Planet 9

from Global Research, June 12, 2009 by William John Cox
From the Airbus to the Spaceplane: The Future of Commercial Aviation

Although pilot control of commercial aircraft had progressed beyond the direct use of cables and pulleys to move aircraft control surfaces by relying on hydraulics and electrical assistance, the introduction of electronic control of commercial aircraft increasingly shifted responsibility from human pilots to computers.

Stevie - I've edited your post to take out the entire story you posted, which is a copyright violation. In the future, please just post the link, and a paragraph. We cannot publish stories from other web site in their entirety here.

This is explained in the posting rules, if you need a refresher, please read them again...

Planet Porsche - Announcements in Forum : General Off-Topic

brad

__________________
22-year PCA Member
PCA DE Instructor

There was a news article a couple days ago regarding US Airways 1549 landing in the Hudson. That plane had fly by wire and the pertinent part was because the computers would prevent the plane from stalling, the pilot was able to just pull back on the stick and concentrate on keeping the wings level.
It seems the jury's out on this but all planes in the future are going FBW.

The liability is shifting to the companies/people who provide the hardware and software (algorithms) that dictate the airplanes performance under all (most) any condition. Whereas this may have been true for sometime now, it seems to be coming under the limelight of attention now.

I wonder in the future if there will ever be pilotless passenger planes, or perhaps the crew scaled back to one pilot and the support crew. Wiith extras (like food and beverage services) also being scaled back already, maybe even that crew will be reduced to almost none, but that is another story.

__________________
Kevin
2007 Cayman S........SOLD.

This tragedy is somewhat reminiscent of the Airbus that lost it's vertical stabilizer over the NY Rockaways. I wonder if the vertical stabilizer is to blame. It seems that in bad crosswinds or violent weather the composite stabilizer is breaking off.

__________________
Gator Bite
PCA Cayman Register Advocate

The NY accident in 2001 was basically because of the Co-Pilot (Pilot flying) flew into the wake of a previous departing airplane and his aircraft had an uncommanded roll with wich he counteracted by aileron and abrupt and full rudder input. He basically went from full deflection left to full deflection right ( i forget which direction he went first, doesn't matter) at the maximum speed at which the controls could move. That overstressed the bolts which hold the vertical stabilizer on and they sheared. The aircraft designers knew of this weak point and had asked that the various airline training departments not teach roll recovery with full rudder deflection. Sad, that the Pilot flying had reverted back to early training (we know best what we learned first) and had done what he was not supposed to do.

I believe that it only happened once and that was in the incident in the post above. And it was not violent weather or crosswinds.

__________________
Cayman Club Outsider

Yikes.... You would think they would make an aircraft capable of handling any combination of control inputs without breaking apart. And, if the aircraft has a weakness like this, you would expect them to put some sort of safety in place to prevent the combination. If the rudder has fly by wire controls, it's as easy as writing a safety into the control program.

__________________
Gator Bite
PCA Cayman Register Advocate

Looks like the answer to my comment is in the article posted by Stevie.

Looks like the rudder is FBW, and it does have a safety in place to prevent fatal stress, but the butthead who programmed the FBW computer decided that at low speeds the limits should be suspened, to make taxi steering possible. You'd think they would base the limit on filght.

__________________
Gator Bite
PCA Cayman Register Advocate

I am not making light of any of this, but this article reminded me of 2001: A Space Odyssey: "I'm sorry Dave, I can't do that...." I would guess that most PP users have read the book, but if not, it includes a fascinating exposition of the relationship between man and machine, and it's over 40 years old.

Well, I may be mistaken, but I have a feeling the limits may be based, at least partially, on flight. Having more rudder authority available during slower phases of flight and landing makes sense (more so for landing, when x-winds must be compensated for rather frequently). I'll admit that it seems strange to program limits and not make Va (maneuvering airspeed -- limit for max control surface deflection) the limit point, but a pilot should know not to make full control inputs (and certainly not a series of vigorous alternating full inputs) above Va to begin with, ESPECIALLY not in severe or wake turbulence.

