Hacker News Reader

Comment by danaliv | Hacker News Reader

Here's the directive: http://rgl.faa.gov/Regulatory_and_Guidance_Library/rgad.nsf/...

This isn't really a grounding, at least not the way the 737 MAX is grounded. Cirrus jets can't be flown until the AD is complied with, but complying with the AD is a (relatively) simple matter of replacing the angle of attack sensors. Once that's done, the aircraft is airworthy again.

The issue is a "quality escape," i.e. the existing sensors weren't built to spec. There are two screws that attach the vane to the potentiometer; the factory didn't torque the screws properly and didn't apply thread locker to them.

So this is a manufacturing error, not a fundamental problem with the design. I say that just to provide contrast with the MAX issues since the problems are superficially similar, but quite different once you get into the details.

View on HN

I used to work as an apprentice for a GA mechanic (never pursued it, went to college instead).

Disassembling something, inspecting it then performing reassembly and properly torquing the screws with locktite per some published procedure is one of those things aircraft mechanics do all day.

Unless the design of the sensor would make it inaccurate after being reassembled (I'm skeptical of this, it wouldn't make sense for a few reasons) this is kind of a dick move to the aircraft owners that they're probably doing in the name of "look how careful we're being"

Any certified mechanic anywhere can do the former with no lead time and if it turns out the threads are close to stripped out from a loose sensor banging around on there then replace the sensor. It doesn't make sense to replace the sensor off the bat unless you just want to be seen "doing something".

> Unless the design of the sensor would make it inaccurate after being reassembled (I'm skeptical of this, it wouldn't make sense for a few reasons) this is kind of a dick move to the aircraft owners that they're probably doing in the name of "look how careful we're being"

I think your cynicism is uncalled for. FAA found an issue that is life threatening to pilots and passengers, just because pilot/mechanics are aware about how careful they are with their planes, does not mean that you invalidate the actual manufacturing issue that is found with these AOA sensors.

"We are aware, we are not stupid, don't tell us what to do." - this is a wrong attitude especially from an Aviation mechanic, unless I am missing something, otherwise please elaborate.

>I think your cynicism is uncalled for. FAA found an issue that is life threatening to pilots and passengers, just because pilot/mechanics are aware about how careful they are with their planes, does not mean that you invalidate the actual manufacturing issue that is found with these AOA sensors.

I'm saying the stupid thing here is that mechanics/owners have to trash the part and wait for a replacement when the fix is to apply locktite and assemble properly (assuming sensor is undamaged) which is something any mechanic can do. Sure it might be life threatening but having a mechanic fix the sensor is no different than having a factory worker assemble the sensor and having the mechanic install it. The "bad" sensors are almost certainly going to be turned in to the company and remanufactured. That process will consist of checking them and if they're good to go then they will be reassembled properly. This is exactly what a mechanic would do only a different person will do each step of the task.

>We are aware, we are not stupid, don't tell us what to do." - this is a wrong attitude especially from an Aviation mechanic, unless I am missing something, otherwise please elaborate.

What you're missing is that if the fix is simply a properly assembled part then having the mechanic disassemble the part, inspect it per some spec then assemble it per some procedure is not functionally equivalent to a properly manufactured part, it it literally equivalent to what would happen to the part as part of being properly assembled at the factor. If there isn't something special about this sensor that requires factory conditions to get right the FAA is basically calling the mechanics stupid here. They're basically saying a semi-skilled factory worker can be trusted to assemble the part but a mechanic who disassembles and reassembles even more "critical to aircraft function" stuff all day can't be.

>I'm saying the stupid thing here is that mechanics/owners have to trash the part and wait for a replacement when the fix is to apply locktite and assemble properly (assuming sensor is undamaged) which is something any mechanic can do.

That would NOT satisfy the FAA Corrective Action listed in the linked document. A part replacement is mandated. [0]

[0]: (g) Corrective Action(1) Before further flight after receipt of this emergency AD, replace the AOA sensor with an improved AOA sensor, Aerosonic part number 4677-03 Mod 1 or Cirrus part number 32159-004 in accordance with section 11. ACCOMPLISHMENT INSTRUCTIONS, paragraphs A, B, and C of Cirrus Design Corporation SF50 Service Bulletin Number: SB5X-34-03, dated April 16, 2019

That is exactly what the parent comment is saying is stupid about the FAA directive. Because parent believes that a mechanic can reassemble the problematic AOA sensor to be at least as good as the "improved" replacement AOA sensor from the factory.

