but I'd prefer to have the ability to kill the automatics if I believe it would save my life.

 

Pilots can kill the laws by pushing a few buttons on the overhead. A jetBlue Captain told me that if AF 447's computers somehow managed to retain normal law the plane would have been in a much better position for the crew to realize their mistake, and the industry also lost 2 757's due to unreliable indications at night and over water.

Pilots can kill the laws by pushing a few buttons on the overhead. A jetBlue Captain told me that if AF 447's computers somehow managed to retain normal law the plane would have been in a much better position for the crew to realize their mistake, [...]

 

You can kill some of it, but not all. See the above posts.

 

[...]and the industry also lost 2 757's due to unreliable indications at night and over water.

 

Please be very careful in how you represent fact here in the forum. I know for a FACT one of those was because there was tape over the static ports. Automation checking pilot actions being on or off would not have prevented that. Additionally, the aircraft did not exceed any flight envelope. It hit the water. That is very, very different.

 

http://en.wikipedia.org/wiki/Aeroper%C3%BA_Flight_603

I'm not sure what your other reference was. There's a Birgenair, but that was over land (also pitot-static related).

Please be very careful in how you represent fact here in the forum. I know for a FACT one of those was because there was tape over the static ports. Automation checking pilot actions being on or off would not have prevented that. Additionally, the aircraft did not exceed any flight envelope. It hit the water. That is very, very different.

 

http://en.wikipedia....perú_Flight_603

I'm not sure what your other reference was. There's a Birgenair, but that was over land (also pitot-static related).

 

All 3 accidents had obstructed sensors whether it was taped over static ports, debris in pitot tubes or iced up pitot tubes. Flight 301 did crash into the Atlantic as did Aero Peru.

You can kill some of it, but not all. See the above posts.

 

What remains if you switch the bird into DIRECT law? I was under the impression that killed all protections as well as all the fancy input processing. As far as I'm aware in DIRECT law, side-stick deflection = control surface deflection. No automatics, just you and physics, just as if you were flying a Cessna.

 

This is also what is mentioned in this link: http://www.airbusdri...us_fltlaws.htm. DIRECT law kills everything.

 

Pilots can kill the laws by pushing a few buttons on the overhead. A jetBlue Captain told me that if AF 447's computers somehow managed to retain normal law the plane would have been in a much better position for the crew to realize their mistake, and the industry also lost 2 757's due to unreliable indications at night and over water.

 

In normal law alpha floor protection would have been active, which would have prevented the stall no? Unless the AoA readings were also invalid of course, anyone who wants to double check that with the BEA report feel free.

 

To return to the main topic, summarising the differences based on the links given above:

 

In pitch the Airbus side-stick always commands a load-factor, whereas the Boeing yoke commands a pitch rate below certain speeds and a load factor above it.

In roll the A side-stick commands a roll rate and the B yoke commands an aileron deflection.

A won't allow the pilot to exceed certain limits, B will let the pilot do that, but the required control forces are very high.

 

All the above only holds in normal law of course.

All 3 accidents had obstructed sensors whether it was taped over static ports, debris in pitot tubes or iced up pitot tubes. Flight 301 did crash into the Atlantic as did Aero Peru.

 

Right, which has nothing to do with computer interpretation of pilot inputs. Sure, it has to do with computer display of data, but even an old altimeter from a Cessna would have errors with pitot-static systems plugged.

 

What remains if you switch the bird into DIRECT law? I was under the impression that killed all protections as well as all the fancy input processing. As far as I'm aware in DIRECT law, side-stick deflection = control surface deflection. No automatics, just you and physics, just as if you were flying a Cessna.

 

I believe you are correct in your interpretation of what protections (essentially none) are provided in Direct, but I don't believe it is as simple to place an Airbus into Direct as people are making it out to be. It's my understanding that a pilot cannot simply flip switches on the overhead to put the plane in Direct. Flipping switches (the FCUs) on the overhead will put it in Alternate. Putting the gear down with the FCUs off will then place the aircraft in Direct.

 

In my arguments for the ability to place an aircraft in Direct in the 0.00000000000001% scenario in which you'd need it. I would argue that the process of getting it into Direct needs to be more simple. On the other hand, I can't think of a situation in which a pilot needed to kill the automatic to save a flight.

Boeing yoke commands a pitch rate below certain speeds and a load factor above it.

 

Load factor; can't find any reference to this in the 777 FCOM?

Load factor; can't find any reference to this in the 777 FCOM?

 

Page 9 of the PDF that was linked above: http://www.davi.ws/avionics/TheAvionicsHandbook_Cap_11.pdf

 

If you have an actual FCOM I would trust that as the more authoritative source though.

I think my main issue with it is that the automatics are only as good as they're designed and written. The people writing the software can't think of every single contingency. The likelihood that you'd ever run into that issue is extremely slim, but I'd prefer to have the ability to kill the automatics if I believe it would save my life. I've spent too much time around computers, and doing my own programming, to simply place my life in the hands of it.

There have been design errors which have been corrected (such as assuming pitch attitude didn't need a limit in altitude capture mode) but the chances of a coding error are remote, since the devices are multi-lane and programmed by different teams in different languages to the same requirement. Any DFCS is rigorously coded and tested to very high standards.

