# Bjorn’s Corner: Fly by steel or electrical wire?

By Bjorn Fehrm

July 25, 2019, ©. Leeham News: Last week’s Corner which dealt with Airbus’ issue with an updated A321neo Fly By Wire (FBW) and how it was unrelated to the issue of the Boeing 737 MAX, gives a good segue to a Corner series about the possibilities of FBW versus classical flight controls when it comes to tuning an airliner’s flight characteristics.

The two different control principles present the designer with very different challenges and possibilities.

The Fly-By-Wire Airbus A321XLR. Source: Airbus.

##### The aerodynamic challenge of the modern airliner

The airliners of today fly a very wide flight envelope. The aircraft and its aerodynamics shall work in a linear way from zero altitudes and 120kts or Mach 0.18 (the speed when rotating for take-off or touching down) and to 41,000ft and 500kts or M0.88 (the typical maximum certified speed/Mach at altitude).

The forces the air exert on the aircraft is dependent on this speed range and the variation of the air’s density. At sea level the air weigh 1.2 kg/m3, at 41,000ft 0.3kg/m3.

If we hold our hand against the wind through the cockpit side window, the difference in felt pressure (the dynamic pressure, also called Q) varies from 2.3kPa/0.34PSI at takeoff to 9.5kPa/1.4PSI when cruising at the highest speed, a change of four times.

If we instead compare the takeoff dynamic pressure and the pressure at highest allowed speed when the aircraft flies at lower altitudes (the Vmo), the difference increases to six times.

##### A problem for a flight control system

It means an aerodynamic surface like the Elevator on the aircraft has an efficiency pitching the aircraft up or down which varies six times from the lowest Q to the highest. Yet, to the pilot, it shall take him about the same force to raise the nose at low speed as at high speed (really Q in both instances). As it’s six times more difficult to move an elevator at highest Q the designers of the high-speed jets were forced to isolate the Pilot and his yoke from the Elevator and give him an artificial force to work against. Moving the elevator was left to powerful hydraulic jacks, controlled from the Yoke over a system of wires, rods, and pulleys.

This takes care of the resistance in the Yoke when changing nose position, but it doesn’t solve the difference in Elevator travel required to generate an up or down movement of the nose. At low speed (a low Q situation) the elevator needs to move six times more for the same generated pitch force as at high Q situations (assuming linear aerodynamics)

The wide variation in needed elevator force and travel to move the nose of the aircraft at different speeds and altitudes must be solved by our flight control system.

When the speed envelope is modest, like for lower speed propeller aircraft, the problem can be solved with a mechanical system with aerodynamic assistance from balance tabs to keep the forces in check.

As the speed increases, we need a hydraulic system to isolate the Pilot from the forces. Our flight control system is now controlling the pressure working on the piston in hydraulic cylinders. It requires modest forces from the Pilot and his Yoke. But we haven’t solved the problem; the movement of the hydraulic piston needs to be large at low speed and minute at high speed.

##### Enters Fly By Wire

Our flight control system needs to adapt its gearing Yoke-to-Elevator when Q changes. Changed gearing in a mechanical system over a wide range is difficult. It was the very wide gearing range needed for the Concorde (which had an even wider speed/altitude envelope) which took the designers to the first civil airliner Fly By Wire system.

With analog computers in the electrical control loop, the gearing of the system could be easily adapted based on Concorde’s Q. But allowing electrical circuits or computers to control powerful hydraulic controls is potentially dangerous. Any hiccup and the aircraft is in trouble.

While FBW opens up more possibilities for gearing changes (and other features) it puts very different safety requirements on our flight control system. The large authority to control the aircraft must be harnessed by a system which cannot malfunction.

