Flight MH370

[Jeff Kelly]

Yeah but with AF 447 they knew pretty early that it crashed and where it crashed. The mystery was how it crashed and that  was only solved once the flight recorder was found two years later.

[calapine]

Yeah but with AF 447 they knew pretty early that it crashed and where it crashed. The mystery was how it crashed and that  was only solved once the flight recorder was found two years later.

Yep, and until then lots of speculations but not one guessed the actual cause. Just trying to say it's all a bit groping in the dark until there is more information available.

Edit (again): As you mentioned fuel running out earlier: Just read that Malaysia Airlines confirmed the machine was tanked for an 8 hour flight.

[01101010]

I always thought black boxes on planes had beacons that triggered automatically in a crash.

[KallDrexx]

I always thought black boxes on planes had beacons that triggered automatically in a crash.

They do but they only have 30 days of battery life, and if the black box is under the ocean the signal won't get very far from where it sunk.

[calapine]

I always thought black boxes on planes had beacons that triggered automatically in a crash.

AFAIK only an acoustic beacon that triggers on water contact, but no radio signal.

Edit: I type too slow.

[MahrinSkel]

I always thought black boxes on planes had beacons that triggered automatically in a crash.

They do, but they are short range, you have to get an appropriate receiver within a few dozen miles, then triangulate.

Anyway, here's a more useful version on the last possible location of the signal to the satellite.

Note that this is not a course, but that when the last "ping" was received, the plane was at a point somewhere on those lines.  No idea what the margin for error is (how wide the lines should be considered to be, could be a hundred miles).

--Dave

[Numtini]

What Im kind of amazed at is that you apparently can turn off the transponder.  Seems like the kind of thing you'd want to make impossible to turn off regardless of what a man with a gun to his head would like to do.  I mean, is there any scenario at all where your want a large commercial jet liner flying anywhere without the ability to easily be tracked?

Patrick Smith/Ask the Pilot addressed that. It's because if there's a short or it's broadcasting incorrect information, you want to be able to turn it off. Also, as I remind my users constantly, if it's not working, you shut it down and restart it.

[MahrinSkel]

I keep coming back to Tajikistan/Kyrgyzstan.  If we assume that the plane was flying the whole 7 hours until the satellite pings stopped, and it was heading *to* somewhere rather than flying some kind of eccentric loops, that's about where you would wind up.  I think it's a fair bet that whoever was shutting down communications didn't know about the engine telemetry system, or did know about it but also knew that Malaysia Air hadn't contracted for the monitoring service from Rolls Royce and assumed that meant the system was disabled.  That implies a level of information that is beyond terrorists or thieves, more like a national intelligence operation.

Which brings us to...the Russians.  Tajikistan or Kyrgyzstan probably couldn't do this (they have virtually no foreign intel capacity), but Russia still has operating bases in both countries (former Soviet facilities they still control).  If we had the charts like above for the previous hours, we could do a pretty good job of plotting potential courses, and I would be willing to bet that a flight over the Himalayas and the wastes of far western China to one of those two countries would emerge as a high probability option.  Tajikistan has closer relations to the Russians than Kyrgyzstan, and is a more likely candidate on those grounds. EDIT: But the whole region is a muddle, with comparatively little information available and borderlines drawn in the Soviet era that make *no* sense from the outside (there are chunks of Uzbekistan in there with no physical connection, and the lines wind around like a snake with a broken spine), so who freaking knows.

And since our national security apparatus probably has that data and has run those calculations, if I could I'd be watching for an unusual burst of activity by the embassies in those countries.

--Dave

[Evildrider]

Some news sources are saying that the plane may have ended up in Pakistan.

[MahrinSkel]

Some news sources are saying that the plane may have ended up in Pakistan.

I doubt it, both because it's not consistent with the satellite data and because both we and the Indians have that area covered by a "no sparrow shall fall" level radar net.  Similar for Iran.

--Dave

EDIT: Looking at reports on it, it seems like people are looking at the "straight lines" version of the satellite data map somebody threw together on MS Paint because the original wasn't ready for TV, then further speculating that since that (erroneous and misleading) map shows a path near Pakistan, it "might" have routed that way.  They're still trying to find a narrative for this, and trying to cram it into the "Terrorist Hijacking" slot even though it makes no sense.

[Jeff Kelly]

Wouldn't it be much easier for a national intelligence agency to simply buy or charter a machine than to steal it, though?

