News for hackers. Stuff that matters.
Re:Explanation of the Asiana SF crash by a former UAL captain by SilasX on Thursday July 11, 2013 @01:54AM (#6024663)
To save you from looking up the more obscure acronyms like I did (summaries where given are my own words):

RTO: Rejected take-off, when you have to abort a take-off (though it seems in the context of the post, the exercise described involved an edge case when you should go ahead with the take-off)

AGL: Above ground level

CAVOK: Ceiling and Visibility are OK: no clouds < 5000 ft AGL, with >10 km visibility

VOR approach: VHF omnidirectional radio range: navigating based on a radio signal, considered "non-precision"

IAF: Initial approach fix http://en.wikipedia.org/wiki/Initial_approach_fix

LNAV/VNAV: lateral/vertical navigation http://en.wikipedia.org/wiki/LNAV

FAF: Final approach fix

UNSAT: unsatisfactory rating (paperwork)

FCOM: Flight crew operating manual

CRM/CLR: Cockpit resource management, a set of NASA-originating practices to limit and contain human error http://en.wikipedia.org/wiki/Crew_Resource_Management

Re:Explanation of the Asiana SF crash by a former UAL captain by rdl on Thursday July 11, 2013 @03:54AM (#6024935)
I was unclear on his example -- their manual says to ignore "aft cargo door" cautions during takeoff? Are those known to be false-positives, or just not that important? I thought the cargo doors were structural parts of the aircraft, and thus having them open would cause both structural and aerodynamic problems. Maybe they just continue takeoff and land again?
Re:Explanation of the Asiana SF crash by a former UAL captain by omegant on Thursday July 11, 2013 @05:49AM (#6025210)
In modern planes there is a filter on the warning computer that doesn´t allow to show certain non vital failures between 80 knots (or 100 depending the plane) and V1. For example an engine fire is going to trigger the alarm, but not an engine EGT (exhaust gas temp) over limit. But the list of warnings not shown, and the ones shown but that must be ignored, are different from plane to plane. That´s why this instructor was doing this test, to make them learn that the 777 had a different list and requirements from the previous one. Mind you that after you´ve been in 2 or 3 different planes it´s easy to react automatically but with the previous plane list.

It´s known that pilots are oversensitive to warnings during take off. A rejected take off is a dangerous maneuver that must be avoided if unnecessary, and is trained in each simulator session. This is more important in wide body airliners like the 747, 777, A340 or A380, as their take off weight is so high that a RTO (rejected take off) will blowout some tires (of the main landing gear) due to the stress or the temperature increase inside the tire created by the brakes. Once the plane is stopped, it has to wait isolated at the taxiway because there are temperature plugs at the tires that may blow out, nobody can get close during that time. If the weight is too high they will even burn as the brakes will be white hot. Look at youtube for a certification RTO of a big plane to see it, or some brake energy absorption test at the laboratory, to see how much energy they have to mannage.

In this case(the simulator story) a after cargo door is one of those warnings that you have to ignore as the plane will flight without problems, then you may turn around and make a landing after dumping some fuel(to decrease the weight).

Hope this helps

Edit: some clarification

Below 80 it is possible to RTO without problems as the speed is still low. Once you reach V1 you will NEVER RTO! as this speed is calculated taking in to account the runway length, and in such a manner that if you RTO at V1, the runway remaining once you have stopped is just a couple dozen of meters. The V1 is so high because that way, when you become airborne, you´ll have a better climb gradient even with an engine fail. All this stuff is calculated with an engine fail, after all it is the most important and probable fail you will encounter (the reverse deployment on take-off is much more severe but thankfully very very rare), and that will affect your climb capability.

Edit 2: added-improved some punctuation.

Re:Explanation of the Asiana SF crash by a former UAL captain by lifeisstillgood on Thursday July 11, 2013 @08:39AM (#6025649)
http://www.youtube.com/watch?v=wXpjBxD0Rhg

An example of the braking test - stopping a 288 tonne 777 travelling at 200mph. The disc brakes were on fire, and they then had to taxi around on wheels at a temp of 3000 deg for five minutes to simulate the time it took for fire trucks to arrive.

Cool

Re:Explanation of the Asiana SF crash by a former UAL captain by trimbo on Thursday July 11, 2013 @12:22PM (#6027024)
The entire 5 hour documentary I think this is from about the 777 is awesome and definitely worth the watch.

It was on PBS, called "21st Century Jet"

Re:Explanation of the Asiana SF crash by a former UAL captain by rdl on Thursday July 11, 2013 @02:35PM (#6028134)
As amazing as the 777 is, the 787 seems to be even better (and there's talk of updating the 777 with 787 technology).

