Disc A/P @ 1000 feet for landing... tips?

[scandinavian13 5,192]

21 hours ago, Ephedrin said:

lol... I love it how you express your endorsement with our admired control freaky EU :D

haha - to a certain degree, I understand it. Given the complexities of the airspace, and the concept of CFMU, with agreements between all of the ACCs between the various countries, it makes some sense to add a little predictability to the system by magenta-lining from dep to dest, and not allowing pilots (and therefore controllers) the variance that comes from pilot-sep visual approaches. We do have the benefit of quite a vast amount of airspace over here, controlled by the same entity, with less population center density (with the exception of the East Coast BOS-NYC-DC corridor, perhaps).

[skelsey 765]

1 hour ago, scandinavian13 said:

haha - to a certain degree, I understand it. Given the complexities of the airspace, and the concept of CFMU, with agreements between all of the ACCs between the various countries, it makes some sense to add a little predictability to the system by magenta-lining from dep to dest, and not allowing pilots (and therefore controllers) the variance that comes from pilot-sep visual approaches. We do have the benefit of quite a vast amount of airspace over here, controlled by the same entity, with less population center density (with the exception of the East Coast BOS-NYC-DC corridor, perhaps).

Quite.

The one thing I would say is that I don't envisage how any benefit would be obtained from running visual approaches -- the limiting factor at most airfields is runway occupancy time, especially with heavies, and innovations like brake-to-vacate on the A380 do help significantly with that. Otherwise, places like Heathrow are packing them in as close as 2.5NM apart on the ILS if wake separation isn't an issue (and perhaps even closer than that these days now we have time-based separation) so I can't imagine that a visual could, from a practical point of view, get the aircraft any closer together without resulting in go-arounds because the previous aircraft hasn't vacated or with the consistency that a skilled radar controller can achieve. Point Merge (which is essentially the "magenta line" type of RNAV operation you are referring to) is slowly gaining more ground and is being sold as providing huge capacity benefits and reduction in holding times, but I think there's still some way to go to convince people it can be better than a skilled radar controller... 

There's even more work going on to get the departures out faster as well -- a controller friend of mine says they will soon be getting a 96x96 matrix of aircraft types, (up from the current 5 broad weight categories) each with specific departure separation times which obviously have to be hit as closely as possible to maximise the throughput -- for example, a 737-200 will now be considered to be in a different category to an NG, so if you depart an aircraft after a -200 you might only need 1'20" separation instead of 1'30" - etc. All of this (as it does now) will have to be essentially memorised and put in to action by the controllers in real time -- it's enough of a feat now with just five wake categories, so it really makes you realise quite how much difference the guys in the tower make and how much is involved -- it is very far from just being first-come, first-served and launching them as they arrive at the holding point!

[jrw4 624]

Just for reference, controllers in FAA-land cannot issue speed adjustments inside of the FAF. A typical approach clearance including the speed constraint (for 31R at KJFK, for example) would then be "ABC123, eight miles from ZULAB, turn left heading 340, maintain 3000 until established, cleared ILS runwy 31R approach. Maintain 170 knots until ZULAB". That last is critical because all previous speed adjustments are cancelled when the approach clearance is issued.

I guess 31R has a fairly typical ILS path (but someone else might have different info), so the FAF is at 5.3 DME and 1888 AGL. At a typical descent rate of 800 fpm on the GS, the crew only has about a minute to stabilize the approach while simultaneously deploying the gear, lowering the flaps to 30, and reducing airspeed by 25-30 kts. I wouldn't be surprised if this is a bit more than most desktop captains can do reliably, but there's no harm in cheating by having the gear down and flaps at 15 at the assigned airspeed prior to the FAF. Yes, it burns extra fuel and creates more noise, but then all you need do at the FAF is lower the flaps two notches and crank back the airspeed to Vapp. 

Issuance of airspeed adjustments on final isn't a magic sauce for increasing a runway's landing rate, since the a/c need to slow down to land and taxi off of the runway, but it does help a controller maintain even spacing. Since the time is short between issuing the approach clearance and the FAF, one can't make up for gross disparities in spacing this way. In the real world, I have landed a C172 in front of a flight of Air National Guard F4s. Tower told me to keep my speed up (hah!), but those guys must have been doing S-turns halfway across the state.

 

[787WannabePilot 12]

Just a small update,

I haven't flown in a few weeks because of work constraints. I decided to take a flight from the NYC region to Fort Myers yesterday (escape the snoowwwwww). Using some advise in this thread, I made one of my best landings (disc A/P at 1000 feet) even with wind at 10 with gusts up to 20. 

