Hacker News Reader

Comment by SmellTheGlove | Hacker News Reader

It's amazing to me that modern aircraft engineering is so good that an engine can lose its fan blade entirely at 35000 feet and the plane can still continue on for 2 hours and safely land. An uncontained failure like this is the worst kind, and I know it could have gone worse, but it didn't and that amazes me.

I wonder if the cowling contained the initial disintegration but was removed in the process. I'm not an engineer, but from the way that looks and the fact that nothing flew off and did further damage makes me at least wonder if that's what happened.

View on HN

The A380 has four engines in total, and can easily tolerate the loss of just one in normal flight. Adding more engines has been a basic way to increase reliability since the beginning of the field; it's not just modern aircraft engineering. There's no other part of the airplane behind the engines either, so nothing to be hit by debris.

What amazes me more is that the newest twinjets are now considered reliable enough to fly six hours away from the nearest landing strip -- that's how far they can coast on the single remaining engine if necessary!

It's not really coasting. A transport category aircraft is required to have enough power to continue climbing if it loses an engine, so it must also have enough power to maintain level flight indefinitely (barring fuel consumption, oil leaks, or mechanical damage from an engine failure)

I meant to say "cruise", if that makes more sense.

If it is any consolation, I knew what you meant. Otherwise, you'd probably have said something like glide and the glide ratio of the big planes isn't very good.

Out of curiosity, I just checked for the A380 and the best guess is 15:1, so it does glide better than I expected, but still not that far.

Google tells me that it'd glide ~130 miles if it had a starting altitude of 45,000'. That's better than I expected.

According to this:

https://en.wikipedia.org/wiki/Gliding_flight#Aircraft_.28.22...

The 747-200 also has a glide ratio of 15:1. I can't tell if that's amazing, good, or average.

A sailplane can have up to 60:1! That's pretty impressive!

According to this: A "clean A320" has 17:1. Neat.

https://aviation.stackexchange.com/questions/14814/could-a-c...

Something I never knew was this:

>As someone said above, you have to remember, pretty well all descents are done with the thrust at idle.

So I imagine the glide ratio is pretty important for all-purposes of flight and not just when engines go out.

You descend at just about the stall speed. It's pretty close to the slowest the individual plane can go before losing lift. I believe the term for coming in with no lift, purely gliding, is called dead stick.

A Cessna 172 is worse, for comparison. I believe they are just 9:1. I guess it improves if you stop the prop entirely. I assume it is less drag.

best glide (converting altitude into the maximum distance) speed is quite a bit higher than the stall speed and found at the lift/drag max. minimum sink (converting altitude into the maximum time) is usually found about halfway between best glide speed and stall speed.

http://learntoflylv.com/drag%20curve.gif

Obligatory Gimli Glider: https://www.damninteresting.com/the-gimli-glider/ http://www.gimlicommunityweb.com/history/gimli_glider.php .

Also (older accident):

https://en.wikipedia.org/wiki/British_Airways_Flight_9

>Ladies and gentlemen, this is your captain speaking. We have a small problem. All four engines have stopped. We are doing our damnedest to get them going again. I trust you are not in too much distress.

That must surely top the list of most awesome pilot announcements (and, simultaneously, best British understatement).

Finesse gives you two things:

Quite directly, from ground altitude, a range within which to look for a place to land. As you mentioned the impressive finesse of gliders gives you more room, but gliders won’t generally be flying as high as cruising powered planes.

Combined with speed, a timeframe within which you have to observe, orient, decide and act. Here, both top speed and stall speed give you some hints, and within which there’s a best speed where finesse is optimal. You can imagine how it’s very different between an A320 and a glider, the former for which time runs quickly, you have to decide where to go in as few minutes as possible, the latter for which decision time brought back at a human thinking time scale makes it less stressful but you have to keep cautious of your goal for a longer stretch of time.

Obviously that’s without counting on soaring, which A320s can hardly benefit from (I suppose they can but would largely not move up and certainly not turn since it would would incur too huge drag).

Fineness, I assume? Although it'd be important to have finesse, too!

Nope, finesse[0] (Other uses) is a synonym for glide ratio. I’m just used to the former.

[0]: https://en.m.wikipedia.org/wiki/Finesse_(disambiguation)

> ~130 miles

That's statute miles (or 111 nautical miles or 205 km). It's quite impressive, really.

It's been done in real life for the A330. 120 km from a starting altitude of 10 km to Lajes Airbase in the Azores. https://en.wikipedia.org/wiki/Air_Transat_Flight_236

>A transport category aircraft is required to have enough power to continue climbing

Can a plane takeoff with just one engine ?

Yes. Twinjet airliners must be able to take off if one engine fails at or after V1.

