Configuration Warnings

[Ibra]

1 hour ago, Blue on Top said:

Takeoff, go-around and moose turns (low altitude, ground-reference maneuvers … looking at your house, girl friend, showing off, etc.)

First hours after getting PPL, I took wife and 2 pax for the 800agl top of the house flying, I remember my wife said something like "can you stop talking while flying or stop flying while talking, you can't do both" 

That was first introduction to distraction (had load of hours on single seats but the family & friend set-up was really new to me) 

1 hour ago, Blue on Top said:

Airlines (and GA) should be teaching 1.3Vs

Yes, I completely agree if you have no POH or lost all your references just fly that "magic speed", that should works with decent approximations for everything including 60 angle bank (of course 1G and -300fpm on VSI) and it also sits in special place vs any other speed points in the flight envelop: min turn radius at 40deg bank, near Vy, bellow Va, bellow Vb, near Vbg, near Vapp, near Vref, decent margin to Vs0, best speed for turbulence, you stall before losing wings, less chance for accelerated stall to go spiral dive, work well for power off, will get you close to max range, get you close to max time, allow max aircraft maneuverability with 60% of excess power available to correct, near best climb gradient/rate and it is the best speed to operate around bad weather, difficult terrain, busy airspace or anytime I feel behind the aircraft....

Why 1.3*Vso? that is the "optimal point" or the "stable speed" for any aircraft where the drag curve being hyperbolic terms near Vso and quadratic terms near Vne (any sub-sonic aircraft has to follow these maths to some extent ), you simply fighting the aircraft when flying other speeds...

The POH should give better speeds for normal operations but the best speed I target to fly anytime I feel I am behind the aircraft for whatever reason

Edited February 27, 2020 by Ibra

[Blue on Top]

8 minutes ago, Ibra said:

Why 1.33*Vso?

@Ibra  If I typed 1.33*Vso, it was an accident/typo.  It should be 1.3*Vso … or 1.3*Vs1 for approach.  On high-powered, newer, twin, turbofan engine airplanes, the AFM may have Vapp or Vref greater than 1.3*Vs because Vmcl (minimum control speed in the landing configuration) might bump the speeds up.  This is also why the pilot is required to use the published/approved numbers and not simply 1.3*Vs.  And, before I get flamed again, yes, some systems on very expensive airliners automatically calculate/know this inccrease.

[Blue on Top]

Just to let people know and trying to get back on topic, scheduled service (airlines) and GA calculate performance differently.  An airline calculates performance based on the runway length that they are flying in and out of.  GA airplanes calculate the shortest possible runway length.  To further breakdown the GA portion, business jets (part 23 and 25) publish in the AFM, the LONGEST of three distances: 1) accelerate-go (engine fails (or master warning) at Vef (several knots below V1) and pilot has decided by V1 to continue to go airborne, 2) accelerate-stop (engine fails at Vef and pilot has already taken action to stop before reaching V1) or 1.15 times the accelerate-go with all engines operating normally.

(back on topic)  For us single engine folks, today we publish the data that was acquired from the prototype airplane.  No new engine.  No higher power engine.  Just the engine on the day the testing was performed.  You should be able to beat the AFM/POH numbers.  The big boys MUST meet the AFM numbers (AFMs are conservative).  Another big factor is that the OEM must cross the end of the runway at 50' on the 3 degree glide slope … which adds ~1000 feet of runway (this is also why you see all the tire rubber at the 1000 foot marker at major airports.  You don't need to be at 50' over the threshold.  Failures (and improperly configured airplanes) are not accounted for in our single engine airplanes. 

[Ibra]

Yes, corrected typo Vapp = 1.3Vstall

I was referring to a "default speed" for stable flying and safe maneuvering of single engines, this to be near 1.3Vs or at least in the 1.2-1.4 range: the drag function for typical GA aircraft combines parasite & lift drag in something along alpha*Vs/(V-Vs)+(1-alpha)*(V-Vs)*(V-Vs) where the minima is roughly in 1.2-1.4 region for touring wings (alpha=0.5-0.15)

Of course one can calculate exact curves in wind tunnels, get exact speeds in POH in test pilots hands and adjust left/right around this for configuration, power, environment, winds...but the rule of thumb is that any aircraft will simply fly safely in a stable fashion around it's min drag speed under 1G with lot of spare power and spin/dive margins, this is roughly 1.3Vs (where Vs is measured by a stall experience and G measured by pilot seat-of-the pants)

The same theory applies to twin engines just the power/drag curves are actually 3D graphs: two engines power outputs and one speed...