Appreciate the advise!

[achutchison 26]

The stable by 1000 feet rule is a little bewildering to me, considering there are airports where you’ll to make the turn to final well after 1000 feet. One example in the US is DCA runway 19. I think its a good thing to have pilots handflying a bit more, especially when workload is low. It’s important to keep that skill alive. It seems like the 1000 feet rule is just to provide insurance for loss in basic flying skills because pilots in the EU can’t get any good practice due to the operating policies.

[skelsey 765]

45 minutes ago, achutchison said:

The stable by 1000 feet rule is a little bewildering to me, considering there are airports where you’ll to make the turn to final well after 1000 feet. One example in the US is DCA runway 19. I think its a good thing to have pilots handflying a bit more, especially when workload is low. It’s important to keep that skill alive. It seems like the 1000 feet rule is just to provide insurance for loss in basic flying skills because pilots in the EU can’t get any good practice due to the operating policies.

Who said anything about hand flying? 

Nothing stopping you from hand flying from FL350 - provided you are configured, on path and on speed at 1000R. Obviously for circling approaches/airports where there is a late turn to final the 'gate' will be modified appropriately. 

It's got nothing to do with hand flying vs automation and everything to do with stopping people from going off the end of runways because they thought they could salvage a rubbish approach.

Read the accident report for almost any runway excursion and you will find that the crew were high/fast/rushed but thought they were good enough to pull it off.

Setting a hard and fast bottom line takes any subjective "I reckon I can do this" out of it: you either meet the criteria or you don't. If you don't, you go around, or your PM will be insisting that you do so. Whether you have the autopilot engaged or not is irrelevant.

[downscc 1,327]

Strongly agree..., if you cannot be stable in proper configuration by 1000 AGL then I'd suggest piloting skills might need polishing.  Given a 3 deg slope and 140 kt GS you are only 3 nm and 2.3 min from landing.  DCA 19 is a fairly unique example, but even there you should be established by JUBOL (RNP19)  at 955 regardless of the heading changes as you approach the runway threshold.  If you are flying the LDA your MAP is at 743 feet and if you not established by then you need to go around.  Tooling around in the single engine Thudpucker you can get away with lower alitudes, depending on winds you might use a rule of thumb such as 300 AGL but you should have a self imposed rule based on your qualifications and stick to it.

[achutchison 26]

53 minutes ago, skelsey said:

Nothing stopping you from hand flying from FL350

RVSM rules do.

 

54 minutes ago, skelsey said:

It's got nothing to do with hand flying vs automation and everything to do with stopping people from going off the end of runways because they thought they could salvage a rubbish approach.

That's poor aeronautical decisionmaking if you try to salvage something you have doubts about.

I certainly wasn't suggesting there should be no stabilized approach criteria, however 1000 feet is very conservative and restricts the options if the event of a late approach clearance. Perhaps I worry a bit too much about automation dependency, however I think it was relevant to mention considering the original topic of the thread. This opinion is reflected in the policies of many airlines in the United States, as well as FAA findings and recommendations.

[skelsey 765]

1 hour ago, achutchison said:

RVSM rules do.

In level flight, yes (although there is some hair-splitting to be done about whether the autopilot needs to be engaged or whether "available" is enough -- though I would suggest the spirit rather suggests it should be on!); in climb or descent, however, as far as I read the letter of the rules you can knock yourself (and the A/P!) out (individual operators may have more specific/restrictive policies, of course).

1 hour ago, achutchison said:

That's poor aeronautical decisionmaking if you try to salvage something you have doubts about.

I quite agree -- however, sadly the accident statistics suggest that people tend to overestimate their own ability (in all walks of life -- if you ask a hundred people to assess their driving ability, the vast majority will tell you they are above average: clearly it is impossible for everyone to be above average!) and aviation is no exception: without a bottom line, people will invariably push the boundaries with the inevitable consequences. One of the big issues of course is the role of the monitoring pilot, who in any country or culture may be a newly-minted FO flying with a 20,000 hour captain: setting some measurable criteria and a hard gate (whether that's at 1000 feet, 500 feet or 50 feet) allows any crew member to make an objective assessment of the approach and have the confidence to call out a deviation from the criteria, rather than succumbing to the "I've/he's done this before and it worked out OK last time" mentality.

