Airliners.net

So, the Gulfstream G500 is coming with an active sidestick. This means that the sticks can be electronically coupled, so they would move with each other. It also allows force-feedback (similar to what Boeing do with their yokes despite their widebodies being FbW), which is something Airbus currently does not have. It is also something very new as traditionally yokes could be coupled, whereas sticks could not.

I understand what people were trying to get at with AF-447, but a lot of "articles" just seem to bash Airbus and the stick (id est, they're useless). Again, A330 has passive sticks so no feedback.

What are your initial thoughts? And does the active stick somehow put Boeing's yoke "in danger"?

Please do not turn this into a ∀B vs B∀ bashing war.

A yoke is right in front of both pilots, very hard to miss its position, good or bad, appropriate or inappropriate. The side stick is out of view of the other pilot, it’s actual movement isn’t nearly as great as a yoke. I like the idea, but I’m not convinced it’s a panacea.

GF

Boeing would be in danger of what exactly? On a Boeing the cross control coordination is strictly mechanical and the artificial feedback is done by electric back drive actuators. I think that electrically back driving the joy stick would make a simple reliable system more complex and prone to mechanical failure although the artificial feel would be a great feature for the pilots.

stratclub wrote:

Boeing would be in danger of what exactly? On a Boeing the cross control coordination is strictly mechanical and the artificial feedback is done by electric back drive actuators. I think that electrically back driving the joy stick would make a simple reliable system more complex and prone to mechanical failure although the artificial feel would be a great feature for the pilots.

To be precise, the backdrive actuators are only on the 777 and 787 (e.g. FBW models). The non-FBW models got control feedback of what the autopilot is doing by the Autopilot Servos driving the control cables back to the Yoke and Rudder.

GalaxyFlyer wrote:

A yoke is right in front of both pilots, very hard to miss its position, good or bad, appropriate or inappropriate. The side stick is out of view of the other pilot, it’s actual movement isn’t nearly as great as a yoke. I like the idea, but I’m not convinced it’s a panacea.

GF

Whether you fly a yoke or a stick (or Hawker handlebars) your first instinct when something is is wrong will be to reach for the controls. Grabbing a sidestick that is deflected will make it much easier to figure out what's going on than one that is neutral.

GalaxyFlyer wrote:

A yoke is right in front of both pilots, very hard to miss its position, good or bad, appropriate or inappropriate. The side stick is out of view of the other pilot, it’s actual movement isn’t nearly as great as a yoke. I like the idea, but I’m not convinced it’s a panacea.

GF

Couldn’t agree more, for maximum crew awareness and feedback control yokes
and back driven autothrottles are as
good as it gets


Having said that I think that linked, active
side sticks are a big advance and what
Airbus should have done originally, along
with back driven AT


The lack of these features has been a contributing factor in AB accidents and
incidents over the years


It’s important to keep the A vs B conversation going

OK. Not exactly "in danger" but Boeing's advantage of being able to implement force feedback on the 777 and 787 is more or less lost if a stick also has feedback

Can of worms... Opening...

The first instinct when something is wrong should be to scan the instruments, PFD and engine thrust in particular, not reach for the controls. The instruments tell you what the aircraft is doing, not the yoke/stick.

Case in point: The outcome of AF447 might have been very different if the pilots had sat on their hands for 30 seconds instead of immediately grabbing the controls. The plane was in trim and flying straight and level.

What should have happened would look something like this:
- Aircraft: Bribbit! Bribbit! (Autopilot and autothrust disengaged). Some ECAM warnings, I would guess "Anti-ice Captain and F/O pitot heat".
- PM reads ECAM.
- PF: "Huh. Ok... I have control. Stand by. Ok we seem to have lost the speed. The pitch attitude looks ok. I'll keep it at 3 degrees and the thrust at 80% N1." At some point during this, reach for the stick and thrust levers. Gently... By saying "I have control", the PF is clarifying that he should be the only one manipulating the stick and thrust levers. Looking at the AF447 data, the pitch attitude at the start of the upset was between 0 and 1.5 degrees, so should be no problem gently nudging it to 3 and keeping it there.
- PF: "Unreliable airspeed checklist..." PF goes through unreliable airspeed memory items. PM checks him.
- There would also have been ECAM items to do... PF: "ECAM actions!"

IIRC, the probe heat came back within a minute. So before they would have finished ECAM actions, they would probably have had valid airspeed indications again.

Would any more pertinent information have been gleaned from the fact that a yoke/stick was deflected in any direction? I doubt it.

Having said that I think that linked, active
side sticks are a big advance and what
Airbus should have done originally, along
with back driven AT


The lack of these features has been a contributing factor in AB accidents and
incidents over the years

There’s a difference between mechanically back driven and software back driven that makes the electronic ones much more difficult in the risk analysis. I don’t know the ins and outs, but several engineers at an OEM told me it creates an order of magnitude bigger risk. This is why after 40 years of operation, BAe is the first with the moving sticks.

GF

Starlionblue wrote:

Can of worms... Opening...

The first instinct when something is wrong should be to scan the instruments, PFD and engine thrust in particular, not reach for the controls. The instruments tell you what the aircraft is doing, not the yoke/stick.

Case in point: The outcome of AF447 might have been very different if the pilots had sat on their hands for 30 seconds instead of immediately grabbing the controls. The plane was in trim and flying straight and level.

