush to test drive.

Use a hydraulic hand pump to slowly pressurize the steering gear.

Watch the values ​​of the two force sensors rise simultaneously.

If there is a difference of more than 2% in the neutral position, rework immediately.

Don't expect the flight controller to correct this difference for you.

The flight controller is not a god, it is just a programmer responsible for wiping the butt.

Tips for article writing.

continue.

This is also the most counter-intuitive point:Regularly "decouple" the parallel servos。

What's the meaning?

Every 200 hours, disconnect the control circuit of one of the motors.

Let the single motor run a complete stroke with the steering surface.

Record the friction curve for this stroke.

Then change to another motor and run the same stroke.

If the two curves do not coincide, it means there are "hard spots" in the mechanical transmission chain.

This "hard point" will be the trigger for future "power disputes".

Don't wait for a fault alarm to be investigated.

That's called making amends.

What we need to do is to "treat the disease before it is cured by the superior doctor".

I know, some people say this is over maintenance.

It says that the manufacturer's manual does not require this.

However, I want to ask you, will the manufacturer's manual bear responsibility for your helicopter operating on a plateau of 4,000 meters above sea level?

The manufacturer's manual guarantees a lower limit, not an upper limit.

The bosses of companies engaged in the general aviation business are reluctant to purchase calibration benches in order to save the cost of regular inspections.

The result?

A forced landing due to jamming of the servo can cost you enough to buy ten calibration benches.

I hope every maintenance brother and aviation school student can understand this account.

Q: Why are parallel servos more prone to "force disputes" than single servos?

A: Since there are two independent drive units, and it is difficult for these two independent drive units to achieve absolute synchronization, even small mechanical tolerances or electrical differences will cause internal confrontation, leading to abnormal wear and heat.

Q: How to quickly determine whether the fault source is the servo itself or the upstream controller?

When a situation occurs, for A, the control harnesses of the two servos need to be swapped. If the fault is transferred after the swap, it means there is a problem with the servos. If the fault location has not changed, then there is a problem with the upstream command or wiring.。

Q: Which parameter is most easily overlooked during daily inspections?

A: It is called the "current difference" under static conditions. Even if the steering surface is stationary, the two motors in standby mode should consume an equal static current. Once the difference exceeds 5%, an early warning signal will be issued.

Q: What is the threat of humid environment to parallel servos?

A: There is a fatal threat. Moisture will cause the insulation of the force sensor bridge to tend to decrease, thereby outputting false error signals, and ultimately causing irregular "ghost jitter".

Q: What is the highest criterion for perfectly replacing the parallel servo at one time?

For A, it is necessary to pursue the absolute coincidence state of "mechanical zero position" and "electrical zero position".If there is any forced installation, stress will be built in, paving the way for future disasters.。

Listen, guys.

A helicopter is not a car and it doesn't get second chances.

Parallel servo is not a black technology, it is just two competing brothers.

Your job is to be the referee.

Let them think in the same place and work hard in the same place.

This requires you to have data in your hands, details in your eyes, and awe in your heart.

Next time, when you tighten the lock nuts on those two tie rods.

You must remember that what you screw on is not just a nut, it is related to the life and safety of tomorrow's crew, and it carries the hope of the child who is eagerly waiting for his father to come home in the hangar the day after tomorrow.

Don't let your empiricism turn into boldface in your accident report.

From today on, treat every parallel steering gear as an independent living entity.

Calibrate it, listen to it, tame it.

This is the destiny of our profession and the meaning of our existence.