Ask The Pilot

[Boris spatsky]

Dear JB, Boris and WT,

Any thoughts on this story from the NZ Herald today:

<snip>

The AAIB report said: "Although the pilots' actions suggested that they were under the impression the autopilot had disengaged at the moment of the lightning strike, recorded data showed that it had remained engaged."
The report said no technical problems were found with the plane, which is now back in service, and that pilot training now included simulations of this incident.
The AAIB investigation is continuing, looking at crew training, autopilot design, and any "human factors".

cheers

I am not familiar with the actual incident that they are discussing, but in this case it would seem (without reading the entire report), that the crew became task saturated and had an expectation that the autopilot had disengaged, when in fact it had not. Ultimately, the appropriate response was to disengage the autopilot and hand fly the aircraft. Even if you thought it was not connected, most pilots would still press the disconnect button in this case to be sure, especially once it started descending at 9500 feet per minute. It would appear that they had some form of automation dependancy which almost caused Controlled Flight Into Terrain (CFIT).

Automation, in most circumstances, will reduce workload; however, in some circumstances it can increase workload to the detriment of the flight. In these cases, it can be more prudent to disconnect the automation. In the end, every plane is a plane and can be hand flown (although I am not an Airbus guy so could stand to be corrected if that is not factually correct on an Airbus level!).

If you are interested, go to youtube and watch 'Children of Magenta', especially the first 3 minutes. It is a lecture by a senior AA pilot about the pitfalls of automation. Informative viewing.

 

[Melburnian1]

This appears to be the link to the 'special bulletin' regarding the above incident:

Air Accidents Investigation: S1/2015 - Saab AB Saab 2000, G-LGNO

Media reports state it occurred on 14 December 2014, but the bulletin summary mentions 15 December.

 

[straitman]

If you are interested, go to youtube and watch 'Children of Magenta', especially the first 3 minutes. It is a lecture by a senior AA pilot about the pitfalls of automation. Informative viewing.

Thanks for the link Boris.

As I read this I remembered this video but didn't have a link to it.

A very worthwhile piece of footage for all to see.

 

[clazman]

A photo from a f18 coughpit in Ottawa space and aviation museum.

Curious on more detail on this button and how it works

 

[jb747]

Any thoughts on this story from the NZ Herald today:

Well the media reports on this one are a bit of a giggle. They cover all of the go to words. Hero, plunged, terror, wrestled, nose-dived etc. The report is more sober, and much more interesting.

Basically this was an 'own goal'. The overall impression is of a crew who were uncoordinated, and not at all in charge of what was happening. Also worth noting that the captain only had 143 hours on type. I wonder what he flew beforehand.

When taking the aircraft off the autopilot, you should always press the disconnect button (twice). If it has disconnected by itself, then this will cancel the warning. The current AB will disconnect gracefully if you don't press the button, but you get the audio alarm. Boeing will fight you for control, but the control loading is so high that it's an instant reminder that you've not disconnected. I don't know how the SAAB behaves but some models are better than others....

This type of event has quite some history though. Whilst AB will disconnect, the 300 and 310 had a couple of quirks. They would only disconnect in the channel that had experienced an input (pull back, and pitch would disengage, but not roll) and they would not disconnect in the approach or go around modes. This variation of the rules, caused three major accidents that I can think of (two 300s, Taipei and Japan) and the Russian 310. Automation has many issues, but the interface between man and machine is one of the biggest problems. Inconsistent behaviour will catch people out.

 

[jb747]

A photo from a f18 coughpit in Ottawa space and aviation museum.

Curious on more detail on this button and how it works

All the information you could possibly want: http://www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA256522

 

[cambriamarsh]

thanks Boris

 

[cambriamarsh]

thanks JB, I'd wondered about that too.

Well the media reports on this one are a bit of a giggle. The cover all of the go to words. Hero, plunged, terror, wrestled, nose-dived etc. The report is more sober, and much more interesting.

Basically this was an 'own goal'. The overall impression is of a crew who were uncoordinated, and not at all in charge of what was happening. Also worth noting that the captain only had 143 hours on type. I wonder what he flew beforehand.

When taking the aircraft off the autopilot, you should always press the disconnect button (twice). If it has disconnected by itself, then this will cancel the warning. The current AB will disconnect gracefully if you don't press the button, but you get the audio alarm. Boeing will fight you for control, but the control loading is so high that it's an instant reminder that you've not disconnected. I don't know how the SAAB behaves but some models are better than others....

This type of event has quite some history though. Whilst AB will disconnect, the 300 and 310 had a couple of quirks. They would only disconnect in the channel that had experienced an input (pull back, and pitch would disengage, but not roll) and they would not disconnect in the approach or go around modes. This variation of the rules, caused three major accidents that I can think of (two 300s, Taipei and Japan) and the Russian 310. Automation has many issues, but the interface between man and machine is one of the biggest problems. Inconsistent behaviour will catch people out.

 

[SFSC]

All the information you could possibly want: http://www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA256522

Well, I'm lost! I take it to mean if you have tried all you have been trained to do, throw the switch and you may not have to eject.

 

[jb747]

Well, I'm lost! I take it to mean if you have tried all you have been trained to do, throw the switch and you may not have to eject.

When it was first introduced (if I remember the tale correctly) the USN managed to lose an F18 during test flying. It was supposedly unspinnable, but the test pilots managed to prove that wrong. But, once in a spin, it was irrecoverable. Engineers got that one backwards I guess. Apparently, the normal spin recovery of full opposite rudder, full into spin aileron, and neutral to slightly forward stick, met with the FBW reversing the pilot's reversed controls, and making the spin worse.