What should have happened would look something like this:
- Aircraft: Bribbit! Bribbit! (Autopilot and autothrust disengaged). Some ECAM warnings, I would guess "Anti-ice Captain and F/O pitot heat".
- PM reads ECAM.
- PF: "Huh. Ok... I have control. Stand by. Ok we seem to have lost the speed. The pitch attitude looks ok. I'll keep it at 3 degrees and the thrust at 80% N1." At some point during this, reach for the stick and thrust levers. Gently... By saying "I have control", the PF is clarifying that he should be the only one manipulating the stick and thrust levers. Looking at the AF447 data, the pitch attitude at the start of the upset was between 0 and 1.5 degrees, so should be no problem gently nudging it to 3 and keeping it there.
- PF: "Unreliable airspeed checklist..." PF goes through unreliable airspeed memory items. PM checks him.
- There would also have been ECAM items to do... PF: "ECAM actions!"

IIRC, the probe heat came back within a minute. So before they would have finished ECAM actions, they would probably have had valid airspeed indications again.

Would any more pertinent information have been gleaned from the fact that a yoke/stick was deflected in any direction? I doubt it.

Agree with all that, the crews reaction
to an event which would have been
a transitory annoyance was disastrous
and to be honest, incomprehensible


As you said, leave it alone, sit on your
hands and run the checklist, normal
airspeed indications would probably
have been back before you reach the
end of the QRH


Land, write it up and go home



That’s how it should have ended



But instead everyone died, lots of
contributing factors of course, number
one being the crews extreme overreaction
to a relatively innocuous event



However, if the side sticks had been linked
the PF inputs would have been immediately
and continuously apparent to the PM, his
default input became mostly nose up
even after they were fully stalled



Of course a conventional yoke would have
been all the way aft if the same inputs were
made and impossible to overlook and take
corrective action


But also, questionable design features, a stall warning that triggered as the stall was
approached and entered but then turned off as speed decreased further


That made no sense, when you’re stalled the warning should stay on until you have
recovered



Problem is the PF believed, as the warning
ended that they had exited the stall and
increased his nose up input, eliminating
a chance to recover


That along with the AB design incorporating normal, alternate and direct
law which has progressively less protection
seems like a logical system but in the heat
of the moment (and this crew was obviously extremely panicked) did they understand that protections that would normally keep them out of trouble
and not allow them to stall were no longer
active in alternate law that had activated with the loss of IAS ?



They may have recognized the transition but if all the changes that come with that
hadn’t fully registered with the PF and PM
they could easily still believe they couldn’t
stall the aircraft no matter what input they
made


Tragic, needless loss of life

BoeingGuy wrote:

stratclub wrote:

Boeing would be in danger of what exactly? On a Boeing the cross control coordination is strictly mechanical and the artificial feedback is done by electric back drive actuators. I think that electrically back driving the joy stick would make a simple reliable system more complex and prone to mechanical failure although the artificial feel would be a great feature for the pilots.

To be precise, the backdrive actuators are only on the 777 and 787 (e.g. FBW models). The non-FBW models got control feedback of what the autopilot is doing by the Autopilot Servos driving the control cables back to the Yoke and Rudder.

Oh. I thought we were comparing FBW to FBW, not FBW to Legacy Boeing mechanical control systems. True though. On legacy aircraft, the autopilot servos physically control the actuators through the primary flight control cables and back drive the pilots controls through the primary flight control cables, however, the FBW back drive actuators only back drive the pilots controls. It is just amazing how simple rigging a FBW airplane is.

Not a pilot, but I was wondering what warning messages the aircraft displayed. Was there an A/S DISAGREE message and an associated emergency check list item for it? Where does the attitude indicator get it's data from and wouldn't the VSI have been believable because it most likely gets it's data from static ports that generally don't ice up? I realize the entire incident was only about 3 minutes long according to WikiPedia.

For the sidestick issue, being that only 1 sidestick at a time should ever be used, what if there was a master caution, aural warning or horn and a warning message displayed: SIDESTICK INPUT DISAGREE when ever both sidesticks are off of neutral at the same time?

My guess is that the sidestick may have contributed to the incident, but human factors and cockpit design were bigger contributing factors.

Max Q wrote:

Starlionblue wrote:

Can of worms... Opening...

The first instinct when something is wrong should be to scan the instruments, PFD and engine thrust in particular, not reach for the controls. The instruments tell you what the aircraft is doing, not the yoke/stick.

Case in point: The outcome of AF447 might have been very different if the pilots had sat on their hands for 30 seconds instead of immediately grabbing the controls. The plane was in trim and flying straight and level.

What should have happened would look something like this:
- Aircraft: Bribbit! Bribbit! (Autopilot and autothrust disengaged). Some ECAM warnings, I would guess "Anti-ice Captain and F/O pitot heat".
- PM reads ECAM.
- PF: "Huh. Ok... I have control. Stand by. Ok we seem to have lost the speed. The pitch attitude looks ok. I'll keep it at 3 degrees and the thrust at 80% N1." At some point during this, reach for the stick and thrust levers. Gently... By saying "I have control", the PF is clarifying that he should be the only one manipulating the stick and thrust levers. Looking at the AF447 data, the pitch attitude at the start of the upset was between 0 and 1.5 degrees, so should be no problem gently nudging it to 3 and keeping it there.
- PF: "Unreliable airspeed checklist..." PF goes through unreliable airspeed memory items. PM checks him.
- There would also have been ECAM items to do... PF: "ECAM actions!"

IIRC, the probe heat came back within a minute. So before they would have finished ECAM actions, they would probably have had valid airspeed indications again.

Would any more pertinent information have been gleaned from the fact that a yoke/stick was deflected in any direction? I doubt it.

