7,800’+/- 
Bonin tried that and look where it got him.
Does the plane automatically return to a straight and level attitude if the crew lets go of the controls?
That’s exactly what it’s designed to do as long as the necessary sensory inputs are deemed valid. The whole problem began when the computer determined that some vital information was not available or reliable (air speed), and switched the autopilot off.
I’m pretty sure it won’t. If the autopilot is off and it is in normal law and you let go of the controls the FBW will hold the attitude that had been set by the pilot prior to releasing the controls.
Thanks.
Small correction: It will maintain the flight path, not the attitude.
That’s what I’d wondered; if an airplane in a downward path would, if the pilots let go of the controls, simply continue plummeting towards the ground.
Yes it will. FBW aircraft are complicated by the control philosophy used, some might maintain the speed while others might maintain the flightpath, none that I know of will automatically return to straight and level flight.
Going back to the “dual input” warning …
One of the other features of modern aircraft is to prioritize the vacuous audio warnings. So as long as the stall warning was blaring, the “dual input” alert would be suppressed.
And as others have said, for some decent fraction of the descent Bonin’s input was the only one.
Can’t the control sticks be made to operate the way that yokes do, and if one pilot moves a stick forward, the other pilot’s stick also moves forward? Synchronized in tandem? There’d be an actual physical tug-of-war if one pilot were pushing forward and another were pulling back. (That *is *how yokes work, right?)
They could be. That’s not really an ideal situation though either and is actually very similar to what happens with dual input to the sidesticks. If both pilots apply an input to the sidestick then the inputs are summed which is basically what happens if both pilots apply input to a yoke except that in the case of the yoke the strongest pilot will win. The advantage of the Airbus FBW is that a pilot can take complete control without having to fight the other pilot for it, the disadvantage is you can’t see or feel what the other pilot is doing.
In some mechanical yoke setups if both pilots fight against each other hard enough the controls will disconnect and each pilot will have control of different bits of the system.
Question:
I heard that some stalls cannot (or are very hard) to be recovered from.
The idea is as the nose pitches up there is a place where the wings block airflow over the horizontal stabilizers and elevators. As such, once in that place, trying to pitch the nose down won’t work.
Was that at issue here? Even if not in this case does that make a plane a lost cause or change how my altitude you need to recover (if it is even possible)?
It took China Airlines 30,000 feet to pull out of a dive in a 747SP and pulling 5Gs.
Has there ever been a case where that functionality saved a plane that would have crashed with linked yokes?
In theory, I can understand the arguments for either system. In practice, it seems the Airbus method has led to more problems. Have there been problems with the linked yokes that we haven’t heard about, and do the pilots in this thread have a preference one way or the other?
It is a difficult question to answer. Accidents that are prevented don’t become public knowledge, they don’t always even become company knowledge, it’s even possible that the crew themselves aren’t aware that disaster was averted.
People typically prefer what they are used to. It is fairly rare to find pilots who have significant experience in both Boeing and Airbus aircraft. I can’t give a preference because I’ve never used an Airbus side stick.
Pretty much what Ricard Pearse said just above. Here’s some more background on related topics.
There absolutely have been flights with control jams where the ability to split the yokes and control parts of the system has been a life saver. They’re not common in modern times, and this whole standard approach to designing mechanical flight controls dates back to the 1940s when surviving battle damage was a major design consideration.
There are a several of different hazards & situations we’re dealing with here. One is what if a yoke or a portion of the control linkages jam? You’d like to have some redundancy / separation to overcome that. Another is what if one pilot is incapacitated and slumped over the yoke? You’d like the other pilot to be able to overcome that undesired input. A third is what if one pilot is nuts & trying to crash? Same problem, same solution, different upstream cause.
A fourth scenario is the simple confusion where each pilot thinks he’s the one actively flying and is actively making control inputs. The effect is the two inputs are fighting each other. A lot of hand-flying is done by feeling the airplane’s feedback into the yoke; if I’m also feeling somebody else pushing & pulling that’s gonna feel real weird, like a malfunction or worse. And the weirder it feels, the more likely somebody is to tunnel-vision in on monitoring the instruments, rather than going wide-angle & noticing the other guy. A four-and-a-halfth scenario is where nobody is flying, each thinking the other pilot is doing so.
Statistically speaking, occurrences of this fourth scenario outweigh the other three by a sizeable margin.
The upside to mechanically linked yokes is the motions are physically large, so I readily can see if the other pilot is moving it. And in side by side cockpits I can look over and easily see if his/her hands are on or off. It’s rare to have either two or zero pilots thinking they are the pilot actively in control. But it happens occasionally for a few seconds.
