Is this something now taught in simulators and is the training widespread?

Not only have they been trained in it, but the AD to fix the potential failure was issued and completed in...1997.

Does this apply to all carriers around the world? Are there any 737 pilots here who have practiced the technique[s] in the simulator? I understand action is required quite quickly. Is that correct?

convair880mfan wrote:

Does this apply to all carriers around the world? Are there any 737 pilots here who have practiced the technique[s] in the simulator? I understand action is required quite quickly. Is that correct?

Did you not just read the above post about the issue being resolved in the late 90s...?

It's not just the 737, it can happen to any airplane in the world... it just so happens that the problem became apparent after 3 737's crashed in similar incidents.

but an Airbus can crash due to rudder hardover as a 737.

The solution was training all pilots (not just 737 pilots) to reduce AoA because at crossover AoA, rudder overpowers ailerons ability to counteract roll.

The entire US airline industry covers this and other specific issues in EET/URT training a minimum of every 2 years,

if unrecognized, departure from controlled flight will happen pretty quickly and if close to the ground (like the 3 737 accidents in question) recovery is all but impossible, but the events leading up to the departure is easily recognized. and it's just a matter of having enough altitude to recover, but even better not to get yourself in that situation to begin with.

The way the demonstration is conducted in the simulator is the instructor has the non-flying pilot deflect the rudder full left or full right with the rudder pedal. - that is the rudder hardover.
you fly the plane normally - it's very controllable even with the rudder fully deflected. you do turns to a heading, turns with asymetric thrust, etc.

Then you increase AoA, by slowing the aircraft, and you see how hard it becomes to maintain heading and altitude as alpha increases, then at crossover alpha, the plane rolls hard in the direction of the rudder, and opposite aileron will not stop the roll. that's when you recover by unloading to reduce alpha and regain control.

GalaxyFlyer wrote:

At high AOA, rudder was how we controlled and rolled the F-100. Get sloppy at the top of a loop, use aileron instead of rudder to pick up a wing, you could come down the back side 45 degrees off the track you went up.

But isn't picking up a wing with a rudder kind of frowned upon in transport category airplanes after AA 587?

Woodreau, could you perhaps elaborate on "all but impossible" Is the "all but impossible" scenario practiced in the simulator?

convair880mfan wrote:

Woodreau, could you perhaps elaborate on "all but impossible" Is the "all but impossible" scenario practiced in the simulator?

Did I not answer in the post above?

Woodreau wrote:

The way the demonstration is conducted in the simulator is the instructor has the non-flying pilot deflect the rudder full left or full right with the rudder pedal. - that is the rudder hardover.
you fly the plane normally - it's very controllable even with the rudder fully deflected. you do turns to a heading, turns with asymetric thrust, etc.

Then you increase AoA, by slowing the aircraft, and you see how hard it becomes to maintain heading and altitude as alpha increases, then at crossover alpha, the plane rolls hard in the direction of the rudder, and opposite aileron will not stop the roll. that's when you recover by unloading to reduce alpha and regain control.

Woodreau wrote:

It's not just the 737, it can happen to any airplane in the world... it just so happens that the problem became apparent after 3 737's crashed in similar incidents.

but an Airbus can crash due to rudder hardover as a 737.

That's a disingenuous way to put it.

Whereas all airplanes with hydraulic flight controls can theoretically suffer from a rudder hardover, the fact is that the 737 had a design fault which made it a lot more probable... As far as I know, no other airliner at that time had to modify their design specifically to prevent it from happening.

The fact that it happened 3 times in a relatively short timeframe on that type and yet has never happened on either the 737 or A320 or anything else since then, despite the global fleet exploding and the number of hours and cycles flown now being orders of magnitude higher is evidence that there was indeed a problem with that particular type.

As for training for control surface hardover events, most operators/regulators don't include it in their syllabuses, at least on the types I'm familiar with.

SteelChair wrote:

GalaxyFlyer wrote:

At high AOA, rudder was how we controlled and rolled the F-100. Get sloppy at the top of a loop, use aileron instead of rudder to pick up a wing, you could come down the back side 45 degrees off the track you went up.

But isn't picking up a wing with a rudder kind of frowned upon in transport category airplanes after AA 587?

It was frowned upon before AA587 as well.

Woodreau, I guess you elaborated enough. I was just hoping for a little more explanation.

I am interpreting what you said to mean that if the aircraft is too close to the ground, anything the pilot would do would not work because the aircraft needs altitude to recover. So I guess it would be like the Delta Tristar accident at DFW which because it experienced a microburst too close to the ground there wasn't enough altitude to right the situation. Or like the 747 that crashed at Bagram AFB because it stalled and it needed altitude [which was not available ] to recover from the stall.

