And potentially a lot more, if they’d not made the runway, given how close residential areas are to that end of the airport.
Question for the pilots - if they’d lost power at high altitude, would it have been feasible to make for another airport (Luton or Gatwick?) where the possibility of falling short would have meant crashing on open ground instead of housing?
Well, the Gimli Glider made something like 45 miles after the engines cut out. They were getting about a 11:1 glide ratio. It would really depend on the altitude at which you lose power.
El_kabong was expressing concern that this might be a flaw with the B-777. The airplane problem you mentioned in Canada involved an airbus, in other words, a totally different model of airplane. It is also believe wake turbulence may have been the cause there, whereas Heathrow incident is very much undetermined at this point. Wake turbulence accidents are a problem of maintaining proper distance (or not), they are not an intrinsic flaw in the aircraft.
I’ll start by saying that my normal source of information for this sort of thing is not available due to a “busy server.” I wonder what they’re all talking about?
Also, “power loss” can mean a few different things. Engine power, hydraulic power, electrical power, etc. A report of “power loss” by the pilot may or may not mean a complete double engine failure.
For the rest of this post I’ll assume it was a loss of engine power, but it’s important to remember that it may not be.
GLIDING AND THE APPROACH
An airliner actually has a very good glide ratio. Something around 17:1 according to this wiki article. This is significantly better than light piston engined aircraft. In fact an airliner is basically gliding from top of descent down to the latter stages of the approach.
One of the worst times to have a double engine failure (if that is what happened) is on late final approach. One thing a jet is bad at is flying slowly. To try and counter this a lot of high lift devices are desgned into the aircraft which are deployed progresively during the approach. These are the flaps and slats which increase the wing’s lift and delay the onset of an aerodynamic stall (where the wing literally stops flying.) As well as increasing lift they increase drag which means that on approach you have a relatively high power setting to produce the thrust necessary to counter all of the extra drag. Lose that thrust and you start going down much faster than you’d like.
So, IF they did lose power on approach then that is bad time for it to happen and they did well to get away with it.
ENGINE POWER LOSS
A double engine failure can really only be a few things. The most obvious is fuel starvation. This may occur because the fuel isn’t in the aeroplane or it is there but for some reason it is not available (pump failures, blockages, etc.) If it’s simply a matter of running out of fuel it may or may not be the pilot’s fault.
There was an aircraft in Australia recently that came very close to running out of fuel (one engine had stopped and the other would’ve a short time after landing) but the fuel gauges showed ample fuel in the tanks and the company’s fuel checking system did not adequately pick up errors in the fuel gauge system. This then was a company system error rather than a pilot error.
Other reasons both engines may fail include flying through a large flock of birds or other contaminates/debris such as volcanic ash, coincidental unrelated mechanical failures (rare but not unheard of), and inadvertantly shutting down the good engine after a bad engine has suffered a fault.
PRAISEWORTHY OR NOT?
We don’t know yet. I’ll reserve any comment until I hear more about what happened. It may be that they did well to achieve a good outcome from a problem of their own making, or they may have achieved a good outcome from a problem not of their own making, or they may have made a minor problem worse than it should’ve been. We just don’t know.
I’m not one to wait two years for the official report before making any comment, but it’s premature to start judging when the only reports come from the media which is notorious for poor reporting of aviation events.
Absolutely. The higher you are when you lose power, the more options you have. Generally you’d aim for the longest runway within gliding distance. You then aim to land about 1/3 down the runway. As you get closer you can work on moving your aiming point closer to the start of the runway, but always preferring to land a little long and running through the end fence at low speed rather than landing short and running through the approach end fence at high speed.
Ah yes, I know the one you mean. It’s still down. Scariest damn site on the Web.
You heard wrong. Airliners make decent gliders, better than some training gliders, although not on the level of competition gliders.
You don’t measure glider efficiency by stall speed but rather by glide ratio. Most of the time I see a figure of 15:1 or 17:1 for big airliners like 707, 737, 747… I couldn’t quickly locate the specific figure for the 777 but I would expect it to be in that ballpark.
In contrast, the training glider I took my one glider flight in had, if I recall, a 12:1 glide ratio. It may have flown slower and had a lower stall speed, but it didn’t glide as well as jumbo jet.