It will be interesting to see what emerges as the investigation goes on. I certainly hope they will find the CVRs, but with the amount of finger pointing that is already taking place, I'm afraid the cover your *** mentality may preclude any true distribution of their contents.

Yeah, no kidding. I didn't even think of that, but while reading that rather long article I felt a chill run down my spine and developed a fear of ever boarding another Airbus. I can't believe how much control they've taken from the pilots, and how many planes have fallen from the sky because of it.

Around here, we have to rehash the usual topics like:

• My BOSE radio sucks, too much bass
• Why doesn't my Cayman have automatic headlights
• The Key is on the wrong side....!?!
• Why can't I hook up my iPod
• etc.....

Somewhere there's a forum dedicated to professional pilots. Those guys are authoring threads like:

• Why did my A320 dive 650 feet three times in a row despite the fact that I was pulling on the stick and screaming like a little girl.....?
• You don't need the nose gear straight to land (Video with awesome spark display)
• Powerbrake maneuver sends me into the wall, again....!
• I sh!t myself while flying again. Anyone found anything that will take the odor out of the cabin?
• My Airbus tried to land itself during an air show flyby.
• Best way to stay calm during an emergency landing
• Anyone installed a Boeing Vertical Stabilizer on an Airbus? How would it effect our warranty?

__________________
Gator Bite
PCA Cayman Register Advocate

Not sure if they could, or they should. Most, if not all, cars still allow you to input until you crash, as do most, if not all, planes. To prevent the pilot from doing such an input, the algorithm has to first decide if this input is dangerous, or not. And then it has to decide if it is just going to stop the input from being processed and output, and how it does that. PID (proportional-integral-derivative) control has to take into account the change, the accumulative effect of the change and the rate of change. Or it has to derive the output response from a map, similar to EFI systems in our cars - which has to be developed from a myriad of possible situations.

Imagine a pilot trying to accomplish a manoeuvre that may be possible in an emergency situation and the maneuver being declined. I wouldn't think that flight control has been progressed to the point where all of these parameters can be developed to a fine point (the 2D aspects of stability control in cars is just getting to that point...)

If anyone knows more about the subject, jump in and add or challenge, I'm basing my comments on process control in petrochemical plants with thousands of initiators, DSC control, and thousands of end devices, PLC's, APC, etc. Flight control, no doubt, would have its own complexities in addition.

__________________
GT Silver CS
1980 Honda CBX
2004 Suzuki SV1000
2006 Subaru Legacy
2002 Honda Insight

Part of the problem is that to make an aircraft structurally sound enough to not break it would be too heavy to fly. There has to be a balance between strength and weight. Carbon Fiber is a material that hold much promise. Today's aerobatic aircraft have fuselages made of it and they can withstand 20 G's (twenty times the weight of gravity). That is well beyond the limits of the human body. A pilot could not make the airplane break. So, we sacrifice strength in order to keep the airframe weight low. So that we may take more passengers, baggage and fuel. Pilots are trained to maneuver the aircraft safely within the design structural limits, even during an emergency. But alas, we are only human and we sometimes act outside of our knowledge base and that is when things usually go wrong. I do not think we are technologically at the point yet that we should let machines make decisions for us. Flying a plane is a very complex activity. There are many variables that effect each and every second of flight. The human brain is amazing at how quickly it can process multiple streams of input and make instant and correct actions. The more we do something the better we get at doing it and the brain is able to increase it's awareness of the task at hand and that's when the subtle nuances appear. That's what I think sets us apart from machines at this point. There is nothing wrong with fly-by-Wire so long as humans retain the ability to do what they need to do when they need to do it.

I just heard that Air France is replacing the airspeed sensors in its fleet of A330s & 340s, but are they just putting a condom on an uzi? If there is a flaw in their overall philosophical approach to flight control, making changes could cost billions. It took 10 years for the Comet’s structural flaw to be recognized and finally grounded. How many more years for the jury to be out on the Airbus?

