Twenty passengers on missing flight 370 worked for Freescale Semiconductors

[kizzap]

The best solution I can think up (following Occam's Razor) is that there was an electrical fire in the plane, which originated in the radio and somehow spread to the radar. The pilots lose the radio first, try to make an emergency landing at the closest airport they can, but then lose the radar, which puts them completely blind, and they aim wantonly for land, not realising that they are slowly turning south, and further and further away from land.

-kizzap

[cimmo]

The best solution I can think up (following Occam's Razor) is that there was an electrical fire in the plane, which originated in the radio and somehow spread to the radar. The pilots lose the radio first, try to make an emergency landing at the closest airport they can, but then lose the radar, which puts them completely blind, and they aim wantonly for land, not realising that they are slowly turning south, and further and further away from land.

-kizzap

Even very hi-tech aircraft such as a B777 have a completely old school 'wet compass' that has no integration with any avionics. Apart from the lighting, it requires no electricity and it cannot fail.

All pilots know how to read a compass.
It is that little box at the very top of this image:
http://farm3.static.flickr.com/2031/2264742995_e10991959b.jpg

[ivan747]

The best solution I can think up (following Occam's Razor) is that there was an electrical fire in the plane, which originated in the radio and somehow spread to the radar. The pilots lose the radio first, try to make an emergency landing at the closest airport they can, but then lose the radar, which puts them completely blind, and they aim wantonly for land, not realising that they are slowly turning south, and further and further away from land.

-kizzap

Even very hi-tech aircraft such as a B777 have a completely old school 'wet compass' that has no integration with any avionics. Apart from the lighting, it requires no electricity and it cannot fail.

All pilots know how to read a compass.
It is that little box at the very top of this image:
http://farm3.static.flickr.com/2031/2264742995_e10991959b.jpg

I think it also has backup insturments independent of the computers like an altimeter, a heading indicator, attitude indicator and an airspeed indicator. But I don't know if the 777 has all of these. It should, I guess.

[AG6QR]

The best solution I can think up (following Occam's Razor) is that there was an electrical fire in the plane, which originated in the radio and somehow spread to the radar. The pilots lose the radio first, try to make an emergency landing at the closest airport they can, but then lose the radar, which puts them completely blind, and they aim wantonly for land, not realising that they are slowly turning south, and further and further away from land.

-kizzap

Even very hi-tech aircraft such as a B777 have a completely old school 'wet compass' that has no integration with any avionics. Apart from the lighting, it requires no electricity and it cannot fail.

All pilots know how to read a compass.
It is that little box at the very top of this image:
http://farm3.static.flickr.com/2031/2264742995_e10991959b.jpg

That's all true.  Furthermore, stars were visible.  As a passenger, even without a compass I can tell roughly what direction a plane is flying at night by watching the constellations.  Most pilots I know have some rudimentary familiarity with the constellations, and even if these pilots didn't, somebody on that flight would have had similar skills, at least good enough to point them roughly northwestward toward the Asian continent instead of out toward open ocean.  They wouldn't have remained completely disoriented for multiple hours with no clue what direction they were flying.  My conclusion is that either the crew was incapacitated or they were deliberately trying to head for open ocean.

Occam's razor is a valuable tool, but it doesn't give one good criteria for determining which is the simplest explanation.  I've heard it invoked to support the "lunatic suicidal pilot" option, and the "electrical fire, causing pilots to shut down systems, and eventually causing breach of fuselage and crew incapacitation" theory. 

The attractive thing about the "lunatic pilot" theory is that, once you've decided that the person at the controls is a crazed and irrational lunatic, you don't necessarily have to come up with answers to question like the motivation for going out to the open ocean instead of simply pointing the nose down and crashing immediately.

On the other hand, if a fire similar to the Egyptair MS-667 occurred during flight, the logical pilot response would be to immediately shut down the electrical systems and turn toward the nearest runway.  If it burned through the oxygen hose and the fuselage like the Egyptair fire did, the pilots would not have supplementary oxygen available, and would very quickly be incapacitated.  The loss of pressure could extinguish the fire, leaving the plane to fly uncontrolled until it ran out of fuel. 