### 80 Comments on “Bjorn’s Corner: Fly by steel or electrical wire?”

1. Good topic to bring up Bjorn. Folks don’t understand automation and often blame it when it has made flight much safer. The idea of ‘steel’ wire systems even those operating hydraulic servo valves in near the flight actuators seems redisclose to me now. Ever since the 1950s we have had hybrid systems. For instance.
1 Jet aircraft require ‘yaw dampers’ that measure the high speed snaking inherent in jets, particular swept wing jets and null it out by applying opposite deflection to the rudder. The B707 era jets all needed it. The German DVA was experimenting with them in 1942 on a Henschel Hs 130 for application to the Me 262. The yaw damper is also used to eliminate the adverse and provers yaw that comes from an aileron or spoiler yawing the aircraft when ailerons are used to roll the aircraft because drag is different in upward and downward deflected ailerons. Although differential deflection can be applied it works differently at different speeds.
2 Low wing aircraft with underslung engines have the principle drag and thrust axis out of alignment. Autopitch can be used to trim the aircraft as thrust levels change, this is essential since it is maddening to adjust ones approach by increasing thrust only to find the nose raises or lowers. Sure you can try the T tail aircraft but then you face the irrecoverable super stall issue of the stabiliser being blanked of in a stall and the engines stalling. You also need to transmit the weight of the engines to the wing.
3 Mach Tuck as the airliner approaches transonic speeds the incipient shock wave transfers the lift to the ear of the aircraft. Mach trim must be used to automatically trim the nose up.
4 PCO Pilot Couples Oscillations. Flutter and oscillations can be transmitted from the flight surface to the pilot and then back out again causing a breakup of the aircraft. Flight controls on transonic aircraft are best irreversible and fully powered. PCO killed Geoffery deHaviland when testing the DH108 due to a shock stall caused by the blunt vampire nose. Aircraft such as the WW2 Lockheed P38 Lightning and Dornier Do 335 used hydraulically boosted ailerons (Only 18% manual and 82% power on the P-38) but 100% power is required to prevent PCO etc.
5 Rudder Roll compensation. Operating the rudder will cause the aircraft to roll so a small amount of aileron should be applied.

All of thee features were already in use in aircraft of the 1960s and were hybridised into the mechanical systems. What is the point of hybrid system. Its best to go for a fully controlled system and integrate it properly where multiple sensors, actuators and control computers can cross check each other and provide redundancy. The lack of FBW caused the MCAS debacle.

Modern FBW FCS can prevent stalls. (plenty of airline pilots have from fatigue and confusion killed themselves and hundreds of passengers through stall). They can also prevent tail strike and limit not only pitch but pitch rate. One way Airbus were able to create the A321 was to increase the rotation speed at take-off while automatically preventing tail strike. The faster and higher rotation allowed the take-off to be reduced.

Aircraft with FBW will be able to generate novel motions. For instance by applying yaw with differential thrust while countering that yaw with rudder a sideways motion can be generated that will compensate for cross winds. Very likely in electric aircraft where engine response is very high.

We are entering an era when pilotless aircraft will come into use. It will begin with the use of oversized drones operating as air taxis. Why have a mechanical system.

• Thanks for the good run down William (you just did the next Corner for me 🙂 ). We will go a bit deeper into the different phenomena you described and discuss how mechanical or FBW handles it. But this is a good start.

• This is why I love this site. Thanks to you all, especially Bjorn.

• Yes to both.

If it was easy anyone could do it.

• A bit of background on the Blackbird SR-71 flight controls. Apparently they rejected the early FBW systems that could have been developed ( Avro Arrow and North American Vigilante had versions them)
They did however have advanced for the time FCS.
They really knew how to do a Stability Augmentation System, unlike Boeing
“In pitch and yaw the SAS uses triple redundant sensors, electronics
and gain-scheduling. There are two dual-tandam series servos for the pitch
axis, each driving an inboard elevon. The yaw axis has four series servos—
a pair for each fin. The roll axis has dual redundancy and a separate
channel is used to drive each inboard elevon. ”

https://www.flightglobal.com/FlightPDFArchive/1973/1973%20-%202336.PDF

• Thanks.
The Blackbird could probably get into trouble very quickly.
I understand that an inlet ‘unstart’ was scary, due to the sudden thrust asymmetry.

• Cant the feedback from FBW control stick movements be ‘overdone’ by some military test pilots leading to PIO -pilot induced oscillations, which have crashed a Saab Gripen and Lockheed F22 during development.

2. “segue”, not “Segway ™”
🙂

• I had to look it up whilst reading the article, and frankly all this sounds like music anyway, so segue works just as well as segway😃

• Miner but relevant, the Pilot is refereed to as Him in the 3rde sentence under

“A problem for a flight control system”

We have female pilots of course.

