5 ms·
The Times seems to quote officials as saying they know this via ACARS messages. However, assuming the primary radar returns are good enough, this is reasonably
The Times seems to quote officials as saying they know this via ACARS messages. However, assuming the primary radar returns are good enough, this is reasonably inferred by its radar track alone - the chances of hitting 4 separate navigation waypoints (IGARI, VAMPI, GIVAL, IGREX) based on random human flying is pretty much nil. So this may be just confirmation of what most investigators were already assuming.
Sad as it is to say, at this point I think the most likely cause is a bizarre pilot suicide. That night one of the pilots likely allowed a dark, dark thought they once had to overtake them for some reason.
First this person programmed a set of remote 5-letter nav waypoints into the FMC without the other pilot noticing - not a difficult task by my understanding. Most likely to hide the evidence for insurance reasons, or just out of shame, to hide his actions from the world (a long flight also ensures the CVR will tape over the relevant bits after its 120min loop - CVR/FDR is one of the few things that cannot be disabled from the cockpit of a 777).
After the radio handoff, the other pilot left the cockpit for a moment. The perpetrator then set the transponder to standby and disabled ACARS messages from being sent via both VHF and SATCOM. Finally, in order to ensure hypoxia took hold quickly, he set the plane to climb as high as the autopilot would take it (which ended up being ~FL450), and then depressurized the plane, quickly and painlessly killing everyone aboard, including himself. Multiple airline pilots in the airliners.net forums have been discussing these possibilities and they seem to agree that all of these things can be performed by a single pilot, in a few minutes time, entirely from the cockpit. The ghost plane then flew the programmed route until running out of fuel.
While this is certainly bizarre, unlikely, and hard to fathom, I posit that it is the least unlikely scenario, because it can be done entirely by one person whose motives we don't understand. Historically, there have been many people who have done bizarre, horrendous things for motives we can't begin to understand - whether due to insanity, sociopathy or zealotry. I truly hope I'm wrong, but I just can't see a grand multinational hijacking conspiracy as being more likely than a sad man wishing to end his sad life in an evil way.
If you're thinking of replying to this with your own theory, let me just add that I say all of this not to add flame to the fire of speculation, or to accuse a potentially innocent man of mass murder, but to try and finally put the whole thing to rest in my own mind - I seem to have been rather obsessed with the whole story over the past few days. While the past week has been a flurry of information and misinformation, it's quite likely we will never know what happened to MH370, at least not for a very long time. We need to find the explanations for ourselves that allow us to come to peace with the incident personally, so that we can collectively move on at some point.
That's the thing. Your "unless..." clause is the key element.
As background, I fly a twin engine piston airplane that's pressurized (Beech 58P). Different level of responsibility from a 777-ER (I have only 6 seats, including my own), but many of the same basic design principles apply.
My cabin is pressurized by air coming off the turbos (the turbos "make" higher than ambient pressure air, and that is fed into the cabin after passing through a restrictor and an intercooler). Turbines (jets and turbo-props) pressurize from bleed air (also high pressure air created in the engine, but via a different physical mechanism.)
If I have an engine or turbo failure, I need to be able to isolate the cabin from the smoke/fumes/fire that may be present on that engine. So, I have a red knob that can isolate the cabin from the feed from that side (and another knob that will isolate the other side).
Similarly, I have a cabin pressure controller, and on that a cabin dump switch. The controller controls a single outflow valve at the rear of the cabin, about the cross-section of a roll of toilet paper, and there is a second, spring-loaded valve of the same size that's designed to automatically open if the cabin were at risk of over-pressurization.
It would be trivial for me to dump the cabin quickly with a single switch. This results in a cabin rate of climb of over 3K feet per minute, which is painful on the ears, and within minutes the cabin will climb to the aircraft altitude. (I test virtually all of the systems annually prior to the annual major maintenance event, such that if a system has failed, I can have it addressed during major maintenance.)
Why is there a dump switch at all? Dumping the cabin is a great way to clear the cabin of smoke, so in the event of an electrical fire or smoke, once the situation is contained, I can dump the cabin, venting all the toxic smoke inside to the outside, with a supply of "clean" higher-pressure air entering at the front. Once the smoke has cleared, I close the outflow valve to re-pressurize the cabin from the engines.
The point of this wall of text is that aircraft are designed assuming that someday something will fail. When it fails, the pilots, assumed to be the "good guy", need to have the best chance possible to react.
All common, and many uncommon, failures have a specific checklist response, sometimes with initial items from memory. (With an engine on fire, you want to perform the initial response from memory, backing it up later with the written checklist.)
No plane is designed to protect the occupants from a pilot intent on killing everyone on board. The pilot has any number of ways to do that, so it's counter-productive to prevent the 99.99% emergencies from being responded to by a well-intentioned pilot in order to prevent the occasional madman from killing all aboard.