[MahrinSkel]

Wouldn't it be much easier for a national intelligence agency to simply buy or charter a machine than to steal it, though?

Yeah, motive is still a problem even thought the operational profile fits.  What reason could they have for taking a risk like this?  More likely to involve *someone* on the flight rather than the plane itself or the cargo, but why would grabbing a whole plane be considered preferable to a straight-up kidnapping?

--Dave

[Evildrider]

Fuck it.. it's aliens.   

[MahrinSkel]

This might be interesting: Tajikistan Foreign Minister met with Deputy Secretary of State last Wednesday.  That appears to be the only "high level" meeting between Tajikistan and State in the last year.  Interestinglingly, a site search of state.gov for Kyrgyzstan returns the high-level meeting schedules for the 12th and 13th (which the Tajikistan meeting is on) in the results, but there is no actual mention of that country on those pages now and there is no cached version on Google.

Hmmm....  Anybody know of a publicly accessible changelog for state.gov?

--Dave

EDIT: The search hits on Kyrgyzstan may just be its presence on the drop-down in that page (every other schedule page comes up as well).  Still leaves the oddly timed meeting with the Foreign Minister of Tajikistan.

EDIT2: Appears to have been scheduled long in advance, part of a tour by the FM that included the UN the day before.

[Merusk]

I've been thinking of this  since day one 

Heh I think of that every time there's a crash or a someone goes missing. The professor who taught my history of science fiction class was roommates with the author of the story the movie is based off of.  Needless to say he was less than pleased with what the movie became.

[Ghambit]

All this talk of hijacking and there's still no definitive way to rule out a cockpit fire.  I believe two 777's have had fires fairly recently.  Egyptair (in 2011) was a total cockpit loss on the ground after 1 minute (since the fire was using the O2 system as an accelerant).   The Swissair flight in the 90's that ironically killed about the same amount of people, was almost identical in circumstance.  Instrumentation fire.  Dead of night.  IFR conditions.  Pilots struggled to maintain control over the flight plan and eventually hit the ocean at 300mph some time later.

A fire behind the panels or down in the instrument bay could easily get out of control and drop systems one after the other, eventually leaving the pilots with steam gauges and dead reckoning in the middle of the night.  Their best bet would be to use the VOR intersections they apparently used and kill time until daylight.  Likely the AP would be the only viable inertial nav. system left (their gyros are not located in the same place), if at all.  They had a full moon so it's possible they had a horizon as long as they stayed above the cloud layer.

Anyways, I find cockpit fire originating at the comm stack a more viable situation them some exotic hijacking scenario.  You could fry an egg on some of those electronics (especially the radios) after they've been on a while.  I am a pilot too btw.  My 2 cents.

[KallDrexx]

Anyways, I find cockpit fire originating at the comm stack a more viable situation them some exotic hijacking scenario.  You could fry an egg on some of those electronics (especially the radios) after they've been on a while.  I am a pilot too btw.  My 2 cents.

Sounds plausible with the oil rig worker spotting too.

[calapine]

Possible.   Or some decompression accident and subsequent flying by the autopilot (until engines died by lack of fuel) as with Helios 552.

The flight path looks very deliberate though, so I'd personally guess human intent, whether by the pilot himself or any hijacker.

[MahrinSkel]

All this talk of hijacking and there's still no definitive way to rule out a cockpit fire.  I believe two 777's have had fires fairly recently.  Egyptair (in 2011) was a total cockpit loss on the ground after 1 minute (since the fire was using the O2 system as an accelerant).   The Swissair flight in the 90's that ironically killed about the same amount of people, was almost identical in circumstance.  Instrumentation fire.  Dead of night.  IFR conditions.  Pilots struggled to maintain control over the flight plan and eventually hit the ocean at 300mph some time later.

A fire behind the panels or down in the instrument bay could easily get out of control and drop systems one after the other, eventually leaving the pilots with steam gauges and dead reckoning in the middle of the night.  Their best bet would be to use the VOR intersections they apparently used and kill time until daylight.  Likely the AP would be the only viable inertial nav. system left (their gyros are not located in the same place), if at all.  They had a full moon so it's possible they had a horizon as long as they stayed above the cloud layer.

Anyways, I find cockpit fire originating at the comm stack a more viable situation them some exotic hijacking scenario.  You could fry an egg on some of those electronics (especially the radios) after they've been on a while.  I am a pilot too btw.  My 2 cents.