3. Presumably the term “linear aerodynamics” is based on thin aerofoil theory.

4. If practical and economical, I’d like to see truly redundant systems built into aircraft. For example the Military A-10’s manual reversion system. NASA did some studies on this after the Sioux City United Airlines 232 crash (they lost all 3 hydraulic systems on a DC-10, but, somehow the Pilot was able to make a crash landing and saved a lot of lives). They called it “Propulsion Controlled Aircraft”. By using differential thrust of the engines.
https://www.nasa.gov/centers/dryden/history/pastprojects/PCA/index.html
It sounds like Airplane makers are trying to incorporate more of this type of truly redundant systems into their designs. I’m guessing Airbus has some truly redundancy in their design from this incident description?
http://www.avherald.com/h?article=48a5ccb7&opt=0
Another good presentation of the Throttles only backup control system

• I suspect propulsion control of aircraft will become the norm .
It will for EVTOL like lilium jet because it’s possible to economically install a dozen electric ducted fans both to provide redundancy and thrust based control. These will be either powered by batteries or a central generator. (EDF or electric ducted fans have exceptionally good power to weight ratios). I’m surprised that ailerons can not be converted into elevons as many aircraft have been lost due to loss or elevator and stabilster control.

• Check out the Lilium-Jet: no control surfaces only engines.

• And that could be certified ?
In reality its what used to be called a ‘paper plane’ and the backers have next to zero chance of getting it to even a flying prototype.
We have vertical takeoff aircraft now , they are called helicopters
There was ‘supposed’ to a proof of concept first flight 2 years ago,…. more like a powered small scale model in flight and a different larger one rising to 2 m off the ground.

• Both Airbus (two projects) and Boeing(at least 1) have EVTOL projects and urban mobility projects. Lilium (Named after the founders hero Otto Lilienthal) has 280 million Euro in venture capital. They are working with EASA and the FAA on certification. A 1/2 Scale Test bed has been flying for 2 year and a full scale prototype has been hovering for a few months albeit without pilot.

Pilots are really undesirable and only there to stop the aircraft landing of a kid runs out onto the pad. They will be used initially.

You can go to an entertainment electronic store and buy a ‘drone’ quadcopter that can carry a GoPro camera for 40 minutes for \$2500.

Its only a matter of time before they scale up to fly people.

Drones are already winching down ‘cappuccinos’ and in tests in Canberra and will soon be delivering spare parts, medicines, blood products and unfortunately pizza. (but relieving traffic).

Helicopters are just too expensive because of the limited redundancy of their highly stressed and complicated drive train. They are noisy, have a severe blade strike danger and require a massive landing foot print.

The EDF “Electric Ducted Fan” that Schubler-de produce is 200mm diameter, produces 25KG thrust, weighs 3.4kg and consumes 11-15.6kw.

The batteries used in Tesla’s Model 3 have a capacity of 248WHr/Kg and will work at the C3 level though C10, C15,C35 is common in the model aircraft word.

• Lilium have may good solutions with vertical T-O and pretty good glide ratio, still it looks expensive buit will drive other designs that are cheaper to make and operate. Like XTI TriFan 600 that have vertical starting with 3 fans and one engine for horizontal flight all powered by GE Catalyst hybrid engine. Other similar designs will use a Safran/Turbomeca hybrid engine, When Honeywell, Thales, Collins fully aircraft automation and autoflight becomes certifiable with no pilot on board (but in a ground stations driven by software from the manufacturer and the EASA/FAA) the safety, economics and productivity will be attractive.

5. Modern Fly-by-Wire control laws are developed mainly using simulation models that relies first on a wind tunnel aircraft model and later with a in-flight matched model. Good models guarantee well adjusted FBW (and Autopilot/Autothrust) control laws. But sometimes not all the effects can be foreseen through simulation and modeling, and sometimes real life operation put the aircraft through conditions not verified during development/certification process. It is an iterative development process.

• Yep, its called tuning. PID. More fun than fighting a hornets nest at times

And as noted, the larger the range of the variable, the harder it gets.

I never got into auto shift of different tuning parameters but we had one system that badly needed that.

• Yep, very funny, specially when you have badly coupled dynamics, thus when you fix one you break the other. The other “fun” part in this adjustments is to have feedback from several flight-test pilots, where each one have their own opinion.

6. Power assisted cables were in use well before FBW. FBW puts the power at the point of actuation. So whatever provides the power is directly attached to the control surface.

The real gains from FBW are speed of response, precision of response and redundency or fail-safe. But it all depends on the servo-actuators.

The A350 uses Moog, world leaders in high speed, high precision servo actuators. It’s an American company.

The A350 is a first generation morphing wing, maintaining cruise trim by minor use of the flaps and minor use of the trim stabiliser. Not possible without whizz bang servo actuators

But to return to the point. Power assist has been there for many a decade – pilots can’t be expected to move the flaps on a first generation 747 using cables, pulleys and wheels! So FBW is all about where to put the power

• The outstanding property of real, state of the art “FBW” is the ability to easily, nonlinearly, role based … mix all inputs into action for all output channels.

