Flying Approaches for Dummies

Version 4, May 13, 2021

TECHNICAL – FOR PILOT’S EYES ONLY

QF32 follower Riccardo Parachini writes:

Hi Richard,

I’d like to ask a curiosity. Cadet pilots starting their career on a big jet (say A320, A330) sometimes have trouble in energy management on approach. Airbus publishes some diagrams which are very good but specific for a “perfect day” and maybe no ATC speed restrictions. Do you have some tips to plan approach good and flap/gear schedule?

Thanks for your question Riccardo. This is a big topic, but I list below my basic thoughts.

Photo by John Feder – The Australian.

Approach Altitude Gates

Aircraft manufacturers published gates are a ideal targets (ATC restrictions permitting).   Know these gates habitually.   Examples for the big jets include:

  • B747 / A380: 250kts / 10,000′ / 40nm  clean. 
  • A320: 250kts / 10,000′ / 30nm clean. (Thanks to my friend Captain Eric “Cap’n Aux” Auxier who writes, “I learned this gate early on, and still use it today – about 10nm less than your A380 behemoth, I’m sure because there’s a lot less mass to slow and get down!
  • Be at FLAPS 1 extension speed at 3,000 feet at 10nm (intercepting the G/S).

Deceleration Rates (clean, idle, level)

Know your aircraft level deceleration rate in knots per second.  For the 747 and A380 it is 1 kt/sec.   This helps you calculate the length of the level deceleration phase during your approach.  Add this deceleration distance when calculating your profile.

Getting to the 3,000′ gate

I have a simple way for you to guarantee arriving at your 3,000 ft gate at 10nm at 3,000′ to start the approach.

Within 25nm, have your systems configured to display the distance to run to the 3,000′ 10nm intercept point. This is an incredibly valuable distance to know. (Airbus pilots: when programming the STAR, enter a discontinuity after the last STAR waypoint, followed by the 3,000′ fix).

If you descend at the standard rate of 3nm per 1000 feet, then, divide this displayed distance by 3 to get the altitude you must lose to get to the 3,000′ point. For example, if you have 10nm to run to the 3,000’point, then you should ideally be 10/3 = 3,300 feet above the 3,000 intercept point, so you should be at about 6,300 feet.

If you regularly recalculate this dynamic “mental dead reckoning” ideal height in your head as you approach the final fix, then you will:

  • know if you are high/low and calculate the ideal instantaneous ROD for your current location
  • be able to update your profile as ATC vectors you around the sky!

Intercepting Glideslope

The many parameters of speed, height, glideslope deviation, configuration, and rate of descent confuse many pilots during approach. 

I reduce this complexity by reducing the number of parameters.  When approaching the cone of the ILS approach, providing you are above the minimum altitudes, get onto the glideslope first and get your self going down at the required rate of descent (half groundspeed +xx…).  When you are locked onto the glideslope, the only parameters left to resolve are the localiser and airspeed.

When flying an ILS to any runway, I pick up the ILS glideslope early and descend on it.  For the remainder of this approach, you just then need to manage thrust, drag and flaps to reduce slowly to the final approach speed.

Follow your manufacturers guidance about flap extension during approach. As a guide for a relaxed approach in an A380, I plan:

  • ILS from 3,000′ – intercept the glideslope with FLAPS 1 running.
  • ILS above 3,000′- intercept glideslope as detailed above
  • ILS from below 3,000′- because there is less time to become stable, consider being partially/fully configured before intercepting the glideslope.
  • If you expereince any malfunction that affects your situation awareness, free mental space or performance, then extend the ILS/GLS intercept out from 10nm to give you more stable time on the glideslope. (During QF32, we required and intercepoted final at 20nmn)

Vertical Offsets

  1. If high/fast, be unafraid to use speed brakes, gear and flaps to slow down (ensuring you do not break any speed limits).  Be unafraid to operate these surfaces up to, but not exceeding, their max speeds.
  2. Flaps and Slats increase the coefficient of drag (Cd) more than they increase Cl.  So if high on final approach, its better to slow down to enable deploying flaps earlier than to speed up and rely on increases drag (at a lower flap setting) to get you down.
  3. Know the restrictions of using speed brake with flaps.   If it is permitted then use them, being aware that speed brakes decrease the wing’s Cl and so increases the minimum speeds.