I do believe that it's made too much of an issue, in that 99.99999999999999998% of the time, it's a non-issue. I do believe, however, that sometimes you have to get inventive when things go terribly wrong. I can't tell you what that situation would be, though, and that's why I believe that the automatics wouldn't know how to handle it, either.

I agree that far too much is made of Airbus FBW "issues", usually by the "If it isn't Boeing I'm not going" fraternity and those who have a very limited understanding of the concept. But if everything is functioning correctly there's no reason for the pilot to "get creative". If systems are not functioning correctly then the DFCS is designed to degrade to lower control law levels appropriately and automatically. It simply is not necessary to manually switch to Alternate or Direct.

Pilots can kill the laws by pushing a few buttons on the overhead. A jetBlue Captain told me that if AF 447's computers somehow managed to retain normal law the plane would have been in a much better position for the crew to realize their mistake, and the industry also lost 2 757's due to unreliable indications at night and over water.

If AF447 could have remained in Normal Law there would have been no mistake to realise. Unfortunately with so many invalid inputs it simply can't be allowed to do so, garbage in inevitably means garbage out.

 

As for manually switching the aircraft into a degraded Law, that can only make the situation worse. An Airbus in Direct Law is much harder to fly manually (and won't fly automatically). The flight control system will automatically degrade if it needs to, but there is no valid reason for the crew to want more than 67 deg bank or to select an AOA above Alpha_CLmax, for example. Airbus FBW allows the pilot to select and maintain maximum lift very easily, why switch that off? Attitude limits are intended to prevent crew disorientation causing excessive attitudes. Switching that protection off would be of no help.

Page 9 of the PDF that was linked above: http://www.davi.ws/a...book_Cap_11.pdf

 

If you have an actual FCOM I would trust that as the more authoritative source though.

 

Thanks. I read the flight controls section of the FCOM from www.smartcockpit.com. I wouldn't say that the document you linked is wrong, as there is nothing in the FCOM that contradicts it; in fact it seems to include additional info not included in the FCOM. I hope the FCOM isn't the final source of info for pilots transferring onto the type in that case!

I think my main issue with it is that the automatics are only as good as they're designed and written. The people writing the software can't think of every single contingency. The likelihood that you'd ever run into that issue is extremely slim, but I'd prefer to have the ability to kill the automatics if I believe it would save my life. I've spent too much time around computers, and doing my own programming, to simply place my life in the hands of it.

 

I do believe that it's made too much of an issue, in that 99.99999999999999998% of the time, it's a non-issue. I do believe, however, that sometimes you have to get inventive when things go terribly wrong. I can't tell you what that situation would be, though, and that's why I believe that the automatics wouldn't know how to handle it, either.

 

Precisely! As you said, "extremely slim". An extremely unusual scenario where there may be some benefit from a manover beyond the flight envelope protection is very, very, very rare.

 

However... the oportunity for flight envelope protection to save lives is far more common.

 

Thus, flight envelope protection saves more lives than the capability to execute crazy manoeuvres.

 

Put it another way... your life is more likely to be saved by flight envelope protection, than the ability to kill the automatics and conduct a super extreme, flight envelope transgressing manoever.

Put it another way... your life is more likely to be saved by flight envelope protection, than the ability to kill the automatics and conduct a super extreme, flight envelope transgressing manoever.

 

I agree. Just to be clear, however, I never argued against flight protection. It definitely has a purpose. I am just under the belief that if it needs to be kicked off, it should be possible.

 

Last I checked, FAR 91.3 states that "[t]he pilot in command of an aircraft is directly responsible for, and is the final authority as to, the operation of that aircraft." If I believe that the automagic is preventing me from doing something that would save the aircraft, I should be able to kick it off. Not being able to do so could arguably be against the line of "[...] and is the final authority as to."

 

Again, I'm not saying that you're going to instantly save people by kicking protections off, but automation cannot cope with every single situation. The protections there are great, and they help out in the 99.99999999999999998% scenario. That last 0.00000000000000002% is the reason we still have pilots up front and not computers.

If I believe that the automagic is preventing me from doing something that would save the aircraft, I should be able to kick it off.

The Habsheim A320 pilot thought he should have been able to pull up the nose but the computers said "no". But had he done so the aircraft would probably have stalled into the runway instead of gliding into the trees. The aircraft didn't have the energy to do what he wanted it to do. The protections in that case almost certainly saved lives which would have been lost in a heavier impact.

The protections in that case almost certainly saved lives which would have been lost in a heavier impact.

 

I'd agree with that. I was about to cite that earlier in defense of my point, and then did a little reading and saw what you're seeing. Reading that report was actually somewhat irritating, especially the complacency of the captain.

 

The next conversation would then be about how many pilots would actually recognize that type of situation? I remember reading an accident report referencing the crew regaining control of the aircraft through lowering the landing gear (I'll try to find it and post a link). It's situations like that where I start to wonder about the instinct versus automation protections. Clearly, in that case, automation probably would've only helped, but it makes you think.

 

There's a pull tab in your trunk in case you ever get carjacked and stuffed in the trunk.

Do you really think you're gonna need that?

 

You're definitely right, though. Had the crew killed the automagic in the Habsheim incident, there would've been a different story.