FBW is not really the proper name for this.

FBW in its basic meaning is actuation by electrical signaling.

Wire actuated hydraulic boost was in a way a step back. 🙂

• The modern sense has computer control for the electrical signalling, either analogue like those of the late 50s , mostly military and Concorde of course.
Then the digital FBW of the A320 combined with envelope protection

• Uwe,

I agree with your definition of FBW. Been trying to convince TW for a long time. What’s at either end of the wire makes FBW interesting. At one end the computers providing software algorithms that moderate pilot inputs as well as sensors, pitots, AoA vanes, radio altimeters, GPs, IRS and so on. At the other end the means of control the airplane, elevators, ailerons, rudder, flaps/slats, spoiliers and so on, and how they are powered for precision and response.

• These are EHA “Electro Hydrostatic Actuator”. They are basically a self contained hydraulic actuator incorporating a pump driven by a variable speed electric motor; by transferring back and forth the fluid from one cylinder chamber to the other, the pump and electric motor achieve the control of the position of the piston connected to the surface. These devices have a “Mode Selector” that allows the pump to be bypassed by a solenoid valve so that an backup EHA can take over control. In this bypassed mode the actuator is said to be an ‘damping mode’ because it functions somewhat as a shock absorber but allows free movement One reason for using EHA as opposed to mechanical actuators “EMA” Electro Mechanical Actuators” is that EHA is much better understood in terms of its resistance to lock up and bypass. An EMA might use a rack and pinion, jack screw etc.

In general the two EHA connected to a flight control surface has 3 kinds of working mode active/active (A/A) mode (both the EHAs are actively controlled), and active/passive (A/ P) mode (one EHA is actively controlled and one passive following the others position). In addition, one working mode refers to the fault damped (DP) mode. When the EHA is in the damping mode, the solenoid valve will break the connection between the actuator and the pump, and the oil passage is connected at both ends of the actuator. In this case, the actuation system is equivalent to a damper.

There are also EHBA “Electro Hydrostatic Backup Actuator” these are exactly the same as an EHA but have an additional valve and servo valve so they can either use hydraulic fluid from their integral pump as in the standard EHA or from the aircrafts main hydraulic system. This is what you tend to get on an airbus: an electrically powered EHA and a EHBA which can selectively be either electrically or hydraulically powered.

• Very nice break down.

We did not have that choice in the 50s. 737 is 50s.

Should we throw it out? Maybe. Should we restrict how often you can update an aircraft (I think it would be a good idea)

But when a 737 Classic.NG has as good a track rerecord as an A320, its hard to argue it can’t be updated to current safety levels (MCAS is a aspect done all wrong, done right not an issue)

Manual. trim? That truly is a rats nest that needs to be looked at.

• Prior to he EBH there was of course the conventional high bandwidth electrohydraulic hydraulic valve to provide for FBW. EBH simplified hydraulic distribution. We might have been looking at 8000psi systems but for that. I think the Concord (1970s), Vulcan Bomber (1960s) PFCU, The German VTOL of the 1950s/60s (VAK191, VJ101,Dornier Do 31), F-16 and from 1994 the B777 used these and of course the 1980s A320 as well as a range of actuators used in rocket motor gimbaling including the V2 exhaust gas vanes.

The B737NG prototype flew in 1997 so Boeing at that time could have used its B777 (first flight 1994, service date 1995) technology in the B737.

I think the B737MAX is the only airliner above 100 passengers still using pull cables as opposed to FBW.

The only time I felt that the Boeing Yoke was better than a Sidestick was the extraordinary AF447 crash where the pilot flying immediately stalled the aircraft after auto disconnect of the autopilot because he didn’t have pitot static information and the pilot monitoring didn’t pick up his mistake (despite duel input warning etc)

That of course was not a FBW issue but an ergonomic sidestick vs yoke issue and at that only a small part of the failure chain.

7. Until I started reading about the MCAS disaster I didn’t realize that the 737 had mechanical controls running the length of the fuselage and I’m astonished. With the benefit of hindsight,(or even foresight?)would I be right in saying that a back up electric motor would have been a much better solution ?

• Grubbie, it’s not that bad. The 737 MAX does have motors but not for manual trim.

Others know better than me, but I think the motor for manual trim was withdrawn with the NG.

Most of the motors don’t meet the definition of FBW. The spoilers are now FBW. Why they did it for the spoilers but not the elevators, ailerons, rudder and flaps? I don’t know. Just one of the many mysteries of the 737 MAX

• They say its used to reduce ‘stopping distances’ as well as the usual weight and maintenance saving
http://www.b737.org.uk/max-spoilers.htm

I also read somewhere that in the A320 series the rudder is the only control surface that is linked to the pilots by a steel cable .