Except: That should not have taken out all of the comm systems (except the one that wasn't officially there) as they are not all in the cockpit in planes like that, and they might not have had instruments but they would have seen ground lights for rough navigation (so you're postulating a flight path that somehow never crossed a significant town or highway).  Also, it was daylight by the time the satellite pings stopped.  You are also still left with the problem of where the damned thing finally wound up, and why there's no debris field anywhere after 7 days of searching.

You essentially have to assume that all of the electronic evidence (radar, satellite comm, etc.) has been mis-analyzed, and that the pilots managed to screw up in exactly the right way to maximize difficulties for the search.

--Dave

EDIT: A lot of these news reports and speculations are assuming a 2250nm limit on how far the plane could have gone and are drawing circles on the map that barely reach the southern tip of Pakistan, that's wrong.  2250 is the distance it was intended to fly (Kuala Lompur to Beijing), it had 8 hours of fuel on board (it's normal for there to be a couple of hours of excess) and could have gone as far as 3100nm from where contact was lost.

[Ghambit]

Eh?  I believe the only comm system that isnt articulated from the cockpit is sometimes a sat-comm from a flight attendant station.  And that's assuming a hypothetical fire or precautionary action didnt take out the supporting systems for that.  ACARS?  Also in the cockpit (in the 'doghouse' access I believe) along with the requisite breaker panels within immediate reach of the pilots (if the breaker/fuse panel goes haywire, so will everything that runs through those busses).  Regardless, apparently the airline didnt subscribe to ACARS engine monitoring.

Also realize, the first thing you're taught to do in midair fire is to pop every breaker you can find.  Even if it was a relatively small hazard, the pilots would've shut off and/or reset every system to curtail it.

Two pics of an Egyptair 777 2 mins. after cockpit fire (on the ground):




Imagine for a moment you're 35,000ft up in near pitch dark instrument conditions (it's midnite) on a plane that size and you lose nearly all instrumentation including most direct flight control.  The cockpit is such a disaster that there's no way you could even land the plane.  You cant talk to anyone either (no comms) and you're a ghost-plane with no transponder in busy airspace.  You have near zero navigation aside from paper charts and maybe some VOR data.  Maybe you've also decompressed and have no shot but to (assuming the AP is still functional) program the plane to fly a different flight plan until daylight... following the Malacca intersections until splashdown.

I know it's a longshot, but they musn't discount cascade failures like this.  This is the way aviation works.  One small thing (or bad decision) leading up to an inconceivable amount of other things.

[MahrinSkel]

Granted, it's a possibility, and it accounts for one thing that the hijacking explanation has to dismiss: The report from an oil worker of seeing a plane that appeared to be on fire.  But eyewitness testimony is often the least reliable form of evidence, and we have only the one guy.

--Dave

[Abagadro]

Would a cascade failure like that result in them taking it up to 45,000 feet?  I guess without instruments maybe they didn't know they were in a climb?

That seems to be a key part of all of this from the latest reports. Some speculation that it was done to intentionally depressurize the plane since it is above the operational ceiling. Knock everyone out so they can't stop you from what you are doing (whatever the hell that may be).

[angry.bob]

Would a cascade failure like that result in them taking it up to 45,000 feet?  I guess without instruments maybe they didn't know they were in a climb?

That seems to be a key part of all of this from the latest reports. Some speculation that it was done to intentionally depressurize the plane since it is above the operational ceiling. Knock everyone out so they can't stop you from what you are doing (whatever the hell that may be).

It could, pretty easily and for a couple of reasons. They might not have known they were in a climb if it was shallow enough and then the plane stalls at 45,000 feet which could also account for it diving straight down. The problem with that is if you lose your instruments in the dark/fog/etc you can get a good idea of your axis orientations by just looking at a liquid in a container and piloting to keep it level until you can see something or you run out of gas.

More likely, if there was a fire it wreaked havoc on the flight controls. The 777 is Boeing's first fly-by-wire plane. Airbus used to have a problem with wires that were bundled together inducing current in each other and causing the control surfaces to move randomly and uncontrollably in flight. Wires are shielded from doing that, so it's not really a problem anymore. But, as Ghambit said, the avionics and whatnot get hot. Really hot, and they cram all that shit into as small a space as they can right in front of and below the cockpit. On a good, old fashioned hydraulic/cable flight control system if you have a fire you put it out and fly on without whatever caught on fire. With the wire controlled system, that might be what caught on fire. Or the insulation and shielding burned off the wire. Or the wiring just melted/burned apart. So now you're riding in a plane that you have no control at all over. Maybe it'll try to do whatever the last input it got. Maybe the uninsulated wires are shorting out randomly and it'll do god only knows what and god only knows when. So, the plane decides to climb until it hits an altitude where the air density isn't letting the plane produce enough lift and it stalls out. The flight controls are out so there's no recovery and stright down it goes. I haven't worked on anything that big, so I don't know what sort of redundancy they have, but since the big feature of wire control is weight savings I doubt it's much. Also, from all I've read the 777 it seems to be a fucking deathtrap with some pretty big design flaws.

