It has been fifteen years since Enstrom developed its light turbine helicopter. In that time it has proven to have an excellent safety and maintenance record and is thus a powerful competitor in the now keenly-contested market for a light turbine

This year, after much anticipation, Robinson’s R66 five-place turbine finally arrived in South Africa. The noise of its arrival overshadowed the fact that Enstrom has, in its model 480B, had a turbine five-place competitor with a loyal following who understand the very persuasive arguments in favour of the Enstrom.

Enstrom helicopters have an almost unassailable reputation for simplicity, strength, stability and ruggedness – all the characteristics that a real helicopter should have.

These qualities become evident when one looks at two key aspects of the long-term operation of Enstroms. Insurance and maintenance costs that exceed fuel burn costs are a huge problem for the operators of light helicopters.  It is remarkable that these two aspects of Enstrom operation are far lower than its competitors. Thus, they have a remarkable safety record and the insurance rates, especially the international rates, reflect this. Secondly, they are built to military standards of strength and quality over lightness and price and so have almost no life-limited parts.

 

DEVELOPMENT

Rudolf Enstrom delivered his first helicopter, the F-28A piston-powered two/three-seater, in 1966. The company changed hands a number of times, and then a new period of growth began in 1984 when famed inventor Dean Kamen, (the creator of the Segway and many other inventions), joined Enstrom as an investor and technical adviser. Kamen was instrumental in helping to improve the company’s products and in introducing the 480, originally conceived as the TH28 (as the military version), to compete for a US Army New Training Helicopter (NTH) requirement.

The NTH contestants had to meet the Army’s exacting quality and performance requirements. Amongst other things, the NTH helicopter needed to be turbine-powered, carry two student pilots and an instructor – and be cost effective. The TH-28, as Enstrom named it, was certainly cost effective - far more so than either Bell’s JetRanger variant or Eurocopter’s AS-350 offering. Although the Enstrom lost to the Bell 206, Army pilots favoured its flying qualities.

Even though it lost the hard-fought NTH contract the competition matured the company, and from the TH-28 Enstrom launched the model 480 in 1994, using the Allison 250-C20W turbine. This engine differs very slightly to the 250-C20J version fitted to Bell’s standard 206B-3 in that the exhaust ports are inverted to allow gasses to escape through the underside of the Enstrom’s fuselage. This does however mean that the helicopter should be operated with care in long dry grass.

ON THE GROUND

The basic Enstrom piston helicopter lineage is clear, but the 480B is much larger than its piston powered forebears. The 480B has benefited from a great deal of work to accommodate the greater power of a turbine. Noticeable at first glance are larger horizontal stabiliser end-plates, a rotor-mast fairing-cum-particle separator housing (standard equipment) and a wider and altogether roomier fuselage. Not so noticeable is a larger diameter tail rotor, heftier undercarriage, beefed-up main transmission with more webbing to stiffen the main casing, a large cooling fan behind the main transmission, and the replacement of Enstrom’s Lamiflex bearings for more conventional TT straps to counter main rotor pitch/torsion stress.

The baggage hold in the tail cone accommodates loads up to 150 pounds. The wide cabin doors facilitate graceful egress. While the high skids allow all manner of FLIRs, NightSuns, loudspeakers, and other law enforcement paraphernalia to be slung below the belly, access to the cockpit is helped by a large step.

From the fit and finish of the airframe and interior it is evident that this is a quality helicopter. The subject of this review, ZS-HDO, has a fairly typical equipment list, based around a Garmin 530 that provides an IFR capability, even though the helicopter is not rated for IFR work. But this is essential for an advanced helicopter that may be used for training.

The engine compartment has easy access from large doors on both sides. The -C20W engine has been fitted along the same plane as the piston engine and drives the main transmission via a large multi-polymer belt. Although traditional turbine pilots may decry a belt-driven set-up, it’s worth remembering that no failures have been recorded amongst the nearly 2000 Enstroms manufactured so far. The 480’s belt is in fact slightly narrower than the piston-engined example and has the same 3 500 hour life as the engine TBO. The engine does however require a 1750 hour hot section inspection as required on all helicopters fitted with this engine (Bell 206, BO 105, MD 500 and others).

