Airbus Flight Controls Question:
 

Looking for some insight into A320 Flight Control Operation specifically during Normal Law, Autopilot Off, gusty wind conditions say on Final.

In "Understanding Air France 447" Bill Palmer writes "One common instance is during approach with Turbulence. Most pilots of conventionally controlled airplanes are used to instantly responding with a wind gust that tips a wing with a corresponding lateral control input. With the side stick in neutral, an Airbus will attempt to maintain a zero roll rate on its own, and will automatically input a roll command in response to the gust. Pilots who react to each bump and gust end up creating their own turbulence by wagging the sidestick back and forth faster than the airplane can respond."

Then I found something online on Airbusdriver.net. (can't post a URL here until I have 8 posts and this is my first I think).

One of the commenters says that the Airbus will look at the last bank angle and try to return to that point.

So this is really digging into the weeds and going down the rabbit hole, and I've noticed myself not thinking about any of this on a gusty final. But it still has me wondering when I'm not in the plane.

So your on final, Normal Law, AP off. Wing tips in a gust. Does the FBW flight control system stop the gust induced bank(because you didn't command it)? And does it actually attempt to return to its previous bank? Now you don't wait for all of this. You correct with sidestick. Are you augmenting the flight controls? Or does the plane sense your input and completely give you back full control? In other words...the plane with full sidestick is a 15 degree/sec roll rate. You correct with half sidestick, so what would be the roll rate command if the system is also responding to a gust?

The same thing could be asked with vertical forces. You hit a downdraft on final...see yourself going below glideslope. Since you didn't command a less that one G force, (since you are in Normal Law-Load Factor demand) with a neutral sidestick, the Flight Controls would seem to attempt to maintain one G flight by pitching up. Again, you see your glideslope going up and you react by pitching up. Are you augmenting the flight controls? Or does the system directly give you what your commanding on the Sidestick?

Thanks!

AB FBW side stick out of neutral in roll commands a rate of roll not bank which is indirect result to fulfill the commanded rate of roll. Similarly in pitch stick in neutral commands 1g. FCC even without pilot input will use necessary flight controls to achieve that. So any pilot input on stick during that stage will augment it leading to pilot induced oscillations. In turbulence corrective pilot inputs should only be made after giving the system a chance to do what it's supposed to do. On approach etc what happens in turbulence is aircraft may automatically stabilize in 1g or zero bank etc but gets displaced from desired course or glide path which needs additional manoeuvring to get it on it's previous trajectory. In manual flight it won't happen automatically.

Airbus used to call this “global corrections”. So on approach, you don’t respond to every lump and bump as you would a conventional aircraft, but you must make “global corrections” to return the aircraft to the correct trajectory.

Just to add to the previous two answers to the OP question:

Your side-stick inputs do not cancel the FBW system. The FBW looks at the ADIRS, your side-stick inputs, the current attitude and feedbacks from accelerometers etc to compute a flight control surface response to all those inputs. With no control inputs, the FBW will (try to) maintain the last commanded attitude from either you or the autopilot. (Within limits).

One way of thinking about Airbus FBW is to imagine that the autopilot disconnect switch is in a different part of the control chain.

A traditional chain with autopilot engaged is :

Navigation guidance - Autopilot - # - Control surfaces.

In a conventional jet like that; the autopilot disconnect switch, '#' is between the autopilot and the control surfaces, so when you disconnect, your inputs go directly to the control surfaces.

With Airbus FBW the chain has an extra element :

Navigation guidance - Autopilot - # - FBW - Control surfaces.

When you disconnect in an Airbus FBW, your inputs join at '#' and go into the FBW, not to the control surfaces*.

*This is a simplification, (and refers to Normal Law), but it should help to appreciate why you can get into PIOs in Airbus FBW. Especially if you have not been trained to use the side-stick properly, (and no TRE ever explained the side-stick properly to me - I had to work it out for myself).

Anyway, just be aware that with the autopilot disconnected, the FBW is still in the circuit and will still be working the control surfaces to maintain the last commanded attitude. So don't stir the stick, just move the side-stick until you have the attitude you want and as soon as it is there, return the side-stick to neutral:

Correct - neutral. Correct - neutral. Correct - neutral etc.

