Machinbird Quote:
Someone mentioned that the SS does offer feel. Yes, it does offer feel, but not feedback. Maybe I'm just "ham handed", but the feel offered by the Airbus SS is of little value to me. Once you break out of center, I don't feel much difference between part and full deflection. But more importantly, the FBW system prevents me from feeling the controls I.e. "feedback". In an MD80, I can feel a wing start to drop (thru the yoke) before it happens. In the Airbus, I have to see the wing drop on the ADI, make a corrective control movement, watch the ADI again to verify the control was successful. IOW, with feedback, I never had to think or process a control input. My hands felt the controls and sent the message straight to flying part of my brain which returned the correct input to the control yoke. No cognitive process involved
.
TTex600, I could be wrong, but what I think you are perceiving is the time lag in your perceive-act cycle. An aircraft like the MD-80, you see the slightest of movement and you apply the slightest of pressure automatically and the wings stay level. If you have to perceive a bit of movement, and mentally process it, before making a correction on the 'Bus, then the response is going to be a bit delayed.
One of the early visual flight simulators for the A-4 Skyhawk had a bit of lag in the visual display of about 0.3 seconds. No problem to fly it 'visually' as long as the control movements and rates were small. But if the rates were high enough, the lag caused a disconcerting overshoot and could generate a PIO, normally in the roll axis. This may be an analog to what the PF on AF447 experienced in Alt 2 law with the higher than accustomed roll rate.
No, I am not perceiving the time lag. When I have a tactile connection to the airplane, I FEEL the aircraft. You are limiting me to sight, just as the airbus limits me to sight, and I reject that condition. Actually, in the Bus, I have no choice to accept it, but I don't like it.
Maybe my problem here is that my swept wing experience comes from opposite ends of the spectrum. Lears, both Longhorn and tip tank, and the DC9 series are both cable controlled, steam gauge airplanes. Then I went to the Airbus 319/320/321. Any reasonable experienced Lear/DC9 pilot can give you a fairly accurate estimate of airspeed using nothing more that control response for input. I thought that was flying 101. Heavy controls = fast, mushy controls = slow. Maybe I expect too much out a transport category airplane, I guess I am just spoiled.
Back to AF447, Why do some of you insist upon damning the pilots because "the instruments were giving normal indications the last three minutes"? How were the pilots to KNOW the the indications were normal? I've worked the UAS/ADR scenario in the sim and it takes far more than three mins to decipher the good data from the bad data. I agree that the PF pulled and continued to pull, but I won't damn him for that even though we can find no reasonable justification for doing so. We know NOW, that the speeds returned to normal and that everything else never deviated from normal, the crew didn't have that luxury. They went from a known attitude with bad indications to an unknown (self induced) attitude with good instruments which added more uncertainty to the equation. Regardless, they had twenty seconds to figure out what we've had twenty months to study.
Yes, if they had just flown pitch and power (lyman appears to believe that pitch info was also suspect which leaves us basically unrecoverable, but I'll assume now that pitch info was accurate),,,,,,if they had just flown pitch and power, they would have had time to decipher the conflicting info and determine their true situation. Be that as it may, he made a mistake thirty some odd seconds into the event and pulled. After that, the combination of UAS - stall warning sounding as speed increased, no tactile feel to provide clues, etc, prevented the crew from understanding the situation. Anyone not familiar with most all modern transport category/high tech airplanes can't really understand, but when one recognizes that everything presented to him is computer generated - including pitch info- and then suffers a failure leading to misleading information, he is faced with two choices. 1. pick the most likely correct instrument and use it to judge the others and hope you choose wisely or 2. assume that NONE of them are correct and methodically work your way through a troubleshooting flow chart (on paper or from memory, either way). Methodical, step by step, procedures take time and require you start from some known. The limited info we have regarding the crews actions the last three mins. would seem to indicate that they were anything but methodical and deliberate. That points to training.
Put yourself in the AF447 cockpit, you've made one mistake, a mistake you don't yet recognize, and find yourself out of control. Your instruments are confusing, your master warning/caution system demands attention, bells and whistles are ringing, You know that something is terribly wrong, you don't know what. What do you do?
In this case, one mistake caused an accident - pulling for no reason. The Airbus philosophy would appear to be one of protecting the airframe from the pilot. Somehow this went horribly wrong. IMHO, Airbus didn't realize that the pilot could damage the airframe in more ways than just flying too fast or too slow, than just pulling too hard, than just banking too much. He could be confused to the point of error and they didn't anticipate that.
If I could go and change one thing and prevent this accident, I would say that would be: train the pilots to ignore everything but pitch and power in the event of any instrument abnormality.
If I could go and change any one thing to have helped them recover from this accident, it would be: put the stall warning system on WOW (weigh on wheels, or air/ground) switch so that it would not bias out at taxi speed and below.(I realize that more needs to be done than just a simple WOW, but I'm trying to keep this example simple) Putting myself in the dead crews place, I imagine that the most confusing input was the stall warning activation that returned, twice, as the airplane trended to recovery.