Agree with all that, the crews reaction
to an event which would have been
a transitory annoyance was disastrous
and to be honest, incomprehensible


As you said, leave it alone, sit on your
hands and run the checklist, normal
airspeed indications would probably
have been back before you reach the
end of the QRH


Land, write it up and go home



That’s how it should have ended



But instead everyone died, lots of
contributing factors of course, number
one being the crews extreme overreaction
to a relatively innocuous event



However, if the side sticks had been linked
the PF inputs would have been immediately
and continuously apparent to the PM, his
default input became mostly nose up
even after they were fully stalled



Of course a conventional yoke would have
been all the way aft if the same inputs were
made and impossible to overlook and take
corrective action


But also, questionable design features, a stall warning that triggered as the stall was
approached and entered but then turned off as speed decreased further


That made no sense, when you’re stalled the warning should stay on until you have
recovered



Problem is the PF believed, as the warning
ended that they had exited the stall and
increased his nose up input, eliminating
a chance to recover


That along with the AB design incorporating normal, alternate and direct
law which has progressively less protection
seems like a logical system but in the heat
of the moment (and this crew was obviously extremely panicked) did they understand that protections that would normally keep them out of trouble
and not allow them to stall were no longer
active in alternate law that had activated with the loss of IAS ?



They may have recognized the transition but if all the changes that come with that
hadn’t fully registered with the PF and PM
they could easily still believe they couldn’t
stall the aircraft no matter what input they
made


Tragic, needless loss of life

Very good points. In my opinion, however, given the actions the crew took they might have had no chance from the start. They never even talked about the unreliable airspeed procedure. Now, I don't know what said procedure stated at the time, but there wasn't even an inkling of either pilot discussing known pitch and power.

Moving sidesticks as a last line of defense? That may have merit. But, again in my opinion, no crew should dig itself a hole quite that deep that it needs moving sidesticks. And yet...



What seemed lacking during the event was an "all of the above" combination of situational awareness, proper application of procedures and proper CRM. Just one example: They didn't respond to the stall warning for 54 seconds. No one on the flight deck uttered the word "stall" even once.

Reviewing the FCTMs very first section of content gives some context.

Airbus Golden Rules for Pilots. First rule.
"FLY". The Pilot Flying must concentrate on "flying the aircraft" to monitor and control the pitch attitude, bank angle, airspeed, thrust, sideslip, heading etc., in order to achieve and maintain the desired targets, vertical flight path, and lateral flight path.
‐ The Pilot Monitoring (PM) must assist the PF and must actively monitor flight parameters, and call out any excessive deviation. The PM's role of "actively monitoring" is very important.
Therefore, both flight crewmembers must:
• Focus and concentrate on their tasks to ensure appropriate tasksharing
• Maintain situational awareness and immediately resolve any uncertainty as a crew.

The bit about "immediately resolve any uncertainty" really stands out. As I interpret it: "figure out what is happening before you act..."

Having said all that, we have the luxury of analysing these events in our own good time while not sitting in the cockpit of an airliner over the ocean, at night, probably fatigued, and under an increasing amount of stress.



Stall warning stopping: If memory serves it is inhibited below 60 knots to avoid nuisance warnings during taxi etc... So once the aircraft slowed below 60 knots, the warning stopped. Obviously not good in this instance. However I think we can chalk this one up to the designers not in their wildest imagination thinking that anyone would not heed the stall warning for 54 seconds, as happened on AF447. How many hints do you need before you push the nose down?

Looking forward, is a revision of the logic warranted? IMHO you want the stall warning to be devoid of most logic because it needs to work reliable. Might adding more logic make it less reliable?


Normal, Alternate and Direct Law indications: Of note is that 17 seconds after the event started, the PM stated "Alternate Law Protections". (The PF did not acknowledge verbally.)

"Active" Law indications are on the PFD and not really hard to spot. Here are three PFDs showing the indications. (Other indications have been removed for clarity.) In Normal Law, pitch and roll limits are shown with green double dashes. In Alternate and Direct Laws these are replaced with amber crosses. In Direct Law you also get "USE MAN PITCH TRIM".



In addition, F/CTL ALTN LAW (PROT LOST) or F/CTL DIRECT LAW would have come up on ECAM. Plus a nice loud ding and an amber attention getter in front of the pilots' noses.

Speed indication also changes when low speed protection is lost. Looking again at our three PFDs:
- In Normal Law, the speed tape has a thin orange line, the top of which is VLS, the lowest selectable speed. Below the amber line is a series of amber dashes, the top of which is the top of the Alpha Protection range. Below that is a red tape, the top of which is Alpha Max, the speed at which no more increase in AoA is permitted. These ranges move up and down depending on G-load, speedbrake use and flap setting.
- In Alternate and Direct Law, the amber VLS line is still shown, and below it are red dashes, the top of which is VSW, the stall warning speed.

stratclub wrote:

Not a pilot, but I was wondering what warning messages the aircraft displayed. Was there an A/S DISAGREE message and an associated emergency check list item for it? Where does the attitude indicator get it's data from and wouldn't the VSI have been believable because it most likely gets it's data from static ports that generally don't ice up? I realize the entire incident was only about 3 minutes long according to WikiPedia.

For the sidestick issue, being that only 1 sidestick at a time should ever be used, what if there was a master caution, aural warning or horn and a warning message displayed: SIDESTICK INPUT DISAGREE when ever both sidesticks are off of neutral at the same time?

My guess is that the sidestick may have contributed to the incident, but human factors and cockpit design were bigger contributing factors.

They did get ECAM messages, one of which was F/CTL ALTN LAW (PROT LOST). As mentioned above the PM acknowledged this ECAM.


The primary attitude indicators get data from the IRs (Inertial Reference Units). The standby gets data from its own independent gyro. AFAIK altitude and VSI indication were correct throughout. About 10-12 seconds into the event, the PF stated "We don't have a good... we don't have a good indication of... of speed". So he at least seemed aware of the unreliable airspeed situation.


Sidestick input. When dual input occurs, there is a very loud "DUAL INPUT!" verbal warning, as well as a warning light right in front of the pilots' eyes. Hard to miss.


