Fast jet FBW systems

[Avionker]

After 521 views, and no responses, of a similar thread in the Questions forum, I thought I would try here. I hope that perhaps someone who missed the original thread may be able to answer my question.

Having never worked on fast jets when in the RAF, I am curious as to whether the FBW systems fitted to Typhoon, F-16, F-18, Rafale, Gripen etc incorporate any kind of artificial feel system?

Having followed some "spirited" debates on PPRuNe, as regards the differing approaches to FBW adopted by Airbus and Boeing, it certainly seems that many people have very strong opinions as to which philosophy is perferable.

I would be particularly interested in hearing the views of anyone who has flown both FBW fast jets and Airbus or 777 (or ideally all 3).

[Easy Street]

Tornado (yes it is FBW!) has a hydraulic-powered pitch feel system which takes inputs from the air data sensors to modify stick force with airspeed and altitude. Roll and yaw feel are provided by simple spring arrangements.

Generally speaking the stick forces are large, especially at the large deflections required for low speed air combat. If you turn the spin prevention system off, the stick force lightens to feel more like that of a Hawk. However, not a recommended technique :eek:

[Avionker]

Thanks for the replies.

I wasn't sure if the Tornado was FBW or not Easy Street, hence I didn't include it in the list.

BGG, I flew in the Nimrod once. I went along to watch the wet guys in action, when I on was on the Nimrod Software Team. As I was on Aco Flt I wanted to see how the AQS was actually used.

Due to some outstanding planning there was no wet team on-board and so began several hours of GH and a few hours of radar homing to surface targets... joy. The DCS was good though.....

@ shy talk

Not Airbus FBW aircraft, which is what prompted the question.

[Courtney Mil]

There are different approaches to FBW and there is a variety of methods in use in types you've listed. Tornado, yes, was FBW, but also had to do more than just react to stick deflection to make the aircraft feel like it was flying properly - the "rate demand system", which increased control surface deflection according to the rate at which the stick was moved. Control stick loading in pitch was varied only according to a/c speed, altitude and stick deflection - plus the SPILS bit mentioned earlier. I say "was" there, because the GR4 may have some differences and I can only talk for the F3.

The F-teen series varied a lot too. F-15 was dynamically stable, but has massive control authority, especially at low speed. As it has a mechanical system as well as FBW, the feel is fairly conventional.

F-16 originally had no feel at all, IIRC. The side stick was solid, worked by force transducers alone. Pilots found that very unnatural so they modified it to produce some movement and to feel some feedback. The stick forces don't vary, which is just as well because the F-16 FCS actually reduces control authority at lower speeds to keep the nose-gunner inside the stability envelope. It would feel very strange.

Typhoon is entirely different to, say the F-15. Being dynamically unstable, the FCS has to work very hard just to stop it departing from straight and level flight because of the canards. Stick feel was one of the aspects we looked long and hard during development, but in that case it may not have much resemblance with what the control surfaces are actually doing, just what the pilot thinks they should be doing.

What I would say is that a lot of a/c, even non-FBW, have some sort of artificial feel. Pure FBW ALMOST always needs some degree of it.

[Fox3WheresMyBanana]

Perhaps worth noting that early FBW, on the F3 at least, the artificial feel gave no indication of any of the limits. Most fast jets also have very little buffet margin, so a frequent scan of the gauges was necessary. SPILS would prevent the pilot over-alpha-ing the jet, but wouldn't stop him/her exceeding the g limits.

[zero1]

All the Tornado marks also had mechanical back up for the primary flight controls (tailerons) in case CSAS failed. And as mentioned before the aircraft was fitted with feedback systems to provide the pilot with some level of control loading vs. speed.

What might be an interesting question to ask is the affects that cause PIO conditions as one or two Tornado's were lost in the early days, plus a few pilots also experience a few OMG moments....

[John Farley]

shy talk

Not all feel systems use Q feel. Q feel means the feel offered is proportional to IAS squared.

[turbroprop]

Thanks John nice little fact. Worked with Q feel systems for many years and never knew Q related to anything in particular. Will now stop calling simple spring systems Q feel.

[Fox3WheresMyBanana]

Some PIO details here (see entry for 1984)
TORNADO (http://www.ejection-history.org.uk/aircraft_by_type/tornado.htm)

The Tornado F3 could, depending on the combination of CSAS mode failures, end up handling like at least 57 different aircraft. Almost none of these were a problem if you just headed home and "flew like a gentleman". Obviously the GR1's had way more stores combinations.

PIO is also possible with the throttles. I found my first AAR in full weapons fit quite tricky with both engines in burner (needed as the tanker was very high (for an F3)). Got home and found the Aircrew Manual said "Yes, it is, don't use more than one burner". Oops!

[BEagle]

Back in the days of the Gnat, we had to learn rather a complicated pitch control system - with various gizmos which these days would be clever little chips in black boxes....