The 480B’s Rolls-Royce Allison engine is capable of producing 420 shp, but Enstrom has chosen to use only a fraction of that power (305 shp for five minutes and 277 shp continuously), which gives the pilot the luxury of having plenty of power in hand and of retaining that power, providing a very capable hot and high performance. The engine is flat rated to 13,000 feet, giving an excellent Out of Ground Effect hover.

Enstrom’s main rotor gearbox is a robust affair, and although the de-rating brings the power parameters closer to the piston engine’s 225-hp, Enstrom has added further webbing to the casing. The main rotor head is mounted on a 9-ton capable 4130 steel tube mast.

The Enstrom mast and rotor system is beautiful in its simplicity and yet strength and functionality. Noteworthy too is the absence of a wire cutter blade to protect the mast. Most helicopters have the cyclic control rods exposed outside the mast, and these are vulnerable to power lines. But wire cutters on Enstroms are redundant because the cyclic controls are safely protected inside the mast.

The three-bladed fully articulated system provides excellent stability and allows the pilot to have some fun without worrying about the aerodynamic problems of a two-bladed system. Aerodynamically the 480B can be looped and rolled, and this has been demonstrated by test pilots.

Enstroms have always been easily trailered, a facility many bush operators like. The main blades take less than 30 minutes to bolt to the main hub, provided there’s muscle-power to lift them.

The key to the Enstrom handling, performance and durability is its three-blade rotor. Like all Enstroms, the 480’s blades are not life-limited. The 480’s blades are identical to the piston-engined version which are notionally life-limited to 96 000 hours.

The two fuel tanks are fixed each side of the main rotor gearbox and contain 45 US gallons each - more than double the capacity of the piston 280FX, and so provide 4.5 hour endurance.

Older Enstroms had a utilitarian quality to their interiors. But the market has moved on and the new 480B is much improved with quality leather upholstery and cabin fittings. The two cabin doors are made from composite material with plexiglass windows and two sliding panels for excellent ventilation. Air conditioning is an option. The instrument panel is finished in a professional-looking light grey with a pedestal containing the avionics sloping down to the floor between the front seat occupants.

Headroom is plentiful and Enstrom offer a casevac facility that enables the helicopter to carry a single stretcher and medical attendant. The Enstrom has a unique seating arrangement that uses great flexibility to fit in up to five seats. The key design need was the military’s requirement to accommodate a third crew member and so three people can fit on a bench-type seat at the rear-right of the cabin and there is a staggered forward front passenger seat for another passenger or instructor.

Normal training operations configure the seating for three by folding part of the rear seat squab against the rear bulk head, allowing the front seats to be placed in parallel. This gives the cockpit a unique feeling of spaciousness. The front seats have a good adjustment range and there is also a fine adjustment facility on the pedals.

 

IN THE AIR

The engine start switch is mounted at the end of the collective. Watching the Turbine Outlet Temperature (TOT), the throttle is opened to the idle stop after the N1 rpm passes 12-15%. The starter is released at 58% N1 as the Rolls Allison becomes self-sustaining.

Just like its smaller piston-engined brother, the 480B’s three heavy rotor blades give the helicopter a brief rocking motion as the turbine winds up. However the skid oleos damp this out as the throttle is fully opened and stabilised at just over 90°c N1. A blipper switch brings the engine up to its recommended 103% for takeoff.

There is no hydraulic system in the 480B, which keeps things simple but also requires a trim system to absorb feedback from the rotor and to reposition the stick datum as required by the pilot. If the trim system fails, the forces required by the pilot to overcome them are about 15 pounds — not excessive, but the pilot would want to land fairly quickly.