During almost the entire sequence of events, only the PF sidestick was deflected. The PM wanted to take control after about 90 seconds, and said "Controls to the left" as well as pressing his priority button. The right seat pilot pressed his priority button two seconds later but did not verbalise. Now there's a problem. There was no actual dual input until about 3 minutes into the event. This did lead to a dual input warning but neither pilot acknowledged it verbally, nor did the captain standing behind them by then.


IMHO the aircraft was telling them what was happening. They were just not listening very well.

There's an excellent AF447 timeline here. It includes CVR transcript, instrument indications, sidestick deflection and so on. If memory serves it was compiled by Pihero.

You can really see how disorientation set in very fast.

https://www.dropbox.com/s/t5a8dd09wxiac9s/AF447JulyReportVer14.pdf?dl=0

Thanks. And that's the thing about all of the digital information. You have to nderstand what it is telling you. I take it ECAM is AirBus speak for EICAS. Sometimes I think we need an Acronym Protection Program or APM to navigate all of this stuff. Let's see. T/R. is that Transformer Rectify or Thrust Reverser. I can never remember.

stratclub wrote:

Thanks. And that's the thing about all of the digital information. You have to nderstand what it is telling you. I take it ECAM is AirBus speak for EICAS. Sometimes I think we need an Acronym Protection Program or APM to navigate all of this stuff. Let's see. T/R. is that Transformer Rectify or Thrust Reverser. I can never remember.

ECAM is indeed Airbus for EICAS.

Don't get me started on acronyms. You should have seen the look on my face when it dawned on me that half the stuff on the A350 has brand new acronyms compared to the A330. (And T/R is thrust reverser, while TR is transformer rectifier... I think. Ok, those two you can thankfully figure out from context.)


Back to your point about understanding what information is telling you, and I would add where information comes from, that is indeed crucial. In the context of AF447, one important tidbit is that the stall warning comes directly from the position of the AoA vanes. It isn't processed in the ADIRUs (Air Data Reference and Inertial Reference Units) or the ISIS (Integrated Standby Instrument System). In other words, when Michelle yells "STALL!" the plane really is stalled.

Regarding AF447:
The Vertical Speed Indicator is a combination of inertial and barometric vertical speed. The loss of measured total pressure in the pitot tubes also affected the altitude and vertical speed indication. At FL350 and 0.8 Mach the indicated altitude loss is up to 300 ft and the indicated vertical speed is up to -750 ft/min. See pages 41 and 42 of the final report.

If in doubt sit on your hands


As has been stated, I heard the same advice decades ago


It is easy to judge them here on the
ground but I just can’t understand the PF actions for the life of me


Inputting a significant pitch up from level cruise into what was effectively a (brief) zoom climb because of what was, granted a temporarily bewildering set of indications just goes against the grain


I’ve not had the same scenario but over the years I have had airspeed discrepancies, in every case I just sat there until it either resolved itself or we resolved it through the use of a checklist


Not for a moment did either of us consider making a significant control input


And that’s no claim of expertise, we just didn’t know what we were dealing with so we left everything alone until we did


It often is the best advice

Starlionblue wrote:

Back to your point about understanding what information is telling you, and I would add where information comes from, that is indeed crucial. In the context of AF447, one important tidbit is that the stall warning comes directly from the position of the AoA vanes. It isn't processed in the ADIRUs (Air Data Reference and Inertial Reference Units) or the ISIS (Integrated Standby Instrument System). In other words, when Michelle yells "STALL!" the plane really is stalled.

Kinda like troubleshooting were the most important tool in your tool box is having a working knowledge of how something works. How does the A330 determine a stall? Is it a calculated value or is it some kind of measurement based on several values? Seems like airspeed and later stall data invalid would have been great messages to have. Of course with airspeed invalid a cross check would have shown a split between the pilot's and copilot's indication, if my understanding is correct. At the lazy "B" IIRC we would have called that a flight deck effect if there is no correlated message.

So the next step would have been "what do I have left?" Keep the last thrust setting and verify VSI, attitude and altitude indications still works like you said, with gentle inputs. Next step, continue flying the aircraft relying on the indication you do have. Seems pretty easy armchair quarterbacking, doesn't it?

Do you have a link to the actual complete final report?

stratclub wrote:

Starlionblue wrote:

Back to your point about understanding what information is telling you, and I would add where information comes from, that is indeed crucial. In the context of AF447, one important tidbit is that the stall warning comes directly from the position of the AoA vanes. It isn't processed in the ADIRUs (Air Data Reference and Inertial Reference Units) or the ISIS (Integrated Standby Instrument System). In other words, when Michelle yells "STALL!" the plane really is stalled.

Kinda like troubleshooting were the most important tool in your tool box is having a working knowledge of how something works. How does the A330 determine a stall? Is it a calculated value or is it some kind of measurement based on several values? Seems like airspeed and later stall data invalid would have been great messages to have. Of course with airspeed invalid a cross check would have shown a split between the pilot's and copilot's indication, if my understanding is correct. At the lazy "B" IIRC we would have called that a flight deck effect if there is no correlated message.

So the next step would have been "what do I have left?" Keep the last thrust setting and verify VSI, attitude and altitude indications still works like you said, with gentle inputs. Next step, continue flying the aircraft relying on the indication you do have. Seems pretty easy armchair quarterbacking, doesn't it?

Do you have a link to the actual complete final report?

The only way an A330 unambiguously determines a stall is through the AoA vanes. "STALL! STALL!" This makes sense since stall can happen at any speed and any attitude, and depends on AoA. It doesn't matter what your speed and attitude are. If you go beyond a certain AoA for a certain configuration, you are stalled.

