CREWLESS PASSENGER AIRCRAFT - BAE SYSTEMS.

[cooperplace]

Same mate told me about the 'sales' flight of one of the first Airbus aircraft to a US airline (possibly Northwestern? -.... Airbus had tweaked the landing control algorithm. The demo flight progressed smoothly until they tried to land, whereupon the engines fired up and forced a go-around; several further attempts had the same result.

As it happened, Airbus had a computer-control guy on board, who was able to hook into the control computer system and change the landing parameters while they circled, and so they all got down safely.

 

That airline did NOT buy Airbus.

very interesting: can you elaborate?

In any case, I thought airbuses could be readily changed into alternate law, as well as another "cessna" mode?

 

I should mention that as a jabiru pilot, I'm unlikely to ever run into these issues.

 

 

[dsam]

I'm not sure whether the story is apocryphal, but I've heard that during WWII, Russian aircraft designers were routinely mandated to be aboard all "maiden" flights along with the test pilot. When those designers knew at the outset that their backsides were airborne & at risk in their new designs, it tended to focus the propeller-head's minds on reliability and safety first & foremost (rather than avant-garde brainwaves). Perhaps the designers of pilot-less airliners need this same "motivation"

 

To my mind, pilot-less RPT is a possibility, provided nobody else (without such equipment) is up there with them (ie. no Rec & GA), and the entire worldwide ATC, Radar & TCAS legacy system is dumped wholesale, in favor of a (yet undeveloped) fully engineered "greenfields" solution to navigation & landing for a pilot-less worldwide system. I'm not holding my breath waiting for THAT to happen any time soon...

 

Dave

 

 

[Oscar]

very interesting: can you elaborate?In any case, I thought airbuses could be readily changed into alternate law, as well as another "cessna" mode?

I should mention that as a jabiru pilot, I'm unlikely to ever run into these issues.

No, I do not have the technical knowledge. But both the AF and the Qantas 380 situations, says to me that Airbus still places its computer algorithms above a pilot's over-riding control. I believe - but do not know, as I have little interest - that the Airbus 'alternate law' means that the 'channel' with the fewest exceptions is given precedence, NOT that it is a 'we computers can't agree, it's over to you chaps' mode.

 

 

[M61A1]

You have to be joking.Even in the '70's, DoT (pre-CASA), had issues with Qantas trying to circumvent regulations that prevented flying with not only primary but also redundant systems inoperable.

Nope, not joking.....the issues you are talking about is humans overriding the computers. When the computer wakes up wrong, and thinks it's faulty, it wont do anything. The human has to reset it.

It's probably the most annoying thing about digital aircraft.

 

 

[biggles]

Bob: you were onboard? That would have me waking up in the early mornings sweating. even now. For anybody with aircraft knowledge, seeing the damage to that A380 would be on the scale of the scene in Pulp Fiction where Jules decides that not being shot was an Act of God.

No Oscar, not onboard that day thank goodness . In saying "we", I was referring to .... the Aussies, and others on board , Qantas, aviation travel worldwide, etc. etc. Having read the book, and some knowledge of digital control systems, the amount of info.being fed back to the flight deck may well have overwhelmed a minimal crew. CRM at its best ..... Bob

 

 

[Oscar]

Nope, not joking.....the issues you are talking about is humans overriding the computers. When the computer wakes up wrong, and thinks it's faulty, it wont do anything. The human has to reset it.It's probably the most annoying thing about digital aircraft.

M6: it depends on the instruction set. I believe there were some early cases where the F22 Raptor was rendered effectively useless by Windows Vista glitches...

 

 

[Marty_d]

So why bother making an artificial human when making a real one is relatively easy and fun?

You don't have kids, do you. The fun bit is the first 3 minutes.... (ok, 45 seconds)

Then there's the rest of your life.

 

 

[dutchroll]

very interesting: can you elaborate?In any case, I thought airbuses could be readily changed into alternate law, as well as another "cessna" mode?