Engine failure on approach is bad because you’re low to the ground, but pilots are trained to, as much as practical, plan the approach so that if you lose power you can still get at least close to the runway (which training may have saved the day at Heathrow yesterday). Take off is bad because you’re low and slow and most runways have stuff at some point past the end that would be painful to crash into. Engine failure on landing is bad, but you were planning to land anyway and the airplane is probably already configured for landing so, ideally, you just modify your plans. Engine failure on take off is worse because you have to completely change your plans, and reconfigure the airplane.
Actually, that’s an illustration of why self-tutoring on a flight sim doesn’t make one a pilot. There’s no “contrary” about it. Your speed is controlled by the attitude of the aircraft, which is typically controlled by the elevator. Once the speed is set more throttle means more altitude, less throttle means less altitude. Elevator is your “how fast?” control and throttle your “how high?” control.
You either slip the plane (not normally done with the big jets but they are perfectly capable of the maneuver) or else it just takes you a longer distance to stop.
Airliners actually are pretty decent gliders.
A nice summation of the problem. The lower you are when the power goes away the more difficult the problem to resolve.
As Gorsnak points out, at least one ‘heavy’ demonstrated an 11:1 glide ratio. That’s pretty good. Landing speeds are high, but if you can make it to a suitable runway (or dry lake bed for that matter) that shouldn’t be a problem. The actual gliding performance is good.
Energy management. Up-elevator increases the angle of attack of the wings, which results in more lift. So one would think that the elevators are used to make the aircraft go up. And they are. But increasing the lift by increasing AOA also increases drag, so it takes more power to climb. If you don’t add power then up-elevator will cause your airspeed to decrease. If you are flying straight and level and add power without changing the elevator you’re creating more lift my the increased airflow over the wings, so you climb.
It’s a balancing act, and all of the controls are interrelated. In a car if you want to turn you turn the wheel, In an airplane you turn the stick or yoke. This banks the wings, so your lift vector is tilted. So you need to add up-elevator to maintain altitude. Pulling back increases drag, so you need to add power if you want to maintain your airspeed. And you need to use the rudder to keep the airplane from ‘skidding’.
Gliding is often referred to ‘trading altitude for airspeed’.
I’ve worked on the FMS (Flight Management System) software on a coupla big jets, and their time/distance/glide/decel calcuation capabilities are pretty sophisticated. I recall that on one FMS (predating the 777) our goal was to calculate the entire descent, deceleration*, maneuver-to-runway, glide and landing such that it could occur with throttles at idle** on normal flights (to save fuel). We tried to predict that exact point where the throttles were closed (called TODD – Top Of Descent Deceleration) so that the runway could be reached with no further application of power.
In the situation you describe, loss of one (or more) engine would prompt the computer to quickly present the pilots with a list of “reachable” airports, and all they would need to do is select the most desirable. Airports were chosen (by the computer) based on their distance (can we make it to the airport, accounting for winds, glide speeds, necessary turns, etc.), runway lengths, (how much distance will we need, based on our predicted weight when we get there), runway suitability (can we use that runway at our predicted landing weight), and various other parameters that I won’t bore you with. Once a choice is made, the FMS busies itself re-arranging fuel within the plane’s tanks***, and dumping it overboard if necessary.
So to answer the question, assuming the computers determined Luton or Gatwick were reachable, they could easily have gone there. Descent path calculations are accurate enough there’s little likelihood of the plane ending up in a field somewhere (assuming any airport was reachable).
*Deceleration was always a concern from high altitude descents, because there are speed limits up there too.
**I’m not a big-iron pilot so I’m not sure what idle is exactly. The engines may be running at some minimum speed to power the various systems in the plane.
***In cruise, some large aircraft pump fuel to aft tanks to reduce drag. (Conventional aircraft have slightly lower drag with an aft center of gravity). Prior to landing, this fuel is either consumed, or moved forward. Thus the need to hurriedly move this fuel during an emergency descent.
[QUOTE=Malacandra]
[quote=Broomstick]
Also, airliners have axillary power sources for just that sort of occasion so the pilots can maintain enough control to get on the ground.
“Wingpits”, actually. 