(I have no experience here, just pointing out something I think you missed.)

It’s probably the right call from an optics perspective so soon after the 737 MAX fiasco.

It’s more expensive than a workbench fix, but cheaper than bad publicity and a loss of faith in FAA governance. It’s a shame the cost is being foisted on GA owners though, owning a private plane is expensive enough already.

I'm not going to shed any tears over CEOs that suddenly have to get first class flights but being thrown under the bus so that someone else can optimize their political situation does not feel good no matter how minor the slight is in the grand scheme of things.

CEOs aren’t flying Vision jets. The ridiculous attitude that general aviation is somehow a class issue is getting old. Cessna Mustangs and other similar types are used frequently for air ambulance operations as well as plenty of non-CEO uses. General Aviation is a lot more than Sergei flying in his Gulfstream. Look at the 1989 earthquake for example and how GA was a huge part of delivering relief supplies to Watsonville. Look at Angel flights and the value that brings to people. GA is also essential to organ delivery.

Eh, those are $2.5 million+ dollar airplanes. It's technically general aviation but these aren't the GA pilots we need to worry about.

Another data point to remind everyone how AoA is one of the most important metrics wrt flying an airplane (airspeed is right up there, too).

Obligatory citation; imho, one of the webs's great resources—

https://www.av8n.com/how/htm/aoa.html#sec-aoa-intro

I am qualified to fly several types of small planes, none has an AoA. I was able to fly with a faulty Pitot reading (airspeed), but I would not do it often; I had never felt the need for an AoA sensor, not even when I was flying in "unusual attitudes" (intentionally).

My second ever solo my airspeed indicator froze :)

Airspeed is mostly, although not entirely, important as a proxy for angle of attack.

> everyone how AoA is one of the most important metrics wrt flying an airplane

Pretty difficult argument to make when most planes do not have AoA sensors!

Almost all new airplanes have them. The Skyhawk got one as standard equipment in 2016.

Here’s a thesis on the topic: https://kb.osu.edu/bitstream/handle/1811/68634/1/AbramsHonor...

Sure, but almost no-one flies brand new airplanes. What percentage of GA flights do you guess contain an AoA display? And the argument's supposed to be that they're a most important indicator for airplanes?

And yet it’s absolutely true.

Note that many light aircraft have stall warning horns, which are just AoA sensors that only detect when the AoA exceeds a certain amount. In many other planes, you can detect excessive AoA by sound or feel.

The "sensor" is often the pilot.

I believe he means jets, which I believe all of them do.

He does not. Let's read as it is written.

But you don't need a dedicated sensor for it. The artificial horizon does that job too.

Edit: To be more clear, the artificial horizon combined with other sources of information available to the pilot (airspeed, stall indicator) provide the pilot with the same level of situational awareness as knowing his/her AOA outright from a dedicated sensor.

The artificial horizon doesn't indicate angle of attack. Please see this comment: https://news.ycombinator.com/item?id=19638250

Or consider an aerobatic airplane at the first vertical point in a loop. The airplane's pitch is 90° but the angle of attack is less than 10°.

You're right but the vast majority of the time the difference between the AI and the AOA is close to zero (plus whatever the angle of the wing is relative to the fuselage which is fixed and should be pretty small anyway). The times when the difference between the AOA and the AI is very different than zero in a bad way the little stall switch on the wing and some sort of chime/buzzer will be more than happy to let the pilot know.

My point is that the AI combined with the little stall switch on the wing seem to be good enough for the overwhelming majority of cases. Add in airspeed and you're more than covered.

We keep trying to add systems to cover whatever remaining edge cases we find but those systems have failure modes too so we take three steps forward and two steps back. These systems pick up the slack when the pilot might be inattentive but the increased complexity and cognitive load poses other risks.

Loss of control during approach and landing because of poor airspeed control kills pilots every year. It's important enough that the EAA has a dedicated "loss of control initiative" to try to figure out what to do. See for example https://www.eaa.org/eaa/news-and-publications/eaa-news-and-a...

Climbs after takeoff are a situation where they will significantly differ, and also one of the times when this information is more important.

Attitude indicator? Look out the window! Stall horn? Just feel the pre-stall rumbling through the controls. You can fly a plane with no instruments. None of them are strictly necessary, but it’s a good idea to have them.

It's really amazing how AA and airspeed are so... difficult for pilots to perceive that they're totally reliant on the sensors to tell them.

Speed I guess makes sense, just amazing how a pilot so easily can slow a plane too much, change the angle too much and have no idea what is up.