However, if you're in Normal Law, you will most likely never stall. (I won't say stalling the aircraft is impossible in Normal Law but darned close.) In Normal Law the PRIMs (FACs in A32x-speak) use ADIRU data to calculate significant AoA values. These can vary widely. For example with slats out the critical AoA increases. With speedbrakes it decreases. With maneuvering (g load increase) you decrease the AoA margin at a given speed. This can be seen on the speedtape as the "red army" moves up and down. If you move down into the Alpha Prot range, the flight control system will change the pitch law from a load factor (a stick deflection results in a proportional g value) to an AoA (a stick deflection results in a proportional AoA). If you go deep enough into Alpha Prot, you end up at Alpha Max, at which point the flight control system will force a pitch command to keep your AoA at Alpha Max or below. Even with the sidestick fully back, your AoA will not exceed Alpha Max.

Crucially, AF447 was in Alternate Law, so this protection was inactive. One of the reasons having a good knowledge of the protections and their conditions is essential knowledge when flying an Airbus.

Unreliable airspeed is a very difficult one for computer logic to figure out. That's one reason it is such a tricky situation. Three separate data sources. Which one is correct? How can we create logic which says one source is correct and the other two not, or two correct and one not? Which is why the meatbags in the seats are forced to do the figuring out. Having said that, if you get a pitot heat fault, I think that is a strong hint airspeed may not be reliable.

Final report is here: https://www.bea.aero/docspa/2009/f-cp090601.en/pdf/f-cp090601.en.pdf

Starlionblue wrote:

Unreliable airspeed is a very difficult one for computer logic to figure out. That's one reason it is such a tricky situation. Three separate data sources. Which one is correct? How can we create logic which says one source is correct and the other two not, or two correct and one not?

Here is an idea (which is probably not that great since it is not implemented) but theoretically:
If the logic in the aircraft monitored the speed and memorized it for a several seconds (or digital samples). Then it would check each new speed sample against the old sample. Do this for all the data sources. If the speed difference between the old sample and the new is larger than what would be during turbulence and normal flight then it would provide a warning and stop using that data sources until it starts agreeing with the other three.

The thing that i am sure of is when airspeed starts to become unreliable does it start decreasing/increasing incrementally or is it a 'fall off the cliff' type of event.

Max Q wrote:

Having said that I think that linked, active
side sticks are a big advance and what
Airbus should have done originally, along
with back driven AT

Back to the subject of back driven controls, and especially in the case of back driven thrust levers, I cannot disagree more.
After 10 years on the Boeing, those moving levers are what I thought I would miss the most on the Airbus. It turns out they are a brilliant, simple and intuitive bit of kit which is one of the things I love most about the airplane.
As for the sidestick, you simply can't beat the ergonomics, and the day I have to go back to having a klunky and unwieldy control wheel straight out of the early ages of flying between me and the instrument panel will be a sad one.
The debate around them being backdriven or not always tends to end up gravitating around the AF447 accident. My view on this is that those pilots were so clueless about what was happening that they would have lost the jet regardless of the flight control system involved.
If Airbus procedures regarding their use are properly trained and applied, It's a great system. Back-driving them would only add to the complexity, cost and weight while degrading dispatch reliability. I personally have not had any bad experiences with it.

For the rest of us, there has been a noticeable pre and post-AF447 when it comes to training. Unreliable airspeed procedure which used to be just another of the dozens of procedures that were never really practiced much to a much more elaborate procedure that gets practiced pretty much on a yearly basis.
Pretty much every pilot now knows exactly what he/she wneeds to do in an AF447 scenario. We know how to recongise the symptoms, which instruments to rely on, what initial actions to take and how to revert to the basics of flying. Airplane systems have been modified and awareness about the issue, which turned out in recent years to be much more prevalent than what was initially thought, has dramatically increased.

To use it to discredit the Airbus flight control system is simply preposterous.

acjbbj wrote:

OK. Not exactly "in danger" but Boeing's advantage of being able to implement force feedback on the 777 and 787 is more or less lost if a stick also has feedback

joystick feedback was a design decision and not a design limitation afaik.

starlionblue wrote:

There's an excellent AF447 timeline here. It includes CVR transcript, instrument indications, sidestick deflection and so on. If memory serves it was compiled by Pihero.

I made that actually, and sent it to Pihero for discussions. I think version 14 is the one where I put the PF and PM in the wrong place... but the info/details didn't change much compared with version 12 or version 13.

It's a pity the AF447 accident report never put the CVR and the relevant FDR data in one place. It would have been a lot clearer to see what went on on the CRM front if they did.

---

Passive sidestick, active sidestick, and a yoke... whichever... just make sure you know how to use it. (Yeah, lame argument, but at the end of the day, that's what's important!)

Even moving thrust levers under autothrust... I prefer Airbus' non moving A/T because of it's simple "slap it full forward on TOGA without the need to press anything else" and Boeing's press the TOGA button is fiddly... but the latter is fiddly only when you're not prepared and being left behind by the airplane... when that happens, the moving levers didn't help to prevent accidents.

I think relying on these backdriven cues is a hazard. But to say they're not helpful when you're paying attention is equally foolish.

WIederling wrote:

acjbbj wrote:

OK. Not exactly "in danger" but Boeing's advantage of being able to implement force feedback on the 777 and 787 is more or less lost if a stick also has feedback

joystick feedback was a design decision and not a design limitation afaik.

The reason there is so much push back from some about AirBus not having joystick feedback is explained in Psychology .

Paradigm paralysis - https://en.wikipedia.org/wiki/Paradigm
Perhaps the greatest barrier to a paradigm shift, in some cases, is the reality of paradigm paralysis: the inability or refusal to see beyond the current models of thinking.[19] This is similar to what psychologists term confirmation bias. Examples include rejection of Aristarchus of Samos', Copernicus', and Galileo's theory of a heliocentric solar system, the discovery of electrostatic photography, xerography and the quartz clock.[citation needed]

E pur non si muove...