I really hope next time they go shopping for Air Force One they take budget into consideration, and the end of the cold war, and ETOPS, and seriously consider a twin engine (ideally a 787 with a reduction in passengers, but probably they'd go with a 777 and continue taking way too many people.) Outside of cargo and the military, four engine passenger jets seem to be a dying species. (and obviously the A380 and other airbus products aren't in the running for AF1. I still haven't gotten a chance to fly on an A380.)

Re:Explanation of the Asiana SF crash by a former UAL captain by rdl on Thursday July 11, 2013 @03:46PM (#6028671)
Re:Explanation of the Asiana SF crash by a former UAL captain by pja on Thursday July 11, 2013 @04:44AM (#6025055)
My understanding is as follows: If you're on the runway then once you go past V1 during take-off you have no option but to get airborne regardless of what else is happening: there isn't enough runway left to stop the aircraft with the brakes + reverse thrust. (You calculate V1 for every take-off)

The pilots he was training were coming from an aircraft with a V1 around 80 knots fully laden. V1 for a 737-400 is around 120 knots so if he triggers an aft cargo door warning at 100 knots then the pilots should abort the takeoff.

He was doing this deliberately to emphasise to the trainees this particular difference between their old aircraft and the new one.

This interpretation is at least consistent, but I'm not sure whether it's correct or not.

Re:Explanation of the Asiana SF crash by a former UAL captain by rdl on Thursday July 11, 2013 @04:52AM (#6025074)
This seems slightly off somehow -- they were used to an aircraft where V1 was 80, were moving to one where it was 120. Event happened at 100; they were initially RTOing but were corrected by the grapevine to not RTO. That seems backwards from the behavior my mental model (which is what you stated as well) would want.
Re:Explanation of the Asiana SF crash by a former UAL captain by pja on Thursday July 11, 2013 @05:13AM (#6025129)
You're right. Given that V1 for a 777 is ~ 150 knots according to the online checklists I found I'm not sure what's going on either! If V1 if 150 knots then a warning at 100 knots should trigger an abort surely?

Perhaps the point is that an Aft Cargo Bay warning isn't sufficient reason to risk aborting the take-off (which presumably carries it's own risks). So the warning alarm would go off & the pilots would abort because they were used to having such minor alarms being disabled about 80knots on their previous aircraft, therefore any alarm at speeds > 80 knots but < V1 is would previously have been cause to abort but on the new aircraft this was no longer the case.

Makes sense? (Although personally I'd have thought that an open cargo bay door would be reason enough to abort a take-off - so that bit doesn't make sense to me at all.)

Re:Explanation of the Asiana SF crash by a former UAL captain by Piskvorrr on Thursday July 11, 2013 @05:39AM (#6025191)
Surprisingly enough, for most issues (barring such situations as the wings literally falling off ;)), it's safer to get in the air to buy yourself some time for analyzing the situation and deciding (even if it's "go around for an emergency landing") than trying to decide during takeoff.
Re:Explanation of the Asiana SF crash by a former UAL captain by mikeash on Thursday July 11, 2013 @11:06AM (#6026456)
It's not too unusual for pilots to turn a minor problem into a major one by trying to fix the minor problem too aggressively. A cargo door opening would be a good example of this. It doesn't threaten the aircraft, so it should be ignored until the situation is calm enough that your attention isn't needed elsewhere. Prioritizing problems like this is a big part of pilot training.

It's a little bit like spilling your coffee while merging onto an interstate in a car. Slamming on your brakes so you can take care of your coffee will likely make things worse. Instead, merge, get stabilized, then figure out what to do with the coffee.

Re:Explanation of the Asiana SF crash by a former UAL captain by SilasX on Thursday July 11, 2013 @11:20AM (#6026561)
>I thought the cargo doors were structural parts of the aircraft, and thus having them open would cause both structural and aerodynamic problems.

I don't know about the other stuff, but I do know that the structure of an aircraft must not be designed to depend on something like a door being closed. That is, if the door is carrying load when closed, "you're doing it wrong". The door frame might carry load, but that wouldn't care whether the door is open.

Generally, something on an aircraft is either intended to carry structural load, or it's designed for such loads to bypass it entirely.

It would definitely hurt the aerodynamics for a door to fly open, but that would probably be a minor concern relative to the other hazards it introduces.