Your inputs are "added" to the FBW commands, so In strong turbulence you might need to hold full deflection for a moment, to 'pick up a wing' for example. But as soon as the wing starts picking up, return the side-stick to neutral.

Airbus FBW helps you and assists with the flying task :ok:

Have a look at this thread: https://www.pprune.org/tech-log/6561...conf-full.html

My advice, after 20+ years flying the A320, is don't over think it.

Thank you Max Angle. That's the advice I am giving myself. I'm totally doing that when not in the airplane. When on final and in the flare? No.

Thank you Sim25. I'll take a look.

Thanks Uplinker. You are right. No one ever explained this to me and the FCOM seems to have no insights into this.

I'm not a pilot but I have twice 'flown' the extremely realistic A320 simulator at Simulator Adventures in Salford (I have no commercial connection). Having only ever flown a Piper PA28 for real, and that was about 40 years ago, I found that I was over-controlling the Airbus when the AP was off - cue lots of pilot-induced oscillation. Second time in the simulator I used a much lighter touch on the sidestick, as per your 'correct-neutral, correct-neutral' comment above and found I was keeping the aircraft in much better shape.

Glad it made sense in practise :ok:

Basically, (and within certain system and atmospheric limits), the Airbus FBW will "stay where you put it".

So if you gently roll on 10° bank, for example, and then return the side-stick to neutral, the aircraft will stay at 10° bank, until you move the side-stick out of neutral again. You could even take your hand off the side-stick, and the aircraft would stay at the new attitude, (within certain limits).

If you are on a turbulent approach, and a gust makes the aircraft roll, the FBW will automatically correct that roll and return the aircraft to whatever attitude you had before the gust, (within reason). So the pilot has to bear in mind that if they put in a handful of side-stick as well, it might be too much. And if they hold the side-stick out of neutral too long, the aircraft will continue banking further and over-shoot. Then the pilot has to correct the other way, and might overshoot again = PIO.

As vilas says; You get used to allowing the FBW to react and correct first before wading in too deep, but this becomes second nature. This is why correct - neutral - repeat is a good technique. As the attitude approaches what you want, put the side-stick back to neutral.

I taught myself to use the side-stick after happening to see a film on TV taken in the cockpit of a Tornado fast jet as it was low level,flying. The Tornado has a FBW system, and the pilot said over the intercom to his navigator; "coming left 5°", and he just more or less bumped the control stick to the left then neutral twice; as quick as it takes to say "bump bump". The aircraft rolled left and turned, staying banked with no further inputs. Ah ah, I thought, I wonder if that technique will work on Airbus FBW side-stick - and it does.

Hello,

First of all, sorry to disappoint, but the normal law in pitch is not an Nz law. It's a C* law, where C*=Nz+Veq*q/g
Where Veq is a speed around 210-220 kt depending on airbus models, V is IAS, q is pitch rate (it's the letter usually used in France to describe an airplane pitch rate), Nz is indeed the load factor.

At very high speed, you require little q to have a big Nz, your C* law is closer to an Nz law. At lower speeds, the q term is more noticeable.
Boeing has the same law on the 777 except it adds a speed stability term C* = Nz + Veq*q/g + k*(V-Vtrim)
By the way if anyone knows where to see this Vtrim speed (trim reference speed) in the cockpit, I'd be amazed.

Imagine you're asking for more lift. You're asking for C* = 0.2g. The airplane has q=0, so it translates at first with Nz = 0.2. It starts pitching up. You're having some positive q that will appear. This term will damp the Nz demand.
Conversely, if you're stable in level flight, and decide to decelerate. You're asking 0g (of variation), so when reducing speed, at first the airplane will have to have some positive q. Since you still want a null sum, the Nz will become slightly negative. It will damp the q, the q will stabilize at 0 again, but with a very shallow descent.