If the Airbus system of non linked
sidesticks and autothrottles that are not back driven is so good then
why did Gulfstream and British Aerospace not copy it ?



It has been obvious since the A320
that improvements were needed, what is bewildering is how it was certified in the first place


British Aerospace took a good idea that was poorly executed and refined it to where it should have
started at


Gulfstream now have active sidesticks with force feedback that
are linked to each other and back
driven autothrottles



The best of both worlds, the space savings of intelligent sidesticks and the feedback and crew awareness that accompany that and back driven AT’s

I'm liking it. The spacious interior of an Airbus plane with the convenience and "feel" of Boeing's force-feedback concept.

acjbbj wrote:

I'm liking it. The spacious interior of an Airbus plane with the convenience and "feel" of Boeing's force-feedback concept.

You should ask for your car steering to be controlled via reigns
and engine health indication via an acoustic whinnie system.
Acceleration will be controlled by whip snapping and tongue clucking
back driven by way of a horse manure dispenser.

Wrapping new systems that work differently in ancient paradigms
to make it feel "like back then" is a long time cul de sac.

At some point you have so many layers wrapped over each other nobody can debug it anymore
( or even understand how it works at the core or give an assessment on safety related probabilities.)

WIederling wrote:

acjbbj wrote:

I'm liking it. The spacious interior of an Airbus plane with the convenience and "feel" of Boeing's force-feedback concept.

You should ask for your car steering to be controlled via reigns
and engine health indication via an acoustic whinnie system.
Acceleration will be controlled by whip snapping and tongue clucking
back driven by way of a horse manure dispenser.

Wrapping new systems that work differently in ancient paradigms
to make it feel "like back then" is a long time cul de sac.

At some point you have so many layers wrapped over each other nobody can debug it anymore
( or even understand how it works at the core or give an assessment on safety related probabilities.)

The word you’re looking for is ‘reins’
those are the leather straps you use to
control a horse


A King or Queen ‘reigns’ over their domain




Flight control systems that keep Pilots
in the loop and fully aware of all modes and phases of automatic and manual flight by providing maximum feedback are not ‘ancient paradigms’ but vital design features that significantly enhance pilot
performance and flight safety



If such a system is so difficult to build
and complicated to understand then how
did BAE and Gulfstream manage to design
and incorporate one (that is universally
praised) that encompasses all that ?

Max Q wrote:

The word you’re looking for is ‘reins’
those are the leather straps you use to
control a horse

A King or Queen ‘reigns’ over their domain

Thank your for your constructive criticism. Nothing less was expected.

Max Q wrote:

Flight control systems that keep Pilots
in the loop and fully aware of all modes and phases of automatic and manual flight by providing maximum feedback are not ‘ancient paradigms’ but vital design features that significantly enhance pilot
performance and flight safety

I am assuming that this affirmative statement is purely of your making and not based on any sort of expert analysis from a standpoint of a human interface engineer? Or, for that matter, that you have never been involved in studying or designing flight control systems?
You are entitled to your opinion on the Airbus design philosophy, but please refrain from making these bold claims without any evidence other than your gut feeling.
If at least you would tell us what makes you feel like you are not getting enough feedback from the airplane when you are flying an Airbus?

Max Q wrote:

If such a system is so difficult to build
and complicated to understand then how
did BAE and Gulfstream manage to design
and incorporate one (that is universally
praised) that encompasses all that ?

Different cost analysis in different eras of technological availability?

And I have no clue where you get that 'universal praise' from, other than, once again, your own, self-amplified conviction.

When it comes to facts, however, I can quite confidently tell you that Airbus has built well over 10,000 airplanes with their brand of FBW, which have accumulated millions of flight hours and accrued statistical safety levels well on par with anything out there that boasts back-driven flight controls, and that it will be eons before Gulstream or BAE achieve that kind of safety or reliability with their aircraft regardless of the flight control system that they use...

I agree, I doubt this is such an advance and Airbus and F-16 experience doesn’t show that “fixed” side sticks are a risk. As I said the yoke in front of both pilots has unmistakeable presence, a side stick, the cross-cockpit one which cannot easily be seen, less so.

GF

Having flown B727, B744, A320, A330, A340, and having watched pilots push/pull/rotate a yoke like they are fighting a buffalo, whilst keeping the platform perfectly still, I reckon its a pointless discussion. Airbus product is great. Boeing product is great. Gulfstream product is great. None will prevent AF449 (A330) or the recent Lion Air accidents.

I don't care what the sidestick or yoke is doing. I care about attitude and energy.

It's like I went back in time to the AF447 thread, featuring the same people arguing with the same American/Boeing fanboy on the sidestick.

T54A wrote:

qHaving flown B727, B744, A320, A330, A340, and having watched pilots push/pull/rotate a yoke like they are fighting a buffalo, whilst keeping the platform perfectly still, I reckon its a pointless discussion. Airbus product is great. Boeing product is great. Gulfstream product is great.

Best comment on a.net today. Thanks.

Francoflier wrote:

Max Q wrote:

Flight control systems that keep Pilots
in the loop and fully aware of all modes and phases of automatic and manual flight by providing maximum feedback are not ‘ancient paradigms’ but vital design features that significantly enhance pilot
performance and flight safety

I am assuming that this affirmative statement is purely of your making and not based on any sort of expert analysis from a standpoint of a human interface engineer? Or, for that matter, that you have never been involved in studying or designing flight control systems?
You are entitled to your opinion on the Airbus design philosophy, but please refrain from making these bold claims without any evidence other than your gut feeling.
If at least you would tell us what makes you feel like you are not getting enough feedback from the airplane when you are flying an Airbus?