• The 737 classic did have dual motors.

And of course its not manual if you have electric motors.

• In a car, it’s manual steering, even if it’s power assisted steering.

• Or a car with parking assist, when the entire wheel movement- and how much- is automated.

• @Dukeofurl:
On A320, A340 and early A330 for instance, the rudder is indeed the only “Primary Filght Control Surface” which is wire commanded.
BUT the trimmable horizontal stabilzer has also a mechanical link when actuaded by the pilots (FBW by computer otherwise)

• Likely because loss of spoilers would be considered as a less than hazardous failure, so a single or dual redundant system would be appropriate, whereas loss of a primary control system (aileron, flaps, rudder) would be more significant, requiring a fully redundant system (which is where the cost and time get much larger). Spoilers systems are also fairly well segregated from the rest of the FCS, so again fairly simple implementation.

• The 747 also has cables running the length of the aircraft. If you walk thought the Flight Test Vehicle at the Boeing Museum of Flight you can see them. They run under the cockpit, up the back wall of the cockpit and then in the ceiling space. They are then connected to the hydraulic valves to create the boosted power.

8. Yep.

To go further servo actuators still use hydraulics. The hydraulics may be self-contained or provided externally (engines). The A350 uses both.

So hydraulics are still central to FBW airplanes.

9. This SLF was lucky enough to see a Flight test briefing in 1959 at edwards. Which included a segment of yeagers discovery of some unplanned- unknown at the time aerodynamic coupling effects. The description of that flight and recovery was suitably impressive. The cockpit film was even better ! I finally found that brefing

The actual film from the cockpit camera during his now famous flight of the X1b starts at about 4:30 in

Thus we learned about pitch, roll, yaw issues 🙂

http://www.chuckyeager.org/news/rare-footage-x1-tumbles-regaining-control/

10. Very interesting. Not being an aerospace engineer, I hadn’t thought of this till now and now I appreciate the challenges. But as a computer scientist, I appreciate the challenges faced by the FBW system developers. Hats off to the people who make these systems fly so reliably. Thanks y’all!

• Hats off to the people who got the harrier and Chinook to work without FBW!

• I don’t get the Chinook ref. Flying Banana did it, seems pretty basic (granted I am not deep into Helicopters)

Harrier, that is another story.

Grippen had two crashes in early days, at least attributed to a software problem.

None of its easy. Other than Bjorn not a one of us could do it.

• Yes, the tandem rotors , as used in Chinook is better aerodynamically. It means no rear counter torque rotor and offers better longitudinal stability than a single central rotor, less power to hover and each rotor would normally have a lower disc loading. The rotor cyclic controls are required anyway and work together or opposed on each rotor to achieve required result.
Chinooks may look like a city bus but they can be just thrown around the sky
The Harrier was really a terrible plane for stability in hover mode, thats why it mostly it used runways like other planes, however short takeoff had advantages. The USMC getting the first carbon fibre wing on the updated version with better lift and range gave the design a fairly long life.

11. The best argument I have ever heard during the FBW x Non-FBW is that it surely has saved far more lives than it has killed due to failures. Full flight envelope protection has made it possible for thousands of “less than great aviators” to fly safely all over the world. The best saying to define this is : “Everybody sees how many times I stumble when I’m drunk , but nobody sees how many vodka, whiskey and beer bottles I drink”

12. A point that seems not to be considered in this discussion is that along with finance guys the airlines have aeronautical engineers, chief pilots, former test pilots, maintenance chiefs, etc on their staff and on the teams that evaluate available aircraft and determine fleet purchases – and they ordered several thousand 737MAXs. If this aircraft is so utterly horrible, disastrous, colossally mis-designed, dangerous, antiquated, stupid, and irredeemable… why did they place those orders?

• None of those knew the full extent of MCAS and most of them were not even told it existed.

• This discussion is well past MCAS however. The assumption is that 737MAX will be fixed and and the grounding lifted, otherwise the topic would be what happens when Boeing files for bankruptcy (Chapter 11 or Chapter 7) and whether another manufacturer can offer competition to Airbus with a clean sheet NSA design. But instead we are discussing design philosophies – which brings us back to the original decisions by the airlines to buy the MAX.