As far as I can tell, there's nothing about either theory that is contradicted by evidence.  I think Occam's razor points toward the latter option, but I've heard others who believe it points toward the first.  They have a different view of which scenario is "simplest" than I have.

Regardless, Occam's razor is only something to fall back on when you have little evidence to choose among alternatives.  As more evidence is uncovered, it may be that the more complex alternative is the only one which fully explains the new facts.

[cimmo]

Even very hi-tech aircraft such as a B777 have a completely old school 'wet compass' that has no integration with any avionics. Apart from the lighting, it requires no electricity and it cannot fail.

All pilots know how to read a compass.
It is that little box at the very top of this image:
http://farm3.static.flickr.com/2031/2264742995_e10991959b.jpg

I think it also has backup insturments independent of the computers like an altimeter, a heading indicator, attitude indicator and an airspeed indicator. But I don't know if the 777 has all of these. It should, I guess.

Indeed the B777 does have backup primary flight instruments (airspeed, altitude, attitude). These do have independant sensor inputs and dedicated displays. However, in the B777 they still require electricity to function and although they do operate off the standby bus (which is powered by the battery and ram air turbine), if the worst case does happen with the power supply (bus short or open circuit, eg; fire in equipment bay), then these displays will cease to function, unlike the wet compass.

These standby instruments are the three small vertically orientated LCD displays between the 2nd and the 3rd large display (counting from the left): (top to bottom - attitude, airspeed, altimeter)
http://wl15www815.webland.ch/travelinfos/images/cockpit/b777panorama/cockpit_B777.jpg

Depending on customer preference, an all-in-one standby instrument is available (same basic location):
http://farm5.staticflickr.com/4135/4792497493_0073782c52_z.jpg

[EEVblog]

The best solution I can think up (following Occam's Razor) is that there was an electrical fire in the plane, which originated in the radio and somehow spread to the radar.

Doesn't need to start in the radio. Just needs to be a sufficiently serious incident for them to be tied up responding to it.
A pilots priority is always in this order:
1) Aviate (i.e fly the plane and keep it in the air)
2) Navigate (so you don't crash into a mountain etc, and are in a position to know and point your plane toward an alternate airport or suitable landing site like the hudson river)
3) Communicate

People wrongly think the first thing a pilot does when a big incident happens is to radio what's hapepned. Nope, that's the bottom of the list!

[EEVblog]

Indeed the B777 does have backup primary flight instruments (airspeed, altitude, attitude). These do have independant sensor inputs and dedicated displays. However, in the B777 they still require electricity to function and although they do operate off the standby bus (which is powered by the battery and ram air turbine), if the worst case does happen with the power supply (bus short or open circuit, eg; fire in equipment bay), then these displays will cease to function, unlike the wet compass.

I fear that some unprecedented incident has occurred that took out all the electrics and they were helpless to do anything at all until they eventually crashed. Like a QF32 incident, but not as lucky.
The very last thing that will fail is the engines, they have their own fuel supply and don't need plane power to operate, they just keep going and nothing will stop them.
Although if that was the case, then I'm surprised it apparently flew so far? As I presume if there was no computer left to drive the hydraulics to the control surfaces to keep it level, then a plane would crash pretty quickly?

[pickle9000]

My thought was a fire, pilots disable power to the appropriate system. No damage to the flight control system or autopilot, massive amounts of smoke in the plane, pilots fly by autopilot for safety (including the final turn), oxygen runs out (crew and passengers die) before any further actions can be taken, plane on autopilot. The fire goes out because of the pilots original action.

[apelly]

Damn, Australia... AC/DC *and* the CVR. Next, you're going to be telling me that Australia gave the world the SCRAM jet or box wine.

Or wi-fi

[cimmo]

....Although if that was the case, then I'm surprised it apparently flew so far? As I presume if there was no computer left to drive the hydraulics to the control surfaces to keep it level, then a plane would crash pretty quickly?