Max Q wrote:

If such a system is so difficult to build
and complicated to understand then how
did BAE and Gulfstream manage to design
and incorporate one (that is universally
praised) that encompasses all that ?

Different cost analysis in different eras of technological availability?

And I have no clue where you get that 'universal praise' from, other than, once again, your own, self-amplified conviction.

When it comes to facts, however, I can quite confidently tell you that Airbus has built well over 10,000 airplanes with their brand of FBW, which have accumulated millions of flight hours and accrued statistical safety levels well on par with anything out there that boasts back-driven flight controls, and that it will be eons before Gulstream or BAE achieve that kind of safety or reliability with their aircraft regardless of the flight control system that they use...

Interesting reply FrancoF


I’ve obviously struck a nerve, you
seem to take any criticism of the Airbus
flight control system quite personally.



While you may not share my opinion it’s hardly unique as concerns the lack of feedback on this now rather old FBW
system



Which is why BAE and Gulfstream collaborated to create a far better one



If the AB system is as good as you imply
it wouldn’t need any improvement but almost everything in aviation moves on, refinements based on feedback from operators and new technology combine
to allow a better system



Of course it will take many years to rack up anywhere close to the number of hours the Airbus products
have accumulated as you mentioned



And there will probably be further tweaks
along the way but that’s a good thing if it improves crew awareness and more importantly, safety


Rather than criticize progress you’d be better off accepting it

crimsonchin wrote:

It's like I went back in time to the AF447 thread, featuring the same people arguing with the same American/Boeing fanboy on the sidestick.

I read most of the final report on AF447 recently and it was very apparent from the crews collection of errors that backdriven joysticks would have made as much difference as rearranging deck chairs on the Titanic. Early on Boeing stated in a round about way that the reason that they stayed with a yoke and column on the 777 and 787 was Paradigm paralysis from "more experienced" pilots. They didn't use the term "Paradigm paralysis", but that is exactly what was meant.

Almost forgot. You might call me a "Boeing Fanboi" but still I like well executed design principles and do consider the AB design well executed. The Boeing design is VG as well except it does add a lot of unnecessary weight and complexity.

The BAE and Gulfstream collaboration is an attempt to fix something that isn't broken. With digital control systems, how the aircraft reacts to pilot input is your feedback. Why complicate something that just plain old works with artificial feel?

stratclub wrote:

crimsonchin wrote:

It's like I went back in time to the AF447 thread, featuring the same people arguing with the same American/Boeing fanboy on the sidestick.

I read most of the final report on AF447 recently and it was very apparent from the crews collection of errors that backdriven joysticks would have made as much difference as rearranging deck chairs on the Titanic. Early on Boeing stated in a round about way that the reason that they stayed with a yoke and column on the 777 and 787 was Paradigm paralysis from "more experienced" pilots. They didn't use the term "Paradigm paralysis", but that is exactly what was meant.

Almost forgot. You might call me a "Boeing Fanboi" but still I like well executed design principles and do consider the AB design well executed. The Boeing design is VG as well except it does add a lot of unnecessary weight and complexity.

The BAE and Gulfstream collaboration is an attempt to fix something that isn't broken. With digital control systems, how the aircraft reacts to pilot input is your feedback. Why complicate something that just plain old works with artificial feel?

That’s a different take on Boeing’s flight
control philosophy, especially the part where you admit you made up a term to describe their design that allows maximum feedback, awareness and aircraft control


It is not at all apparent that AF447 would have suffered the same fate if both sticks
had been moving in synchronicity


That might have been enough to make a difference to the awareness of this very confused and frightened crew



Of course they also had to deal with the unexpected switching to an unfamiliar control law where several of the protections that normally ‘kept them out of trouble’ were gone


Did they really understand this ?



It doesn’t seem like it as the PF inputted full back stick nearly all the way down ‘trusting’ in envelope protection that was
no longer there



This is where all that ‘weight and complexity’ of back driven controls is worth every penny



The brief loss of indicated airspeed did not
cause this accident but a panicked overreaction to that followed by an almost
total lack of awareness of their energy state by the crew sealed their fate



Conventional yokes and / or linked sidesticks may well have made a difference

Max Q wrote:

Rather than criticize progress you’d be better off accepting it

I have, at no moment, criticized the new flight control design by Gulfstream and BAE. I'm sure it'll work very well.

I am also aware of the limitations and quirks of the Airbus FBW along with the functioning of more conventional flight control systems and would be happy to discuss them with someone who has at least some degree of knowledge and exposure to them as well and who isn't clouded by bias.

So at the end of the day, what annoys me is not criticism of the Airbus design, it this the ridiculous declarations and proclamations regarding what constitutes good or bad flight control design without the slightest effort at providing some meat to them. You seem to just be airing grievances and so far have not substantiated any of your bold claims.

No attempt to disguise that as a fanboy war is going to change the fact you haven't answered my question: what about using the Airbus FBW did make you feel like you weren't getting enough feedback?

Max Q wrote:

While you may not share my opinion it’s hardly unique as concerns the lack of feedback on this now rather old FBW
system

So it is just an opinion. What makes it of any value? And who else exactly does this 'concern' and why?
You seem to operate under the illusion (delusion) that it is common knowledge that the Airbus flight control design is flawed. That's called confirmation bias, if I'm not mistaken.

Max Q wrote:

Which is why BAE and Gulfstream collaborated to create a far better one

Again, why is it better? How, in which way? Not saying you're wrong, just trying to get some reasoning out of you.

Max Q wrote:

If the AB system is as good as you imply
it wouldn’t need any improvement but almost everything in aviation moves on, refinements based on feedback from operators and new technology combine
to allow a better system

What needs to be improved? Again, better why? Opinion?
You keep speaking as if moving controls are necessarily better just because they move. Can you delve a bit deeper than that?
For instance, how would the feedback work on a modern day FBW which operates the surfaces according to complex algorithms and not necessarily in a way that would feel natural or logical to the pilot?
If we are going to put sophisticated computers between the pilot and the surfaces, do we want the pilot to know what the surfaces are doing, or do we let them concentrate on what the airplane itself is doing instead?