Rate of Descent during final approach

During the approach, given the glideslope angle and your approach speed, you should be able to dynamically calculate the required rate of descent during that approach.

It’s important to be able to calcualte your expected descent rate. Because when you set and maintain an attitude (the first loop to maintain this calcualted ROD, and then to keep on the glide slope), then the aircraft will be more stable than if you “chase” the glideslope position index.

Note 1: If you imagine the ROD as a “velocity”, and the glideslope index as a “position”, then you can see why flying to set the derivative of position is much more accurate than flying to set a “position”.

Note 2: On an Airbus aircraft on a smooth day, you can be incredibly accurate flying by taking your hand off the sidestick, and hitting the sidestick to make tiny inputs to the flight control software. This is because Airbus sidesticks have an extraordinary integrator function, that converts a short-sharp lateral input into a reequest for a small change in bank angle (yes, bank angle, not roll rate). This integrator calculates a desired bank angle according to the size & duration of the stick input (Ref 1). I found that two hits to the stick laterally are converted into a roll change of 1/2 degree. Think of these “hits” as the third derivative of position, that is acceleration. Note, this feature is not documented in Airbus manuals.

Note 3: Certainly, the new flight directors smooth out the required inputs to deliver a constant trajectory, but they only work when correctly coupled to the correct flight mode at that time, and sometimes, they fail.

Given a 3 degree glideslope, the target rate of descent is equal to the addition of:

  • GS x 5 (or halve GS then add a zero), plus at least
  • GS / 3

To keep it simple (but to 99% accuracy), just remember one of these rules for your aircraft type.  If your normal approach ground speed is:

  • 80 kts, then ROD = 400 + 25 feet/min
  • 100 kts, then ROD = 500 + 33 feet/min
  • 120 kts, then ROD = 600 + 40 feet/min
  • 150 kts, then ROD = 750 + 50 feet/min

Note: These ROD figures are correct at the threshold. To be accurate, due to the curvature of the Earth, to remain on the G/S, increase the required ROD pro rate with distance from the threshold – from 0% at the threshold to 5.3% at 10nm, to 7.7% at 15nm to run.

Here is my spreadsheet of data for a 3 degree ILS/GLS approach flown at 150 kts. Notice the increased required ROD due to the Earth’s curvature. Contact me if you wish to receive data at a different glide slope angle or ground speed.

Dist to Rwy (nm)0123456789101112131415
Alt (flat Earth ft)0318637955127415921911222925472866318435033821414044584776
Alt (ft (curved Earth))0319640963128816141942227226042937327336093948428946314975
ROD (ft/min)794798803807812816820825829834838842847851856860
Incr ROD0%0.6%1.1%1.6%2.2%2.7%3.2%3.7%4.3%4.8%5.3%5.8%6.3%6.7%7.2% 

Stable Approach

When you stand in a court of law explaining why you were involved in an accident, playing victim to air traffic control is not an option.

Most landing accidents occur at the end of unstable approaches.  

Know your requirements for a stable approach, and go around if they are not met.   The good airlines have a culture of no jeopardy for go arounds (though they might like to know the reasons for their records).

If your airlines ‘stable” requirements vary for Instrument (higher) and Visual (lower) conditions, then (to maximise situation awareness) it’s a good idea to verbalise your stable status when passing the higher of the two limits

Air Traffic Control or Air Traffic Service?

I think Air Traffic Control should be renamed to Air Traffic Service.

Refuse any instructions that you cannot achieve or that you think will compromise safety.

When things go wrong, tell ATC what you want to do. ATC wants to help you, so your requirements make it easy for ATC to help and clear the skies for you.

PAN PAN PAN, Qantas 32, engine failure , maintaining 7400 and current heading…… we’ll keep you informed and get back to you in five minutes…

(QF32 page 163)

You are in charge of the aircraft, so take control of it. Do not become a victim to ATC. When you stand in a court of law explaining why you were involved in an accident, playing victim to air traffic control is not an option.

Summary

These are my simple rules that I use when planning an arrival and approach.

In my 45 year flying career, I observed many pilots flying missed approaches for many good reasons. Though I have no problems with, and applauded many others for going around, during my 33 year QF career I proudly watched just two of my first officers go around (because Dubai’s runways were occupied). Curiously and (especially) fortunately, I never needed to fly a missed approach outside the QF simulator.