 

 

Edited February 27, 2020 by Ibra

[Andy95W]

1 hour ago, Blue on Top said:

@Andy95W  Thank you for your level-headed explanations.

1.  I agree with your statement.  But, if this were known and trained, why did the crew continue to keep the airplane in the stall?  AOA told them the aircraft was stalling/stalled (until airspeed went below the low airspeed threshold and invalidated AOA, too, but it too came back on the way down … when the pitot tube heaters caught up with the ice pellets and cleared again).

2.  Yes, again.  As a very educated guess, I would highly doubt that simultaneous failure of 3 pitot tubes was ever analyzed before this event.  It would also have to be a failure (icing) that completely blocked the pitot tubes, but at sometime the drain holes were open(ed) to bleed off the indicated airspeed.

3.  So I can learn, are you saying that either "normal law" allowed the airplane to stall or that the airplane left "normal law" and allowed the airplane to stall?

Thanks! -Ron

PS.  Pitot heat on or off is not relevant to the cause of this accident.  All 3 pitot tubes were overwhelmed with ice.  Pitot tube certification regulations (I know in the US) have been changed to handle this new requirement. 

@Blue on Top- thanks for providing such good discussion.

The pitot tubes that iced over were the ones that fed information to the air data computers.  So the airplane was indicating an increasing airspeed.  Apparently, there is evidence to show that the third, standby pitot tube was actually not iced over.  There were also indications that only the first officer in the right seat was applying pitch up commands to the airplane, which was poor airmanship- but perhaps understandable due to the increasing airspeed indications, even though he had to ignore the stall warnings to do so.  After the pitot tubes cleared, he must have maintained back pressure because of the high descent rate- again, poor airmanship.

We actually train to recognize these scenarios, now.  And I am here to say- high altitude stall recoveries are no joke.  Watching 15,000 feet bleed off the altimeter before you can recover is an eye opener.  Obviously, the low air density is the primary cause, just to get enough air molecules flowing to affect a recovery.  

But the other factor is less intuitive- it is the engine thrust itself that can prevent recovery.  With engines that are mounted below the wings, the thrust itself gives a pronounced pitch up moment.  At high angles of attack and decreasing airspeed at very high altitude, there simply isn't enough elevator authority to pitch the nose down after you've applied full thrust.  That's why we pitch down first, accept a huge altitude loss to regain flight control authority, then slowly bring power back in.

3.) as to your question- the airplane either stalled in normal law due to the bad date being fed into the air data computers, or the airplane reverted to alternate law, where the pilots are able to stall the airplane.  Regardless, with the increasing angle of attack, autothrust would have applied full power- which may have exacerbated the situation, as I described above.

[GeeBee]

1.4 is based upon a  bank overshoot. The avocation is to fly minimum 1.4 while maneuvering then slow to 1.3 for landing when pattern maneuvering is complete. 1.4 allows up to a 55 degree bank in the event the pilot overbanks the turn onto final. The concept is based upon "minimum maneuvering speed used in transports.

[Blue on Top]

And @Andy95W thank you, too.  Wonderful explanation.  Ironically, and I don't want to go there, but I believe that this is the reason MCAS was given such power.

[Ibra]

1 hour ago, Andy95W said:

At high angles of attack and decreasing airspeed at very high altitude, there simply isn't enough elevator authority to pitch the nose down after you've applied full thrust.  That's why we pitch down first, accept a huge altitude loss to regain flight control authority, then slowly bring power back in.