Knowing more than this would require to have in depth knowledge of the design office.
However I'm not inventing anything as you can see here :
https://www.engineeringpilot.com/pos...n-unknown-star

For the roll axis it's a completely different story. The flight control law will have you control a roll rate. If your stick is neutral, you're asking for 0 roll rate. So the aircraft will damp any roll rate. But the resulting roll won't be corrected.
A technique that I found useful when flying the bus was to consider doing a half correction of the gusts.
- When you correct for a gust, on a conventional airplane :
You counter the roll rate (p = roll rate), by applying some ailerons (p' = roll rate rate) for a certain time. Let say dl times dt
When you reach roll rate 0, you apply dl times dt again (or dl times dt twice as long) to get the airplane back with an opposing roll rate.
You get the wings level, and apply dl times dt again to erase the roll rate, at a zero roll angle.

- On an airbus, the airplane counters the gust, and the orders are added to yours.
The airplane counters the roll rate by applying dl times dt.
You just have to apply dl times dt (but only once, or for one single dt period, as opposed to two in the previous example). Then when the wings are level again you can fly your wings level as you know.

So, when counteracting a gust, telling myself I would only correct half as I would think I should worked very well to avoid overflying the plane.
You still have to counter the gust. Just counter it half as much as you would feel doing.

Final note : on the roll axis, it's not entirely true to say that an aileron input is strictly proportional to a roll rate rate. It is, at first, that is to say, this approximation works very well for the initial phase of controlling gust.
However, if you perform large bank variations, with an established roll rate, you need some ailerons to counter the difference in lift that occur on the wings due to rotation. The rising wing has a lower angle of attack. The lowering wing has a higher angle of attack. It results in a roll moment that is opposed to the roll (roll damping, or Clp coefficient ...). So you need some aileron applied in a steady state in order to get a roll motion.
However, it's also true that at first the aileron inputs are proportional to a roll rate rate.

That's a lot of equations ! They are making my brain hurt :{

Just simplify as follows :
Once the gust has made the airplane depart, arresting the roll rate is half the job. Reverting the roll rate is the other half of the job.
Airbus does the first half for you. The second half is up to you.

As for the C* law, if you ever notice your airbus not holding exactly exactly level flight while you're decelerating or accelerating, just know it's expected.
Regarding the OP question about longitudinal gusts (having an impact on lift in a symmetrical way), I never noticed the same thing as with roll. 'Theoretically' according to airbus lie about the pitch law being a load factor law, yes, but knowing the real equation, no.
At approach speed, the law equates maybe to a law that's a quarter load factor and three quarters pitch rate. I never noticed the quarter load factor part.
Maybe airbus test pilots do, when they compare with other laws, but we don't really get to do that.

Excellent understanding and a very good explanation, CVividasku . Thanks

That's not correct. Roll law is a roll angle target law. Pilot's order is integrated... If disturbance causes the plane to leave the target it will go back, without pilot intervention.

Not shown anywhere as far as I know. Have had a dig through maintenance data in flight and not seen it.

Only place it’s shown is on the IOP in the simulator. I’d quite like it to be shown on the PFD, as I’m never entirely sure what speed the aircraft is trimmed to when disconnecting the autopilot in gusty conditions. Speed at moment of disconnect? MCP target? Vref + 5? Some sort of long term blended average? All four?

I manage flying without thinking about any computer equations.

I know the FBW, (and auto-thrust* and GS Mini), are there. I know what they do and how they help me, and I know how to operate the side-stick without creating PIOs. Then I just fly the 'plane by feel, observation and experience, and it all works very well.


*older A330 auto-thrust can be 'lazy' near the ground in certain conditions but you simply deal with it.

Regarding the c* law:
https://cimg9.ibsrv.net/gimg/pprune....312787b622.jpg
Source: Flugregelung - Brockhaus, Alles, Luckner
https://cimg1.ibsrv.net/gimg/pprune....56185359ff.jpg
Source: Flugregelung - Brockhaus, Alles, Luckner

The above shown structure is of course simplified, but this is pretty much it...

So you’re saying if in a 5 degree bank, a gust pushes you to 15 degrees, the FBW will return you to 5 degrees?