Unlike a Robinson or JetRanger, Enstrom’s cyclic should be gripped rather than held lightly between two fingers. This helps impart the feeling of stability, solidness and clean handling on a par with the McDonnell Douglas MD500. A coolie hat trim switch at the top of the cyclic is used to trim the main blades to neutral and once the pre-takeoff checks are completed we’re ready to lift-off.

For this flight test I flew as part of a conversion to type for Henley Air’s Andre Coetzee. Well-known South African helicopter Designated Examiner Des Dumbleton was the instructor and he occupied the right front seat as the Enstrom is normally flown from the left seat.

Enstrom has done an excellent job designing the 480B’s control system so that hydraulics are not required, which means pilots never have to worry about a failure, or inadvertently switching the hydraulics off. The trim system is fast and easy to use, allowing for hands-off flight in smooth air when required to grab a map or adjust the radios. The collective is also nicely balanced to allow for hands-off flight in the cruise when needed. Both the cyclic and collective have a smooth and solid feel that makes the 480B a blast to fly and exceptionally responsive. André took his hands off the controls to sample its hands-free handling and the Enstrom remained commendably straight and level. Then Des slapped the cyclic hard a number of times from both sides and again there was minimal disruption of the helicopter’s steady progress through the sky, demonstrating the 480B’s inherent stability.

The turbine is fully governed. Once engine RPM is set in the flight idle, there is no further requirement for throttle input by the pilot. In addition the 480B incorporates ‘a power anticipator’ on the collective which provides immediate power when the collective is raised. Many turbine helicopters have a noticeable power lag, which can be disconcerting and sometimes dangerous on short final when power is required immediately. The 480B anticipates turbine lag by measuring how fast the power is being adjusted and so helps the helicopter provide an instant response. Interestingly, the same collective response is also featured in the two smaller Enstrom piston models.

It was a particularly warm summer’s day with an 18 knot northerly wind blowing across Rand Airport. We lifted off easily and headed towards the south west for a brief evaluation. Visibility is excellent, cabin sound levels are very good, even with my headset off, and the cabin really does have a wonderful feeling of space - it takes an outstretched arm to reach across the middle seat to the other occupant. With two on board and 300 lbs of fuel, we set up a climb of 80 knots at an unfamiliar (for an Enstrom) 1000 fpm.

The 480B is little different to a piston Enstrom - except it is more powerful and willing. The helicopter requires constant trim changes, especially in turns, but it comes naturally after a few minutes. In the slightly turbulent air, the helicopter felt solid and manageable, riding easily through the bumps. At our weight and temperature Vne was just 97 knots indicated, due to the possibility of retreating blade stall. A large Vne reference table is mounted centrally above the instrument binnacle. The 480B’s cruising speed is reduced at higher density altitudes, as is its speed when a door is removed for observation or photography.

We flew a few circuits at Henley’s training runway at Kromvlei before returning to Rand for Andre Coetzee to sample the autorotation. Thanks to the high inertia main rotor blades autorotation has always been an Enstrom strength. After rolling the power off, the helicopter settled into a comfortable and smooth autorotation at 60-knots. For insurance purposes we recovered a good 50 feet before touchdown. 50-knots is the number for minimum rate of descent after a power loss while maximum glide distance is at 80-knots. Best rpm for an autorotation is 334 and Enstrom warns that rotor rpm is non-recoverable below 240. Des Dumbleton assured me that a full ground contact would be a routine training exercise, especially with the massive shock-absorbing qualities of the substantial skids.

Heading back to the landing pad, it was evident that Enstrom has taken great care in ensuring as smooth as possible a transition to the hover, and despite the need for constant trim changes, the lack of any artificial control damping is hardly noticed. There was little vibration and with the increased tail rotor-diameter, anti-torque control was more than adequate. Even under difficult quartering winds the 480 is not prone to Loss of Tail Rotor Effectiveness (LTE). Andre had no problem manoeuvring the 480 through full 360 degree turns for the camera. The Enstrom performed these handling tests effortlessly.