See also:

https://calculators.io/curvature/

Curvature of earth (miles depression) = r – sqrt(r^2 – dist^2) = 3863 – Sqrt(3863^2 – dist^2) (Simple to calculate using Pythagorean theorem)

Here is an excellent SKYbrary article listing flights that experienced approach and landing errors:

References

Ref 1: Captain Jacques Rosay, Airbus Chief Testr Pilot, presentation at A319/A320/A321 Operational Liaison Meeting – Seville 2000

6 comments

  1. qantasboeing747 · · Reply

    Good morning Richard, recently I have been looking at QF32 and I have noticed there is (now) an airport 0km away Changi, called Hand Nadim airport, with a slightly longer runway that changi airport’s. Why (if it was there at the time) not go there?
    Also, do you have flight simulators? Personally I recommend Xplane11 if you have $60 to spare. Oh, and you should buy a joytick and blah blah here I go again…..

    1. You are right. Batam (Hang Nadim) (WIDD) airport was underneath us, biut it is a simple military base with:
      – just a 45m wide runway
      – Dubious 22 VOR DME (3.45 deg!)approach for rwy 22.
      – Minor ammount of support for an A380 (in fact A380 Engineeers would need to boat to Batam if we landed there)

      I did raise the possibility of flying to other airports (including Batam) before we decided to return to Singapore. Everyone agreed Singapore was the best operational choice.

      I don’t have time to fly the flight simulators. I prefer the real thing!

      1. qantasboeing747 · ·

        The other day, when I was in X-plane 11, I forgot to put flaps on, and took off a 737-800 without flaps. How is this possible? Technically you could, but what makes it impossible to takeoff without flaps in real life?
        If you have a flight sim (the ones used for training pilots) can you ask/try taking off without flaps if possible please?
        I really think you might.

      2. Challenges to taking off with no flaps are:

        1. the wing’s Cl will be low, so speed will be high and the tail might impact the runway when the aircraft is rotated for takeoff
        2. Runway length is limiting. Vr^2 = 2as So for a constant acelleration (a), the ground roll is proportional to the Vrotate squared.
        3. Brakes might not be able to absorb sufficient energy to stop in the event of an aborted takeoff
        4. Tyres might exceed their maximum speed
        5. One of the takeoff commit speeds mnight have to be below Vmcg, that means if an engine fails below Vmcg, the pilot would have to find something to crash into to stop the aircqraft. (See the chapter on Wombat Airlines in QF32 for more details) Commercial aircraft are never permitted to operate under these (RAAF) conditions.

        I guess that the Clean stall speeds are 130% the takeoff flap stall speed. That means that the Clean takeoff ground distance is 169% of the normal takeoff flaps distance.

        The Airbus A380 Takeoff Configuration Test simulates the application of Takeoff Power and triggers any of 5 Warnings and 17 Cautions if necessary. The test is applied again when takeoff power is set, checking:

        1. Slats/Flaps
        2. Rudder/Pitch Trimn
        3. Speedbrakes
        4. Sidestick
        5. (any) Door open
        6. Brakes
        7. Generators
        8. Oil Temp
        9. Flight Control Computers
        10. FMS takeoff speeds
  2. Captain,

    Thank you for a very thorough analysis of jetliner approaches.

    As a 27-year vet of the A321 cockpit, I agree with your analysis and advice. Even to this day, I remember my struggles as a brand new FO on my first jetliner, and came to realize my energy management concept was backward. That is, I came from turboprops, where to “go down” you needed to “slow down.”

    In jetliners, to “go down,” you need not be afraid to pitch down and speed up!

    Also, in the A321-family, I’ve found that getting “dirty” to Flaps 2 will help immensely to “go down.” Typically, I like to be at Config 2 on my base-to-final when intercepting the ILS.

    Also, I agree–don’t be shy with the boards (spoilers), flaps and gear to aid in “getting down!”

    1. Great advice Eric, thank you!

      Your comments about pushing the nose down and speeding up to lose height fast is very apt. When descending from high and intermediate levels, converting height to speed chews up a lot of altitiude.

      For those who cannot remember high school physics:

      1. Potential Energy is proportional to Height.
      2. Kinetic Energy is proportional to Speed squared.
      3. Cd increases dramatically as the speed increases towards Vmo/Mmo.

Leave a comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.