Which, incidentally, means that twins are quite powerful - a twin has 100% more engines (thus, power available) at takeoff than needed, while a quad has 33% more.

Thus a twin has 50% more power than a similar 4 engine. At least conceptually-ish.

Actually, adding engines increases resilience, or the ability to tolerate failure.

The extra complexity of extra engines actually reduces reliability, since there are more components that can fail.

As the saying goes: If you have an engine failure in a twin, the remaining engine has enough power to take you to the scene of the accident.

> There's no other part of the airplane behind the engines either, so nothing to be hit by debris.

The fans spin very fast, and will fly out radially, not just fall back. They'll slice right through anything they contact. The Airbus flight was very, very lucky.

The engines are designed so that a detached fan blade stays contained within the cowling. Here's a video of a fan blade containment test on a Rolls Royce Trent 500: https://youtu.be/RIOWjhjC1jM

I guess they use really, really precise timing and explosives to knock one blade loose in order to do the containment tests. They want it to come loose at a specific point in the revolution.

The documentary didn't say why they wanted it to be at such a specific place, only that they did. I've long since forgotten the margin of error, but it was very small.

You probably noticed that the individual blade is painted a different color and is different colors. My recollection is that they use the multiple colors because they do very high definition recording and the colors are all in different areas of the spectrum, so they can observe it better as the thing tends to be quickly obscured by fire and smoke.

I probably do watch too many documentaries.

> I probably do watch too many documentaries.

You are not alone. I actually watch more and more documentaries than movies or TV series. Reality is even stranger than fiction often.

It has been great to see the proliferation of documentaries on the web. Streaming services have been a great thing for me.

I watch documentaries pretty much to the exclusion of all other genres, except American football. Even if I'm not actively paying attention, I'll frequently have a documentary playing in the background.

It's not a scholastic pursuit. It is purely entertainment. I watch them because I enjoy them. Learning is incidental and not the objective.

I like to queue up playlists of documentaries and let them play through. The 'net has been excellent at facilitating this.

One reason I love documentaries is the people in them are often truly intelligent and knowledgeable, where as tv and the movies are full of people who are supposed to be that but don't talk and act like it. And even if the people in the documentaries are not particularly intelligent, at least they are real human beings confronting real situations, instead of phony imitations.

> The engines are designed so that a detached fan blade stays contained within the cowling.

Yes, but uncontained failures can and do happen. Qantas Flight 32 is an example of an uncontained engine failure on an A380, though with a Rolls Royce engine instead of Engine Alliance (GE + P&W). [0]

I do think they're very lucky that this uncontained engine failure didn't cause more damage. QF32 was heavily damaged by debris coming out of the engine, so the fact that this seems to be relatively contained to the engine itself is fortunate.

[0] https://en.wikipedia.org/wiki/Qantas_Flight_32

As far as I have picked up on that subject, passenger aircraft have multiple layers of protection against blades separating from the engine. The cowling is supposed to contain it. If that does not work, the fuselage in the area of the wings is hardened to prevent the separating part(s) from entering the passenger compartment and doing damage there. This is typically implemented using a Kevlar lining on the hull.

The Quantas flight is remarkable in that engine parts were not contained successfully and managed to damage critical systems inside the wing not related to the engine itself. Some reports say that the flight crew was overwhelmed by over 50 individual failure reports resulting from this damage. I was actually quite surprised when I learned that it was so easy to damage that many major systems with a seemingly small puncture in a wing.

Keep in mind that the Quantas failure, while horrible, resulted in zero fatalities.

Very remarkable indeed, considering the damage to the landing gear, flaps, fuel tanks, and engine controls. The airplane at the time it landed was way out into the experimental flying zone.

When people talk of pilot-less airliners I have to think of incidents like this. Pilots do make mistakes and people die as a result, but when equipment fails and nobody dies because a pilot does what is needed to land safely despite the issue, it rarely makes news. Sometimes people on board don't even know.

> Sometimes people on board don't even know.

That's the ideal general case. You really don't want hundreds of panicked people trapped in a broken tin can thousands of feet above ground while you're trying to handle the situation.

At least you (as a pilot) are behind a solid locked door in that scenario.

However, that brings up another question I've been wondering about: Could the pax/cabin crew of an airplane collaborate in bringing a jet down, by e.g. all together going to the very front of the plane (near the cockpit door), moving the trim up, then all running to the very end of the plane, bringing it into an uncontrollable stall?

I haven't really seen a satisfactory answer to that question. See Quora and Stackexchange (where I attempted an answer):

https://www.quora.com/Would-a-passenger-jet-stall-if-all-the...

https://aviation.stackexchange.com/questions/2507/can-passen...