We elected not to do any run-on landings, but Enstrom test pilots love to show how the 480 can be run-on at as much as 25 knots, accompanied by full anti-torque pedal inputs before the aircraft has slowed down. The landing gear is wide and its shock-absorbing oleos go a long way to keeping it smooth.

Depending on equipment, the 480B has a useful load of nearly 1,200 pounds. Although it has five seats, it will rarely use this capacity. Not only would a full cabin be cramped for even the voluminous Enstrom, it would be uncomfortable. For charter or passenger carrying, operators either remove the front seat or place two people on the bulkhead position. (See diagram on next page.)

 

LONG TERM

Initial acquisition cost is over the long term a relatively small part of helicopter operation. It is noteworthy that the Enstrom’s $1.048m price tag puts it in a unique place in the market. It is significantly more expensive than the Robinson R66’s typical $860 000, but is far less than an MD500E at $1.8m. Unlike the R22, R44 and R66, the Enstrom models do not have life-limited frames, hence they do not require a mid-life rebuild. Enstrom also maintains its value far better than those helicopters with life-limited components. Thanks to its almost unlimited parts life and lower insurance costs, Enstrom claims that the 480B costs 30 percent less to operate than a Bell 206 or Robinson R66.

In South Africa at present there are two Enstrom-approved maintenance facilities – Nicholson Helicopters at Grand Central airport and Trio Aviation at Lanseria. Agent Safomar Aviation Aircraft Sales and Leasing Division (formerly known as Aerosales Africa) is the spare parts supplier for South Africa and holds more than enough inventory to support regional Enstrom fleet operations.

There are three flight schools currently training on Enstroms – Babcock Flying Academy at Grand Central, Henley Air at Rand Airport and Dave Gove at Petite Aerodrome.

Enstrom may have lost the contest for the NHT contract, but since then the company has gained an extensive training market in the various militaries, air forces and law enforcement agencies around the world. More than eighty 480Bs were sold in 2011.  In less frenetic civilian operations, the 480B will suit the operator wanting turbine reliability, but without the high acquisition cost.

Enstrom backs up the inherent quality of its product. All Enstrom helicopters come with a two-year or 1,000 hour warranty that covers everything from tip to tail.

The Enstrom Model 480B’s blend of forgiving flight qualities, turbine power and an almost six-foot-wide cabin helps make it one of the most surprising light helicopters on the market today. Pilots love it, for all of the reasons listed, as well as for its low operating costs and an exceptional safety record since its introduction in the early 1990s.

 

Briefing

Flight Control Automation The recovery of the flight data recorder and cockpit voice recorder from the Air France A330 Flight 447 provides much food for thought about Fly-by-Wire (FBW) and flight control automation. This article explains how the Airbus flight control computers confused the pilots

 

Introduction

Asking computers to manage a pilot’s flying is not new. The first flying machine to use fly-by-wire successfully was the Lunar Lander as steering it above the moon on top of a column of flame was considered too difficult even for Neil Armstrong.

The French then pioneered FBW for commercial aircraft by using it in the Concorde. The more conservative (and litigious) Americans first used FBW on fighters, as the commercial sector was less enthusiastic. 
While Boeing continued to rely on conventional control systems for its 757 and 767, Airbus went ahead and, building on its Concorde experience, introduced full FBW in its A320. It was only on the 777 that Boeing belatedly introduced a limited FBW system. Boeing claims pilots should have the ultimate say, meaning that the pilot can override the flight control computers and so Boeings have ‘soft limits’ on their flight computers. Some have argued that if the Airbus A330 of Air France Flight 447 which crashed into the Atlantic off Brazil had the Boeing 777 system the pilots would not have crashed.

It now appears that the FBW systems became too much for the pilots of Air France Flight 447 when it suffered pitot tube icing one dark and stormy night over the Atlantic.

