Turkish plane crash

[RogerS]

From what I can gather, the Boeing 737 has two altimeters - the left hand one drives the auto-pilot. Which was how this plane was being flown to land automatically ..normal practice.

But....big but....in fact, a bloody BIG but...on this occasion the lefthand altimeter indicated an altitude of -8 ft when the plane was actually around 2000ft in the sky (shades of Die Hard?) and so the autopilot thought that the plane had landed....so shut the engines down...which the pilots did not pick up on for about 100secs by which time it was a bit too late.

Not sure which is more scarier....Boeing designing an airplane that doesn't seem to have any sort of failsafe or backup or arbitration between the two altimeters....or the pilots not realising that the engines had throttled back to idle.

 

[mailee]

Were the pilots asleep!! :shock:

 

[The Bear]

Don't get me started on altimeters. On the aircraft at work the light in the standby altimeter went the other day. The standby altimeter itself works fine, just need a torch to see it at night. The CAA give you 3 days to fix it.

If the actual standby altimeter itself stops working (ie you have no second altimeter) how long do you think you have to fix that?

CAA say 10 days - where's the logic in that?

Mark

 

[devonwoody]

Well Roger they should fly A.A or something, their pilots can land on water.

 

[newt]

RogerS, that is a worry if true. I would expect the 2 altimeters to talk to each other and if there is an error the autopilot would disengage. Was this aircraft fitted with a cat3 autoland system, if so they are triplex, was autoland selected or was it a manual land with an approach on autopilot. Bloody big worry and surprised there has not been a grounding. Also part of auto flying requires manual cross checks of altimeters and airspeed indicators. Even more concerned.

 

[Ironballs]

Not only that but Boeing had issued an advisory about the altimeters previously as this wasn't the first time the problem had occurred. Given that an altimeter is on of the most important items in the cockpit at what do you issue a recall/no fly until fixed notice.

I was in Germany with work the day that happened and had to fly back the next day. The old sphincter was certainly well clenched when taking off and landing

 

[CHJ]

RogerS, that is a worry if true. I would expect the 2 altimeters to talk to each other and if there is an error the autopilot would disengage. Was this aircraft fitted with a cat3 autoland system, if so they are triplex, was autoland selected or was it a manual land with an approach on autopilot. Bloody big worry and surprised there has not been a grounding. Also part of auto flying requires manual cross checks of altimeters and airspeed indicators. Even more concerned.

I would hope that we are seeing some serious mis-reporting newt, if not someone on the flight deck was being very negligent. As you say true autoland needs triple system verification but a lot of airlines decline to fit it due to cost of maintaining. (keeping the integrity of three systems up to the same standard can be time/equipment costly) As you say if it was on ILS approach then surely one of the pilots should have had hands ready monitoring of the stick and the other hand monitoring of throttles, flaps landing gear etc.

If they were on ILS with a manual decision point surely there should have been some reaction to glide path position and altitude reporting error. Surely they are not autolanding on ILS alone, where was the analogue/rad alt comparison, was the QFE set correctly.

 

[Grinding One]

Who was flying the plane then???
More training I suppose....

 

[newt]

CHJ you make some valid points, I spent some of my aviation time on perfecting autoland system and this accident has me worried particularly if the report is correct (I hope it is not). I do not believe you can perform a ILS coupled autoland without a triplex system. I guess on a 737 you can perform an ILS autopilot/autothottle coupled approach, but even that requires cross reference for airspeed and altitude. This sudden change in altitude would have disconnected the autopilot as there are limits set on inputs. The report states that the trottles went to idle because the system thought it was on the ground, this indicates an autoland. Really very strange indeed.

 

[newt]

Grinding one, the system I guess and not very well.

 

[devonwoody]

Perhaps terrorists messed up the ILS signal?

 

[CHJ]

newt, I left auto land development behind in 1985 when I lost contact with the Blind Landing camp at Thurliegh. (748-BA111) But my years endeavoring to help pilots keep research aircraft out of the ground or sea when flying at or below 100ft tells me someone is either cutting corners or adopting procedures that are using equipment for tasks it was not intended for.

 

[Grinding One]

Always bring it in with a for real pllot...self landing planes...lazy people.

 

[newt]

CHJ, Bac111,748 and a Trident bring back some interesting memories. Getting the one million successful autolands before it could be used for fare paying passengers was certainly heady days. Im with you on this one, either they were messing about or the reports are incorrect. I have a couple of mates in the AIB they may know more. By the way there is a 111 still flying at Boscombe down for the ETPS.

 

[newt]

CHJ dont suppose you knew Roger Beasley he was a test pilot and did some autoland work.

 

[CHJ]

By the way there is a 111 still flying at Boscombe down for the ETPS.

If it's the one from Bedford I was seconded to the BLU to prepare trials wiring diagrams for it in 1976.

First time I've worked with a section that waited for dense fog to do trials flights. :shock:

XX105

 

[p111dom]

We had a 111 based at Edinburgh util about 3 years ago which had a eurofighter nose grafted on the front for radar trials. It used to chase an HS125 out in the North Sea after spendin many an hour sitting on the secondary runway threshold irradiating us all.

As for the crash, you would have thought that rad alt would have kicked in but it's not uncommon that it doesn't. I may be one of these occossions where all the holes ion the safety net lined up. A coincidence of multiple layers of safety critical equipment failing at the same time.

 

[devonwoody]

Who was on the plane?

How many airliners have crashed because their approach system failed?

 

[p111dom]

I wouldn't have though many have crashed due to approach systems failure but plenty have crashed due to other issues on final. I would have thought the largest cause of crashes on finals was over confidence that the pilot can complete an approach in poor weather conditions. Failure to see visual markers and yet continuing when they should go around, micro bursts and wind shear have all been contributing factors in the past. Approach systems failures are only usually a problem when they are lost as part of a wider systems loss such as the 777 that came up short at Heathrow.