Whislt technically correct (the best kind), attributing this to adequate engineering given scope of damage is difficult. I've reviewed this incident myself several times -- not a statement of my qualifications (nil), only familiarity -- and tremendous credit must be given the flight crew which included the presence of a check team (there were five pilots in the cockpit).

One engine failed completely, two others were seriously degraded, aircraft control was minimal, fuel could not be dumped (leading to an overweight laading). The impression from one documentary was that the first officer spent much of his time overwhelmed with system reports which had to be cleared from his displays. Systems and information design and flows seem a net negative for this incident -- there is such a thing as too much information.

Overall, many reserve margins were exceedingly tight.

Your comment seems to me to grossly understate the severity, and overly credit Airbus and Rolls Royce, in this case

I'm fairly sure there are many other things that'll kill me first (hello, lack of exercise), but I do wonder where a detached turbine blade might cut through the fuselage and any passengers contained therein, and thus try to avoid the rows parallel to the engine.

However, I'm wondering - would the blade go straight lateral? Or, relieved from having to pull the whole damn plane forward, would it fly forward and hit the fuselage in front of the engine? Or, slowed down by the air resistance, would it hit the fuselage behind the engine?

Best, I suppose, to sit up front in first class...

The blades will come out radially.

If you're concerned about which seats are safer, the ones in steerage in the back are the safest. They tend to arrive last at the scene of the accident :-)

Good shoes can save your life, because you can walk out over hot, burning wreckage.

Wear cotton, not synthetic, clothes, and you'll have less severe burn injuries.

And, of course, pay attention to where the exits are and the flight crew's safety instructions. They really do know what they're talking about, and they're well trained.

Oh, I never take off my shoes until a few thousand ft AGL - I figure at that point it doesn't matter much anymore :-)

> They really do know what they're talking about, and they're well trained.

I wonder, btw, whether people on Saudia 163 would have survived if they had stormed the emergency exits after landing against (presumably) the orders of the crew...

https://en.wikipedia.org/wiki/Saudia_Flight_163

Speaking of flying out radially, at least here in Norway, commercial prop planes had a broad red line painted on the interior cabin wall where prop blades would come knocking in case they were somehow torn from the hub.

I have never heard of an incident where prop debris penetrated the cabin, but avoided the seats on that row if at all possible, just in case.

http://gawker.com/passenger-injured-by-broken-air-canada-pro...

https://en.wikipedia.org/wiki/Reeve_Aleutian_Airways_Flight_...

You are assuming the blades separated from the hub. If the fan separated as a unit (someone suggested the shaft may have failed), I would guess it would go forwards.

Update: in 1981, the fan of the rear engine on an L-1011 separated, and chewed its way through the inlet duct ("The 'S' duct was damaged internally from next to the fan blades forward about 16 feet") before exiting the fuselage sideways, sending shrapnel into a lavatory and doing serious damage to the flight controls: "The displacement of the fan module in the course of the engine failure sequence caused loss of hydraulic systems A, B, and D and jammed the captain's and first officer's rudder pedals in the neutral position."

The aircraft landed safely, but with a little more damage, the airplane would have been uncontrollable.

https://www.faa.gov/about/initiatives/maintenance_hf/library...

An engine-out situation does get interesting though when the aircraft is low and slow, such as when taking off and landing. That's where good training is needed to recover quickly and successfully.

What amazes me more is that the newest twinjets are now considered reliable enough to fly six hours away from the nearest landing strip

Yes, there's also a short video about that: https://www.youtube.com/watch?v=HSxSgbNQi-g

It's very interesting since ETOPS[1] affected the financial bets that Boeing made on 787 (point-to-point) vs Airbus with the A380 (hub-and-spoke).

[1] https://en.wikipedia.org/wiki/ETOPS

I wasn't really commenting on adding engines - I know that concept has been around a long time. This wasn't just an engine going down, this was catastrophic failure, and my amazement is in that no other part of the aircraft was affected outside of that one engine.

> There's no other part of the airplane behind the engines either, so nothing to be hit by debris.

No, but there's a wing and the fuselage nearby. Again, I'm not an engineer, but I'm assuming that in this type of failure, the part of the engine that's spinning might want to come laterally (is that the right word?), and that seems like it could be trouble. I'm amazed that it didn't.

> What amazes me more is that the newest twinjets are now considered reliable enough to fly six hours away from the nearest landing strip -- that's how far they can coast on the single remaining engine if necessary!

Totally amazing. I'm not sure, but is that only the A350XWB for now, or have others been certified past ETOPS-360? It's basically certified to the design limit of the aircraft, so I'm not sure what else has been designed for that. Even 5+ hours is nuts though.