Commercial pilot here, while possible (almost everything is possible at this moment), I don´t think it was a suicide, they even say at the article that if this is a suicide it´s inconsistent with previous cases where the pilot just pushed the controls and dived to the ground.
Why it´s too complicated?:
-To kill everybody via hypoxia you don´t need to climb to Flight Level 450 (45k feet). FL350 works just as good, the difference in the time of useful consciousness is 10 to 15 secs (from 35' at 350 to 15' at 450), not worth the trouble of climbing and possibly stalling the plane. After that time you either pass out or you start not knowing what you are doing.
-You don´t need to turn and fly for 6+ hours to random points, just keep your flight plan and selected altitude and the airplane will behave like the Helios (flying till the end of the flight plan and then holding till running out of fuel)
-Why flying for 6 hours if you want to suicide? passenger oxygen only works 15 min (time enough to reach a breathable altitude), the portable oxygen tanks for the auxiliary crew will work for 30 min more or less.. Look at this video of hypoxia test at 25k feet to see how fast you pass out http://youtu.be/hSrGfElyfVE http://youtu.be/hSrGfElyfVE
-why disconnect the communications? you only need to ignore them. why bother with the ACARS?
So far all the things are more consistent with a hijacking done by someone with some knowledge of flying a plane (like asking pilots to shut down coms, climbing descending, turning, introducing a waypoint), but not enough to know that flying at FL450 it´s so unstable for a plane like this, or that you´ll eventually will run out of fuel.
Also I must say that all theories are taking in to account good and skilled pilots and rational decisions at the cockpit (maybe from the hijacker). The captain really seems to be a skilled professional, but I don´t know about the flight officer. Friends of mine who are flying at Asian companies complain all the time how awful the instruction and flying skills are there. For example the pilots of the Asiana that crashed at SF just lacked basic flight skills, like using the thrust levers to do a normal landing. At china the flight officers are NOT allowed to take off, land or touch the controls below 1000' (How can you be a pilot without knowing how to land?, what happens if the captain has a heart stroke?). But I don´t know how good it´s pilot instruction at Malaysia Airlines.
I´ve seen very strange results during simulators when trying to solve emergencies, and all where very capable pilots. After all airplanes are very complex systems working in very different scenarios. Corrected procedures and system fixes are constant even on models that have been flying for 20+ years.
What I mean is that yet another possibility, is that the weird behavior is caused by some kind of emergency that went wrong somehow while trying to fix it, maybe due to a momentary bad decision that made everything more difficult, or just plain incompetence.
Also you can not trust all the theories and data they are releasing, it might be a bad interpretation, or just plainly wrong. For example, look at the Malaysian government, they told everybody that they lost track of the flight at one point. Everybody supposed that it was the actual point where they lost ALL contact (primary, secondary and acars), but it was just the secondary radar...and it took them almost a week to realize this. As this point is the ABC of Search and Rescue knowledge, I can not even imaging how many things they are screwing during this search..
edit to finish an incomplete sentence
edit 2 missing words and improving some sentences
> I´ve seen very strange results during simulators when trying to solve emergencies, and all where very capable pilots. After all airplanes are very complex systems working in very different scenarios. Corrected procedures and system fixes are constant even on models that have been flying for 20+ years.
The initial behaviour of the plane around the time the transponders went out seems very possible to explain in terms of a less-than-perfect response to a sudden emergency. But how could one square a) making a series of direction and altitude changes over a longish period of time, well after the initial incident, b) no attempt to communicate at the same time and c) at most, limited damage to the plane's systems with d) no foul play? It seems that at least one of those has to give...
Agreed.
Additionally, none of the mechanical failure theories explain the aircraft flying towards Antarctica / Australia.
I still think that it´s some kind of hijacking, but there is the possibility that the crew started screwing themselves while working on an emergency, and then getting lost at sea.. Not the most probable but possible.
You´ll be amazed how fast you can start screwing things at a cockpit once you have an unknown emergency or even worst a known emergency that you have misunderstood. Just deselecting the wrong button (for example disconnecting a generator and a cross tie connector), will put you in manual control, with all the cockpit lights and half the instruments off, several alarms ringing... not a desirable situation even for an experienced pilot. You are able to forget to communicate, to navigate, you are able to crash to a mountain because you are looking at a flashing light, I can´t find a good example outside of aviation to make you understand the feeling to look at a panel and not knowing what the hell is going on.
At simulators we practice all the normal emergencies, that will cover you 99,5 of the times. Most emergencies are simple, but some times something that is not known even to Boeing happens. Or it´s a simple emergency but you take the wrong steps. This happens more frequently to pilots who passed a lacking instruction method (like the one I describe that´s happening at china), but can happen to anyone.
The Air France crash was due to a no emergency situation (they just needed to keep altitude, and engine thrust selection to keep a normal flight and recover the instruments) that was converted in to a crash due to a misinterpretation and wrong piloting skills.