I’ve often heard mentions by pilots that “this baby can practically take-off and land by itself.” Airbus designed its flight control system to eliminate common all pilot errors, by eliminating the pilot. But just like HAL it was not as smart as it thinks. And that may be its downfall. It seems that Airbus has a system that flies the plane with the assistance of humans. While Boeing has a system that allows the pilot to fly the plane, and gets corrective assistance from computers.

It scares me to death to think that same computer that can cause nose wheels turn at 90 degrees before landing; is the same one that may cause and rudders to over-extend when it should not. Changing sensors seems like a cheap solution. And, with the upcoming Paris Air Show and, United’s big multi-plane order on the block, the timing for Airbus could not be worse.

-Stevie


PS: Gatorbite, you going to DE at Lime Rock?

I have been involved with the design and flight testing of fly by wire systems since the late 70's and the airframe design ramifications that are the result of our constant quest for higher performance and increased safety. I also have personal experience in the operation of such systems being the first person to land a fully fly by wire aircraft (F/A-18A) on and aircraft carrier at night and the first person to ride a fully automated carrier landing in a FBW aircraft. I also led the design team that designed and certified the first system that uses adjusted engine power as a directional flight control. So I am fairly familiar with the issues being discussed here. I'll say up front, however, that I have no insight into the investigations that are ongoing in the altitude loss incident nor do I have insight into the exact Airbus implementation of the flight control laws. For these reasons, I will discuss general design criteria and will state that the incident investigation and root cause correction process has worked fairly well for a number of years.

First, the vertical stabilizer fracture over New York. The design criteria for rudder deflection is primarily determined by airplane control in an engine out situation, both during the takeoff run and in initial climb. The size of the rudder is the driving contributor to minimum engine control speed (the speed at which an engine failure can be controlled and the takeoff continued). An engine failure below that speed results in a rejected takeoff, above that speed, the takeoff is continued. In flight, the design must accomodate the the maximim deflection of roll devices and still maintain balanced flight (basically, keep the nose into the wind). On the other hand, the rudder cannot be so large as to cause excessive sideslip in a crosswind landing situation. The structural design of the vertical stabilizer/rudder combination is set by expected flight loads and maintaining aircraft control during upsets. I believe that the aircraft involved has an augmented flight control system (similar to power steering with the airplane equivalent of PSM) rather than a full fly by wire (computer in 100% control of flight control surface position). To put it in our car perspective, we have an augmented system. The PSM will help you avoid trouble, but if you ignore the signals that the car is giving, the car will not prevent you from overcontrolling the car losing control of the car (departure from controlled flight in airplane parlance). In a real drive by wire system, the car would alllow you to go to the limit of adhesion, but no farther no matter what you did to the steering wheel. But I digress, back to airplane structure:

All modern airplane have hydraulic actuators that move the control surfaces. These are sized by the amount of force they need to deliver to meet the design criteria. Let's say that the most critical design maneuver for a particular airplane is failure of the left outboard engine and the need to maintain balanced flight at 175 knots airspeed with the other 3 engines at full power. With the rudder size and deflection set by the continued takeoff criteria, the rudder actuator needs not be stronger that required to meet the directional control criteria. That criteria sets the strength of the acrtuator mounts and the aerodynamic loads under that most critical condition determines the entire vertical stabilizer loads. Now, quite frankly, there are numerous airplanes in service that are perfectly capable of producing flight loads that will break the airplane. Designers do things like rudder deflection limiters that are activated by airspeed or flap position to make sure that even a crazy pilot cannot deflect the rudder enough to cause a catastrophic failure. To this day there are few airplanes that limit pitch control in such a way that the pilot is incapable of pulling hard enough to break the wings off, and every pilot knows that beyond a certain speed (called corner speed), you will overstress the airframe before the wings stall. The same is true of the vertical stabilizer. I would venture to speculate that there are probably no transport airplanes that are designed to withstand fulll opposite rudder deflection while in a high sideslip condition. As to designing to alll possible conditions? How would you react if Porsche came out with a mandatory mod to add 400 pounds of steel to every car because of the possibility that you would crash into a tire barrier and be unable to drive the car away undamaged. Back to airplanes: There are numerous airplanes that are certificated with pilot warnings and training as part of the certification. An FAA certification does not include only the aircraft, but the way it is designed and made, the way the maintenance and pilot manuals are written, and the training that every pilot must have. In this case, there are statements in the public media that the pilot flying commanded maximum opposite rudder in a high sideslip condition. I have also heard that the flight manual contains a caution against such control inputs and I have heard conjecture that the training did not emphasize these limitations enough.

