Posts about: "FBW (Fly By Wire)" [Posts: 35 Pages: 2]

The side stick fitted to 201 (in 1977 according to Flight International), was very much proof of concept for what would be the future A320 programme sponsored by the French government, very very limited hours were flown using it apparently (like 10!). The next step I believe in the development was to fit it onto a FBW modified A300 in the early 80s, but for the initial tests to see if a pilot could fly with a "computer joystick" it had to be done on Concorde as this was the only a/c with a suitable FBW system.

It would be good to know how it was done, but I guess it could have been wired into the emergency flight control system, which relied on electrical strain gauge inputs to move the control surfaces if there was a control column jam.

I can't add too much as I was only on the fringes of this.


I think it might have been done by fitting a D/A converter to substitute the digital signals from the sidestick for the normal Concorde stick resolver output. The 'laws' could then be treated as a special case of pitch damper etc. inputs so that the standard Concorde electrical signalling system could be used downstream. I don't know this for sure, it is just a thought - maybe Christiaan could comment on its feasibility.


I think they were looking at at least the rate control but probably not all the envelope protections. Certainly at that time we were having discussions with them about the merits of their FBW laws against the laws we (BAe) were trying out on our BAC1-11 flying test bed. No prizes for guessing who won that argument.


And before anyone asks; no I don't remember (if I ever knew) how we were implementing FBW into the 1-11!


CliveL

I think Gordon probably meant "airliner" rather than "aircraft".

The implementation of FBW on the F-16 was intended to solve a very different problem than that of the A320 series. The F-16 was one of the most maneouverable fighters of it's day, but the way General Dynamics achieved that was by having an aerodynamically unstable airframe that *required* constant computer correction to keep her airborne and going in a straight line. The A320 was designed to be as aerodynamically stable as any other airliner, but the FBW was simply designed to assist the pilot by easing the workload when it came to actually controlling the thing, as well as provide safety features as backstops when things got hairy.

The only Western analogue FBW aircraft of the time of comparable size to an airliner (IIRC) were the Avro Vulcan and Concorde. France had no Vulcans, so the airworthy test Concorde they had to hand was the obvious choice.

"The only Western analogue FBW aircraft of the time of comparable size to an airliner (IIRC) were the Avro Vulcan and Concorde. France had no Vulcans, so the airworthy test Concorde they had to hand was the obvious choice."

The Vulcan wasn't pure fly-by-wire. It's control surfaces were remotely signalled but had a physical connection rather than electrical. Sort of fly-by-stiff wire :-)

One of the spin-offs of having Concorde at Manchester is that it is used by many different organisations for events. One a while back had a discussion about Technology benefits due to Concorde, part of that was about Concordes "1sts". First FBW was challenged as the feeling was that there must have been another by then. Nobody could think of another so it stood the debate. Is anyone aware of any others ?

If you limit consideration to aircraft that entered serial production, you may be correct. If aircraft that were prototyped and flew are included then the Avro Arrow was probably the first, as it was designed, flew a partial test program, and got cancelled all prior to 1960!

Thanks, johnjosh43 .

Not a hundred percent legible (lack of resolution), but I'll try to decipher it.

I agree, no trace of any electrical signalling ("FBW").

CJ

As I recall, they referred to this research project as a CCV (Controlled Configured Vehicle) design study. It would be great if we could get this confirmed, but they talked about subsonic drag reductions of 10 to 15% by flying (not taking off!!) with a far more aft CG than the norm. The 'system' I seem to remember, as a result naturally commanded some down elevon, which increased lift. As the aircraft could then fly with less alpha, I guess this is where the drag reduction comes from. (Clive, I wonder if you could find out through one of your contacts if this was true?).
I'd still personally like to know how the sidestick was integrated into the flying control system, I've been thinking and can not now believe that sidestick inputs could be simply input to the flying control system 'at resolver level'. Remember that the concept of the FBW system on Concorde was that resolvers were utilised as simple 4 wire synchros, and the pitch and roll axis utilised a CX/CDX/CT chain, which produced the error signal to the ESA's in the Autostab computers. Using a sidestick completely breaks up the chain, and my guess is that a seperate digital unit contained the flight rules which were summed against PFCU CT position and sidestick input would have been necessary. It is possible then that an analog output from this 'box' could be fed to the Autostab Computer ESAs and hence drive the elevons. I'm probably completely wrong, but I'd surely still love to know the truth. As you say Clive, ideal stuff for Concorde 2.

