People keep asking if me if I’ve read Langewiesche’s Fly by Wire. So I finally did and drafted a book review of Fly by Wire. Comments would be appreciated.
13 thoughts on “Review of the Fly by Wire book”
Comments are closed.
A posting every day; an interesting idea every three months…
People keep asking if me if I’ve read Langewiesche’s Fly by Wire. So I finally did and drafted a book review of Fly by Wire. Comments would be appreciated.
Comments are closed.
You say: “The Airbus could have had a big switch saying “I am ditching”. This would disable landing gear extension, something that pilots might be inclined to do by habit, and close some of the pressurization outflow valves to prevent water seeping in. But the Airbus A320 doesn’t have anything like this and the crew was required to remember not to extend gear.”
It is my understanding that the A320 does have a “ditch switch” which performs the aforementioned functions, but was not actuated on Flight 1549.
“The Airbus could have had a big switch saying “I am ditching”.”
Actually, I believe several articles have mentioned that the Airbus (as well as most commercial jets) do have a ditch button which closes up the aircraft. I’ve read that pilot routinely press this during de-icing, so as to not ingest any nasty de-icing compound.
Where to begin.
First I have not read the book so I am making some inferences based upon your review.
With respect to the Cali crash you state; “By then it was too late to climb over the mountain without perfect technique and the two Air Force veterans were not quite as good as the fly-by-wire system of an Airbus.”
I find it hard to believe that Langewiesche left out the fact that the 757’s spoilers remained deployed from the initial GPWS warning until impact and if memory serves the aircraft impacted terrain less than 200ft. from the peak. Though the airbus originally only auto-retracted spoilers if the aircraft approached the high angle of attack regime it was later modified to auto-retract if TOGA power is selected. Thus a current model airbus would likely not have duplicated the Cali disaster.
I can’t speak to Langewiesche’s interpretation of the 1988 Airbus crash. Though it has been a long time, I did read the final report. Suffice it to say, though the pilots were confused about what mode the aircraft was in, modifications were made to the FADEC system after that accident.
The best glide speed of an airbus is somewhere between 230 to 270 Kias depending upon model and weight. This speed is NOT displayed on the tape. What is displayed is best lift over drag for optimum single engine climb. I don’t know if this is your mistake or his.
You state: “The computer system could easily have been programmed with the gliding capabilities of the A320. Did it offer any suggestions about the nearest glidable airport and runway? Did it highlight those airports in a special color on the moving map? No.” The airbus navigation display does show a gliding range, depicted as a dashed circle showing the pilot how far the airplane can glide.
Though bleed air would have been needed for an engine re-start attempt the APU bleed sw was never depressed and thus no APU bleed air was ever available for a starter assisted re-start.
You state that Sullenberger seemingly flared too high but provide none of the author’s evidence. At 144 Kias the airplane should have been well above stalling speed in the configuration the crew chose. (CONFG 2 Slats at 22deg / Flaps at 15 deg) Therefore Sullenberger never approached the pitch limits, which is a good thing as the FBW system would have lowered the nose to prevent the stall resulting in a pitch down into the water.
Flaps and slats on airliners will automatically move in certain circumstances in quite a few airplanes. (Slats in a 737 approaching a stall will auto extend in certain configurations and the flaps on an airbus will auto-retract in certain conditions.) As I pointed out earlier the spoilers would have retracted, had they been deployed, as the aircraft approached a critical angle of attack.
Further the airbus is equipped with a switch that ensures the aircraft is configured for a water landing. In fact it is called the DITCHING switch and it closes all openings below the waterline – outflow valve, pack valves, ventilation inlets and outlets etc. Crews are quite familiar with this switch as it is commonly used when the aircraft is de-iced/anti-iced. (It was not activated in this event)
I am now tempted to read the book just to see if the author is even close to accurate.
Kurt, Greg, Jon: The Airbus’s “ditch switch” is not a workload-reducer for pilots, but rather increases their workload. It is something that they’re supposed to find at the end of a long checklist in the QRH. It doesn’t help with flaps or gear. The switch that Airbus labels “ditch” closes valves to try to seal up the pressure vessel. Other airliners have similar switches (not always as comprehensive), but they are labeled differently.