With the exception of a late generation fighter jet that is designed with negative stabilty, all aircraft generally exhibit positive natural stability. Once trimmed for a certain speed then any aircraft can be flown hands off for a long time. Yes, any gusts will disturb the flightpath, but the aircraft will naturally seek to return to the pre-gust condition. Pitch upsets will result in what is called a phugoid, a long period vertical oscillation - eg, if the gust results in a climb, then the airspeed will decrease. As the airspeed decreases then the aerodynamic download generated by the tailplane reduces, lowering the nose. As the nose lowers, then the climb becomes a descent, airspeed increases, download on the tail increases, nose comes up. Repeat many times. Eventually this damps down to essentially level flight again.

Roll stability is provided by the dihedral of the wings (upward bending). Many aircraft are fairly neutral in roll stability and a large gust upset can lead to a roll angle that won't naturally restore (typically > 5 or 10 degrees), leading to what is known as 'the graveyard spiral'.  But since a B777 has a very flexible wing and a lot of dihedral in cruise, I suspect that a B777 has quite a strong dihedral effect/roll stability.

Swept wing aircraft are also prone to a coupled yaw/roll oscillation called "Dutch Roll" and often require active yaw damping. But more modern designs usually exhibit this behaviour a lot less than earlier aircraft. Dutch roll in a modern airliner is usually more of an inconvenience for pax than a serious risk of becoming divergent.

So, as long as the aircraft remained clear of thunderstorms/convective turbulence and jetstream related clear air turbulence, I can conceive it remaining in the air for many hours with no active control, although it will wander up and down and left and right. Essentially the natural aerodynamic forces of a well designed aircraft provide a negative feedback loop.

This scenario may even explain how it was initially heading west and ended up going south - a slight fuel imbalance may have resulted in a very slight left roll bias and gentle turn to the south. Speculation of course.

I hope this isn't too long an explanation of basic aerodynamics.

[EEVblog]

So, as long as the aircraft remained clear of thunderstorms/convective turbulence and jetstream related clear air turbulence, I can conceive it remaining in the air for many hours with no active control, although it will wander up and down and left and right. Essentially the natural aerodynamic forces of a well designed aircraft provide a negative feedback loop.

Nice, thanks.

[CaptnYellowShirt]

Roll stability is provided by the dihedral of the wings (upward bending). Many aircraft are fairly neutral in roll stability and a large gust upset can lead to a roll angle that won't naturally restore (typically > 5 or 10 degrees), leading to what is known as 'the graveyard spiral'.  But since a B777 has a very flexible wing and a lot of dihedral in cruise, I suspect that a B777 has quite a strong dihedral effect/roll stability.

Airplanes have no natural roll stability with respect to the ground. Dihedral only functions in what is called 'uncoordinated flight' which deals with the orientation of the aircraft with respect to the relative air stream. Dihedral is added to make airplanes 'easier' to fly, not to give them self-righting ability.

Aerodynamics for Navy Aviators (PDF)
http://www.faa.gov/regulations_policies/handbooks_manuals/aviation/media/00-80T-80.pdf
pg 300: "If the airplane is unrestrained and sideslip is allowed, the affect of the directional stability and dihedral effect can be appreciated."

Also "graveyard spirals" are a pilot induced phenomena. They have nothing to do with the stability analysis of an aircraft. Ironically, they occur precisely because pilots are not trained to understand and respond to the inherent lack of roll stability found in all airplanes.

[cimmo]

Roll stability is provided by the dihedral of the wings (upward bending). Many aircraft are fairly neutral in roll stability and a large gust upset can lead to a roll angle that won't naturally restore (typically > 5 or 10 degrees), leading to what is known as 'the graveyard spiral'.  But since a B777 has a very flexible wing and a lot of dihedral in cruise, I suspect that a B777 has quite a strong dihedral effect/roll stability.

Airplanes have no natural roll stability with respect to the ground. Dihedral only functions in what is called 'uncoordinated flight' which deals with the orientation of the aircraft with respect to the relative air stream. Dihedral is added to make airplanes 'easier' to fly, not to give them self-righting ability.

Aerodynamics for Navy Aviators (PDF)
http://www.faa.gov/regulations_policies/handbooks_manuals/aviation/media/00-80T-80.pdf
pg 300: "If the airplane is unrestrained and sideslip is allowed, the affect of the directional stability and dihedral effect can be appreciated."