I guess in high altitudes a big machine will have a very tight flight envelope between high power over-speeds (30deg nose down) and high power stalls (on 30deg nose up) with bigger time lags due to low density compressible aerodynamics, engine response and heavy mass inertia...if the pilots are confused which regime they fly (failed instrumentation or crew coordination) they could easily en-up oscillating between the two corners and quickly burn all of the FL330 down to 10000ft, but I always thought under low thrust regimes stall/dive recovery can be done bellow 10000ft before the ground as long as wings and tails are still attached?

Edited February 27, 2020 by Ibra

[GeeBee]

The bottom line to AF 447 is the pilot could not attitude instrument fly for squat. I had the scenario sprung on me several times in the sim both before and after 447. You don’t go yanking back beyond 7.5 at those altitudes. Put the dang. Airplane at 2.5 up, set the power at 92% GE, 84 P&W and sit on it. If you don’t know what pitch attitude to use, drop down the FPV and put it on the horizon. 
 

We can sit here and discuss the qualities ir lack there of with regard to automation and different implementation of it but it is an absurd argument if you cannot attitude instrument fly the airplane. Af447 is a poor example to cite for automation, because the at the controls was unqualified to start with.

[carusoam]

The challenge with pointing out somebody behind the controls was unqualified...

Some system said they were qualified...

It seems that when they were qualified... the training didn't cover all of the issues that needed to be covered... or they were too complex to really remember? Or it got forgotten, and there was no E-procedure to read while trouble shooting... ?

Or the check list took longer than Alt/descent rate...

2020 rear view mirror kind of thing...

 

In GA... we can opt to skip transition training...

Discussions like these help reinforce the value of getting good training... and really knowing the documents...

Continue... 

Best regards,

-a-

[GeeBee]

The problem for foreign carriers is they produce ab-initio trained "wunder kind" who have barely enough time to occupy the seat, let alone the savy to handle to handle challenging situations. Such was the case here. Not to worry, it is coming to a US airline near you.

Around 2008 my carrier was hiring hot and heavy and we found some new hires being assigned a low seniority base on transoceanic equipment. Once they completed 75 hours they were off low time restrictions and could be paired together. I asked the boss if he was comfortable with the Captain going on break with two guys with less than 75 hours handling the flight deck He told me they were qualified and I had to take my mandated break. Well..... thank God it was not me but.....two wunder kind in charge of a 767 had a low oil pressure light come on. One ex-Navy, the other 5000 hour civilian started an immediate divert to BIKF. Trouble was, pressure gauge and oil temp was just fine. Captain non plussed. 

These two AF pilots were wunder kind, and while they had the time, they passed the minimal training, they were incapable of hand flying an airplane in the dark over an ocean.

[Blue on Top]

3 hours ago, Ibra said:

I guess in high altitudes a big machine will have a very tight flight envelope between high power over-speeds (30deg nose down) and high power stalls (on 30deg nose up) with bigger time lags due to low density compressible aerodynamics, engine response and heavy mass inertia...if the pilots are confused which regime they fly (failed instrumentation or crew coordination) they could easily en-up oscillating between the two corners and quickly burn all of the FL330 down to 10000ft, but I always thought under low thrust regimes stall/dive recovery can be done bellow 10000ft before the ground as long as wings and tails are still attached?

@Ibra  I would click to like your comment, but I have been too emojial today .

Although thrust at altitude is significantly lower than at sea level, all the forces are lower (which is why we can go faster on less thrust up there).  But, as I believe that you were getting to, the N1 (thrust) has to remain high (percentage-wise) to keep the engine running at altitude … especially at really, really low Qc (dynamic pressure, airspeed).  I know a biz jet that flamed out both engines at 51K doing stall tests up there.  Although they were both re-lit by 30K, both seats came back re-contoured..  Then we got a letter of discontinuance on the way down (no more flight testing until we figured it out … the concluded reason: our engines weren't designed to fly that slowly up there.