There was a 'cam' which was known as 'Cam K' whose function was to provide a non-linear output from the control column to the pitch motor - the infamous 'Hobson motor' which moved the slab tailplane. This ensured that small control column deflexions causes less sensitivity than at larger deflexion.

But we needed an artificial feel system, so a spring in a tube gave some force vs. deflexion feel. However, there was also a need to compensate for high speed, to give the same stick force per 'g' throughout the speed range as far as could be achieved - so a plunger moved in response to dynamic pressure, varying the spring feel mechanical advantage. This was known as 'Q-feel'. Pitch trim was achieved by 'feel trim' switches on top of the control column which moved the spring system to a 'load free' point. But the neutral position varied with speed, so it was important to have a 'feel trim' position indicator.

Due to the way the undercarriage lowered, a piece of bike chain connected the main landing gear to the Hobson via a system known as 'datum shift' - which automatically compensated for the aft CG movement when the gear came down by tweaking the tail plane incidence up by about 3º.

If the feel trim packed up, you could unlock small elevators at the rear of the tailplane which compensated to some extent. But there was an associated speed limit to prevent the onset of flutter. You pulled a small lever, checked for 2 clicks, a white mark on the lever and the 'ELEV' caption.

But worse was to come. The Hobson motor was driven by hydraulics, so if you had a HYD failure it could only be moved up from the failure point by electrical power using 2 other 'standby trim' switches. If you didn't get the tailplane incidence to the right value quickly enough, you wouldn't be able to land the thing. To help, you used the feel trim to position the system at the 'safe/ideal' setting on the feel trim position indicator before exhausting the tailplane accumulator pressure.

The first part of the 'STUPRECCC' hydraulic failure drill was to reduce speed (I think it was 400/M0.85?), trim to the safe/ideal setting and unlock the elevators. Then turn the hydraulic power cock off and raise the standby trim switch guard. 'Speed, Trim and Unlock' were well known to all Gnat mates and weren't just a firm of dodgy solicitors! You could at least fly the thing then, but next came a need to exhaust the tailplane and aileron accumulators, making VERY sure that the tailplane incidence was at the correct setting (I think it was 1½º-2½º gear up?). Then you checked control response to the control column in pitch and roll and that the standby trim was working, illuminating the TRIM caption. Remember the elevators? Well, they then became the pitch trimming system in manual, with primary pitch control achieved with the little standby trim switches at the back of the left cockpit side panel. To make things slightly less awkward, standby trim control could then be transferred to the now redundant feel trim switches on top of the control column by selecting yet another set of switches, known as the 'Mod. 399' switches just next to the standby trim switches.....

You then flew carefully home keeping the control column 'load-free and central' with a combination of standby trim and elevators. But the tailplane didn't move very quickly in response to standby trim, so to make sure you had sufficient control to flare during the subsequent flapless landing, after lowering the gear on the emergency system you pre-set the tailplane incidence using the TPI gauge, allowing for the fact that the datum shift didn't work in manual either, then checked that you had a push force - which you subsequently relaxed into a slight back pressure on landing. Whilst groping for the tail brake chute lever.

Complicated? Ooh yes. But we spent so much time STUPRECCC-ing that it became second nature. Simulated engine fires were even more fun, for you had to do the fire drill, then the hyd failure drill and then the AC/DC failure drill which also involved turning a static pressure selector cock to turn the cabin altimeter into an actual altimeter as well as waiting about 30 sec for the emergency UHF radio to become available!

Yes, we had artificial feel and a lot of other systems on the little beast. But when it all worked it was an utter joy to fly, knifing along at low level like a three dimensional sports car. And if a bumbling idiot like me could be let loose on the Gnat, then anybody could!

[Eminence Gris]

Q feel means the feel offered is proportional to IAS squared.John/shy talk,

:= Strictly speaking Q is proportional to EAS squared. Of course at sea level EAS=IAS, but not at altitude.

EG http://images.ibsrv.net/ibsrv/res/src:www.pprune.org/get/images/smilies/nerd.gif

[BEagle]

Indeed, Q=½ρV² :hmm:

[blimey]

Of course at sea level EAS=IAS

Once calibrated and rectified.

[Courtney Mil]

BEagle,

A great post, thank you. I thoroughly enjoyed the read. I arrived at Valley after the Gnat had left, but remember seeing the Q-feel K-cam display in ground school and hearing the stories. Now, only 34 years later, I find out what they were all talking about!

Courtney:ok:

[BEagle]

No problemo, Courtney!

When I did a Hawk refresher at Valley, I don't recall being given anything like as much gen about the aircraft as we'd had to know about the Gnat.

The old Gnat longitudinal control system training rig was indeed put up on the wall in the Hawk groundschool - although I was a bit miffed that some arty-farty type had twisted it into a form of sculpture.