Also "graveyard spirals" are a pilot induced phenomena. They have nothing to do with the stability analysis of an aircraft. Ironically, they occur precisely because pilots are not trained to understand and respond to the inherent lack of roll stability found in all airplanes.

"Airplanes have no natural roll stability with respect to the ground"
I believe that to be an unnecessary complication. And unless there is some factor causing a sustained sideslip it is not relevant to this general discussion.

Page 295 of that document you quoted says this on the topic of lateral stability (roll stability):
"The principal surface contributing to the lateral stability of an airplane is the wing. The effect of the geometric dihedral of a wing is a powerful contribution to lateral stability."

Yes, in practice there is always roll/yaw interaction, especially with a swept wing configuration. I  was trying  to distill the topic from a few hundred pages to a para or two and I still assert that basically my simplified explanation is correct.

Your quote of page 301 is in the section marked 'control' not 'stability'. The two concepts are highly interactive, but in this case we are talking about an aircraft that has no control inputs, therefore only the intrinsic stability is relevant.

"Also "graveyard spirals" are a pilot induced phenomena."
They can and usually are exacerbated by pilots pitching up before rolling out,  but even with no pilot interaction an aircraft will exhibit spiral divergence:

Page 299:
"Spiral divergence will exist when the static directional stability is very large when compared with the dihedral effect. The character of spiral divergence is by no means violent, The airplane, when disturbed from the equilibrium of level flight, begins a slow spiral which gradually increases to a spiral dive. "

Further on the topic: (page 300)
"The choice is then the least of three evils. Directional divergence cannot be tolerated, Dutch roll is objectionable, and spiral divergence is tolerable if the rate of divergence is low. For this reason the dihedral effect should be no more than that required for satisfactory lateral stability. If the static directional stability is made adequate to prevent objectionable Dutch roll, this will automatically be sufficient to prevent directional divergence, Since the more important handling qualities are a result of high static directional stability and minimum necessary dihedral effect, most airplanes demonstrate a mild spiral tendency. As previously mentioned, a weak spiral tendency is of little concern to the pilot and certainly preferable to Dutch roll."

I would add here a note that a 'weak spiral tendency' is fine if there is someone/something in the loop to rectify it. In practice very small bank angles as typically encountered in mostly smooth air will not diverge - the sideslip will be dealt with by the fin (this will lead to dutch roll), but a larger gust will pass the neutral point and cause spiral divergence.

Anyway, the point I was trying to make is that any aircraft's intrinsic natural positive stability (unstable fighters excepted) makes the immediate "stall, spin, crash, burn, die" event unlikely - especially for an airliner where maneuverability is not a high priority and even more so in smooth air.

[CaptnYellowShirt]

Anyway, the point I was trying to make is that any aircraft's intrinsic natural positive stability (unstable fighters excepted) makes the immediate "stall, spin, crash, burn, die" event unlikely - especially for an airliner where maneuverability is not a high priority and even more so in smooth air.

You're correct with respect to an aircraft's pitch attitude. Most airplanes (including most fighters) exhibit a natural pitch stability which extends to all modes of flight -- normal, stalled, and spinning. Some modern designs (namely UVA's) are unstable in pitch which makes them more maneuverable and more efficient to operate but more difficult to control. You either need special training (e.g. the Wright Flyer) or a sophisticated computer aid to fly a pitch-unstable airplane.

But, again, airplanes have no stability in roll. Some can be highly damped, which is what you are talking about. But no airplane is self-righting in roll.  A control loop must be added to keep the plane's wings level -- the pilot looking at the horizon, the pilot looking at the gyroscopic instruments, or an autopilot.

But back to the point about MH370 zombie'ing across the Indian Ocean -- no pilots and a broken control system. 1) The engines would probably keep turning. I'm not familiar with the engines on MH370, but in general turbines are designed to fail in the 'on' direction under the notion that more thrust is better than no thrust. 2) The plane probably wouldn't stall. As we've both noted, most airplanes have a natural pitch stability. So baring some extreme event -- like a thunderstorm -- I don't think it would stall. But... 3) Airplanes have no natural roll stability. How far a plane could get without functional feedback (e.g. autopilot or roll control surfaces) is a matter of statistics. But for light aircraft flying in convective conditions accident studies have shown it to be 178 seconds (oddly specific huh?).