In this briefing, SAA Captain Stefan Poprawa looks at the basic structure of the Airbus FBW systems. Specifically he looks at the flight control computer’s logic or laws and how they respond to failures in the information they receive. Captain Poprawa takes us through the various levels of flight control laws and explains how they affect the controllability of the aircraft. He also deals with the question of whether the resulting crash was avoidable.

Captain Poprawa is eminently qualified to tackle this complex subject as he is SAA’s Chief Technical pilot and Head of Fleet Management. A professional engineer, he holds three Masters degrees.

 

Using the information gleaned from the flight recorders, the French Civil Aviation Authority (the BEA) was finally able to conclusively establish the fateful chain of events that ended when the aircraft impacted the water in a deeply stalled condition (angle of attack around 35 degrees) at a descent rate in excess of 10,000 feet/minute (more than 100 knots vertically). The impact is comparable to driving a car at close to 200 km/h into a wall. At such speeds there is little difference between colliding with water or a solid surface.

The chain of events that led to the accident started while the aircraft was avoiding thunderstorms. Erroneous pitot probe readings led to an autopilot disconnect and that was followed shortly by the autothrust as a result of non-computing speed indications. The pilots were forced to control the aircraft manually.

The pilot’s subsequent handling of the aircraft initiated a very healthy and necessary debate on “automation addiction” and “automation dependency”, that quite possibly is long overdue.

So what are the implications when some of the automation is lost? Does this differ between conventional and fly-by-wire technology? Is the loss of automation so severe that the pilot is faced with a completely different aircraft with completely different handling characteristics?

Automation in this sense refers to the broader use of technology to assist in controlling an aircraft. It goes beyond the loss of autoflight systems (autopilot and autothrust). Loss of automation can also occur on systems that continuously function automatically behind the scenes, without any flight deck crew intervention. Examples of these are the systems that automatically assist moving the flight controls.

We focus here on the specific challenges pilots of large transport aircraft (conventional and fly-by-wire) would face under the circumstances faced by the Air France crew.

Prior to expanding on this, though, it needs to be highlighted that transport category aircraft are by design essentially stable platforms. Left alone they will ride turbulence without any major deviation from the intended trajectory. At worst there may be some slight dynamic instability which is easily controlled.

 

CONVENTIONAL FLIGHT CONTROLS

Modern conventional transport aircraft have highly advanced autoflight systems, capable of controlling the aircraft in all phases of flight after takeoff, up to and including automatic landings and automatic roll-out on the runway after landing, even when one of the engines has failed.

But when the autoflight system disengages, the flight controls on a conventional aircraft revert to a displacement response of the control surface (aileron, elevator or rudder) to pilot input, i.e. the flight control movement is directly proportional to the control wheel (or stick) movement.

As a result, when flying at 150 knots true airspeed a pilot would see, for example, a roll rate of five degrees per second in response to a certain control wheel input. Flying at 300 knots in the same configuration with the same control wheel input the roll rate now becomes 20 degrees per second, four times higher, even though the ailerons were displaced by an identical amount. These divergent aircraft responses are further aggravated with increasing height, as the gap between Indicated Airspeed on the displays and the True Airspeed of the aircraft increases.

If necessary spoilers assist with turn coordination to avoid the secondary effects of aileron deflection, so that the need for any rudder input by the pilot is not required, other than during asymmetric flight following an engine failure or during crosswind landings. Yaw damping is automatic.

To prevent damage from excessive control movements at high speed some of these conventional aircraft have speed-dependent limiters on the control surfaces (e.g. locking out of the outer ailerons, limiting the maximum rudder displacement), designed to limit the loads on the aircraft when the aircraft is flown manually at high speed, to prevent overstressing the aircraft structure.  A classic example of such damage is the Airbus A300 that crashed in New York after suffering the loss of its vertical stabiliser from excessive rudder inputs.

However, control limiters for speed generally are not found on pitch control. One would not want to limit pitch control for a number of reasons, stall recovery capability being one.