• Boeing sold 4000 MAXs in the first 6 months. The “low to the ground” design Boeing used in the 1960s for the B737 is still valued by Airlines. 1 You don’t need to wait for a conveyor to load/unload baggage, can be done from the back of a luggage trailer. 2 You don’t need stairs (not so much a concern nowadays), 3 engine is accessible from the ground which can’t be said for aircraft with engines in the rear or taller undercarriage. Boeing have stayed with steel as opposed to carbon fibre brakes and although less effective it means the cool of period is less before luggage handlers and service personnel may approach the tyres. In addition the B737 has a larger wing (enlarged from the classics) than the A320 and climbs faster and cruises at higher altitude although the neo has much better climb rate to the ceo.

Offcourse the promise of the MAX was that nothing much would change from the B737 except much better fuel consumption and economics. Unfortunately this wasn’t possible and MCAS was required and it wasn’t done properly. In a way the customers dictated these unfortunate contradictory specifications.

• If Boeing had to include flowcharts describing all software on the plane, to the airlines, maybe you’d have a point. But, as far as I know, they don’t. I think it should be required. The software is flying the planes now. I’d think that airlines would be interested in how their planes are programmed to fly. I’d expect that if a pilot had a question on how something worked, they could look it up, or have someone if their organization answer their question. Right now, they don’t even have a clue, that the system even exists, unless Boeing tells them it does. Is there an NCAS, or OCAS, or PCAS or .. whatever in the software now? Is Boeing required to tell the FAA or the airlines what systems are buried within their software? If Boeing had to publish a flowchart change for the last minute change to MCAS, to the Airlines, it may have been given more visability and been fixed before the plane was flown too much. With software controlling more and more functions, it needs more and more visability, to the FAA, airlines and pilots.

• One key knowledge for pilots is to know when the logic changes like altitude mode, speed mode, TOGA, engine alternate modes, autopilot alternate mode and combinations of these, like the 777 speed trap. Early A320’s could be flown “too slow” and thus sunk quickly without wing folding either right or left fooling fooling some pilots thinking they had other problems. So making the FBW flying the aircraft “too perfect” can cause problems.

• That was a different issue , that test should not have happened below FL140
https://www.flightglobal.com/news/articles/crashed-a320-attempted-slow-flight-test-at-low-altit-323028/
And not mentioned in the preliminary report above is the real cause, two of the 3 nose AOA sensors had been filled with water during a pre flight pressure wash and the water later froze. For some reason the computer code ignored the remaining working sensor which let to the stall and crash, which was at the wrong altitude for that test anyway.
https://en.wikipedia.org/wiki/XL_Airways_Germany_Flight_888T

The FBW computer code was far from ‘perfect’. Even the very first A320 crash doing a slow flyby over a grass airfield which then crashed into the surrounding forest was well below the minimums for that sort of demonstration.

• For the Habsheim crash that brand new “less than perfect” FBW worked well enough to allow very good survival chances. death toll was minimal.
737 crash ahead of the Polderbahn killed quite a bit more people.

Most of the examples brought up against FBW are those where the involved persons really made an effort to create a dangerous situation.

• I think you are referring to Air France 296. The A320 prevents stalls in two ways: firstly it maintains a minimum flying speed that will approximately maintain level flight (Alpha Floor) and Alpha Limit (stall or angle of attack protection).

Alpha Floor is a low speed protection (in normal law) which is purely an autothrust mode. When activated, it provides TOGA thrust. As the aircraft decelerates into the alpha protection range, the Alpha Floor is activated, even if the autothrust is disengaged. Activation is roughly proportional to the rate of deceleration. At low speeds, if a rapid avoidance manoeuvre is required to avoid terrain, windshear or another aircraft, it is safe to rapidly pull the sidestick fully aft and/or bank and hold it there. The aircraft will pitch up to max Alpha, engage TOGA thrust and climb away.

In the Case of AF296 the A320 was carrying passengers for the first time (French Journlists and their Families and Children) and was to do a 100ft low pass along a runway at spectacular high angle of attack to showcase its FBW protections.

Unfortunately when the aircraft came around some terrain the pilot found the crowd lined up on a different grass runway and conducted some aggressive manoeuvres to get there.

As part of this the pilot disabled alpha floor to get a steeper slower descent.

The terrain was much higher ground than expected and the aircraft was heading for trees. Because of the over ride of alpha floor the aircraft was too slow to climb even with slat retraction and because engines were at idle they would require over 9 seconds to spool up to full power (as opposed to 1.6 seconds)

Because of the poor flight planning and actions the pilot was sentenced to jail and on appeal the sentence increased.