[Fox3WheresMyBanana]

How much do aircrew actually need to know?
This is very tricky. There was a guide to the Tornado flight control system on the F3 OCU, as the Aircrew Manual wasn't exactly easy to digest. The Aircrew Manual had, I remember, a fold-out A3 page of a wiring diagram that looked like 400 spiders on Acid had drawn it, titled 'Roll Channel (simplified)'. Then we found we needed a 'guide to the guide' because the guide still wasn't understood.
I thought "Sod this" and rewrote the whole set of lectures/guides, getting it down to about half what it was.
I think there comes a point where understanding the whole system is neither possible (given the other 4 million things aircrew need to know) nor useful. Maybe that point for control systems was the Gnat.

[Courtney Mil]

I don't recall being given anything like as much gen about the aircraft as we'd had to know about the Gnat.

That would be, as I've always said, because the Hawk only has two moving parts: the throttle and the pilot's mouth. Splendid aeroplane.


Fox3,

Yes, I still have my copies of the OCU FCS guide. The whole thing on two sheets of paper, IIRC.

[212man]

Complicated? Ooh yes. But we spent so much time STUPRECCC-ing that it became second nature

It must have been a 'barrel of laughs' determining that procedure during the development test flying phase!

[BEagle]

I always thought that someone at Folland knew he was about to be fired, so dreamed up the most complicated system he could think of late one Friday afternoon before leaving the office for the final time!

Often these systems are designed without catering for the idiocy of certain senior officers. One of whom found that if he moved the Gnat control column fore and aft very quickly, the TPI could ge out of phase with the stick position. Why anyone would wish to do something so daft is another question, but it was subsequently ordained that 'scissor restrictors' must be added to the Hobson to prevent this situation being reached.....:(

On the Hawk, the groundcrew wouldn't lower the canopy hood completely. They would lower it almost completely, then they'd turn round and bend over. This quaint ritual was the result of some tart of a senior officer stowing his SD cap in the back, which then fired the MDC when the hood was closed, giving the groundcrew a nasty shock.

The Hawk is a nice jet, which has decent range for a trainer and is much easier to fly than the Gnat - which was a trciky little beast even to taxy due to the brake response valves! But it lacks the pure swept wing feel of the Gnat and doesn't have the students' friend, the wonderful offset TACAN box. Neither does it have the superb throttle response of the Orpheus or the almost 'fly by thought' control response. However, it is a far safer aircraft generally and much more reliable than the Gnat ever was.

[FCSoverride]

For the system I am familiar with (on a pointy, grey aeroplane), there is no artificial feel. Basically, the more you pull or push, the harder it gets to do so. If I recall, the relationship between stick force and displacement is linear. I think there is a damping force too. There is a breakout force to get away from centre stick.

The FCS is designed such that centre stick means no manoeuvre is commanded, i.e. the aeroplane flies a gravity corrected load factor that keeps the flight path constant. The pitch stick extremes command the positive or negative alpha or load factor limits for the particular mass, flight condition and configuration. There is a variable "dead band" at the pitch stick extreme to keep the stick force per g constant for the variable g limits. For roll stick, centre commands no roll rate. Full roll stick commands the roll rate authority for the particular flight condition, config and mass. Roll rate vs stick force/displacement is a nonlinear function.

One other thing to note is that there are alpha limits for the "auto trim". So, if a pilot was to fly straight and level and let the aeroplane decelerate, the flight path would remain constant (the aeroplane would increase alpha to ensure this). Once the auto trim alpha limit is reached, the aeroplane sort of behaves like it has conventional static stability and will nose over at the trimmed alpha limit. The pilot has to pull to get more alpha.

Designed to be as simple and easy to fly as possible.

[Avionker]

For the system I am familiar with (on a pointy, grey aeroplane), there is no artificial feel. Basically, the more you pull or push, the harder it gets to do so. If I recall, the relationship between stick force and displacement is linear. I think there is a damping force too. There is a breakout force to get away from centre stick.

Would that not be regarded as artificial feel then, if it feels as though the stick is actually connected to something? And surely the increasing force required to move the stick further from the centre point is simulating aerodynamic forces to some extent?

[FCSoverride]

There is no feedback to the stick. It is just a spring and damper. Is it artificial feel? I leave that to be answered by pilots.

[John Farley]

You are right of course but I was trying to use an acronym that many people are familiar with rather than one that could easily turn them off so that they learned nowt.

I guess to keep he all knowing happy I should have said:

"Q feel means the feel offered is related to IAS squared."

I must do better next time.

[Dan Winterland]

I (think) knew what you meant John.

But how does it feel in a hover when the IAS/EAS is zero?

[John Farley]

Dan

Dunno...bit like a funny dream without the mess I guess.

[Obi Offiah]

One other thing to note is that there are alpha limits for the "auto trim". So, if a pilot was to fly straight and level and let the aeroplane decelerate, the flight path would remain constant (the aeroplane would increase alpha to ensure this). Once the auto trim alpha limit is reached, the aeroplane sort of behaves like it has conventional static stability and will nose over at the trimmed alpha limit. The pilot has to pull to get more alpha.

This is why the high alpha passes witnessed at airshows, are flown with the pilot applying back pressure to the stick.