Best regards
Dude

They weren't looking for cruise drag reductions; just takeoff climb improvements which would have required genuine relaxed stability - CGs back at the aerodynamic centre etc.
This entailed introducing artificial stability terms that would have been difficult in a purely analogue system such as the basic Concorde controls, so they decided to go digital.
Sidestick or conventional control column doesn't come into that of course - see Boeing vs AI FBW systems; but no doubt the French government saw the opportunity to get two for the price of one .....

CliveL

OK.. It is a FBW aircraft, right? Did the flight control software include this artificial damping?

I would not expect that it has a ,,direct law'' type FBW where a control deflection gives a control surface deflection, right?

Thanks CliveL

Sorry all, this is now forty years back, literally....
So no, I don't have all the block diagrams and circuit diagrams in my head any more... I'll have to look through what I still have in the way of documentation.
i'll try to give some quick answers.

Da-20 monkey ,
Yes, Concorde had "artifcial damping", or "autostabilsation" as we called it, on all three axes (pitch, roll and yaw), even if it could be flown without it.
On the prototypes there were three separate computers (one per axis).
On the preprod and production aircraft the A/S function was 'compressed' into a single unit (I still have one).

CliveL has given the basic answer.
Don't confuse the Concorde FBW (which we referred to as "electrical signalling") with the current "Airbus" digital FBW.

One, rather than in previous-generation aircraft, the pilot no longer pushed and pulled on cables and rods to move the control surfaces. Instead, when he moved the controls, those movements were translated into electrical signals that were sent to the electro-hydraulic control surface actuators (even if in the Concorde days there still was a mechanical back-up).

Two, the entire system was "analogue". A concept difficult to explain in these days, where nearly everything is digital.....
Very briefly, you can convert 'physical' data, like control positions, or altitude, or pitch or roll rate, or Mach number, into 'analogue' electrical signals. You can then perform all kinds of 'computations' on those signals, like filtering them, or add or subtract them, or even multiply them, using electronic circuits based on 'operational amplifiers'.

In digital systems you go one step further.... you convert all those data into digital values, and use a digital computer to perform all your calculations, in accordance with the 'system software', then convert all the results back into physical data, such as control surface commands.

In analogue systems there is no "software". The entire system is defined by 'control laws ' (not the same thing at all as in the Airbus FBW aircraft) that are fixed in terms of 'transfer functions' of the various control loops.
Those in turn are determined by the values of the components in the various electronic circuits (resistors and capacitors mostly). So in those golden days.... we didn't re-write and re-program software.... we changed resistor and capacitors, and re-wired logic circuits.

I admit, you almost have to have been there to understand it....

I'm not sure whether it's worth starting an entire new thread on 'analogue computing' (maybe there's something on wikipedia, I haven't looked)....

CJ

A collection of links to videos, articles and much more on the rise and fall of analogue computing should prove interesting.

For those cutting their teeth in the computing field in the last twenty years, the analogue concept was already well and truly buried. As we all know, the Concorde project epitomized analogue computing which peaked in the decade between 1960 and 1970. One can only but wonder how much weight could have been saved and performance improved if current FBW digital computing was available at the time. That the aircraft type soldiered on well past its "use by date" is a credit to the designers and engineers who implemented the "state of the art" systems of the time.

Following on from what EXWOK was saying, you've got to remember that the spec was hammered out in the late '60s - so it's not surprising that the flight deck could *appear* antiquated by 1980's standards. But in this case, as alluded to, appearances are deceptive. While the gauges and switches are very definitely of that vintage, the systems behind them were very much bleeding-edge technology (by aviation standards) in a contemporary sense. Even when Concorde entered production, the most complex digital displays available to aviation were of the 7-segment LED type (as used in the Apollo Guidance Computer), and they were both wildly expensive and of limited use. The flight and engine controls were in fact a pioneering kind of analogue FBW - way in advance of any other type, even those making their debut at the start of the '80s (though FADEC was becoming more widespread by then - with the advent of the B757 and 767).