• Yes, you also had the A330 AF447 where one pilot did not know what the other was doing due to its sidesticks are not connected as they now are on the A220 and latest Gulfstreams. All alternate modes including TOGA were involved in this accident.
“In response to the stall, first officer Robert said “controls to the left”, and took over control of the aircraft. Robert pushed his control stick forward to lower the nose and recover from the stall; however, Bonin was still pulling his control stick back. The inputs cancelled each other out and triggered a “dual input” warning.”
One can argue that all Airbus aircraft should get the A220 modern sidesticks. Most likely will not happen as the accident rate is too low today.

• The AF447 A330 there was only a few seconds of both pilots manipulating the controllers at the same time, long after the stall had ocurred. That conflict was just a part of the larger confusion in the cockpit.
A Tristar crashed on takeoff at JFK when the co pilot ‘just let go of the control column’ immediately after rotation when they received a ( erroneous) stall warning, and the plane at max weight settled back on the runway. The captain didnt have a chance to take the control column to keep the plane in the air.
(https://aviation-safety.net/database/record.php?id=19920730-0 this needs updating to reflect the co pilot ‘letting go prematurely’, this database is full of errors, for some reason in air engine failures become ‘runway incidents)

13. I invite y’all to read this:

Wow. Not just the ears ringin’

Not sure i believe Q419/Q1 20.

Production is likely to abate some.

In a previous analysis, i said BA was solid enough to deal with 6-12 months of MAX stop. Beyond that the \$\$ is not sustainable.
AB could come out winning but even that should scare them. Regulatory stuff will come back swinging. They’ll be affected.

Fun for everyone 🙁 The plot may thicken for real aside from partisanship/emotions we read sometime of this blog (Uwe/brits).

• Thanks a lot for the link. A very good article it seems, but the content is not pretty.

• Unfortunately the NYT article is behind a paywall but the heading suggests the NYT article will say the things you would expect of the NYT: regulation good, Republicans bad, Democrats good, republican Congress deregulated or cut back on FFA bad. Perhaps I’m wrong about NYT pushing a political agenda, I don’t care as I’m not American but I everyone knows how far to the left this publication leans. (Perhaps you can provide a summary) I suggest it’s far simpler. 1 The FAA approved MCAS but then Boeing changed it radically without telling the FAA because it was considered a minor modification That’s foolishness and even unethical by Boeing but it’s hard to blame the FAA for a deliberate act of omission. 2 Boeing fart arsed around launching an NSA and by 2011 was forced to go with MAX after American Airlines placed and order for 100 B737 with LEAP 1 engines. This left the engineers without a FBW system or ECAM and a direction by sales not to change anything that triggers retraining. The MAX was impossible to do under these these constraints. 4 Finally we’ve entered an era of software being part of the certification process and it is hard to monitor and easy to alter compared to a mechanical or hydraulic system. 6 Finally there is a failure chain here and an unknown root cause. What did the individual engineers think they were doing?

• Thanks, I was able to get to it by a google search though the direct link doesn’t work. It’s nowhere as biased as I’d expected.

• William:

Its not left wing to think regulation can be good. We owe clean water, clean air to that and most of our progress.

Crash of 2008, Maconda blow out ad nasuea.

You can argue if a reg is too strict, but arguing its left wing for that is nonsense. Market crashes are bad for every0one.

• @Transworld. What I was trying to say, I didn’t do a good job, was that someone at Boeing was doing the wrong thing and not submitting MCAS to self analysis or approval. The problem wasn’t the amount of inspectors at the FAA or the fact that the Inspectors would pass back particular safety analysis assessments back to Boeing. The problem was that Boeing was keeping the FAA in the dark, either deliberately or through some kind of mass psychological delusion. Extra resources at the FAA probably wouldn’t have picked this up. That is why the FBI, Grand Jury and DoJ are involved. Im sure with this amount of high end legal fire power the issue is beyond regulatory anomalies and oversight issues and a bunch of Boeing executives and middle managers will have wire taps on them.

The regulator process seems to me that either:
1 Boeing submit their engineering to the FAA, the FAA assesses it with statistical safety analysis and regulatory conformity and approves it.
Or:
2 Boeing submit their engineering to the FAA, the FAA then pass it back requesting Boeing do a statistical safety analysis and check it of for regulation and show their working and then pass it back to the FAA for approval.
This seems to be what self certification is. The problem was that Boeing just wasn’t submitting documents in regards to MCAS beyond the initial ones.
My own work in process plant automation to cover my legal back is that you develop and then conduct an FAT (Factory Acceptance Test) and get the client to approve and witness or part witness the test. You then do an SAT (Site Acceptance Test) in the same way. Both client and supplier OK documents before and after.
After a certain amount of free for all debugging and At a certain point after the testing has removed any significant bugs you start logging software or design changes formally.