On light aircraft without hydraulic assistance the different forces needed to displace the control wheel at different speeds are noticeable. The forces needed to displace the flight controls on large aircraft, however, can be beyond human capability so that hydraulic actuators are needed. As these typically work on 3000 psi hydraulic pressure, any feel for different flight control loads from different speeds is largely eliminated, thus removing a natural tendency to initially apply less control input during higher loads / speeds.

To overcome this, a speed-dependent artificial feel is introduced to the control column, to simulate and re-introduce the effect of aerodynamic forces, typically in pitch only. At higher speeds the control column feel becomes heavier. The speed is measured by the pitot probes. So if the pitot probes provide incorrect readings the pilot would be faced with “unrepresentative” stick forces for the true speed and altitude of the aircraft.

If, for example, the speed indication erroneously drops to, say, 60 knots indicated from previously 280 knots as a result of incorrect data, the artificial feel would simulate control forces as for a slow flying aircraft. Couple this with synthetic stall warning activation and the flight crew is left with multiple pointers towards the aircraft flying slowly, when in fact it is not.

Further, in manual flight the aircraft has to be manually re-trimmed following any control column pitch input. The trim becomes increasingly sensitive with increasing true airspeed.

 

FLY-BY-WIRE FLIGHT CONTROLS

When the autoflight system is not engaged, the aircraft is always flown in “Normal Law”, unless some degradation occurs. From “Normal Law” there are two reversion laws, “Alternate Law” and “Direct Law”. For all three laws, the direct links between, in this case, the side stick (which replaces the control column) and the flight control surfaces are removed. Instead, numerous multiple-redundant flight control computers process pilot (and autopilot) inputs according to these Normal, Alternate or Direct flight control laws to move the flight controls through the hydraulic actuators.

There are two further control laws: “Abnormal Attitude Law” allows for recovery from unusual attitudes such as when the aircraft has rolled inverted. “Back-up Law” allows for manual control of the aircraft during complete electrical failure, through manual pitch trim and rudder control, until at least some electrical power is restored.

 

NORMAL LAW

In pitch, Normal Law provides a load factor (G load) demand proportional to the side stick deflection, independent of aircraft speed. The flight control computers do not directly require any speed information here. Rather, accelerometers provide the required data for pitch control and pitch trim is automatic. With the side stick at neutral and the wings level, the system will maintain 1 G in pitch corrected for pitch attitude.

(During the takeoff rotation and during the landing flare the pitch reverts to a direct stick-to-elevator relationship.)

The roll rate provided by the flight control computers is thus proportional to the side stick deflection. Independent of aircraft speed and altitude, the roll rate will always be the same for the same amount of side stick deflection, up to the maximum roll rate.

Normal law provides the following flight envelope protections:

• Load Factor Limitation to prevent structural overload

• High and Low pitch attitude protection

•High angle-of-attack protection against stall and wind shear events

• High speed protection

• Bank angle protection

• Low energy warning during approach

All these protections will disengage the autopilot or override any side stick inputs in order to keep the aircraft within the certified flight envelope.

 

ALTERNATE LAW

In flight, the Alternate Law pitch mode provides a load-factor demand much as in normal law, except that some of the protections are lost or reduced. The aircraft’s handling in pitch thus does not change. Automatic pitch trim remains active.

But, as happened with Air France 447, with a loss of reliable speed measurements, lateral (roll) control becomes a direct stick-to-control surface relationship. This then is much like a conventional aircraft except that in this case there are some gain adjustments, set automatically according to the slats/flaps configuration, which also serve to limit the maximum roll rate achievable.

The protection features are reduced depending on the cause of Alternate Law reversion. When there is a loss of reliable speed measurement all the protections are lost except that Load Factor Limitation to prevent structural overload remains active.

 

DIRECT LAW

For a fly-by-wire aircraft to revert to Direct Law requires multiple failures of the same type of equipment, depending on aircraft type, e.g. the complete loss of inertial referencing.

