QZ8501 goes missing Surabaya

Forums › Travel (planes, train, cruise) › QZ8501 goes missing Surabaya - Singapore

1 | 2 | 3 | 4 | 5 | 6 | 7

Batman

Mad Scientist
29760 posts

Uber Geek

Trusted

Lifetime subscriber

#1439227 2-Dec-2015 12:39

I think if we did look up all airlines this is quite likely very common across the board. But I'm only guessing. In a plane that has about 1 gazillion bits that can get faulty over time ... times number of planes per airline ...

PolicyGuy

1726 posts

Uber Geek

ID Verified

Lifetime subscriber

#1439241 2-Dec-2015 13:04

On my reading, 155 passengers and 7 crew were killed by

1. substandard maintenance practices - this was the twenty-fourth(!!!) unfixed occurrence of the same fault - followed by

2. pilots so poorly trained that
a) they powered-off both FACs at the same time - obviously ignorant of the effect this could have - by pulling their CBs, and
b) were then unable to maintain control of the aircraft in a not-completely-automated state, as the control system degraded from Normal Law to Alternate law

Would I fly Air Asia? Not a chance, no matter how cheap the tickets

Another thing that interests me is that, as far as I've seen so far, there doesn't seem to have been an airworthiness directive requiring inspection of all the RTLU (rudder travel limiter unit) cards in service to check for badly soldered joints which were the initiating cause for this sequence.

[edited for formatting]

Technofreak

6530 posts

Uber Geek

Trusted

#1440524 4-Dec-2015 13:59

No doubt airline company training and maintenance practices are a big factor but I have to wonder that the Airbus philosophy around flight controls and flight control software is part of the problem.

AF447 was a similar situation, and I think most people would consider Air France to be better than average so far as training and maintenance go.

Both accidents were in Airbus aircraft. There have been other similar Airbus incidents.

Disclaimer: The comments below are my understanding of the general way Airbus controls work. While I am a pilot I have never flown an Airbus so some of my understanding may not be entirely correct.

Aibus differ from traditional control designs.

One big difference is that the Captains and Co-Pilots control sticks are not interconnected. It is not easy to see what inputs the other pilot is making, or if the one pilot has their hands on their controls it's isn't possible to feel what the other pilot is doing as both control sticks do not move in unison. This is at odds with almost every other make of aircraft the pilots will have flown.

One other factor is that if one pilot pushes the nose down and the other pulls the nose up the computer will average the two inputs. In this accident the Captain was trying to input the correct control inputs but the Co-Pilot was pushing/pulling the other way and the Captains inputs were cancelled out. There is a way for the Captain to override plus standard operating procedures should have determined who was flying the aircraft to prevent this, this didn't happen unfortunately.

The Airbus flight control computer controls the movement of the flight control surfaces (ailerons, elevators etc). It stops the surfaces from being deflected in a manner that would over stress the aircraft. Even if the pilot tried an abrupt manoeuvre for any reason (e.g. avoiding a near miss) the computer would only allow control deflections that would not exceed the aircraft's safety parameters.

This extends to stalling the aircraft, in theory it is impossible to stall them. I think Airbus use this feature as a selling point. However any time the computer is down graded these protections may longer be present.

I suspect it doesn't stop someone who is used to the protections being in place 99.9% of the time from instinctively expecting them to be still there even though the computer is known to be down graded.

Sony Xperia XA2 running Sailfish OS. https://sailfishos.org The true independent open source mobile OS
Samsung Galaxy Tab S6
Dell Inspiron 14z i5

wasabi2k

2096 posts

Uber Geek

#1440592 4-Dec-2015 15:18

Technofreak: No doubt airline company training and maintenance practices are a big factor but I have to wonder that the Airbus philosophy around flight controls and flight control software is part of the problem.

AF447 was a similar situation, and I think most people would consider Air France to be better than average so far as training and maintenance go.

Both accidents were in Airbus aircraft. There have been other similar Airbus incidents.

Disclaimer: The comments below are my understanding of the general way Airbus controls work. While I am a pilot I have never flown an Airbus so some of my understanding may not be entirely correct.

Aibus differ from traditional control designs.