Max Q wrote:

This is where all that ‘weight and complexity’ of back driven controls is worth every penny

Now you're mixing up flight envelope protection / flight control laws and sidestick feedback. Those are two distinct things. How would moving sticks have removed their confusion regarding what law they were in?

Max Q wrote:

The brief loss of indicated airspeed did not
cause this accident but a panicked overreaction to that followed by an almost
total lack of awareness of their energy state by the crew sealed their fate

Conventional yokes and / or linked sidesticks may well have made a difference

Another non sequitur. How would linked sidesticks have been able to make the crew aware of their energy state?

The brief loss of indicated airspeed did not
cause this accident but a panicked overreaction to that followed by an almost
total lack of awareness of their energy state by the crew sealed their fate



Conventional yokes and / or linked sidesticks may well have made a difference

I would have thought the large upward pitch input followed by the equally large upward rate of climb and rapid falloff in Mach would have been all the information necessary to fly the plane. Who, at optimum flight level, would think those indications aren’t wrong and will rapidly deteriorate the flight path?

GF

stratclub wrote:

You might call me a "Boeing Fanboi" but still I like well executed design principles and do consider the AB design well executed. The Boeing design is VG as well except it does add a lot of unnecessary weight and complexity.

The BAE and Gulfstream collaboration is an attempt to fix something that isn't broken. With digital control systems, how the aircraft reacts to pilot input is your feedback. Why complicate something that just plain old works with artificial feel?

I agree with your post otherwise, but wanted to point out that it is possible that BAE/Gulfstream also have a well executed design. I don't think there is just one possible answer (and certainly Boeing and Airbus both have reasonable approaches), so maybe BAE/Gulfstream are simply looking at what makes sense for them, in 2018, and what possibly also differentiates them from others. Doesn't mean that Boeing or Airbus design would be any worse. But we have no information to conclude that BAE/Gulfstream design would be abdly executed...

I have no doubt that the BAE and Gulfstream collaboration will be done in the best possible way in regards to safety. With aircraft, we always try to error on the side of safety. So really they are offering another option that will be just as safe or safer than anything we do today. Is the idea sound? IDK. We will just have to see how the idea plays out.

The trouble with a feed back system is that it will rely on the flight parameters that are inputed to it. So if you do lose IAS or AOA, the computer will do or prompt you to do actions that may end in a disaster. The ability to recognize when a computer is wrong should be moved up to the front for pilot training.

That would go hand in hand with a pilot understanding about how an airfoil functions. Thrust, lift, gravity and drag. If you don't have a memory understanding about how they work, you should not fly an airplane. In any situation or any aircraft. In a non normal situation, job one should be to just fly the airplane. With training involving flight computers somehow that basic concept has been lost.

One thing we are told at my areospace company in Ethics training and other areas is to not criticize the competitor. Obviously we think our design philosophies are superior to the competition or we wouldn’t be doing them. That doesn’t mean the other guys’ design philosophies don’t have merit too. They are just different.

Sometimes bad days happen. AF447 was an unfortunate bad day. We don’t - or shouldn’t - get satisfaction over something like that happening.

You don’t throw stones when you live in a glass house. Just as easily as you can say AF447 wouldn’t have happened on a Boeing airplane, there unfortunately may be an accident where someone can say that wouldn’t have happened on Brand A.

I’ll always enjoy a debate on A.net about economics, Pilot satisfaction, sales numbers, etc for A vs B, but I don’t think it’s appropriate to suggest that one design is less safe.

AF447 was truly a sad thing. But not the first with similar circumstances. There is this from the days of "steam gauges" and 3 man crews well before FBW ever existed.

https://en.wikipedia.org/wiki/Northwest ... light_6231
December 1, 1974 in Harriman State Park near Stony Point, New York, just north of the New York City area. The Northwest Airlines 727 had been chartered to pick up the Baltimore Colts professional football team in Buffalo in western New York.

All three crew members on board died when the aircraft struck the ground following a stall and rapid descent caused by the crew's reaction to erroneous airspeed readings caused by atmospheric icing. The icing occurred due to failure to turn on the pitot tube heating at the start of the flight. This was one of two Boeing 727s to crash in the United States that day; the other was TWA Flight 514 in northern Virginia, northwest of Dulles airport.

From the NTSB report's abstract:

...the probable cause of this accident was the loss of control of the aircraft because the flightcrew failed to recognize and correct the aircraft's high angle of attack, low-speed stall and its descending spiral. The stall was precipitated by the flightcrew's improper reaction to erroneous airspeed and Mach indications which had resulted from a blockage of the pitot heads by atmospheric icing.

The crash site was nearly perfectly circular about a 100 yards in diameter, chilling to view from the air.

GF

AirlineCritic wrote:

Doesn't mean that Boeing or Airbus design would be any worse. But we have no information to conclude that BAE/Gulfstream design would be abdly executed...

To make sense the backdriven sidestick(s) would have to show a significant improvement on those incidents and accidents that
relate to the airplane command interface.

Einstein: Make your model as simple as possible ... BUT not simpler.

At the moment the Airbus sidestick arrangement looks simple but not too simple.
Savetywise it seems to perform at least on par to the B layer cake.

All the Boeing backdrive and split actuation stuff added to
the ancient pulley and rods system is then unnecessary complexity.

And it should also be mentioned that the MC21 is also coming with active sticks. In a commercial airplane. Hope Airbus offer active stick as an upgrade for 320, 330, 350, and 380.