Ask The Pilot

[jb747]

the attitude indicator is the instantaneous 3D position of the aircraft in space? - but relative to what?

Surface of the earth. So, the simple things it tells you are you instantaneous pitch and roll. Heading is displayed separately. Those three things tell you your attitude in the three axes.

What are the raw data?

Raw data encompasses a number of things, and can be used to describe some facets of navigation instruments, in particular ILS. In the case of flight directors, it means ignoring them (or at least not slavishly following) and looking at the actual attitude data which is displayed. As a generalisation, F/Ds are great where small changes, and finesse, are needed, but less so for large changes. They make manual ILS approaches easy and extremely accurate. When people refer to raw data for ILS approaches them mean looking at the actual ILS display and making changes based on it, rather than following the F/D.

And the flight director indicates the attitude change to needs to occur to achieve its next intended position?

The various modes and settings that people put onto the MCP (mode control panel....across the top of the instrument panel), include things like speeds, heights, rate of climb or descent. That's also where the autopilot switches, and flight director switches reside. Nothing that is on the MCP has any direct effect upon the aircraft unless an autopilot is engaged. All of the settings apply simultaneously to the autopilots and flight directors. If both are turned off, the MCP does nothing. If an A/P is engaged, you can make the aircraft do things via the panel. The F/Ds, when turned on mimic what the A/P is doing.

If the A/P is off, and F/Ds on, they suggest what you should do to make the aircraft do whatever is on the MCP. Note though, that they do not show absolute targets. If the F/D drifts to the right of centre it wants you to roll right, but it does not indicate that it wants any particular attitude. It will go back to the centre when it thinks you have enough roll. Better technique says that you roll to whatever you think is the correct attitude and then see what it does. F/Ds are basically stupid, and will move further and further from the centre if you don't immediately follow them... you're generally better off to look through them to the raw data, do what you want, and let it come back to you. Or, in a Boeing, just cycle it on/off, at which point it will always centre.

747-400 PFD. The F/D is the large white cross in the centre. It's really magenta, but overexposed. The entire cross does not move. The vertical and horizontal bars are separate elements.



A380, on the ground. F/D is turned off.



A380 in flight. The F/D is the green cross. A/P is also engaged, but the F/D is showing a command for slight left roll. So, it's basically been in a turn to the right, and now wants to go back towards wings level.


The important thing is that the flight director does absolutely nothing unless you manually follow it.

 

[Quickstatus]

F/D is showing a command for slight left roll

Thank you, let me see if I understand what you are saying.

The roll command or “A/P roll intention”? is the triangle at the top slightly pointing left of the vertical?

So the F/D is displaying what the A/P would do to the the aircraft’s attitude if the A/P was engaged based on the MCP settings - assuming the MCP settings are correct?

And if the A/P is not engaged, the F/D is just displaying a suggestion of what the aircraft attitude should be?

And the whole system is dumb because it does not know when it is wrong - only the pilot does, except when the pilot thinks the F/D must be right?

But then it can also be dangerous to assume the F/D and attitude indicator is wrong - when in IMC and the inner ear is saying to do something else?

Surface of the earth

Gyroscopically?

 

[jb747]

The roll command or “A/P roll intention”? is the triangle at the top slightly pointing left of the vertical?

No. The triangle at the top is called the 'sky pointer', and it is literally pointing at the sky. It is simply showing the actual bank angle. There is another instrument inside it, but not relevant to this discussion.

The roll command from the flight director is the tiny bit of displacement between the vertical green bar, and the yellow circle in the centre. I don't have any pictures of it with more displacement, though if you were flying it then it shouldn't ever move far.

So the F/D is displaying what the A/P would do to the the aircraft’s attitude if the A/P was engaged based on the MCP settings - assuming the MCP settings are correct?

As the A/P will always follow the MCP settings then if both the A/P and F/D are engaged (as they are normally), then the F/D is pretty well always centred. It’s really for the times you’re manually flying, and is simply the aircraft telling you what it would do if you were letting it.

And if the A/P is not engaged, the F/D is just displaying a suggestion of what the aircraft attitude should be?

Remember that it doesn’t display absolute attitudes. It tells you that it would like to pitch up, by moving up. As you approach the attitude that it wants it moves back to centre. Same for roll.

And the whole system is dumb because it does not know when it is wrong - only the pilot does, except when the pilot thinks the F/D must be right?

That applies to autopilots too. I said it was dumb because of the way it will increase the deviation from centre, if you don’t follow it exactly. For instance if it wanted a roll angle of 20º, but you only gave it 10º, then you would still get to the heading it wants eventually. But, what it would do is move further and further from the centre, basically trying to get your attention. Then to get it back to centre, you either have to apply much more roll than you really need, or, as I mentioned, cycle it on/off, which will bring it back to the centre. This is quite an issue in pitch, when it could be commanding silly attitudes after take off, simply because you were a bit slow at following it initially. This is often when you’ll hear people talking about ”looking through” the F/D…or basically ignoring it.