How far could MH370 in such a condition? Its anybody's guess, but its not on the order of 'hours'. Personally, I'd put the mark at ones to tens of minutes. If it flew for any longer than this, something on board the plane was keeping the wings level.

[pickle9000]

Perhaps the stupidest questions of the day.

- If a pilot turns on the autopilot, is that a single button affair? Could it be invoked by anyone to maintain level flight?
- Is the Autopilot the same as the Flight Management System (I'm thinking no)?

[CaptnYellowShirt]

Perhaps the stupidest questions of the day.

- If a pilot turns on the autopilot, is that a single button affair? Could it be invoked by anyone to maintain level flight?
- Is the Autopilot the same as the Flight Management System (I'm thinking no)?

I've asked my fair share.

To get a good answer you'd need to ask someone that's type rated for the B777. However, certain modes of lighter aircraft autopilots can be switched on with 'a single button'. Typically, this would be a wing leveler mode (see previous post), or magnetic bearing track mode -- anything more involves programming the nav computer (e.g. FSM).

But how all that stuff is hooked up -- how its integrated, what bus its powered from, what shelf or box its in, what wire looms it shares -- is beyond my pay grade. These are exactly the kinds of questions you get when you sit for a type rating: If Bus B trips, will the autopilot kick off? Will it default into a certain mode or just shut down completely? If you suspect a fire in the _____, should you rely on the autopilot? Why or why not?

That being said, every large plane has *a very specific* enumerated list of actions to take during an emergency of any type. These actions are rehearsed by the crew to be performed at a moment's notice. There's no room for 'creativity' in these events. Pilots are trained to fly the checklist, and baring some crazy crazy event -- like the second coming of jesus on this particular flight -- that's what these pilots did. If you really wanted to know if the autopilot was (or should have been) switched on an the event of an emergency, you'd need to see what the emergency checklist specifies.

[Wytnucls]

Not really, the autopilot is always used in an emergency (if it is available), except in some specific situations, like TCAS and GPWS warnings.
They would have been on autopilot in the cruise and it would have been left on throughout the emergency they encountered, the pilot taking over controls only if the autopilot drops out.

[pickle9000]

Ok, so you'd use the autopilot to give yourself time to evaluate the problem (if possible).

So how would you evacuate smoke from the aircraft? Switch off potential causes, lower the altitude so the air is breathable when the masks run out.  Then what? Not like you can open a window.

I was in a training situation years ago with full face protection and an air tank, the room was so full of smoke I couldn't see my hand 2 inches from the mask. Even so I was barely able to see for the next 10 minutes (and I had full face coverage), I could imagine without eye protection you'd be in real trouble.

[Wytnucls]

There is a lengthy smoke checklist. If the smoke can't be brought under control, the aircraft is taken to 10,000 feet, depressurized and one of the cockpit side windows is opened.
The cockpit oxygen mask is a full face mask to protect against fumes and smoke.

[CaptnYellowShirt]

Today (in the US), NPR aired an interview with Richard Branson and the CEO of Delta:

http://www.npr.org/2014/04/03/298779764/transatlantic-duo-looks-into-the-future-of-flight?ft=1&f=1006

They were asked about transponders and ELT's -- specifically why after 9/11 nothing was done to improve the technology of either. They didn't mention ADS-B, but I think the intent of the questions was more about how we can loose a plane with our current state of technology.

Does anyone know if MH370's ELT was set up to broadcast on 121.5Mhz or did they dump that completely in favor of 406?

[Wytnucls]

All three:
121.5MHz (Civil)
243MHz (Military)
406MHz (COSPAS-SARSAT satellite system)

[cimmo]

You're correct with respect to an aircraft's pitch attitude. Most airplanes (including most fighters) exhibit a natural pitch stability which extends to all modes of flight -- normal, stalled, and spinning. Some modern designs (namely UVA's) are unstable in pitch which makes them more maneuverable and more efficient to operate but more difficult to control. You either need special training (e.g. the Wright Flyer) or a sophisticated computer aid to fly a pitch-unstable airplane.