Ergonomically speaking, both engineers and pilots of the era write of Concorde's flight deck being the best possible balance of form and function available at the time - sure it looks cluttered to the modern eye, but everything was placed in a logical manner and the sheer number of systems used in the aircraft made the accessibility of all that information a basic requirement. It's worth bearing in mind that even those not particularly well-disposed to Airbus will grudgingly admit that the flight deck ergonomics on those types are extremely good - and a lot of the lessons learned were from cramming all that information into Concorde's limited space.

As for the cabin, again appearances are deceptive - I have sat in one of those seats and they are extremely comfortable for the size. Also one must bear in mind that unlike the subsonic Atlantic crossings, these were happening in about 3 hours - so no need to be particularly wide or convert into a bed like we see in Business and First today - not to mention less chance of a queue for the WC!

I have to thank EXWOK for explaining the windows - but I'll add the more prosaic reason that you don't need a particularly large window to see the curvature of the Earth in all its splendour - which is for the most part all you'd be seeing during the flight!

[EDIT : I should also confirm that EXWOK is also correct in stating that BA had Concorde turning a profit from the early-'80s onward, and it took a combination of a financial downturn and the fallout from the terrorist attacks of September 2001 to end the service.

While Concorde herself never recouped the development money granted by the governments of the UK and France, the infrastructure and R&D her development put in place paved the way for the Airbus project which, as we know, ended up becoming a leading player in airliner design and manufacture in the West. ]

Hardware-wise, maybe. In most other aspects, absolutely not - otherwise the transition from analogue to digital would not have happened.

"Triplication"? I'm unsure as to what you're referring to. If you're referring to the two disparate software implementations used in the Airbus FBW systems of the A320 and her descendants, then there were only two - not three - distinct implementations, and they were not so much a necessity as a "belt-and-braces" failsafe, given that the A320 was the first implementation of its type.

Software likewise, as AirborneAgain alludes to.

Not necessarily - see AirborneAgain's post.

But in a software-based system, the logical functions can be replaced simply by replacing a ROM IC or by re-writing to an EPROM IC - a much less problematic process than re-jigging discrete hardware across hundreds of airframes.

Airbus/Boeing FBW systems use hardened versions of obsolete commodity hardware - the suppliers won't stop making them as long as there's a demand.

Airbus/Boeing FBW systems use hardened versions of obsolete commodity hardware - the suppliers won't stop making them as long as there's a demand.

IC Part obsolescence is actually a big problem in aviation - our market is too small to justify keeping these components in production when they are decades obsolete in consumer products.
The best option is 'life time buys' - where the vendor stockpiles what they hope is a life time supply of the critical components (IC chips, ASICS, basically any logic devices). Of course, life time buys are not foolproof - not only is it dependent on accurate forecasts of need, but other things go wrong - crates go missing, warehouses burn down, bean counters dispose of what they think is excess inventory, etc.
The second option is to periodically certify hardware packages where they update components and re-certify. This is difficult and expensive - it takes extensive testing and analysis - even subtle changes in things like throughput timing can turn a digital system on it's ear. But it is done (in the last 10 years or so we've had FADEC parts obsolescence updates on both the PW4000/94" and CF6-80C2 FADEC controls - which date back to the mid 1980s).
The third option is just do a clean sheet of paper new device - really expensive and difficult, and often means having to update the associated s/w as well (this is what Pratt did with the PW2000 FADEC, coming out with completely new control - hardware and s/w - around 1995 to replace the original that dated back to about 1980).


But in the end the airplanes keep flying

To add to this, there's also the ability to create a hardware abstraction layer (similar to virtualisation) which allows you to swap out the underlying components whilst maintaining a uniform look and feel to the systems they support. I would expect (if it's not already being done) this to be one of the drivers for newer aircraft, as they can then forklift the proven systems from one aircraft type to the next, and provide redundancy/resilience and obsolescence-proofing.