[Ibra]

1 hour ago, Blue on Top said:

Then we got a letter of discontinuance on the way down (no more flight testing until we figured it out … the concluded reason: our engines weren't designed to fly that slowly up there....I know a biz jet that flamed out both engines at 51K doing stall tests up there

Indeed, but they should magically restart at 15000ft assuming pilots, engine fans, aircraft wings and tail surfaces are still around (and they should as not much air density to blow things up), once you get outside "high altitudes compressible aerodynamics" engine thrust and aircraft dynamics should magically go back to normal  here is another example:

https://ops.group/blog/enroute-a380-wake-flips-challenger-604-upside-down/?fbclid=IwAR3EuvrFqQvXxpAeh-z9INTowB3vobY5AH2BqlUEC9Vqqd6Del3_Ew6m4CM

Yes, the hard bit for engineering and training is that you can't test much engines & aircrafts & pilots apart from near earth surface but that is all you really need: the only training one needs for flying at FL510 is how to make FL120 quickly without losing it  I am sure AF447 crew had good chance to catch it in low altitudes: they had a serviceable engine & aircraft all way to the ground...

Ona side note: the challenger/airbus example is interesting as many did think that under low angle of attack and Mach cruise speeds, wake turbulence risk is low and the main risk is collision, that was true until it happens between heavy (A380) and light (Challenger) flying under RVSM, why turbulence happens and persist at high speeds and low angle of attack relate to high altitudes flying (so it is not just turbulences from stalled wings on heavy flaps landing) 

Edited February 27, 2020 by Ibra

[GeeBee]

" I am sure AF447 crew had good chance to catch it in low altitudes: they had a serviceable engine & aircraft allway to the ground..."

Nope. Airplane was lost out of FL270. I don't know how many times I have to say it. The stab trim was too far up to allow for the nose to be pushed down even with full down elevator. The Airbus stab trims automatically to elevator position. The elevator was up so long the stab was trimmed full nose up. If at Fl280 the elevator had been reversed to full nose down, the airplane would have been recoverable with just a few thousand feet to spare.

As to engine recovery at altitude, even if the fires stay lit, in general the average twin spool engine in the middle 30's requires 20 to 30 seconds to accelerate from idle to full power. In that time you only have pitch to recover the speed if your speed is inadequate.

 

 

[Ibra]

7 hours ago, GeeBee said:

Nope. Airplane was lost out of FL270. I don't know how many times I have to say it. The stab trim was too far up to allow for the nose to be pushed down even with full down elevator

Interesting to hear, you mean the trim position at FL270 was not at that altitude? Or not recoverable any altitude? is that hard to cut-out and manually reset? (well it should be set for FL120 flying not FL270 by the time you finish rotating...)

On configuration warning (which is the original topic), my guess that one 1m$ Airbus cockpit should be able to calculate/warn/reset trim position for a stable minimum drag flight at 1G (cruise/decent on low/zero power), it will be hard for the crew to manually rotate the big trim wheel for that position but at least this is a case where automation can help?

Obviously, this does not depend on instrumentation outputs or pilot pitch/roll inputs, only environment (temp, pressure, altitude, speed of sound...) and aircraft (surfaces, weight total & distribution, engine outputs...)

The aircraft would naturally oscillate around that trim/speed/config with load of margins away from any maximum drag flying regimes (e.g. spins, stalls, dives, side-slip, steep turn, steep climb, steep decent...), then a pilot should do nothing if those oscillations decay by themselves or just counter them with opposite controls

Note that in high energy-drag flight regimes, aircraft flight controls do get reversed, or don't commute (order matters), or just hit  boundaries: pulling on stall, pulling on dive, rolling on spin, yaw on dive, power on turn, ailerons before pitch, yaw before ailerons, pitch before yaw, pitch before power...so even an experienced pilot does get confused on what is happening there, especially under instrumentation or crew failures

Edited February 28, 2020 by Ibra

[GeeBee]

At the rate the airplane was descending at FL270 there was insufficient time for the stab trim to roll to a nose down position such that it and the elevator in combination could recover the airplane. 