Angle of Attack doesn't refer to the angle of the plane relative to the horizon but the angle of the plane relative to airflow.

To illustrate further using the most extreme example I can think of - flying a loop:

The angle of attack will remain in its usual very narrow range at any point when you fly a loop. The critical or stalling angle of attack is typically around 15° - 20° for many airfoils (so your AoA should be less than that). Note that if you fly upside down (negative-G capable aerobatics airplane) you will still have a positive AoA, just on the other side of the wing.

> If the airplane you’re flying is equipped with an angle-of-attack indicator, it will illustrate an abstract concept that pilots learn to recite during ground school but rarely fully comprehend—which is that an airplane can stall in “any attitude” and at “any airspeed.” For me, it took stalling on the back side of a loop (while pointed straight down at a high power setting) for that idea to sink in.

https://www.aopa.org/news-and-media/all-news/2013/may/pilot/...

Eh, that would generally be considered a negative angle of attack, since we arbitrarily label one side of the wing as positive direction.

As a private pilot who has done a bit of aerobatics, I disagree. It still is a positive AoA. Using a negative would be in conflict with what you would mean when you mean.. well, a negative AoA, for example when people ask questions such as this one: https://aviation.stackexchange.com/questions/51477/do-negati...

I would use AoA in relation to the direction of flight, not the up or down orientation of the airplane. Even upside down I still fly quite normally (in straight & level upside-down flight), and the lift created is positive: Up/down in relation to the earth don't change, so the direction of lift does not change. In this context it would be a bit confusing to use negative numbers as you suggest, IMO.

However, since you can set your relative coordinate system and context however you like this is a topic we could discuss this ad infinitum without convincing one another if we insist on our respective PoVs. When faced with the linked question you would have some more explaining to do though, because there are negative AoA and no upside-down flight (now imagine adding that to the question to make it even more confusing...) :)

The linked example provides a good reason you should use negative AoA. It would be ridiculous to have a discontinuity in AoA jumping -5 to +5 just because the direction of lift changes slightly.

The direction of lift does not change at all. It still is UP. That is the point.

No, as long as you are pulling positive G's, your angle of attack is positive. It has nothing to do with pitch angle.

Edit: Oh you meant flying upside down, not doing a loop?

Yeah, negative AoA is not the same as positive AoA. It makes sense to distinguish them because, unless the wing uses a symmetric airfoil, it will behave differently in the two cases. Given that you also calculate the lift vector as a function of AoA, and for anything to make sense the lift vector is antiparallel in the two cases, the AoA would also be negative.

Yeah, flying upside down.

It depends very much on the circumstances. In most slow GA airplanes (and yes, the Vision jet is a slow plane), one can perceive both the airspeed and the angle of attack by the control feel ("mushy" vs "firm"). Many older airplanes (for example the fastest GA piston twin, the Aerostar 601P) did not have even stall indicators. That airplane used to scare the crap out of me during a takeoff on a hot day. Everything had to be done just right ...

You mostly need a stall indicator or an AoA during tricky procedures like really short field operations (which in the Aerostar world translates to "just don't"). Jets are a slightly different story however I do not believe the VJ gets fast enough for that to matter.

Under VFR what you say is true, under Instrument conditions the AoA indication is critically important. Once you’re in a cloud, you literally can’t tell which way is up without instruments and the AoA is an essential part.

Not every airliner even shows AOA, it is just used as input for stall warnings and the flight recorder.

It should be possible to fly safely using pitch and air speed indicators. And even a loss of air speed should be survivable by using known good configurations.

Most GA pilots are not reliant on sensors, we learn to fly with no instruments. Most airline pilots fly on instruments and almost never without. My instructor told me in the school I will have to learn to know the speed without looking anywhere but feeling it from the way the plane flies. He is right. Flying too much on instruments is taking away the feel of the plane, there are skills you can train and hone and there are skills you loose due to lack of use.

Helped on a ferry flight of an aerobatics plane from Denver area to Chicago area. The front half I was in had a ball compass, altimeter, and airspeed indicator. It has been a long time since I've been in something without any electronics. The dumb and ugly was strong when I started to chase the compass rather than look outside. :)

Is this the same manufacturer as the sensors on the MAX?

Good question. I don't think so but I can't tell for sure. They make pitot-static probes for Boeing and Airbus jets but I don't know if their AoA sensors are used in commercial aircraft.

Exactly. The AD says that the plane can be flown once the part is replaced.

Reply on news.ycombinator.com