For pitch, Direct Law is a direct stick to elevator relationship, i.e. elevator deflection is proportional to stick deflection, as for a conventional aircraft. There is no automatic trim, i.e. the aircraft has to be trimmed manually. There is no artificial feel and the elevator response becomes very sensitive at high speeds.

Roll control is, as for the Alternate Law, also a direct stick to aileron relationship. All protections are then lost. This means a fly-by-wire aircraft has now effectively reverted to being a conventional aircraft with hydraulically actuated flight controls, except that there is neither artificial feel in pitch nor any gain control in roll.

 

SUMMARY

In both conventional and fly-by-wire aircraft the autoflight system will disconnect when the airspeed indication becomes unreliable, forcing the flight crew to manually fly the aircraft.

A comparison of the conventional and the fly-by-wire airliner handling during manual flight is best tabulated, specifically with reference to normal manual flight, and the subsequent effect of the loss of reliable airspeed information. The pitch and roll axes are listed separately.

With fly-by-wire, the aircraft flight control laws degrade to alternate law primarily because many of the flight envelope protections (high speed and low speed protection) can no longer be reliably provided.  All the warnings are still active as on a conventional aircraft.

Similarly, the flight control laws are unable to provide a specific roll rate as they need the actual aircraft speed to calculate the required aileron displacement. Roll control thus reverts to a conventional aileron response to side stick deflection, adjusted according to the aircraft configuration (there is lower aileron displacement for a given side stick deflection in clean configuration than with flaps extended).

Manually controlling an aircraft with conventional flight controls in pitch is more challenging at high speed or high altitude as the aircraft response to even small control column and trim inputs is more rapid, making precise flying such as maintaining altitude more challenging.

If the air speed indications become unreliable and affect the artificial feel, an added difficulty would be introduced as the control wheel feel the crew might be used to can become unrepresentative of actual speed, as previously discussed.

Flying a fly-by-wire aircraft manually in pitch, on the other hand, will not seem all that different, whether at high speed and high altitude or flying an approach. Loss of speed information does not noticeably affect aircraft handling in pitch other than the loss of various protections as the response to side stick deflection is purely a G-load demand. The aircraft remains in trim at all times and requires no flight crew intervention.

The work load is less on the fly-by-wire aircraft and the aircraft handling in pitch will feel more familiar since the flight crew regularly flies the aircraft manually during approaches for landing. Achieving the requirement to fly an aircraft to a certain pitch value and power setting in the case of unreliable airspeed, as per manufacturer checklists, appears easier to do on the fly-by-wire aircraft. Further, the aircraft remains in trim at all times.

The aircraft pitch axis is arguably the more difficult to control during high speed / high altitude flight.

Manually controlling an aircraft with conventional flight controls in roll is more challenging at high speed / high altitude as the aircraft response to even small control column inputs is more rapid.

The fly-by-wire roll control is a rate control in Normal Law. Releasing the side stick will command a zero roll rate and the aircraft will maintain the current bank angle if within normal flight parameters.

With the loss of reliable airspeed the roll reverts to a control surface displacement control, adjusted for aircraft configuration. This is not something the fly-by-wire pilot is normally exposed to. However, the aircraft would naturally tend to maintain the required bank angle or wings level just as the conventional aircraft would. And whether turning at 15 degrees or 20 degrees, angle of bank is not that material when trying to turn, whilst a small pitch change can quickly result in a significant altitude deviation which may be problematic, especially when operating in RVSM airspace.

Automation then, even in degraded mode, can still assist the flight crew in retaining control of their aircraft when forced to fly manually following a failure of some sort. Whilst the impact of a complete loss of all automation intensifies with increasing automation levels, this discussion illustrates that the loss of automation is in fact only partial, and does not alter the handling characteristics of the aircraft to a level where they are fundamentally different to normal flight.  Certainly, the large transport aircraft remains a stable platform. This understanding should have made the loss of Air France 447 avoidable.