Air India Crash
- Thread starter Grim5
- Start date
Caporegime
- Joined
- 8 Jan 2004
- Posts
- 32,913
- Location
- Rutland
It would have to be a pretty epic software fault to turn off both engines.
Last edited:
Caporegime
- Joined
- 8 Jan 2004
- Posts
- 32,913
- Location
- Rutland
It won't have been overweight, it wouldnt have had a full fuel load as it wasn't doing anywhere near its max range.
No, not even close.
Well any dual engine failure would be epic. The odds are just insane on it happening. But if it happened it does explain RAT.
Last edited:
Caporegime
- Joined
- 22 Oct 2002
- Posts
- 28,890
- Location
- Boston, Lincolnshire
That is the problem with modern planes. The safety is written in blood unfortunately. In aviation we use TEM (Threat and Error Management) and within that is what they call the swiss cheese model. It could well be that the pilots did something that allowed all these scenarios to line up and create this accident. It might well be a problem with the plane (Boeing have previous) but the sad reality is the vast majority of accidents are caused by pilot error.
Hopefully what ever caused the accident they will prevent it from happening again.
Associate
- Joined
- 19 May 2010
- Posts
- 1,877
- Location
- Cheltenham
The fact it’s taken 14 years to happen would suggest the opposite - it’s some innocuous line of code somewhere which combined with set of circumstances and is a trillion to one chance of happening.
The fact it happened at the change between ground and air mode is another red flag for software.
Well a software fault could quite literally do anything to anything. Any unexpected condition where the software has literally had a "fault" will mean it has broken free of any expected boundaries and you would expect the unexpected. Things are off limits then. It could be something super specific and rare, which caused such a crash (logical) that the internal system behaved strangely.
I'm actually surprised how many crashes come down to pilot error when you would think that they would build systems that simply do not allow for such inputs to be allowed.
Just as one example, I remember the one where the pilot allowed his kid to fly the plane whilst it was on auto pilot seemingly not actually adjusting the course at all. The kids inputs were held for such a period in one direction, that the auto pilot disengaged and it pitched the plane to such a steep banked turn, that it went almost to 90 degrees. The G force caused one of the pilots to barely be able to get back into his seat for a while. This was beyond that which the aircraft was designed for. A stall followed, spin and eventual crash into the ground. Everyone died. Why would the system allow a turn to be made beyond system design? Why allow manual human control to be even able to perform such inputs?
What happens when those systems fail, take over the aircraft and cause it to crash despite pilots doing everything they can i.e. MCAS on the 737 Max 8?
For almost every time you design a system to prevent "pilot error" in one circumstance you are potentially preventing a pilot from doing the same thing to solve a problem.
You can't design a system to be reliable enough and prevent every possible "pilot error", because sometimes the system itself is faulty or it's getting dodgy information so the system thinks the pilot is doing something wrong when the pilot has noticed that there is a problem with say a set of sensors and is trying to do the right thing.
As an example if some of the sensors for air pressure etc are blocked you can have the computers thinking that the aircraft is flying in one way, whilst the pilots can see it is flying in another, so the humans will be doing the correct thing but to do so they need to override the computers (and possibly do it very quickly).
MCAS was IIRC trying to solve a problem and basically overriding the pilots (repeatedly) to prevent a problem that could occur under some conditions for example.
Caporegime
- Joined
- 22 Oct 2002
- Posts
- 28,890
- Location
- Boston, Lincolnshire
14 years is irrelevant if the series of scenarios was there from the beginning it might just be now that it has happened.
Some checklist might have been missed, combined with something else and those chain of events caused the plane to crash. There are so many redundancies in planes now that a "computer bug" will not bring down a plane.
Aeroperu flight 603 crashed due to pilot error in some respects but the main culprit was all the static ports were taped over due to maintenance with silver tape which were hard to see as they were the same colour as the fuselage and were missed by the pilots on visual inspection before take off. The pilots had no reference to altitude whilst flying at night because altitude is calculated by the static pressure around them and airspeed is also calculated by the dynamic pressure and static pressure so that was wrong too as rate of climb would be too.
This tape solution was the known servicing method at the time and after this accident was changed.
Boeing 757 was introduced in 1983 and this accident which was waiting to happen. Happened in 1996. 13 years later.
Last edited:
Yes, the aircraft will be preprogrammed with a destination. It then detects how much fuel is in the tanks and if it's not enough, it cuts power to both engines mid air to let the pilots know.
Soldato
- Joined
- 9 Nov 2005
- Posts
- 9,547
- Location
- Southampton, hopefully not too hot, or too cold
That's not what I asked.
I was wondering if it was possible for the flight to have taken off on fumes from the flight to India, or would the plane computer prevent attempted takeoff.
Mate it's Friday night, you should know by now that I'll only post absolute nonsense most of the time.
Anyway, I have no idea about your actual question. There must be some form of warning system in place I'd have thought.
That was an awful lot of fire for a plane with no fuel.
Associate
- Joined
- 19 May 2010
- Posts
- 1,877
- Location
- Cheltenham
A computer bug absolutely could do something like that. If it happened at altitude it wouldn’t be as much of an issue with glide distances and the opportunity to attempt engine restarts etc, but with 400ft to play with they never stood a chance.
I’ve seen very weird things happen with aircraft and electronic boxes when they’ve been fitted with out of date or incompatible software, and an A400 was lost due to incorrectly loaded engine control software:
2015 Seville Airbus A400M crash - Wikipedia
Time will tell if this was the cause or not.
Incidentally, taping over the static ports for aircraft washes is very common - I’ve done it myself many a time as that’s what the procedures call for. These are usually pre-written processes which will also have entries for the blanks to be removed. If not, any capable engineer will leave an open entry on the job card for the it.
I don’t believe you can design a completely foolproof system, especially where computers are concerned. I could very well be wrong - I’ll admit that right now. I just can’t see anything else that fits all the details so far.
Prevent take off? No. Give you a warning? Probably. You’ll absolutely get low level warnings when you’re getting close to the bottom of the tank, and even if fuel wasn’t sufficient it wouldn’t cause the loss of both engines at the same time.
Associate
- Joined
- 19 May 2010
- Posts
- 1,877
- Location
- Cheltenham
The 787 compares the selected position to that programmed into the Flight Management Computer. Assuming it was all calculated and entered correctly, they would have been given a configuration warning if they weren’t in the correct position.
It’s also very doubtful you could even take off in the first place.
No conspiracy just in my opinion it's a strange way to act and a near impossible situation. Truly staggering.