But, again, airplanes have no stability in roll. Some can be highly damped, which is what you are talking about. But no airplane is self-righting in roll.  A control loop must be added to keep the plane's wings level -- the pilot looking at the horizon, the pilot looking at the gyroscopic instruments, or an autopilot.

But back to the point about MH370 zombie'ing across the Indian Ocean -- no pilots and a broken control system. 1) The engines would probably keep turning. I'm not familiar with the engines on MH370, but in general turbines are designed to fail in the 'on' direction under the notion that more thrust is better than no thrust. 2) The plane probably wouldn't stall. As we've both noted, most airplanes have a natural pitch stability. So baring some extreme event -- like a thunderstorm -- I don't think it would stall. But... 3) Airplanes have no natural roll stability. How far a plane could get without functional feedback (e.g. autopilot or roll control surfaces) is a matter of statistics. But for light aircraft flying in convective conditions accident studies have shown it to be 178 seconds (oddly specific huh?).

How far could MH370 in such a condition? Its anybody's guess, but its not on the order of 'hours'. Personally, I'd put the mark at ones to tens of minutes. If it flew for any longer than this, something on board the plane was keeping the wings level.

1: In  general, modern FBW fighter aircraft (gen4 or 5) ARE pitch unstable and cannot be flown without artificial stability.
2: "But, again, airplanes have no stability in roll." Some do not, but this statement is NOT correct for all aircraft, especially a transport category airliner. Why do you insist with this false information?

Quote from Flight Training handbook:
" If a momentary gust of wind forces one wing of the airplane to rise and the other to lower, the airplane will bank. When the airplane is banked without turning, it tends to sideslip or slide downward toward the lowered wing. Since the wings have dihedral, the air strikes the low wing at much greater angle of attack than the high wing. This increases the lift on the low wing and decreases lift on the high wing, and tends to restore the airplane to its original lateral attitude (wings level); that is, the angle of attack and lift on the two wings are again equal.
   The effect of dihedral, then, is to produce a rolling moment tending to return the airplane to a laterally balanced flight condition when a sideslip occurs.
   The restoring force may move the low wing up too far, so that the opposite wing now goes down. If so, the process will be repeated, decreasing with each lateral oscillation until a balance for wings level flight is finally reached. "

From Aeronautics Learning Laboratory website:
"The dihedral angle is the angle that each wing makes with the horizontal. The purpose of dihedral is to improve lateral stability. If a disturbance causes one wing to drop, the unbalanced force produces a sideslip in the direction of the downgoing wing. This will, in effect, cause a flow of air in the opposite direction to the slip. This flow of air will strike the lower wing at a greater angle of attack than it strikes the upper wing. The lower wing will thus receive more lift and the airplane will roll back into its proper position."


"But for light aircraft flying in convective conditions accident studies have shown it to be 178 seconds "
I have no doubt about that at all - turbulence and a low inertia aircraft is not conducive to stable flight.
But this is not the same as a 200 tonne airliner with a very high dihedral angle (have you seen how high the wingtips are in the cruise of a B777?) potentially flying in glass smooth air (speculation, but possible).

All I am saying is that technically it IS possible for an aircraft to fly hands off for some considerable period of time - the larger the aircraft the better as it will have higher moments of inertia. The length of time would be  determined primarily by atmospheric conditions. If those conditions ARE conducive, then minor upsets will NOT result in the aircraft entering a spiral dive.

[cimmo]

Perhaps the stupidest questions of the day.

- If a pilot turns on the autopilot, is that a single button affair? Could it be invoked by anyone to maintain level flight?
- Is the Autopilot the same as the Flight Management System (I'm thinking no)?

1: Yes and no. You will get stable flight with a single button push to engage the autopilot. This switch is on the Mode Control Panel (on the glareshield located between and in front of the pilots) -  but if the aircraft was in a turn then this bank angle will be maintained and if wings were level then it will maintain the current heading.
The vertical speed of the aircraft at moment of engagement will be maintained.