So I am guessing that simulator training is for rudder hardovers that are not close to the ground, rudder hardovers close to the ground being hopeless situations. Is that correct?

convair880mfan wrote:

Woodreau, I guess you elaborated enough. I was just hoping for a little more explanation.

I am interpreting what you said to mean that if the aircraft is too close to the ground, anything the pilot would do would not work because the aircraft needs altitude to recover. So I guess it would be like the Delta Tristar accident at DFW which because it experienced a microburst too close to the ground there wasn't enough altitude to right the situation. Or like the 747 that crashed at Bagram AFB because it stalled and it needed altitude [which was not available ] to recover from the stall.

So I am guessing that simulator training is for rudder hardovers that are not close to the ground, rudder hardovers close to the ground being hopeless situations. Is that correct?

No. Everyone above has indicated the purpose of the training is for crew to recognize the onset of a hardover condition early enough to take action so that it does not become unrecoverable at low altitude.

And the 747F at Bagram stalled because the cargo load shifted and exceeded CoG limits for the flight controls. The outcome would have been bad regardless of altitude.

You might get a better ‘feel’ for the subject by contacting a local flight school and arranging a demo flight. Ask the CFI to demonstrate a stall/spin so you can feel the change in control response.

At least in the US, the FAA now requires air carriers to conduct EET/URT training every two years.

When the rudder hardcover events happened to the 737s, pilots were presented with an aircraft rolling due to the rudder. Under normal circumstances, when the plane rolls uncommanded, you counteract with opposite aileron. Opposite aileron didn’t work in the 3 accidents. As the airplane continued rolling the nose drops and pointed towards the ground. When the nose drops, normally you’d stop the nose from lowering by pulling back on the yoke/stick, which increases alpha. that compounded the problem. Pilots at that time didnt have the time to fully appreciate what was happening.

You can trade altitude for time. How much altitude do you need? that depends…

For the crossover AOA demo, I believe we did them at 5000ft. During upset recovery exercises, some scenarios were as low as 3000ft with the plane inverted when the plane is turned back over to you to recover.

AF447 had 37,000ft to recover from their stall…. 37000ft wasn’t enough.

Rudder hardover close to the ground is not an unsurvivable problem. You have to understand what makes the plane uncontrollable is when alpha exceeds crossover alpha. So keep alpha below crossover alpha. You’ll know you’ve reached it when the plane starts rolling uncontrollably and the nose drops. To reduce alpha, you lower the nose and/or increase airspeed.

When you are close to the ground you have less time to do the correct action. But it’s not “nothing you do will change the outcome.” Sure you can give up and do nothing because you think it’s hopeless. Pilots back then didn’t know what to do. pilots today know or should know that the correct response to uncommanded roll due to rudder hardover is to reduce alpha to regain aileron control. so the only correct action is to push the yoke/stick which is counterintuitive because your nose is already pointed towards the ground.

Todays pilots have the benefit of accidents and the resulting investigations to learn from the accidents.

convair880mfan wrote:

Does this apply to all carriers around the world? Are there any 737 pilots here who have practiced the technique[s] in the simulator? I understand action is required quite quickly. Is that correct?

Any 737 flying in the world with the old Parker Hannafin rudder set up that can fail in that way would be flying illegally. That AD was issued in 1997.

As Woodreau said, pilots of transport category aircraft with the potential for this failure are now trained to deal with uncommanded rudder hardovers. The pilots of Eastwind Flight 517 did successfully deal with an uncommanded hardover and their experience is what led to the AD. The AD that was, once again, issued in 1997.

Aaron747 wrote:

convair880mfan wrote:

Woodreau, I guess you elaborated enough. I was just hoping for a little more explanation.

I am interpreting what you said to mean that if the aircraft is too close to the ground, anything the pilot would do would not work because the aircraft needs altitude to recover. So I guess it would be like the Delta Tristar accident at DFW which because it experienced a microburst too close to the ground there wasn't enough altitude to right the situation. Or like the 747 that crashed at Bagram AFB because it stalled and it needed altitude [which was not available ] to recover from the stall.

So I am guessing that simulator training is for rudder hardovers that are not close to the ground, rudder hardovers close to the ground being hopeless situations. Is that correct?

No. Everyone above has indicated the purpose of the training is for crew to recognize the onset of a hardover condition early enough to take action so that it does not become unrecoverable at low altitude.

And the 747F at Bagram stalled because the cargo load shifted and exceeded CoG limits for the flight controls. The outcome would have been bad regardless of altitude.

Not only did they exceed center of gravity limits for the flight controls, they had no flight controls altogether. The load shift resulted in severed hydraulics lines along with the jackscrew for the horizontal stabilizer.