But then it can also be dangerous to assume the F/D and attitude indicator is wrong - when in IMC and the inner ear is saying to do something else?

Anyone who uses their inner ear for anything in flying is simply an accident going somewhere to happen. You can never assume the AI is wrong, but if in doubt, it’s the reason that there are three in the coughpit, all with separate information sources. If the F/D is wrong, it means you’ve commanded it to do something incorrect. So fix that, or turn it off.

Gyroscopically?

Three laser ring gyros. A gyro with no moving parts.

 

[jb747]

There was a question here a day or so ago about the EK incident in Dubai. There's no hard data out about it, and I suspect that there's unlikely to ever be a report that we can all read.

The gist of it seems to be this:
MCP altitude incorrectly set to zero for take off
Captain noticed during the roll, and called for it to be set
The MCP was not reset, but coughpit confusion set in and the aircraft was not rotated until in the runway overrun area, at 216 knots
Captain screaming (!) for MCP to be reset, FO missunderstands and engages autopilot.
Aircraft levels off at 100' or so, with high thrust setting, and exceeds all flap speeds.

What should have happened.

MCP error noticed during take off roll. Ignore it. Fix after take off.
Rotate normally to precalculated target. IGNORE F/D.
After gear retraction, set MCP to correct target.
At 1,000' climb thrust, and FLCH or VNAV
Pitch down a few degrees to get the clean up acceleration started
Engage A/P.

It should have been a total non event. As it was, it's almost as bad as that airline that did the 777 go around without power.

 

[harvyk]

So I realise this is calling for pure speculation here:

Knowing that certain airlines have a policy of zero hand flying (I don't know if EK fits this description):

Could this has been the result of pilots who just are not familiar with hand flying and who are trained to turn the automatics on as soon as the aircraft leaves the runway and therefore rotating without immediately turning on the AP is considered bad (according to the company), or;

Was is just lots of confusion at a critical stage of flight, and even in the "no hand flying" companies hand flying would have been allowed in this situation?

 

[mjt57]

In the paper today, article about pilots and they making mistakes.

Hopefully people (esp. JB and AI) can read it.

Is it more of the media and its penchant for making non-problems sound cataclysmic?

Qantas pilots making errors after long periods without flying

The airline is finding that pilots who have not flown for long periods due to the pandemic are making errors such as commencing take-off with the park brake on.

www.theage.com.au

 

[jb747]

Knowing that certain airlines have a policy of zero hand flying (I don't know if EK fits this description):

I don't know what their policy is, but they've had some amazingly bad incidents attempting to use automatics when it would be so much easier to just look out the window.

Could this has been the result of pilots who just are not familiar with hand flying and who are trained to turn the automatics on as soon as the aircraft leaves the runway and therefore rotating without immediately turning on the AP is considered bad (according to the company), or;

Was is just lots of confusion at a critical stage of flight, and even in the "no hand flying" companies hand flying would have been allowed in this situation?

Autopilots can't be engaged until you reach a couple of hundred feet, and this event was going wrong well before that.

I see it as people who simply cannot fly, and for whom the automatics mask that inability. I suspect they cleaned out the expats, and that included most of the actual pilots.

In the paper today, article about pilots and they making mistakes.

Hopefully people (esp. JB and AI) can read it.

Is it more of the media and its penchant for making non-problems sound cataclysmic?

Interesting timing. On no planet is the EK event just about lack of recency, and the timing of this article makes me wonder.

From discussion with people still flying, QF put out a memo talking about things are are easy to get wrong, and emphasising that until things return to normal that everyone should be taking their time, and following the most conservative path possible. The errors that are mentioned sound like the sorts of things that used to come up in standards reports, and it's worth noting that they only monitored the threat. For instance if an FO missed a switch in preflight, and the Captain subsequently noticed and corrected it, then it would still be a recorded error. All flights are full of errors. I always promised myself that if I ever did a flight that was error free, I'd retire on the spot...and look how long my career ended up being. Errors are not a problem, as long as they are corrected.

 

[flyingfan]

What would you dread to be told over the radio (especially by but not just from ATC)? "Stand by to copy a phone number to call..." or "Indefinite hold" or "Clearance cancelled"?

Been watching some Mayday recently and got me thinking

 

[D747]

Qantas are currently doing some A380 flights that might be for crew training. If that is the case what sort of load would they be carrying? I ask because if very lightly loaded would it be less likely to provide the full gamut of training? I believe JB once wrote on this forum about the tendency for the A380 to float on landing during a lightly loaded ferry flight. Flight 6017 seemed to get airborne very early on 16R out of YSSY today.

 

[jb747]

What would you dread to be told over the radio (especially by but not just from ATC)? "Stand by to copy a phone number to call..." or "Indefinite hold" or "Clearance cancelled"?

Been watching some Mayday recently and got me thinking

"We're sending fighters to escort you to ......"

ATC are an advisory service. They can't make the aircraft do anything. If lack of ATC clearances starts to box you in, then declare an emergency, and do whatever you need to do.