Jon: Langewiesche does mention the spoilers on the 757 in Cali. Retracting them was part of the “perfect technique” that the pilots did not have and that a computer would have had (in fairness to the AA crew, their sim training on responding to a GPWS warning did not include retracting the spoilers; they did everything that they’d been trained to do). On some modern business jets the speedbrakes will automatically retract if the thrust levers are advanced beyond the 85% N1 position (as they would be any time a pilot wanted to climb out of something ugly).
As far as the best glide speed being displayed, that’s what Langewiesche says. I have not flown an Airbus so I can’t offer an opinion. Best L over D would be minimum sink with no engines, correct? Not best glide?
Thanks for the information about the glide ring; Langewiesche did not mention that.
As far as the final landing speed, I don’t think it was 144 knots. Langewiesche used MPH. So it was probably closer to 125 knots.
I wouldn’t fault them for landing downwind; this is probably close to the minimum airframe damage that could be expected regardless of details about the configuration and the environment.
Phil,
I disagree with your assertion that the ditching switch increases the workload. On must aircraft closing the outflow valve manually is at least a two step process. The system must be switched into an alternate or manual mode and then a pickle switch needs to be held against a spring to actually close the outflow valve. This can take up to a minute or two. (On most aircraft manual control of the outflow valve operates much slower than automatic control – this is done so that manual control can be accomplished smoothly. The downside is a long time to fully close the valve.)
Second, though the QRH is available and should be used in most circumstances, it is not a substitute for systems knowledge and planning. In this particular incident the QRH was of little or no value. Unfortunately circumstances placed a pilot new to the airbus in the FO seat that day. Optimally, at USAirways, the captain will turn the aircraft over to the FO in an emergency as soon as possible so that he can focus on planning. Given the nature of the emergency and the FOs lack of experience in type the captain felt this was inappropriate in this case.
As far as setting flaps and gear automatically; Aircraft configuration changes require one pilot to command and the other to execute. The reason this has been standard practice is to ensure that no inappropriate configuration changes occur. For the landing gear to be inadvertently lowered it would take two pilots making the same mistake. I can’t imagine a scenario where one pilot makes such an obvious blunder let alone two.
With respect to the flaps; there is no way to automatically set them for a ditching because the requirements are going to be different in every scenario. The recommended setting, for a power off ditching, is COFG3. However, if one or two hydraulic systems are out that may not be the case. The flaps will operate slowly when powered by only one of their two drive systems.
If the aircraft is down to one or two hydraulic systems the pilot may not want to rob half of the primary flight control jacks of power to move the flaps. (Or even with all three, with two operating through the PTU)
Though I have not talked to the man I am guessing this is why the captain stated that “flaps two was good enough” when asked about the configuration prior to impact. He probably did not want to reduce the power to the primary flight controls that late in the game.
Even prior to the Cali accident it was almost universal technique to never remove one’s hand from the spoiler lever if they were deployed. The Cali scenario had long been on manufacturer and training dept. minds. Learjets had a red spoiler warning light on the annunciator panel, Falcons had a big red light that shined into the captain’s eyes, etc.
Though some modern designs have incorporated a TLA switch to retract spoilers this was not possible in earlier jets. These aircraft sometimes required spoilers and power for decent in icing conditions.
On the matter of touchdown speed I will have to access the actual reports and look at it again. This does not change the fact that in a ditching a pilot would have to ensure that touchdown occurs prior to the FBW system reducing pitch to preserve airspeed. In a ditching the FBW stall protection is sub-optimal. A full stall at 12 inches would be preferable to a forward impact at flying speed.
I have strong doubts about the inherent goodness of automated flight-envelope protection systems. As a 40-something programmer, I’m well aware of the sausage-factory demons hiding inside every complex body of code. Coupled with the reliance of such systems on fallible sensor technology (which is itself reliant on proper maintenance), I’m extremely hesitant to put my faith in any flight automation system. The recent B2 bomber crash in Guam did little to alleviate my concerns about the reliability of such systems.
On the other side of the equation we have the crapshoot of possibly exhausted and/or inexperienced humans who often, in emergency situations, seem incapable of doing much of anything beyond exactly what their training and/or checklists tell them to do. (The 757 crew forgot to retract spoilers? Really?)
In the absence of convincing evidence that either alternative is significantly more reliable than the other, where am I supposed to place my faith?