Re:Explanation of the Asiana SF crash by a former UAL captain by omegant on Thursday July 11, 2013 @07:28AM (#6025425)
As you took the time to explain all the acronyms I´ll explain a bit more about the context of the post.

Commercial pilots begin flying small planes like Cessna 172, then they fly a small multiengine with instruments, they also take lessons in small simulators. After this you must have a good (or very good) manual handling in visual and instrumental flight to obtain the commercial pilot license. Once you get more flight experience flying small cargo or aerotaxi, you get to flight medium turboprop commuters like a ATR-72, and then if you are lucky you get to fly in a mayor airline with 737 or airbus 320 before you can fly the big birds like 777, 747, 767, A340 or A380.

Once you are on a big plane, flying manually is almost optional as you can have the autopilot engaged from shortly after take off on climb, to 1000´ to the touchdown. Usually pilots enjoy flying manually, we do it whenever we have the opportunity, of course if there is a lot of traffic congestion and/or bad weather we keep the autopilot connected because the workload builds quickly (other traffics, avoiding thunderstorms, checking all the instruments for an auto-land, control instructions, etc..). In my case I was lucky to be able to start flying a B737 when I had only 200 total flight hours. My instructor was afraid that I will not be able to deliver due to my lack of experience, but I had no problem doing so (thankfully!)

It seems that in the Korean instruction method, obedience and memorization is more important than situation awareness and flying the plane. When this instructor tried to teach and measure the performance using the proper and internationally accepted method he was chased and punished by the airline establishment.

Just to give you some flavor of the difference between flying manually or with the autopilot I have this analogy, is not perfect but it transmits a bit the sensation between one and the other: Flying manually is somehow similar to how you ride a bicycle, you have to "feel" it, if you are going to be able to take a curve, or to brake on time, etc.. once you are comfortable with that, you don´t even have to think consciously any more about it. Going with autopilot, well is like going on rails, if the rail is not broken you know where is going and how, you only have to advance how to brake and accelerate (I Know that driving trains is more than that, I just mean a straight simple railway). Autopilots are a huge advantage in aviation, it allows to fly in much more congested airspace with much more safety, and with all the computerized systems, we are relieved from a lot of tedious monitoring.

It seems that the Korean pilots (this is the OP´s version, I don´t really know) have very limited experience and proficiency flying manually. He is surprised because most of the senior pilots were ex-military with experience in F4, F5, F16 fighters, etc.. I am not very surprised. I have flown with ex-military (Spanish) and the kind of plane they had flied previously wasn´t an indicator of dexterity. After all, combat flights can be very technical and planed (all points prearranged and studied). Thankfully I never had an experience nearly as bad as the ones he describes.

The flight data released by the NTSB shows a high approach (above the normal flight slope), first when you flight a ILS is very unusual to have a situation like that (constant 3º descend) because everything is controlled step by step 12 miles out of the airport. Second, if you are used (and like) to flight manually, disconnecting the A/P(Auto pilot) is what you are going to do (and enjoy), probably also the A/T (auto-throttle for boeing, the thrust levers move; or auto-thrust for airbus, levers doesn´t move), above all this, if you have normal skills flying you are perfectly capable of flying an approach from above or from below, is not a big deal, if you see you can´t do it you just turn around and try it again. But a pilot that is afraid and insecure of his capabilities, will try to do it with the A/P or with all the automatic systems available till the last moment.

Technically if you are above the normal slope, you have an excess of energy (altitude), to correct it you set the throttles manually back, that way you are able to descend faster (higher vertical speed) and capture the glide slope "visually" with the VASIS (white and red lights that let´s you know when you are high or low on the glide).

But it´s dangerous to land with the engines at idle, it is recommended that at 1000´ you must be on the glide slope and fully configured to land (gear down, full flap, speed at Vref[Vref is the reference speed for landing, is the minimum speed you will flight]). If not you must go around and try again.

DISCLAIMER, this is my opinion based on previous similar accident reports, the OP thoughts, and my own experience. The investigation may tell a very different situation.

It seems that the crew at the Asiana flight, kept trying to set the A/P or at least the flight guidance + A/T to descend and catch the visual path, this is very dangerous at low altitude, because you loose the situation awareness, AND you put all your attention inside. In that moment it becomes more important for you to see what´s happening at the panel (which is a big mistake), than to what´s happening to the aircraft trajectory (which is of course much more important). I can imagine a kind of fight against the flight guidance buttons to try to descend the plane. Once they got it on descend mode (I don´t know the 777 modes, but it´s not that important in this situation, in fact you must avoid them and flight manually), they realized that they had to slow the descend for the landing. But once you are outside the loop (flight awareness) it´s very difficult to get in again.