My money is on a Fire too. I'm not a pilot but I've had an A & P license since 1989.

[Lightstalker]

I'm not sure how much should be made about that altitude reading - both on the accuracy of the number actually hit and the implications of hitting that number.  Back in college Boeing was still doing this to full aircraft (not just representative mock ups); the operational limits on an aircraft are well under the failure limits of that craft.

You wouldn't want to run near the limit anyway, lest the unexpected happened and you had to / were forced to exceed that threshold.

[MahrinSkel]

Not knowing exactly what the Malaysian military was using, all I can say is that in general, military radars only give reliable altitude for planes they are actively attempting to track, and as this had to be reconstructed from recorded data, they presumably weren't explicitly tracking it.  However, they seemed pretty confident in it, and there are some newer radars that would give that kind of detail without explicit tracking, so it's possible (the original source did say the report wasn't definitive because of range, that implies a radar designed for close-in air defense operating way outside its design parameters and someone trying to calculate what the recorded signal means, an informed guess at best).  The same reports have it stabilizing at 23K and returning to normal cruising altitudes, which isn't consistent with uncontrolled flight.

There are a few other possibilities at the limits of chance, like the fact that Malaysia Post is one of the few postal services still shipping lithium batteries in aircraft.  But there would be little reason to take them from Kuala Lumpur to Beijing (most are bound from Hong Kong or Singapore to the US or EU).

It all comes down to the debris.  It's nearly impossible that the plane could crash without leaving a big mess, unless it was soft landed on the ocean and then sank to the bottom.  In that case, you'd expect that at least some people would be floating in the Indian Ocean.  If nothing ever turns up....

--Dave

[KallDrexx]

It all comes down to the debris.  It's nearly impossible that the plane could crash without leaving a big mess, unless it was soft landed on the ocean and then sank to the bottom.  In that case, you'd expect that at least some people would be floating in the Indian Ocean.  If nothing ever turns up....

i was thinking about that last night and a soft water landing (like the Hudson river landing) is the only thing that makes sense in my mind.

If it crashed in the ocean at high speed it sounds like there should be a lot of debris everywhere (the chairs and a lot of other pieces are made to float for just that very reason).   If the plane was hijacked it seems extremely unlikely (of course not out of the realm of possibilities I guess) that it landed on land somewhere in tact.  It would take far too much coordination to not only fly without identification codes (and without getting shot down) in a country's air space, then get access to land on a runway that supports a plane as big as a 777.

Then when it lands they have to make sure no one notices a random Malaysian Airlines 777 land when its already been reported that one is missing.  Furthermore, what do they do with the passengers?  Either they are all alive and someone has to have the resources to keep, feed, and hide 239 passengers + crew or they are all dead and no one has noticed a 777 with 238 dead people on board.

[IainC]

Flight 447 gives a pretty sobering account of how a cascade of failure from a very minor issue, can cause pilot error that puts an otherwise perfectly OK aircraft into the ground. If the pilots of MH370 didn't trust their instruments for whatever reason, there are a lot of very small mistakes that can have devastating consequences.

[Jeff Kelly]

As far as the timeline of events is concerned I don't see how a technical issue could be the explanation. If we'd still be talking about a missing plane that was last seen at location X, yes sure. Not when we're talking about a machine whose tracking and transponder systems were disabled deliberately, that deliberately changed course and that deliberately flew in a way to avoid tracking and radar systems.

[satael]

(one more) infographic on the MH370 flight stuff known so far:

[angry.bob]

As far as the timeline of events is concerned I don't see how a technical issue could be the explanation. If we'd still be talking about a missing plane that was last seen at location X, yes sure. Not when we're talking about a machine whose tracking and transponder systems were disabled deliberately, that deliberately changed course and that deliberately flew in a way to avoid tracking and radar systems.