This is where Boeing fell apart. They were years behind in submitting documents.

I don’t see the problem as ‘self certifying’ but that Boeing was not actually following the self certification process by not lodging completed documents with the FAA. One wonders whether they even did the tests and signed of on them. Why the FAA didn’t have the balls to stop them is part of the mystery.

Toyota produce some of the most amazingly high quality cars in the world and have developed incredible quality processes mostly under self supervision.

Obviously we need standards and some regulations but I’m somewhat concerned that by making this process too unwieldly we add expense and reduce safety and the innovation that drives safety.
Cost reduces safety, inefficiency reduces safety.

• William:

All well put.

I do not think you can atribute this to just one aspoect. Ture safety is layered.

Inspector no longer reporting to FAA would be one of those safety alert areas.

Boeing pulling wool is clearly another.

Boeing setting up the organization so no one has a full picture of the process is another one as well.

Every aspect should be addressed as well as the should you be able to grandfather forever? Or if you do should you still have to do a full recert and meet the current standards?

I do think its a major mistake to compare this to the Auto Industry.

That is truly a dog eat dog industry and if Toyota went down, a year later no one would know a major mfg was out of it, others would have filled the space. Its very nimble industry.

Aircraft are not nimbler nor easy to adjust to with a roll of the dice risk aspect and the massive investment in a single model.

• It’s not behind a paywall for me. So not sure what’s happening.

It’s a good article. The article does make the point that Boeing did increase the severity of the trim stabiliser AND deflections late on in the development process. The reason given, paraphrased: to make it behave like it’s predecessor.

Don’t know why you can’t see the article.

• I’m surprised you haven’t noticed “to avoid a stall ” in relation to why MCAS was supercharged. We’ll have to wait and see, but it’s what you have been saying all along!

• Thanks for saying it for me. It’s always been by deduction. There can be no other reason for what MCAS does.

• NY Times has a paywall after a few free articles and you are ‘invited’ to create a userid for for that priveledge.
Turn off browser javascript, likely what you have done, and those restrictions disappear.

14. Ben, agreed. But Boeing do know how to do FBW for primary control systems: 787, 747-8, 777.

To be clear about the 737 MAX. All primary control systems are hydraulically driven. Actuation is though cables.

This then comes to the trim stabiliser. It is electrically driven using electric motors. I think actuation is through cables. Manual trim no longer as an electric motor. So it’s brut force from the pilots.

This comes to a question that I’ve posed many times. Is the trim stabiliser a primary control system or a secondary control system like the spoilers?

If the regulators say it’s a primary control system because of the function of MCAS, Boeing are done for. Primary control systems must be fail-safe, which means they must be fully redundant. No part of the trim stabiliser action/reaction loop (feedback loop) meets the standard of fail-safe, a fully redundant system.

As a footnote, primary control systems on Airbus airplanes are driven by hydraulics but actuated by wires. But there is a move to self-contained hydraulics driven by electric motors (A350 and Moog). Apparently there are some exceptions – the rudder on the A320 is cable not wire, as posted elsewhere. Interested to know other exceptions.

NG removed one.

I don’t think that is in keeping with the intent, but the road to MCAS is paved with small changes that in the end change the whole picture.

Keep in mind single was allowed though, single jack screw failure on the MD-80s series and AK Airlines crash.

I also do not agree with it but supposedly it was bullet proof (where have we heard that before)

On the 737NG and MAX, you have to break loose the clutch on the motor to move it if the motor or electrical circuit has gone south.

Mesh that with they changed the Simulator (who did it and why) and the wheel worked fine there, pilots had not a clue in real world you could get speed lock up or you could not break out the clutch.

15. An interesting debate about “Boeing’s safety dollars”,should statistical reality be allowed to trump the aviation authorities rules?Ofcouse it should.

16. William
July 29, 2019
I think you are referring to Air France 296. The A320 prevents stalls in two ways: firstly it ………..

claes
July 29, 2019
Yes, you also had the A330 AF447 where one pilot did not know what the other …………….

Where do you get these nice narratives from that have tenuous contact to reality ( actually invented from whole cloth… afics ) ?

Working towards ghostwriting for Loren Thomson?

17. FBW = fewer human errors and very very few FBW related issues and Airbus FBW is 1000x better then 737MAX frankensystems.