To get proper level flight you must also press the Altitude Hold button and if not already wings level, you press the Heading Hold button. Then rotate the dials associated with those functions to fly the aircraft to a different heading or altitude - together with FLCH button and Autothrottle (A/T arm).

2: The FMS and the AP are highly integrated, but you can operate the AP directly through the glareshield controller. The FMS keyboard/displays are located near the pilots knees and with a lot more button pushing you can program or modify an entire flight, both course over the ground (lateral navigation) and altitude variations (vertical navigation).

If this is set correctly, then the AP can be coupled to the FMS by pushing the LNAV and VNAV buttons on the mode control panel as desired.

source: B777 Flight Manual

In these images you can see the mode control panel:
http://www.flightglobal.com/assets/getasset.aspx?itemid=33315
https://www.metabunk.org/data/MetaMirrorCache/761fa24f12c6e98528b1e4006f7230d0.jpg

And this is the FMS (one for each pilot):
http://www.fs21.com/images/Rimg0052.jpg

[cimmo]

Ok, so you'd use the autopilot to give yourself time to evaluate the problem (if possible).

So how would you evacuate smoke from the aircraft? Switch off potential causes, lower the altitude so the air is breathable when the masks run out.  Then what? Not like you can open a window.

I was in a training situation years ago with full face protection and an air tank, the room was so full of smoke I couldn't see my hand 2 inches from the mask. Even so I was barely able to see for the next 10 minutes (and I had full face coverage), I could imagine without eye protection you'd be in real trouble.

B777 Flight manual emergency checklist: (there are two other pages for SMOKE / FUMES / FIRE ELEC and SMOKE / FUMES AIR COND)

SMOKE / FUMES REMOVAL:
Condition: Smoke / fumes removal is required.
Oxygen Masks And Smoke Goggles (If Required) .................................. ON
Crew Communications (If Required) ....................................... ESTABLISH
Flight Deck Door................................................................................... CLOSE
Prevents smoke / fumes from penetrating onto the flight deck.
Recirculation Fans Switches...................................................................... OFF
Stops recirculation of smoke / fumes and increases fresh air flow.
Equipment Cooling Switch........................................................................ OFF
Attempts to discharge smoke overboard by using the equipment cooling
override mode.
Note: After 30 minutes of operation at low altitude and low cabin
differential pressure, electronic equipment and displays may fail.
Do not accomplish the following checklist:
EQUIP COOLING OVRD
IF Most Smoke / Fumes Is In Cabin Forward Of Mid-wing And Outflow
Valve Manual Control Available:
Aft Outflow Valve Switch................................................................. MAN
Aft Outflow Valve Manual Switch.............................................. CLOSE
Position outflow valve fully closed.
Concentrates smoke / fumes in the forward part of the aircraft and
attempts to remove it through the forward outflow valve.
Do not accomplish the following checklist:
OUTFLOW VALVE AFT
* * * * ~
IF Most Smoke / Fumes Is In Cabin Aft Of Mid-wing And Outflow Valve
Manual Control Available:
Forward Outflow Valve Switch ................................................... MAN
Forward Outflow Valve Manual Switch ................................. CLOSE
Position outflow valve fully closed.
Concentrates smoke / fumes in the aft part of the aircraft and
attempts to remove it through the aft outflow valve.
Do not accomplish the following checklist:
OUTFLOW VALVE FWD

[cimmo]

Today (in the US), NPR aired an interview with Richard Branson and the CEO of Delta:

http://www.npr.org/2014/04/03/298779764/transatlantic-duo-looks-into-the-future-of-flight?ft=1&f=1006

They were asked about transponders and ELT's -- specifically why after 9/11 nothing was done to improve the technology of either. They didn't mention ADS-B, but I think the intent of the questions was more about how we can loose a plane with our current state of technology.

Does anyone know if MH370's ELT was set up to broadcast on 121.5Mhz or did they dump that completely in favor of 406?

ELT’s are an integral part of the slide / rafts at doors 1L and 4R.
·   Automatically activated upon deployment of slide raft.
·   May be manually activated by pulling pin at transmitter base and touching contacts with damp finger.
·   Frequencies that are used are VHF 121.50, UHF 243.0 and UHF 406Mhz.