Honestly, it's the most likely explanation. There's no evidence that indicates any of that stuff was intentional instead of accidental. A fire or other mishap in the avionics bay would be much more likely than a movie-level series of events. Here's a picture of the avionics bay in a 777:

While that is actually a luxuriously spacious avionics bay, it's still iall in one small area below the floor directly underneath the front passenger hatch. Pretty much everything involving electricity on the plane is either directly in there or is controlled by something in there. The only exception would be the flight recorder. Those are put in the tail since that's usually hit the ground last and takes the least damage. As far as avoiding radar, it doesn't sound like anyone with radar actually gave a shit about tracking it until after the fact. There's an old saying about apathy, incompetence, and sloth being more likely than a brilliant plan flawlessly executed. I don't remember it, but you get the idea.

As far as no debris, the ocean is big. BIG. And stuff in it, even the size of an intact plane, is hard to see. Really, really hard. Like if you don't have a location pinned down pretty tightly, don't even bother hard. And in this case they have the area narrowed down to a band that stretches from the Crimea to the southern coast of Australia. That's nearly as bad as "somewhere near or in Asia". Actually, that's pretty much exactly what it is.

[Jeff Kelly]

The satellite transponder they used to triangulate the projected path and that reported its last ping 6 1/2 hours after the machine vanished doesn't work once it's submerged under water or after a crash. Firstly because it's just a satellite transponder that is not powerful enough and because it draws its power from the engine.

That would mean that the plane continued on, on the wrong path, for six hours. If that had been the case wouldn't it have been more likely that the autopilot would have continued on its way to Beijing?

[Hawkbit]

I had a weird feeling that when they didn't find wreckage it was flown nose first into the ocean like United 93 did into land.  It's a weird situation though and I'm glad F13 has a crack-investigation team working day and night on it. 

[Merusk]

That would mean that the plane continued on, on the wrong path, for six hours. If that had been the case wouldn't it have been more likely that the autopilot would have continued on its way to Beijing?

I could be mistaken but doesn't Autopilot only keep a heading, not adjust for a destination or the original magnetic direction? i.e. If you're headed 122d 15' but then manually adjust course by 22d it will keep you going at the new heading of 100d 15'

[calapine]

The satellite transponder they used to triangulate the projected path and that reported its last ping 6 1/2 hours after the machine vanished doesn't work once it's submerged under water or after a crash. Firstly because it's just a satellite transponder that is not powerful enough and because it draws its power from the engine.

Yes, still receiving signals for 6 hours is what makes this so strange. Else one could reasonably assume: "It crashed somewhere of course, and, well, its a big ocean, so..."

That would mean that the plane continued on, on the wrong path, for six hours. If that had been the case wouldn't it have been more likely that the autopilot would have continued on its way to Beijing?

Think so too. I mentioned the Helios 552 flight before, and that is exactly what happened there. Crew unconscious, AP continues towards destination (Athen), actually executes approach procedure for the airport, but stays at the same flight-level, passes the Airport then turns into the designated holding area and stays there doing loops for the next 1 1/2 hours until it crashes due engine flameout.  

The official account was a really chilling read.


Going back to the signal and knowing the plane didn't crash. If we assume mechanical failure, that would require an accident that:

A) Knocks out the Transponder.
B) Knocks out the ACARS, but
C) Does Not disable the satcom which is used by the ACARS
D) Knocks out all navigation in such a way the pilots weren't able to do a straight 180° turn and fly back to the starting airport, but
E) Leaves enough of the other systems intact to enable pilots to stay in the air and fly controlled paths for the next 6 hours.

It also requires that when the pilot crossed Malaysia again they failed to notice being over land and continued past the coast and that whatever course they took was so unfortunate that they never ended up over land again before running out of out fuel.

 

[calapine]

That would mean that the plane continued on, on the wrong path, for six hours. If that had been the case wouldn't it have been more likely that the autopilot would have continued on its way to Beijing?

I could be mistaken but doesn't Autopilot only keep a heading, not adjust for a destination or the original magnetic direction? i.e. If you're headed 122d 15' but then manually adjust course by 22d it will keep you going at the new heading of 100d 15'

AFAIK (unlike Ghambit not a pilot): Depends on settings. Simple AP in Airbus is just some dials: one for heading, one for FL. I don't think that adjusts if you manually change course and then re-enage the AP. The more complex way are complete routes, which are stored in the Flight Management System, which relates its orders to the AP.

Edit: Here is an image: