News:

Precision Simulator update 10.173 (24 February 2024) is now available.
Navburo update 13 (23 November 2022) is now available.
NG FMC and More is released.

Main Menu

G/S beam capture at hard constraint at high OAT

Started by Hardy Heinlin, Mon, 3 Jun 2019 11:51

Hardy Heinlin

Moin,

as we know, when the current OAT disagrees with the ISA temperatures, QNH altitudes are not true altitudes; when we follow a barometric flight plan profile, and the air is colder than in the ISA, we will be below the planned height above the ground, or, respectively, above it when the air is warmer.

More and more approaches will be designed as continuous descents with no level-off segment before the G/S capture. How do you solve the temperature problem when you approach a hard altitude constraint while the G/S needle remains a dot below your barometric profile due to hot weather?

Example: MMMX ILS05R, airport OAT 25°C, continuous descent to 8800 ft at which the G/S beam will be 120 ft below the aircraft due to non-ISA OAT. 


Regards,

|-|ardy

emerydc8

I don't know how they're dealing with it today, but 20 years ago on really hot days, the feds were violating guys for following the glideslope on the CIVET arrival into LAX RW25L. If I recall correctly, CIVET was 52DME on the ILS and we had to cross it at or above 13000. If you picked up the glideslope before that and followed it in (a pretty common practice), it would sometimes put you below 13,000. The only solution was to level off at 13,000 until you crossed CIVET.

It wouldn't surprise me at all if the chart designers gave exactly zero consideration to this issue. Why should they bother now? Most of the arrivals they've designed require speed brakes at some point just to get the airplane to comply with the speeds anyway.

Hardy Heinlin

Maybe ATC will report a modified QNH, temperature corrected for the hard constraint at the G/S capture point?

But then during landing the lower altitudes and the MDA will be wrong.

emerydc8

Before there was VNAV (747-200, DC-8, DC-10, etc.), we used to do the equivalent of continuous descents on virtually every leg just by doing the math for a 3-to-1 descent and using V/S. Now that we have VNAV, hardly an arrival goes by that doesn't require the use of speedbrakes.

Maybe I'm looking at it wrong, but wouldn't any use of speedbrakes on a descent actually be a descent where fuel is wasted because you stayed at cruise longer than you otherwise would have? If you stayed at cruise for one extra minute and burned 300 pounds of fuel during that minute, you could have spent that minute descending at idle thrust with a much lower fuel burn and still been in the same position without having to use speedbrakes. So, the way I see it, the chart designers are creating charts that cause VNAV to keep you high and/or fast and actually waste fuel. I thought the original selling point of VNAV was to save fuel.

Jeroen Hoppenbrouwers

I am pretty sure that VNAV and steeper (with speed brakes) approaches are not related, just happen to occur at the same time (same decade). VNAV most certainly wants you to descend with no speed brakes. That is its primary function. But busier airspace with all kinds of traffic above you and under you, necessitates more 'square' profiles to assure you get over A and under B, for all kinds of aircraft. The heavier ones will need speed brakes most of the time, alas. It's the vertical variant of zig-zagging over VOR beacons while RNAV/GPS allows you to go direct.

Probably you still burn less fuel when being kept high a bit longer and then dive for it, compared to being talked down 100 nm early to get under the departures.

"NextGen will solve this all with 4D trajectories."


Hoppie

emerydc8

#5
If your planned descent speed is the same in both cases (300), I don't see how being kept higher and using speed brakes to arrive at the same position saves you fuel , versus an idle descent at 300 knots without the need to make a dive for it.

In any case, I'm not even convinced these ridiculously complex arrivals are necessary to keep up with the traffic flow and density. There are only so many airplanes that can land at an airport within a certain time period anyway.  I think the chart makers have gone over the top on a lot of these arrivals. Maybe they should be required to fly them after they've been up all night. Miss an altitude or speed restriction? Go two months without pay and have a letter in your file for life.

skelsey

Quote from: emerydc8 on Tue,  4 Jun 2019 01:29
If your planned descent speed is the same in both cases (300), I don't see how being kept higher and using speed brakes to arrive at the same position saves you fuel , versus an idle descent at 300 knots without the need to make a dive for it.

Exactly - from a theoretical point of view at least any use of speedbrake is 'throwing away' excess energy which can only have come from burning fuel.

Far be it from me to defend the chart designers but I had heard about the issue at LAX in the past. My understanding if I recall correctly is that in this specific instance the CIVET restriction is airspace related: i.e. if you cross CIVET 'low' (per the altimeter) you potentially may end up in conflict with, say, a VFR aircraft flying quite legitimately (though undoubtedly foolishly!) below -- obviously they will be referencing their altimeter to ensure they do not bust the CAS and care not about the geometric position of the LAX G/S which is fixed in space but not in relation to the altimeter. I stand to be corrected if my memory is playing tricks though.

I do sympathise with your overall point, however!

cagarini

That's why we, in Meteorology, also use QFF, unfortunately not adopted in Aviation, other than to complicate the live of the poor ATPL students in their exams :-)

I'd say that there should be a minimum barometric level bellow which you should be allowed to follow the G/S indications, irrespective of the geopotential height effect due to T.

Hardy Heinlin

Say, a well-designed constant descent has brought you to the final hard constraint where you are supposed to capture the G/S. But because it's a hot day, the G/S is 100 ft below you.

Under such hot conditions, will ATC allow you to capture the G/S a bit earlier? Say, 1 nm before the constraint?