The eventual fix was to have the controls revert to a more direct response once a threshold yaw rate was sensed. Basically, if it sensed too much yaw, the controls stopped being smart, and simply did what they were told. I think the switch is a manual way of making that control law change.

 

[defurax]

So who is responsible for making sure that all baggage handlers are out of the cargo hold before closing the door :shock:

http://blog.alaskaair.com/2015/04/13/flight-448/

US flight worker 'trapped in hold' http://www.bbc.co.uk/news/world-us-canada-32296170

 

[under the radar]

So who is responsible for making sure that all baggage handlers are out of the cargo hold before closing the door :shock:

Statement on Alaska Airlines flight 448 | Alaska Airlines Blog

US flight worker 'trapped in hold' Alaska Airlines plane 'flew with worker trapped in cargo hold' - BBC News

i do apologize for intruding in on the experts that contribute to the 'ask the plot' thread....but my 2cents worth is the ACTUAL EMPLOYEE is responsible for their 'actions'....THE GUY FELL ASLEEP!!!! ON THE JOB!!!! if you are that tired, and you don't 'realize' it...YOU are an idiot..sorry to be so blunt....do airlines now need to introduce an 'in n out roll call' for baggage handlers (and any other employee who services an a/c???)... tick off/swipe the IDs of each employee servicing a particular a/c...and if the 'in' doesn't match the 'out'...start a search for the employee

 

[eric2011]

i do apologize for intruding in on the experts that contribute to the 'ask the plot' thread....but my 2cents worth is the ACTUAL EMPLOYEE is responsible for their 'actions'....THE GUY FELL ASLEEP!!!! ON THE JOB!!!! if you are that tired, and you don't 'realize' it...YOU are an idiot..sorry to be so blunt....do airlines now need to introduce an 'in n out roll call' for baggage handlers (and any other employee who services an a/c???)... tick off/swipe the IDs of each employee servicing a particular a/c...and if the 'in' doesn't match the 'out'...start a search for the employee

Yes but, surely there is a ramp supervisor who is overall charge of that side of the operation and whose job it is to ensure the safety of the people who work under him and also to ensure the integrity of the aircraft

 

[Dmac6345]

Yes but, surely there is a ramp supervisor who is overall charge of that side of the operation and whose job it is to ensure the safety of the people who work under him and also to ensure the integrity of the aircraft

Hi Pilots.
Drinking and daydreaming in a lounge at the moment and watching aircraft coming and going. Why is it that aircraft don't actually touch down at the start of the runway. Watching them land it seems they are aligning to a single white line either side of the centre line. Why not aim for the start? Is it room for error?
Thanks in advance.

 

[jb747]

Hi Pilots.
Drinking and daydreaming in a lounge at the moment and watching aircraft coming and going. Why is it that aircraft don't actually touch down at the start of the runway. Watching them land it seems they are aligning to a single white line either side of the centre line. Why not aim for the start? Is it room for error?

We aim for a visual impact point about 1,000-1,200 feet into the runway. That means that if we didn't flare at all, that's where our eyeballs would impact. The act of flaring moves the touchdown point about another 100-500 feet in.

The aircraft sits below our eyes. If we don't flare, the gear would impact about 300-500' before our eyeball aim point...so short of the runway if we aimed at the end.

There isn't much margin as there is...wheels cross the runway end at about 25'. That doesn't leave a great deal of room for the varying conditions.

 

[Kerrodt]

JB, what do you mean when you talk about "flaring"? Thanks.

 

[eric2011]

In layman language. Think about when you are watching a plane which is just about on the ground, the nose is high and the tail is low. Usually the power is off, the nose is high to reduce speed and the plane floats to the ground until the main gear touches down followed by the nose wheel. That was flaring.

( I think )

 

[jb747]

JB, what do you mean when you talk about "flaring"? Thanks.

Landing approaches are generally a -3º path that intersects with the ground about 1,000 feet into the runway. At the speeds we normally fly, that -3º path gives us a descent rate of about 700 feet per minute. But, if we just drove the aircraft into the ground at that sink rate, you (passengers) would hate us, and the aircraft wouldn't last very long. So, in the very last few feet of the approach we reduce the sink rate back to almost nothing. If we get it right, you get a smooth touchdown.

Basically a flare involves increasing the pitch attitude by about 2º (we don't look at the exact attitude change, just eyeball it until the sink rate looks ok). At some point in the flare, the power is also reduced to idle. Upshot is a gradual speed reduction, with the aircraft settling on to the ground.

 

[Kerrodt]

Landing approaches are generally a -3º path that intersects with the ground about 1,000 feet into the runway. At the speeds we normally fly, that -3º path gives us a descent rate of about 700 feet per minute. But, if we just drove the aircraft into the ground at that sink rate, you (passengers) would hate us, and the aircraft wouldn't last very long. So, in the very last few feet of the approach we reduce the sink rate back to almost nothing. If we get it right, you get a smooth touchdown.

Basically a flare involves increasing the pitch attitude by about 2º (we don't look at the exact attitude change, just eyeball it until the sink rate looks ok). At some point in the flare, the power is also reduced to idle. Upshot is a gradual speed reduction, with the aircraft settling on to the ground.

Thanks JB. Much appreciated. Makes a lot of sense. Sounds very elegant and graceful too, which is how it looks and often feels.

 

[Major]

Since we are talking about landing speeds, just what are they ? Do the larger aircraft land at a higher speed than say a 737 ?