Uh, yeah - the RAT is an auxillary power source. Were you expecting the Energizer Bunny or something? The RAT is designed to deploy automatically to give you the minimum power needed to control/guide the airplane to the ground safely. Which it did in the case of the Gimli Glider. I’m not sure why you find that confusing.
There are a lot of variables to consider, such as how high the failure occurred, glide capabilities of the airplane in question, distance to other fields, winds and weather, whether or not control systems are affected… As a general rule, “closest field” is usually the safest option though there are exceptions. For a big airliner pavement is also usually the best option because they tend to sink into bare ground and then parts start coming off. Of course, yesterday’s landing demonstrates that even if you do land on turf and the plane starts to disassemble you can still wind up walking away from the mess with no or minimal injuries.
Certainly, the higher the altitude at which failure occurs the more time you have to analyze the situation and make a decision, which usually improves the odds of a good outcome.
What, you’ve never had to sit in the middle seat in the very last row of coach, the row right next to the lavatories where the seats won’t recline? That’s the armpit of the plane…
I was coming in to say that same thing.
It seems pretty obvious that the aircrew achieved something better than the bare minimum for a landing - everyone seems to have been able to leave under their own power.
I want to believe that the air crew didn’t create the situation, but I have to admit that I’m enough of an NTSB junkie to be aware that the single most common cause of incident is “operator error.” Having said that, even if the cause of the loss of power was operator error it doesn’t mean that it’s impossible for the air crew to have subsequently reacted in a brilliant manner after that.
I can echo this. My dad was giving me a lesson in a Cessna 172 once. We had the plane perfectly trimmed for straight-and-level at 80 knots, took our hands off the control yokes. I pushed the throttle in and we started to climb, pulled it back out until we leveled off, out some more and we descended, then add power back to straight-and-level; hands-off and 80 knots the whole time. It was an enlightening lesson. I mean, stepping on the gas makes you go faster, right? Not so.
I’ve got Flight Simulator 10 at home, I should try that and see how accurately it responds. (Might be tricky; trimming the plane with no sensory feedback is hard to do.)
I thought that was the colon of the plane.
Yes I have sat there.
Engage autopilot and let it trim the plane, and then you can turn George off and play once you’re settled. Works fine on FS2004.
Flight simulators can educate as often as deceive; until I played MSCFS it had never occurred to me that throttle controlled height and elevators controlled speed, though I’ve been conversant with the theory of flight since quite a young age (RAF background, donchaknow).
The question of what controls airspeed is an old one in flying. It seems obvious to many that it should be the throttle, but it doesn’t work that way in most cockpits. A good clue comes from looking at a glider, which typically has no engine and no throttle: it’s quite easy for the glider pilot to control airspeed, but he has little direct control over altitude. I’ve heard it said that a hot-air balloon is an example of the opposite: it has a throttle (the burner) which gives good control over altitude, but there is absolutely no control over (horizontal) airspeed, which is basically always zero.
The enlightened view is that the elevator controls angle of attack and the throttle controls the rate at which energy is added. You can take your energy in a rather wide range of airspeeds and altitudes - you make the choice by controlling angle of attack.
Indeed, the aircraft did, too. There was some damage to the nose, due to the fact that the nose gear didn’t lock down and thus collapsed. But repairs were done, fuel was added and the plane flew out of Gimli two days later.
Wow; consider my ignorance fought. I’d often heard that big airliners were terrible gliders and drop out of the sky like bricks if they lose all thrust. Glad to learn differently. They’re really pretty amazing machines.
One question that I have sometimes thought of when looking at airports, and again on looking at the aerial photograph of the crash site:
Wouldn’t you get a slight additional margin of safety by building the physical runway right to the perimeter fence, instead of short of it (according to the photograph, the touchdown point was about 200 m short of the runway)?
Of course pilots would still need to consider the ‘logical’ beginning of the runway to be where the ‘physical’ runway begins now (i.e. the not-to-be-used part would need to be conspiciously marked as such and touching down on it would need to be punished as a severe violation of safety rules), but wouldn’t that help planes who through some technical emergency touch down short?
This concept exists, and indeed is rather common. It’s known as a displaced threshold. In the link’s image, an aircraft landing left to right is not supposed to touch down short of the solid white line (at the end of the black section) that crosses the runway.