 

No, I do not have the technical knowledge. But both the AF and the Qantas 380 situations, says to me that Airbus still places its computer algorithms above a pilot's over-riding control. I believe - but do not know, as I have little interest - that the Airbus 'alternate law' means that the 'channel' with the fewest exceptions is given precedence, NOT that it is a 'we computers can't agree, it's over to you chaps' mode.

I am both a current Airbus A330 pilot and also a mate (from military days) and colleague of mine was one of the pilots in the QF32 cockpit when that incident happened.

The Airbus has 3 basic flight control laws (actually it has some in between which are used for takeoff and landing but never mind about those) which are applied in certain circumstances by the flight control computers - "Normal", "Alternate", and "Direct" in order of priority and reversion.

 

The plane is flown in normal law. With certain system failures, it will revert to alternate law or direct law. Each reversion causes the loss of certain flight control functionality including the loss of flight envelope protections, and progressive loss of flight control feedback and gains until with direct law, it's simply "control deflection proportional to stick deflection". It's not particularly easy to fly in direct law, because there is very limited tactile control feedback through the side-stick to the pilot.

 

You cannot "select" alternate law or direct law. The only way of doing this is to deliberately induce certain failures which will then cause the reconfiguration to take place in the flight control computers.

 

Alternate Law changes the pitch rate feedback and gains. Pitch attitude protection is lost. Alpha floor protection (stall protection) is lost. Bank angle protection is lost. Overspeed protection may or may not be lost depending on the failure. Yaw damper authority is limited. Roll rate is reduced.

 

In direct law there are no flight envelope protections, pitch trimming is manual only, and control deflection is directly proportional to side-stick deflection. As I said above, it's actually not that easy to fly in that mode (especially given the number of failures which lead you there in the first place) - at least not smoothly.

 

 

[dutchroll]

^^^ Having said that reversionary flight control laws can't be "selected", there is actually a way of quickly forcing the plane into alternate law by "simulating" a multiple failure. It has consequences for flight data display for the pilots though.

 

It is designed to be used in the exceedingly rare case that a fault in the stall protection system causes it to activate falsely and force the nose down. By getting the plane into alternate law this protection is disabled and control is regained. It's basically tricking the computers.

 

I'd be interested to know how they figure that out in a pilotless plane.

 

 

[planedriver]

My prediction is that they will prove beyond any doubt that the pilotless plane is a perfectly safe and viable option.Until it crashes.

Just after the automated anouncement that "to reassure you all, we wish to advise that everything has been so thoroughly tested and retested, there is not the slightest chance that anything can go wrong, go wrong, go wrong"!

 

 

[planedriver]

Having just viewed a slide-show of raising the Airbus from the Hudson River, the last page really says it all. "Can we ever imagine how lucky those people were that it happened so close to the city, and more so that they had a seat of the pants pilot on board"

 

 

[Oscar]

^^^ Having said that reversionary flight control laws can't be "selected", there is actually a way of quickly forcing the plane into alternate law by "simulating" a multiple failure. It has consequences for flight data display for the pilots though.It is designed to be used in the exceedingly rare case that a fault in the stall protection system causes it to activate falsely and force the nose down. By getting the plane into alternate law this protection is disabled and control is regained. It's basically tricking the computers.

 

I'd be interested to know how they figure that out in a pilotless plane.

Dutch - thank you for that very full and very clearly able to be understood, even by me - explanation. I assume that that 'tricking the computers' operation is what was done on the Northeastern demo flight I mentioned.

 

But it raises a question - to me - that may be completely on the wrong track, perhaps you can set me straight here. Please bear with my very limited knowledge!

 

If you have to introduce 'apparent' failures to force the computers to relinquish control, is there any way of the pilot knowing what has ACTUALLY failed and what has been induced? Again, my lack of understanding of the complexity of a 'heavy' control systems may make this a nonsense question, but it seems to me that if you have to disable certain control functions, you may be losing control capability that actually exists but its condition is being misreported?