I mean, a G/S capture within 9 nm from the runway. Not 20 nm. Of course, the farther away, the greater the altitude error. You shouldn't cross a hard constraint 200 ft below it. At 10 nm it's just 100 ft below. At 5 nm it's 50 ft. A lot less, but still great enough to miss the G/S capture.

And ... do you -- or ATC -- anticipate the fact that a short "dive" will be necessary, and therefore reduce the airspeed a bit earlier if you know the weather is hot? Or is this question too academic? :-)


|-|

Hardy Heinlin

The same problem actually occurs in cold weather, when the G/S is too far above the barometric VNAV path. In that case not even a level-off segment prior to the capture point will help.

You are at the final hard constraint, the G/S is above and not captured, and you now start the VNAV on-approach descent. The G/S will be captured only shortly before landing. The closer you get to the airport elevation, the smaller becomes the altitude error.


|-|

Jeroen Hoppenbrouwers

Quote from: emerydc8 on Tue,  4 Jun 2019 01:29
If your planned descent speed is the same in both cases (300), I don't see how being kept higher and using speed brakes to arrive at the same position saves you fuel, versus an idle descent at 300 knots without the need to make a dive for it.
True, however this is not what I meant.

There should be quite a difference between gliding down late with speedbrakes, and gliding down early followed by a long segment of level flight at low altitude.

If you must choose between boards out versus throttle up, both for more than just a few seconds, because the idle glide without speedbrakes has been made impossible due to air traffic restrictions, I suppose the late descent is preferable. Is it?

I'm not a controller at all! I presume that some of the procedures have been designed to simplify controller's life at the expense of fuel economy and pilot's load. Probably inspired by the idea that one controller has 15-20 planes to take care of while two pilots just do one. Tunneling all aircraft through the same bent pipe in a way that is "guaranteed to work for each type" relieves controllers from per-aircraft details, but it comes at a cost. The NextGen system is supposed to automate the piping mess and to allow aircraft freedom of path and profile as long as they stick to their promise. I'm not saying this will work as advertised, mind you.

Just guessing.


Hoppie

Hardy Heinlin

Sure, 5 nm level flight at optimum altitude is more economic than 5 nm level flight at lower altitudes. (If the wind is not a factor.)

So if I had to decide whether to start the descent 5 nm after the T/D or 5 nm before, I would take the late descent.

emerydc8

QuoteIf you must choose between boards out versus throttle up, both for more than just a few seconds, because the idle glide without speedbrakes has been made impossible due to air traffic restrictions, I suppose the late descent is preferable. Is it?

Not to me. From an operational point of view, if the STAR had to be screwed up on the top end or the bottom, I would rather start down a few miles early and have a level five-mile segment at the bottom where I can think about configuring rather than using speed brakes (waste fuel) at the midpoint on descent to make the altitude.  You can't add flaps above FL200 anyway. Putting the gear down at FL260 for drag and to cool the brakes is just a survival technique from the old Classic days going into KDH, KBL, and BAG. Nothing like pitching more than 10-degrees nose down to maintain 250 knots at FL260.

Rarely are you going to get a constant descent all the way to touchdown anyway, as much as the chart makers dream about making this the standard. They aren't standing in the control towers at JFK,  ORD or ATL.  At busy airports, you are going to be given vectors, speed and altitude assignments all over the area as they line you up in the conga line, notwithstanding the "fuel efficient" chart you just came in on.




Hardy Heinlin

Can we conclude that from the pilot's operational point of view, an early descent is better than a late one; and regarding fuel saving and noise abatement, a late descent is better than an early one; and for the best compromise, the T/D is the optimum starting point? :-)

emerydc8

I would rather be down 5 miles early than 5 miles late. One go-around during a career from being too high would probably be more than all the fuel you would spend descending 5 miles early on every descent. Just my 2 cents.

Jeroen Hoppenbrouwers

Just to keep nagging and whining, I understood that in many cases early descents for ATS reasons are more like 25-50 miles early.


Hoppie

emerydc8

#16
Unless they need you down to get you under some conflicting  traffic, if they start you down 20 or 30 early you can just use altitude intervention or press DES NOW and execute. This will give you a 1250 FPM descent until you capture the original path. If the rate is not high enough for ATC, they'll let you know, in which case FLCH SPD and running the MCP speed up to close to BP will usually do it.

andmiz

There's two good airports that highlight this problem (realworld).  25R ILS into Hong Kong, and as Hardy mentioned, the 05R ILS into Mexico City.

HKG has a 4500' final approach point where you must have captured the glideslope by; active awareness that a higher than ISA temperature will leave you above the glidepath and cause an inadvertent ALT capture and kill any necessary rate of descent required is important to prevent leading yourself into a missed approach.

Going into MEX we fly the base to final turn in LNAV/VNAV prior to arming the LOC/GS.

Hardy Heinlin

When you are aware of this ISA deviation, will you disconnect the A/P a mile or two before the FAF, ignore the VNAV F/D, and move the nose towards the actual G/S?


|-|

andmiz

Generally no; we want to follow FD commands.  If it's not doing what we want it to do, then the solution is to correct it with an appropriate mode such as V/S (unless it's obvious that manual flight is a safer option).  Disconnecting to ignore the FD would otherwise increase workload, as the PF will not only have to then fly manually, but should also direct the PM into the correct mode settings to refollow the FD.

The solution is to be aware of the ISA deviation and have a shallow descent segment perhaps a few nm long prior to the FAP where you'll be below the projected glideslope, so that a capture from below is assured.  That won't occur in VNAV; active intervention with another mode is required.