I am not familiar with the area, but I would guess that potential water-rescue resources are more plentiful south of the George. That could have been (consciously or unconsciously) part of the decision to turn left instead of right.
Don: It would definitely have been possible to land the plane in water with less damage, though perhaps not without at least one engine. Much of the damage was caused by the fact that the vertical speed was greater than the design spec. According to Langewiesche, airliners must be certified to land in water without breaking up. The conditions, including descent rate, are essentially those that are easy to achieve with one engine spinning. The certification authorities have not updated these regulations despite the fact that no airliner has ever ditched while one engine was still running (everyone who ditched had run out of gas on both sides, had all engines fail, etc.).
Rob: There are certainly more boats in the area adjacent to Midtown Manhattan, but that was probably something that the pilots would have wanted to avoid. They would not have wanted to hit a boat. Mostly, I think, they turned south because that was the direction of all of the nearby airports: LaGuardia, Teterboro, Newark. They had not yet made the decision to land in the river.
Jon: I will have to defer to your experience with the Airbus. In the CRJ that I flew we had a switch to close the outflow valves. It was as quick and easy to push as any other switch in the airplane. And as with the Airbus, it did not do anything to help the pilot make a successful landing in the water.
Airbus only has a finite amount of time to develop an airliner, so also only a finite amount of time to develop, integrate and test all these extra things that would be nice to have. The “common cockpit” policy also limits how much extra they can put into newer models compared to the original A320.
They will have to make a break from that policy at some point in the interest of moving forward.
Bas: I did not mean to imply that the Airbus was deficient, certainly not compared to other airliners. My point was that any Airbus versus Boeing differences were not relevant to the Hudson River landing.
I really enjoyed this book, as I have most of Langewiesche’s writing. But it was intriguing that he missed one opportunity to more deeply analyze a situation were flight envelope protection may have actually hindered pilot response. In his discussion of the Gimli Glider episode, he points out (as do contemporary accounts) that the airplane arrived high and fast at approach. Without sufficient hydraulic power to deploy spoilers or anything else, the captain was forced to sideslip the 767 to lose altitude and airspeed. Correct me if I’m wrong, but I’m pretty certain that no airliner’s envelope protection is going to allow a sideslip! Without the ability to slip, what would have been the captain’s alternate option?
WL came tantalizingly close to this analysis, even laying all the groundwork in his Gimli narrative, so I guess there was a conscious choice not to contradict the ‘thesis’ of his book. As Phil and others have pointed out, the ‘thesis’ that FBW w/ envelope protection greatly reduced risk on Cactus 1549 may be a little weak. But in the larger sense, Langewiesche’s points about modern airliner technology and flight discipline were fascinating and I didn’t mind using the A320 as the vehicle to carry them through a book.
I really enjoyed the technical observations made by your contributors and can only add that my letter to the editor, published in the New Jersey Star Ledger, pointed out that Airbus automation benefits actually posed a drawback here because Sully was prevented from approaching as close to full stall speed as he might have liked. Therefore, I will proceed to vent my pique at the book for WL’s gratuitous characterizations of airline pilots and their profession. When Sully laments that “the best and the brightest” might in the future avoid the airline piloting profession, WL retorts (prologue) that they never have been attracted to it because of the job’s supposed “monotony.” What a slap–especially coming from one who never worked in the profession. (WL may have piloted professionally–but never as an airline pilot.) Personally, I was never bored due to the fact that I was always looking for something to go wrong and always in the midst of making an alternate plan–just as the profession and common sense demand. As for the “brightest,” I never met a dumb airline pilot, headstrong maybe, opinionated definitely, but never dumb. As I reflect on the relative firepower of my former law school classmates and my fellow airline pilots, I would have to opine that the two groups were probably not separated by more than two or three IQ points–with the pilots on the high side. Finally, the WL comment that absolutely takes the cake is his declaration that the “most desirable trait for airline pilots” would be “placidity” (prologue). I can think of a lot of desirable traits, and “placidity” would not be among them. How about intelligence, drive, reality-based, strong responsibility quotient, high capacity to forego short term for long term goals, health consciousness due to the demands of the profession– I could go on. My surmise is the WL probably wanted to be an airline pilot but never made it, for whatever reason, with his subtle jibes forming an unconscious compensation. R.D. Truitt