 

I have a fair bit of faith in the competence of (most) pilots to judge conditions ( aerodynamic conditions, anyway) from the feel and reaction of the aircraft - though that may only be realistic for small and uncomplex aircraft - and to evaluate what is possible to achieve and therefore how to handle the situation. As a crude but true example: a family member, when test-flying a Jabiru for certification, had the elevator control cable disconnect at the stick-end clamp ( Bolts not properly tightened). However, in a Jab., the elevator trim provides a secondary elevator control (albeit of somewhat limited authority), and he was able to make an uneventful landing, by advising a straight-in approach and playing the throttle and flaps a little bit differently to normal to minimise pitch changes and effect a round-out..

 

What would happen in an Airbus aircraft, if (say) a limit-switch for the flaps on one side goes feral and mis-reports the flap position? I imagine that the computer(s) would rationalise that unequal flap extension is dangerous and a) refuse to deploy the opposite side flaps other than as balanced to what is being reported for the 'faulty' side, and b) calculate the landing profile (if that is the correct term) for that condition? Do you guys get the chance to actually test out and determine by feel and experience, what is ACTUALLY happening?

 

P.S. - have you gotten your engine problems for the Pitts sorted?

 

 

[dutchroll]

Dutch - thank you for that very full and very clearly able to be understood, even by me - explanation. I assume that that 'tricking the computers' operation is what was done on the Northeastern demo flight I mentioned.If you have to introduce 'apparent' failures to force the computers to relinquish control, is there any way of the pilot knowing what has ACTUALLY failed and what has been induced? Again, my lack of understanding of the complexity of a 'heavy' control systems may make this a nonsense question, but it seems to me that if you have to disable certain control functions, you may be losing control capability that actually exists but its condition is being misreported?

Not sure about the demo flight but I do know in the test flying phase when they first went to fly-by-wire they certainly did have problems landing it and had to change the software.

The ECAM display is pretty good at telling you exactly what has failed. Plus, the particular way it is induced is also indicated on the overhead panel. So there are different indications for a failed item, and one which has been deliberately switched off.

 

What would happen in an Airbus aircraft, if (say) a limit-switch for the flaps on one side goes feral and mis-reports the flap position? I imagine that the computer(s) would rationalise that unequal flap extension is dangerous and a) refuse to deploy the opposite side flaps other than as balanced to what is being reported for the 'faulty' side, and b) calculate the landing profile (if that is the correct term) for that condition? Do you guys get the chance to actually test out and determine by feel and experience, what is ACTUALLY happening?

There are three separate pickoff units which provide flap position data. Flap asymmetry is sensed by left and right asymmetry position pick-off units which have no other purpose in life. These units apply wing tip brakes to lock the flap position if asymmetry is detected and they can't be released in flight. An instrumentation pick-off unit provides separate position data to the centre display. Feedback position pickoff units provide data to the slat/flap control computer. So if the instrumentation pickoff unit went awry and gave a bogus position indication in the cockpit, the flaps would still extend just fine. If the asymmetry pickoff unit gave bogus information, they'd lock in position and what you have is what you have, but the indication in the cockpit would accurately show that the position is not asymmetric. You'd just be going "why the heck did they lock?" The only answer is that for some reason an asymmetry was detected but is not shown.

The ECAM message, if there is an actual fault, guides the computation of the landing performance data if the flaps are in a non-landing position. Otherwise you'd go with what is shown. That data is all computed on our company iPad using a special Airbus app which does takeoff and landing performance for all normal and non-normal configurations.

 

P.S. - have you gotten your engine problems for the Pitts sorted?

Unfortunately no. Still waiting for the engine people to decide what to do in the USA. Probably new cylinders and pistons under warranty is most likely, I think. At the very least it'll be new pistons and rings.