Once they started pulling back, the speed decreased, usually (this is surely what they expected), the autothrust would have engaged to keep the speed at the normal Vref (landing speed). But as they had touched the modes, the A/T didn´t engage and the speed started decreasing towards the stall. Then it becomes a matter of managing your energy, if you want to keep altitude you are going to end stalling, and if you want to keep speed you´ll descend too soon and too fast. Surely they started noticing something was odd (too high nose, too low trajectory) but when they finally realized that they where going to crash and/or stall, and/or fall short (not necessarily knowing why), they set the throttles to full power and tried to go around, but big engines need a lot of time to go from idle to full power (usually you engage them slightly in advance to when you are going to need them), and it was too late to avoid the crash.

This may happen even to good manual flyers, if you are not comfortable flying like that it´s much more probable to have an accident.

Edit: added some explanation to acronyms, and improved a bit the text.

Re:Explanation of the Asiana SF crash by a former UAL captain by drivers99 on Thursday July 11, 2013 @07:50AM (#6025491)
I still had to figure out a few acronyms so I will include them here:

Took me a while, but I figured A/P = autopilot, and A/T = auto throttle?

Vref = Landing reference speed or threshold crossing speed. Huge table of "V speeds" here including extra info about Vref: http://en.wikipedia.org/wiki/V_speeds

Re:Explanation of the Asiana SF crash by a former UAL captain by omegant on Thursday July 11, 2013 @08:15AM (#6025563)
Ups sorry! I ended typing in autopilot..
Re:Explanation of the Asiana SF crash by a former UAL captain by losvedir on Thursday July 11, 2013 @01:58PM (#6027778)
Wow, utterly fascinating! Thanks for this wonderful comment.
Re:Explanation of the Asiana SF crash by a former UAL captain by raverbashing on Thursday July 11, 2013 @03:27AM (#6024878)
Yes

RTO is basically when you abort the takeoff by not taking-off, that is, hitting the brakes before the aircraft leaves the ground

The other option is taking off and returning to the airport, when a problem happens later

Re:Explanation of the Asiana SF crash by a former UAL captain by yuvadam on Thursday July 11, 2013 @03:39AM (#6024903)
Hitting the brakes before the aircraft reaches V1, specifically, which is the takeoff decision speed [1]

[1] - https://en.wikipedia.org/wiki/V_speeds#V1_definitions

Re:Explanation of the Asiana SF crash by a former UAL captain by grarok on Thursday July 11, 2013 @04:10AM (#6024978)
Thank you for the acronyms, but it's still missing a lot of info. The forum post is shit-garbage. Can you please rewrite the post and make it readable by a non-asshole? UAL, Standards Captain, -400, Asiana, "right seat, left seat", "upgraded by phenomenal growth" - jesus christ what the fuck is this.
Re:Explanation of the Asiana SF crash by a former UAL captain by rdl on Thursday July 11, 2013 @04:22AM (#6025009)
It's not unreasonable that in a forum for airline pilots they use the technical jargon of those airlines to discuss things among themselves. If you want to read something for the mass market, read a mass-market publication. It wasn't much more jargon-filled than your average HN post, just different industry jargon.
Re:Explanation of the Asiana SF crash by a former UAL captain by brettnak on Thursday July 11, 2013 @04:29AM (#6025025)
This should help a little:

-400: Likely the 747-400. One of the variants of the 747

UAL: United Airlines

Asiana: This is the airline of the plane that crashed at SFO.

Right Seat: First Officer / Co-pilot's seat.

Left Seat: Captain / Pilot's seat.

Standards Captain: Probably means he was one of the people who would have final say in whether or not a pilot was certified to fly a specific aircraft. I believe there are two different standards, one for pilot and one for copilot, but I'm not sure.

Upgraded by Phenomenal Growth: The airline grew so fast they didn't have enough pilots so they had to promote copilots to pilots faster than they were probably qualified.

*edited to correct information based on the comment below.

Re:Explanation of the Asiana SF crash by a former UAL captain by rdl on Thursday July 11, 2013 @04:32AM (#6025031)
-400 is almost certainly a Boeing 747-400, i.e. the model of 747 flown by virtually everyone now. There isn't a 777-400 yet.
Re:Explanation of the Asiana SF crash by a former UAL captain by ceejayoz on Thursday July 11, 2013 @08:18AM (#6025568)
Pilots would have as much difficulty with stuff like "AWS EBS latency" that gets thrown around here.