There are a three things at play here. First the Airbus stab trim system. You don't "trim off pressure" in an Airbus. In fact you feel no pressure, there is no control pressure feedback. If you pull back or push forward on the stick, the elevator moves, then the stab trim follows and trims until the elevator is faired with the stab. You do not have a stab trim switch on stick. It is all automatic. When Jaque Jet Jock started yanking 15 degrees nose up at high altitude he got the stab trim full nose up, to the stops. Second, the Airbus trim wheel is small, by design. It is small so that should all that quadruple redundant fly by wire stuff quit, you can still fly the airplane with the stab trim, BUT your inputs need to be small and precise. I've practiced it a couple times in the sim. You are lucky to get it on pavement let alone the touchdown zone of 10,000' slab. The reality is such a failure is one quarter the chance of both engines quitting over water. The fastest way to roll the trim forward is automatic electric actuation. Third issue. For fuel economy, fuel is transferred in cruise, starting around 25,000 feet on climb out to the fuel tank in the stab, which holds about 11,000 pounds. At cruise I have seen a CG as far aft as 39.5% of MAC. So if you stall the airplane at high altitude, you are going to need a lot of pitch authority to get the nose down.

If at the first indication of stall they had executed standard upset recover, push, roll, power, stabilize the loss off control would have been survivable. They did not believe the stall warning because their airspeed was invalid......except stall warning come from the AOA sensor, not airspeed. Which is why I am a big advocate for true AOA indicators in all airplanes.

 

 

 

[Ibra]

3 hours ago, GeeBee said:

If at the first indication of stall they had executed standard upset recover, push, roll, power, stabilize the loss off control would have been survivable. They did not believe the stall warning because their airspeed was invalid......except stall warning come from the AOA sensor, not airspeed. Which is why I am a big advocate for true AOA indicators in all airplanes.

Yes agree AOA all the way (but needs constant speed & 1G to stay meaningful)

3 hours ago, GeeBee said:

Third issue. For fuel economy, fuel is transferred in cruise, starting around 25,000 feet on climb out to the fuel tank in the stab, which holds about 11,000 pounds. At cruise I have seen a CG as far aft as 39.5% of MAC. So if you stall the airplane at high altitude, you are going to need a lot of pitch authority to get the nose down.

Someone did told me a joke about how Concord (no aerodynamic trim) would recover from stalls during flight testing vs with passengers, always better with with passengers! at least you can PA asking everybody in economy pax to move first class, but as there are not that many flying economy you should start venting fuel from the tail

Edited February 28, 2020 by Ibra

[GeeBee]

G loading can be biased in the display, 99.99% of the time, AoA is highly reliable in a transport aicraft.

 

[Blue on Top]

10 hours ago, GeeBee said:

1. If at the first indication of stall they had executed standard upset recover, push, roll, power, stabilize the loss off control would have been survivable.

2. They did not believe the stall warning because their airspeed was invalid......except stall warning come from the AOA sensor, not airspeed. Which is why I am a big advocate for true AOA indicators in all airplanes.

@GeeBee

1. I was told that standard upset recovery (normal law) in an Airbus is throttles full forward and stick full aft.  Isn't that exactly what the crew did?  Is this how the stabilizer drove to full nose up position (in cruise, just before the event, the stabilizer trim position would have been at almost full nose down …. which is slightly LE up)?

2. Bullseye on this statement, or as the British would say, "Spot on, Chap."  Honestly, almost all Part/CS 25 and higher end Part/CS 23 airplanes do have AOA indicators.  One doesn't need the separate indicator.  The information is right "beside" the vertical tape airspeed indicator in the form of the green carat, yellow band and red band (different OEMs display this data a little differently.  But and although the information is displayed "on" the airspeed tape, it is independent of airspeed.  (note: this is very difficult to explain.)  Airspeed is only used as a "reference" number, so the airspeed must be displayed value.  IOW, if the airspeed were stuck at 100 knots, the vertical colored bands would still move appropriately (without the airspeed moving).  In fact on the Airbus airplanes, the airspeed on the PFD can revert to AOA.

So looking forward to receiving the book.  -Ron

[GeeBee]

I don't know who told you that but it was wrong before 447 and it is wrong after. You slam the thrust levers full forward on a set of under slung engines and I will guarantee secondary and tertiary stalls and the stick full aft will aggravate that further. It is true that recovery training now is more aggressive, i.e. get light in the seat, but it has always been taught nose down, feed in partial power, control the pitch up, continue increasing power while maintaining pitch stability.