 

 

[Oscar]

Wow, thanks for that. Gives me a bit more confidence if I HAVE to fly in an Airbus aircraft...

 

Poo on your engine problems - is this likely to be the automotive pistons? Mark Willard sometimes stretches his Numchukka's feathers roughly over my place to shake out the mud-wasps... and the gentle rumble of the roundy-roundy is a welcome change from bloody choppers thrashing overhead sometimes at lower than 500 feet (Alan Jones, one day I may have the rifle out, and return your pain-in-the arse comments with something more tangible). He's very happy with the standard engine rather than your hot-dog version, reckons he just turns on the fuel, loads it with all the oil he can carry, pats his whimpering wallet fondly, and off he goes.

 

 

[facthunter]

I wouldn't run a worked over radial unless it was absolutely required to beat the competition. Military engines are short life compared with the derated civvy versions of the same engine. High revs at low boost is bad for radials. Puts a giant load on the big end bearings The pistons on each master rod have a lot to do too, so can't possibly last as long as the others do. Love them otherwise, but they are designed to cruise. Nev

 

 

[dutchroll]

It's not something they've experienced on other similarly modified ones by the same company (as far as I know) so I think there has been some debate about what has gone pear-shaped. Essentially what was found was excessive piston-cylinder wall wear, causing the engine to be very hard to start when warm and getting worse as it ran in. Interestingly all the oil changes (of which there have been several) have been extremely clean with virtually zero metal (filters pristine). Engine parameters have been consistently right down the middle of the green ranges. Rings are currently a prime suspect but they're looking at all angles.

 

 

[facthunter]

Metal has to go somewhere if it's getting excess clearances. HAS to show up in the filter, unless there are galleries it could lodge in. They MUST have measurements. IF the cylinders are Nitrided they rarely wear. Pistons can collapse (lose dimension) if they aren't up to spec. A loose piston will barrel face the rings and you will have lot's of blow by. Measuring blow by is a very effective way of knowing the engine's health. IF it's breathing heavily, it's sick. Nev

 

 

[dutchroll]

4 oil changes. There's only been a very small number of extremely fine particles in the filters. I've seen them myself, cut wide open and rolled out flat. These have been checked by three licensed engineers who were all impressed by its cleanliness. But there has been material deposited onto the cylinder wall. Piston & ring specs are being checked by the manufacturer. Sorry for thread drift!

 

 

[Phil Perry]

4 oil changes. There's only been a very small number of extremely fine particles in the filters. I've seen them myself, cut wide open and rolled out flat. These have been checked by three licensed engineers who were all impressed by its cleanliness. But there has been material deposited onto the cylinder wall. Piston & ring specs are being checked by the manufacturer. Sorry for thread drift!

You are allowed a bot of drift on my thread Dutch, as the fount of a host of educational stuff on here. . .! ( Ass licking ?. . .wot ME ? ? ? ? ? )

 

 

[dutchroll]

( Ass licking ?. . .wot ME ? ? ? ? ? )

Jeepers Phil. Get your tongue out of there!

 

 

[facthunter]

I'm not buying into this. Though Flattery does seem to get you everywhere . (so I'm told). Nev

 

 

[DWF]

I've been told the crew of the next generation of airliners will be a man (or woman) and a dog.

 

The man (or woman) is there to feed the dog ...... the dog is there to bite the man (or woman) if they touch anything!

 

 

[Birdseye]

I've been told the crew of the next generation of airliners will be a man (or woman) and a dog.The man (or woman) is there to feed the dog ...... the dog is there to bite the man (or woman) if they touch anything!

Showing your age David with that one.....

 

 

[Yenn]

I recently found a video of an engine problem on an airbus of some sort. German crew and the temps were rising too much. It seemed that the crew spent most of their time looking at info on the glass cockpit, wondering what is happening and then calling in another crew member to discuss what to do. Obviously not a serious problem but the crew didn't seem to be really on top of the problem.