Now to fly by wire: The above scenario would probably not have happened because rudder deflection commands would be modeled to prevent overload - IF the airplane is equipped with sideslip sensors. While I do not know the specifics of the altitude loss incident, I have heard enough to speculate that the independent airspeed sensors iced over at essentially the same rate such that the comparative algorithms that compare the readings of each channel to all the others in order to detect a failure could not detect the common error. As the pitot probes iced, the total pressure would reduce that would indicate decreasing airspeed. At some point the computer begins to force the nose down in order to prevent a stall. In that case, an algorithm designed to increase safety may have actually contributed to an incident.

I look at FBW in airplanes like a PDK in a Porsche. It controls the clutch and transmission better than any man can and it completely ignores any command that would result in an overspeed. And you could argue that it increases safety by giving the driver more mental capacity to deal with detecting track conditions and feeling the car.

My professional opinion is that FBW increases the efficiency and safety of modern airplanes. In the military, we are already operating completely autonomous air vehicles for 24 hour missions controlling them from half way around the world and flying them from Guam to California. The industry is now dealing with the requirements for operating autonomous (unmanned) vehicles in the national airspace.

The challenge of developing FBW airplanes is the system engineering to make sure you consider and mitigate all risks and that the software is developed and tested under the most rigorous of circumstances. I'm really proud that the first FBW military fighter that I helped develop has the best carrier landing safety record in history, but the development darn near killed me more than once.

__________________
Kenny

2006 Arctic Silver CS 6M, PASM, Sport Chrono,, short shifter, 19" wheels, Pirrelli tires, Milltek Catbback, Softronic SRP and 4.2 Flash.

From my experience, American designers include more flexibility in the systems than do European designers. That leads to some interesting situations: In an American airplane, disengaging the autopilot overrides autonav also. in the early versions of the Airbus, pilots got "caught" in holding patterns because the logic prevented them from disconnecting the nav guidance until an Expected Approach Clearance time was reached, While their system works, ours was more flexible - at the cost of much more complex software.

__________________
Kenny

2006 Arctic Silver CS 6M, PASM, Sport Chrono,, short shifter, 19" wheels, Pirrelli tires, Milltek Catbback, Softronic SRP and 4.2 Flash.

Kenny:

Thanks for sharing your knowledge and experience . . . that's an example of one of the reasons I love this website!

__________________
Andrew My Blogs My Garage My Articles
2006 Cayman S Tiptronic (Year-round Daily Driver & Part-time Track Toy)
Mods: SRP, DREAM-1, VentureShield, Xpel Headlamp shields, 35% Tint, MotorDrive Front Mesh Grills, Bumperplugs..
Track Prep: Pagid Yellows, Champion RS98s/NT05s, Deka ETX14 Battery, CG-Lock, Extinguisher, Rennline towhook, BBI Harness/Rollbar

Awesome reply Ken.... Thanks....

I don't have any issue with FBW. I just don't trust the computer control they are putting on top. That wouldn't be so bad if they allowed the pilot to easily override it, like Boeing does.

__________________
Gator Bite
PCA Cayman Register Advocate

OK, that's just plain funny!

Kenkg's write up on flight control systems was one of the more interesting and informative things I have read in quite a while. Thank you for that!

I'm now going to have to send it along to my buddy at NAS Patuxent who works on avionics for the F/A-18 to decipher it for me...

__________________
'08 CS - Meteor Gray, Cocoa, PASM, Short Shift, Sport Seats, Sport Steering Wheel, Bi-Xenon, and a few other goodies.

I love this car!!!