After Boeing screwed up the elevators on the 737NG (yes, they messed up that 737 too) we had to fly a full stall profile on them after they got new tail feathers. Flaps up at FL410, Full flaps at 10K and 20K. We had the new AOA indicators on the PFD and I can report they are rock solid accurate. Airplane broke every time at the red tick. The problem with airspeed tapes is they quite often are GIGO because they are based upon ZFW or GW data entered in the FMC. More than once I have been at min maneuvering fir a given flap position or on final at Vref+5 only to notice the AoA saying something different. Subsequent load audit reveals the airplane heavier than the garbage fed into the box. AoA does not lie. It is how the Navy nails the 2 wire.

 

[GeeBee]

I also just noticed you said “upset recovery in normal law”. The airplane should not be upsettable in normal law. It can only happen in alternate law or below. Slamming the throttles forward and stick full aft in normal law is terrain warning procedure. It will simple climb at minimum speed in normal law.

 

[Ibra]

On 2/28/2020 at 10:45 PM, GeeBee said:

G loading can be biased in the display, 99.99% of the time

Yes, G loading is on aircraft displays is never reliable (it is calculated by 2nd order differentiation in accelerometers, so less smooth signal to display)  

Although, G loading as measured by human body is highly reliable, works pretty well for natural 1G when our heads are not tilted 

I saw few 1G barrel rolls on paraglider, gliders, C152A, M20C, SR22, T34, Aero Commander, 707, Concord...

My guess none of them was on any instrument or visual references? 

Edited March 1, 2020 by Ibra

[Blue on Top]

20 hours ago, GeeBee said:

The problem with airspeed tapes is they quite often are GIGO because they are based upon ZFW or GW data entered in the FMC.

I'm confused …. but wanting to learn.  Are you saying that airspeed is a function of weight?  Or airspeeds to fly/reference are a function of weight?

Thanks, Ron 

[GeeBee]

Min maneuvering speeds for various flap positions, vref, min clean speed, stall speed are all weight dependent, based upon the weight entered in the FMC. The tape is also g dependent. I had an F/O (ex fighter type)in the sim decide to come out of an over bank by kicking the rudder over and go nose low. As the airplane dived, he pulled up the speed brakes (stall speed increase) I watched the red zipper at the low end jump up. Then he started to pull g's to recover. The low end red zipper increased more. At Vne he started to pull up except the g loading increased the stall speed. The red low end zipper met the high Vne and the entire airspeed range was in the red. So there we were, at Vne with the stick shaker activated, the entire airspeed indicator in the red zipper. So yeah, the markings change with weight, g loading and even altitude if the envelope so dictates. 

I said "your only way out is to drop the boards and exceed Vne". He did that and it took 15,000' but he got it stabilized. Real world, I would guess we might have shed some parts off the airframe.

 

[Blue on Top]

Just finished reading the wonderful book "Understanding Air France 447" by Bill Palmer.  It not only supported what I previously knew, I also learned a couple more things about Airbus control laws (one in particular that is poor - only IMHO.  Really good book and highly recommended … especially for our Mooney pilots that happen to have real jobs flying airplanes, too.

Note 1)  Oh, I do take a slight exception to my profession being called "eggheads" as I typically don't refer to pilots as "trained monkeys'" or "the nut holding the wheel/stick/yoke" … at least in a public, permanently stored record.  Yes, I am a pilot, too.

Note 2)  The book refers several times to the airplane as being "deeply stalled".  I've never heard of an airplane being "shallowly stalled" or "lightly stalled".  If the author is trying to say the airplane was in a "deep stall", there is a definition for that, and AF447 was never in a deep stall (it was always recoverable (with enough altitude).  With an AOA of 45-60 degrees, the wing was definitely fully stalled.

@GeeBee or anyone else that would like to add to my knowledge/experience base (I'll never stop listening or learning), my email is fly-in-home@att.net or solutions@blueontop.com (which just goes to fly-in-home@att.net  Thanks, Ron