Somewhat related, but check out the Chilean regs on flying from the mainland to IPC/PPT. As far as I know, they still have the one aircraft rule.

The real concern here is not loss of thrust, but rather, what other parts of the airframe are seriously damaged or destroyed by the debris. In the accident record I believe there is precedent for separated engine parts seriously damaging wing and tail surfaces, causing at least partial loss of control.

Agreed. When I saw the amount of damage, but little on the wing edge (there may be damage on the underside), I assumed the engine maintained confinement long enough to dissipate a lot of energy. If the cowl was damaged during this, then it may have then come off due to aerodymic stress. Either that or they were very lucky. From the photo I saw, it looks like they lost the forward cowl, fan and maybe a compressor stage. If I had to guess, failure of the low-pressure shaft.

> I assumed the engine maintained confinement long enough to dissipate a lot of energy.

Not unless the cowling was armored, which it wasn't at least back in the 80's. That thin sheet of aluminum won't even be noticed by the high strength fan spinning at tremendous speeds.

> Not unless the cowling was armored, which it wasn't at least back in the 80's.

AFAIK part of engine qualification is proving that the nacelle will retain a fan blade breaking off at full throttle. There are remarkable youtube videos of these qualification tests.

The entire fan is another matter though. I assume the nacelle will offer some resistance, but not contain it entirely.

And the turbines as well, they don't even attempt to retain those.

> qualification tests

When did that become a requirement?

> And the turbines as well, they don't even attempt to retain those.

I knew about the turbine blades, I just assumed the fan blades were as uncontainable. I think my information is out of date. Thanks for the correction.

I found it:

http://federal.elaws.us/cfr/title14.part33.section33.94

Looks like it became a requirement in 1984, after I left Boeing. So that's why I didn't know about it.

Thanks for taking the time to dig your way to the facts, and then report back with what you found.

As it turns out, I learned something as well. Contrary to what I said, that same passage requires failing turbine blades to be retained as well. However, they seem to be drawing the line at complete fan or compressor assemblies or turbine disks.

I suggest you watch the blade separation test videos. The entire point of the cowling is to contain fan blade loss.

The main purpose of the cowling is to improve the efficiency of the fan by preventing the air from being slung off radially.

https://en.wikipedia.org/wiki/Ducted_fan

Agreed, but failure containment is another of its purposes, at least on contemporary turbofan engines.

Did you ignore what I said and not watch the blade release tests?

Indeed. https://youtu.be/736O4Hz4Nk4 is a bit long, but they detonate some charges on a blade, watch it take out all the others and keep it all inside.

The thing about it is, that these sorts of things get designed because of evolution. Some engine fails in some way, and it gets redesigned. Planes fail, ditto.

And after >100 years of that, we have pretty safe planes.

What most amazes me is that despite the scientific knowledge we "have", we can't design a new engine from first principles directly. We do it through trial and error. Lots of hilarious/horrible error.

It is kind of nest how they have progressed, as well. It is sometimes little things that make a big difference. It iterative process, I suppose.

Rolls Royce prides themselves on hiring interns. The little cone piece in front of the jet engine had issues with icing up. This had plagued jet engines for years.

They tried all sorts of ways to fix it, including complex methods of supplying heat.

One of their interns just happened to hear about this problem and suggested they simply make it out of rubber. Sure enough, it worked and the problem with that pointy bit freezing was solved.

I'm sure it has a formal name and that it is probably not pointy bit or little cone piece. I don't actually know anything about jet engines, I just watched a documentary about Rolls Royce's jet manufacturing.

Anyhow, that wasn't something that could probably have been designed with first principles. Someone, pretty much in passing, pretty much had to have that creative spark at just the right time and with just the right information.

It's things like that story that make me think humans are going to be in the loop for a long time still and that AI taking over the world is still a long ways away.

>I'm sure it has a formal name

Spinner.

> It's things like that story that make me think humans are going to be in the loop for a long time still and that AI taking over the world is still a long ways away.

I've found that the world continues to turn contingent on tremendous amount of what I call "dark wisdom" - crucial pieces of knowledge not found in any server, spec sheet, or manual. Usually inside the noggin of one or a few people. It just seems like a consequence of maintaining complex systems. In order for AI to fully take over, it would have to control and account for every corner case of every system, from mining the ore, to fabbing chips, to replacing wires chewed by rodents.

There was also this much discussed incident

http://avherald.com/h?article=46a8f70a

where they just shut one engine down above canada and continued on 3 accross the atlantic to kuwait (they couldn't safely - with enough fuel reserve for multiple approaches and a diversion - reach dubai)

I am fairly sure this was a contained failure. The reports I saw (unofficial) said no blades detached or hit the aircraft.

Reply on news.ycombinator.com