In fore-aft separate cockpits like the USAF/USN F-4 and USAF T-38, B-47, etc., there’s a lot more possibility for confusion since you can’t see the other pilot. For that reason there’s a standard “ceremony” for exchanging control and for double-checking there’s no confusion. Such aircraft absolutely have crashed for that documented cause: two or zero pilots were flying and they didn’t correct the confusion before the situation got beyond recovery.
In the Airbus style cockpit there’s an intermediate opportunity for confusion about zero, one, two, or who. Unlike an F-4, you can see the other pilot, but unlike a yoked aircraft you can’t easily see his inputs or feel them at all.
OTOH, each pilot has an absolute veto over the other pilot’s input, which greatly simplifies recovering from crazies, or 1000x more commonly, incapacitation. Consider a 300 lb. fatso having a heart attack and slumping over the controls next to a lightly-built 110 lb. other pilot. Muscling the deadweight off the controls may be more than he/she can do. The Airbus approach means one button-push solves that problem.
And the concern about mechanical jams is almost completely ameliorated in the Airbus because there’s so much less mechanism to jam. And for jams of the input stick itself, there’s still the veto button. The mechanical equivalent, split-yoke operations, still leaves an adverse input into the total aircraft from the jammed side.
OTOH, now you gain all the other risks of FBW: computer malfunctions and failures.
Bottom line: It’s different. And so it’s exposed to failure modes the other approach isn’t. We can’t legitimately call out those new failure modes without also giving credit for the other failure modes the engineers removed by not using the alternative system.
There are just now becoming available FBW side sticks which have motion feedback, where small electric motors are controlled by the FBW computer and push back against the pilot’s hand. The intent is to provide a more natural feel when hand-flying, but a side effect is the motors now enable adding software to the FBW system to move the sticks in sync, providing that visual and tactile feedback about any inputs the other pilot is making.
Sounds good, but the side-stick is well out of your forward field of view, and the physical range of motion is tiny, 1/2" total at most. Meaning most actual inputs are very very hard to see. So when I’m flying I could readily feel any unexpected inputs from the other side, but when I’m not supposed to be flying I can’t effectively monitor what the other pilot is doing by eye; I’d have to put my hand on the stick & feel it. Which in turn would be disruptive to his/her feel.
And now you gain all the failure modes associated with those motors. What if one goes stupid and starts pushing? What if the extra software contains a bug? etc.
So far these new sticks are being introduced on the latest generation of ultra high end business jets. AFAIK Airbus (and Boeing) have no plans to include them in any current products.
My personal bottom line: Moving sticks are an incremental improvement, but they don’t fully eliminate the incremental potential for confusion over a yoke. But more importantly, confusion-avoidance is not the only measure of merit for a control system.
Thank you both for your observations. I guess I find it odd that for a field as widely studied as flight crew performance that there’s no consensus on how best to control an airplane.
To me, the Airbus system of averaging the inputs from too sticks seems too clever by half (I think that’s the expression). When things start going wrong (as they did on Air France 447), and you’ve got a complex problem to diagnose and solve, seems like you’d want to know exactly what commands the airplane was getting; not through some compromise algorithm, but actually feel it in your hands. It would be one less thing to worry about when the shit hits the fan.
There’s a great deal of ongoing research in this area by NASA as well as the avionics companies and aircraft manufacturers.
The F16 and A320 FBW were bleeding edge new tech when invented in the 70s and 80s. And to a certain extent were done just because they could be. And subsequent Airbus products have all used essentially the same system in the name of inter-type compatibility and reduced training costs.
The next generation cockpits will be the first that have really had the benefit of extensive in-service ops with multiple competing high-level approaches to cab design. The B787 actually is fully FBW like the Airbus, but they chose traditional yokes, mostly for reasons of training familiarity and inter-type compatibility.
Cockpit instrumentation is likewise undergoing a research revolution. In many ways what we have today is simply a computer-generated replica of traditional mechanical instruments whose design was largely dictated by the limitations of WWII-era mechanical technology.
With all these things we’re feeling our way into the future with a lot of concern for backwards compatibility with at least the workforce, if not existing hardware / software. And given the reliability standards involved in the whole man-machine system, we’ve got to be very sure any major changes in presentation or operating theory not only work well in normal ops, but are robust in partial and total failure modes. Plus not contribute to operator confusion.
It’s a very tall order. One of the ideas to the fore right now is that the real safety problem is having humans in there at all. There are large-scale well-funded research programs to remove one of the two pilots from airliners. With the longer-term goal of removing that other goofball eventually.
IMO if there wasn’t weather it’d be technologically doable in 10-20 years. Given the realities of weather, and more significantly, the difficulties in machine detection and understanding of weather, it’ll take longer.
Removing copilots will save about $1 per ticket. Management throughout the value chain is pretty convinced people will gladly choose the cheaper alternative.
I’ll be retired before it happens. But I’m not sure I’d want to be a 20 yo aspiring professional pilot right now.