One big difference is that the Captains and Co-Pilots control sticks are not interconnected. It is not easy to see what inputs the other pilot is making, or if the one pilot has their hands on their controls it's isn't possible to feel what the other pilot is doing as both control sticks do not move in unison. This is at odds with almost every other make of aircraft the pilots will have flown.

One other factor is that if one pilot pushes the nose down and the other pulls the nose up the computer will average the two inputs. In this accident the Captain was trying to input the correct control inputs but the Co-Pilot was pushing/pulling the other way and the Captains inputs were cancelled out. There is a way for the Captain to override plus standard operating procedures should have determined who was flying the aircraft to prevent this, this didn't happen unfortunately.

The Airbus flight control computer controls the movement of the flight control surfaces (ailerons, elevators etc). It stops the surfaces from being deflected in a manner that would over stress the aircraft. Even if the pilot tried an abrupt manoeuvre for any reason (e.g. avoiding a near miss) the computer would only allow control deflections that would not exceed the aircraft's safety parameters.

This extends to stalling the aircraft, in theory it is impossible to stall them. I think Airbus use this feature as a selling point. However any time the computer is down graded these protections may longer be present.

I suspect it doesn't stop someone who is used to the protections being in place 99.9% of the time from instinctively expecting them to be still there even though the computer is known to be down graded.

There was a 99% invisible podcast about AF447 and the Airbus control systems. From memory they have a number of states and it isn't necessarily immediately apparent what state it is in.

The state will control how stick and other actions are handled.

It also addressed the conflict between the desire for more automation in the sky and reduced pilot experience (not necessarily training). If you are hands off for large portions of your job do you have the experience and intuition if you have to deal with an emergency? Are we eventually going to have fully automated planes?

It is an interesting issue, but we must be careful not to jump to conclusions on what is a complex topic.

Technofreak

6530 posts

Uber Geek

Trusted

#1440599 4-Dec-2015 15:39

It also addressed the conflict between the desire for more automation in the sky and reduced pilot experience (not necessarily training).

I fear some in the airline industry, in some parts of the world, do not place enough value on experience and training. This is an issue that it starting to show through in some countries.

If you are hands off for large portions of your job do you have the experience and intuition if you have to deal with an emergency?

It's not just experience it's also recent experience. Use it or lose it. I remember a very highly regarded specialist surgeon saying he needed to keep in regular practice at his specialist surgery to ensure he kept his skill level at its peak. He said he could notice a drop in skill after a period of leave.

Are we eventually going to have fully automated planes?

Not in our lifetime. The cost of the automatic systems currently outweigh the costs of pilots.

While drones aren't fully automated (they're flown by ground based pilots - remotely piloted) their accident rate is at least as high as for piloted aircraft.

I'd never get on an aircraft where there wasn't a human on board to sort out any issues. I think there'd be plenty of consumer resistance. If you've got that human there you need to ensure he/she is current and able to manually fly the aircraft. Kind of defeats the purpose of automation.

Sony Xperia XA2 running Sailfish OS. https://sailfishos.org The true independent open source mobile OS
Samsung Galaxy Tab S6
Dell Inspiron 14z i5

hairy1

3332 posts

Uber Geek

ID Verified

Trusted

Lifetime subscriber

#1440615 4-Dec-2015 16:10

You could start a whole thread on Boeing vs Airbus philosophy. Having flown both I am not going to harp on here about which is better (I could to wind up @sbiddle I suppose). It is similar to comparing apples with oranges. Both have good and bad parts.

The human - machine interface is a critical component of the aircraft. Problems occur when aircraft system malfunctions do not present themselves to the human part of the equation. From early in a pilots career we are taught to always believe our instruments. For example, the "leans" is a common occurrence in new pilots in cloud where the pilot feels like the aircraft is not flying straight and level. We are taught to look at our instruments and believe what they are telling us. In both cases (Air France and Air Asia) the crew had to disregard and "look through" what the machine was presenting (due to malfunctions) and were unable to do so.

Training, training and more training is the answer methinks. A completely intuitive flight interface would be a huge help but I can't see that coming any time soon. Certification, commonality and costs will ensure that.

My views only.

Cheers, Matt.

My views (except when I am looking out their windows) are not those of my employer.

1 | 2 | 3 | 4 | 5 | 6 | 7