Now, we can argue about whether 1000R is the right 'gate' -- as I say, it's not for all approaches by any means but the principle of setting a suitable gate remains a solid one. As I say, many airlines will use 1000 ft IMC and 500 ft VMC anyway which gives some flexibility. Even with something like the LDA to 19 at DCA, there's no reason why one shouldn't be able to comply with standard stable approach criteria by being:

• Established laterally on the LDA
• On the published glide path
• In the landing configuration
• At Vref + additive with engines spooled & speedbrakes stowed
• With all checklists complete

By 1000R (i.e. around WEVPU). Why would you need to still be steaming in at 160+ kt and running out flap as you go round the final turn at 500 ft?

1 hour ago, achutchison said:

Perhaps I worry a bit too much about automation dependency, however I think it was relevant to mention considering the original topic of the thread. This opinion is reflected in the policies of many airlines in the United States, as well as FAA findings and recommendations.

I don't fundamentally disagree with you re: automation dependency and I do see where you're coming from in so far as by no means should "stable at 1000R" mean "leave the automatics in until 1000R" as perhaps may have been inferred earlier up the thread. However, I also think it's important to make the distinction between the desired outcome (stable approach) and the method (which may or may not involve automatics).

As I've probably said before: for me, the most important thing when it comes to the automation is that it is viewed as a tool to be used appropriately and I would suggest most (not all) enlightened airlines share the same view. There are some situations where it is inappropriate to be faffing with the automatics (aeroplane not pointing where you want it to point being chief amongst them) and clearly pilots should be comfortable hand-flying the aeroplane. However, there are also situations where it is inappropriate to be hand-flying for the sake of it: in increasingly busy skies there is a lot of thinking and decision-making to be done in an airliner flight deck and in a multi-crew situation hand-flying saps both crew-members' capacity: one bloke is concentrating exclusively on poling the thing around whilst the other can end up working like a one-armed paper hanger doing everything else. Not always a terribly good division of labour or conducive to maintaining awareness of the bigger picture!

[achutchison 26]

2 hours ago, skelsey said:

As I've probably said before: for me, the most important thing when it comes to the automation is that it is viewed as a tool to be used appropriately and I would suggest most (not all) enlightened airlines share the same view. There are some situations where it is inappropriate to be faffing with the automatics (aeroplane not pointing where you want it to point being chief amongst them) and clearly pilots should be comfortable hand-flying the aeroplane. However, there are also situations where it is inappropriate to be hand-flying for the sake of it: in increasingly busy skies there is a lot of thinking and decision-making to be done in an airliner flight deck and in a multi-crew situation hand-flying saps both crew-members' capacity: one bloke is concentrating exclusively on poling the thing around whilst the other can end up working like a one-armed paper hanger doing everything else. Not always a terribly good division of labour or conducive to maintaining awareness of the bigger picture!

I completely agree with this, beautifully said too!

[scandinavian13 5,192]

14 hours ago, downscc said:

DCA 19 is a fairly unique example, but even there you should be established by JUBOL (RNP19)  at 955 regardless of the heading changes as you approach the runway threshold.

Definitely not.

1. Try it in the sim
2. Come visit and I'll take you to Gravelly Point Park and you can see how much of a handful it really is, and how late the turn is really made. All day.

[downscc 1,327]

6 hours ago, scandinavian13 said:

Definitely not.

1. Try it in the sim
2. Come visit and I'll take you to Gravelly Point Park and you can see how much of a handful it really is, and how late the turn is really made. All day.

I misappropriated the term established..., I do not mean established on runway or final course.  Of course not, and yes...I have flown the simulator in/out of KDCA and been there real world but that was years ago when I was still on active duty.  My assumption is that established means the aircraft configuration is in landing configuration.  The FCTM allows this to be as low as 500 AGL for a visual approach but I advocate that in most cases the final 1000 ft and 2 min should be stabilized.  Obviously the DCA 19 arrival turns at the last 200-300 ft simply because one has to avoid the prohibited areas just North of DCA.  I assume the aircraft is in landing configuration before making that final turn into the flare.

[JoeDiamond 67]

Stable is typically defined as configured, on speed, power stabilized and sink rate within limits.  You can make turns below 1000 AGL all day long as long and be considered stable as you meet those criteria.  Most guys flying the approach to 19 in DCA are stable well before 1000 AGL, I usually shoot for 1500.  There is a lot going on during that approach.  Getting stable early gives you one less thing to worry about later.

[Ephedrin 418]

Fly it like a Cessna. Seriously. You’re in a Simulator, just fly around... first with A/T at 220 kts, flaps 5 or so and then conpletely manual... that‘s how I get a feeling for every new addon. You don‘t learn anything if you just glide it to the runway im landing config. Learn to fly your plane as a pilot, not as a passenger.