 

[jb747]

Qantas are currently doing some A380 flights that might be for crew training. If that is the case what sort of load would they be carrying? I ask because if very lightly loaded would it be less likely to provide the full gamut of training? I believe JB once wrote on this forum about the tendency for the A380 to float on landing during a lightly loaded ferry flight. Flight 6017 seemed to get airborne very early on 16R out of YSSY today.

Lightly loaded the aircraft does behave differently, but you'll see days with light loads, even in normal times. They can carry fuel to act as ballast if they really want to. It doesn't get really floaty until below about 320 tonnes. Easy enough to have fuel to keep above that.

 

[D747]

I have no idea how accurate this data is but is this a long landing?

 

[Captain Halliday]

I have no idea how accurate this data is but is this a long landing?

When I first saw the screenshot, I assumed it was the SQ landing from last week.

 

[jb747]

I have no idea how accurate this data is but is this a long landing?

On first glance, yes, but I'm a little suspicious about the data. The height at the threshold is quite high, more so than you'd ever accept. So that makes me wonder about the accuracy of the height data.

Looking a bit further...there were three flights yesterday. The last of the three is quite normal, and I found a video on youtube, showing a nice approach and landing. The other two were the earlier of the three, and both show a quite similar profile, with a high threshold height, longish landing, and a turn off further down the runway than normal. One possibility is that there were two different instructor pilots, and the one involved in the first two flights may have given the 'trainee' an imaginary displaced threshold. Basically he's said, "they're doing runway works, and the first 1,500 feet is not available". In the real world, that happens quite regularly, and it's a sim requirement for the Captains just about each time they visit.

I'll ask and see if there's any more to find out.

 

[D747]

On first glance, yes, but I'm a little suspicious about the data. The height at the threshold is quite high, more so than you'd ever accept. So that makes me wonder about the accuracy of the height data.

Looking a bit further...there were three flights yesterday. The last of the three is quite normal, and I found a video on youtube, showing a nice approach and landing. The other two were the earlier of the three, and both show a quite similar profile, with a high threshold height, longish landing, and a turn off further down the runway than normal. One possibility is that there were two different instructor pilots, and the one involved in the first two flights may have given the 'trainee' an imaginary displaced threshold. Basically he's said, "they're doing runway works, and the first 1,500 feet is not available". In the real world, that happens quite regularly, and it's a sim requirement for the Captains just about each time they visit.

I'll ask and see if there's any more to find out.

Really appreciate it JB.

 

[jb747]

A little more on the training that's being done. On completion of the sims, each of the Captains is doing two observation trips, and two sessions in the aircraft.

 

[Quickstatus]

Comments appreciated.

How much of a disruption will this be.

Just remind me again -
When is auto land typically used?

1)Autoland training
2) visibility below minimum decision altitude
3)
4)

 

[ayebee]

Comments appreciated.

How much of a disruption will this be.

Just remind me again -
When is auto land typically used?

1)Autoland training
2) visibility below minimum decision altitude
3)
4)

re the 5G interference question ...
Juan Browne (B777 pilot) on blancolirio youtube channel did quite the deep dive on that a while back.

My basic assumption is the FCC (ie mobiles) will go ahead because of the $$$, and the aviation industry will need to suck it up.
Maybe the technology can solve frequency bandwidth filter leakage, but that could be expensive for the existing fleet.
A specific risk identified was from 5G phones on board, so would expect the requirement to have all devices on flight mode enforced more rigorously.
In my last travels in US (early 2020), my observation was that many ignored requests to turn off cell phones as "what possible harm can it do & of course I'm special" so maybe will take some doing to get the freedom loving passengers to comply.

I would be interested in the view of the pilots here about the doubts this may introduce when you can't see the runway.

 

[jb747]

How much of a disruption will this be.

Just remind me again -
When is auto land typically used?

1)Autoland training
2) visibility below minimum decision altitude
3)
4)

It's used in poor weather. At night. When you feel like it.

It has it's own minima, which may be zero, but is generally 100-200' on an RA.

A low vis approach is not the same thing as an autoland.

The radalt is needed to measure the minima, and also to control the flare sequence. Loss of RA will result in a go around.

I would be interested in the view of the pilots here about the doubts this may introduce when you can't see the runway.

If something affects the radalt, it most likely won't affect all of them equally, so you'll most likely end up with a detected system error. The aircraft response to that could be almost anything, including dropping the autopilots completely. Potentially dangerous, but with a low incidence. Just like the Max I guess.

Seems to me that ATC needs a 5G kill switch. If they start autoland ops, then they kill 5G within few miles.

 

[Saab34]

I was watching a video with that United twin which had engine issues/parts seperated and was shaking fairly violently on the wing. Also the AirAsia A330 which flew for hours with a vibrating almost ‘jumping’ engine on the wing.

My question is, if the engine gives way and falls off, what happens to the aerodynamics of the aircraft. Would it be expected to just spiral out of control, or is glide still available (although one wing more heavy)