Westland Lysander

Over the past few years I’ve said some quite scathing things about the Westland Lysander. I know I’ve offended one or two good people, and to them I apologise. To offend was not my intention. What I tried to do was to describe the Lysander as I see it. It’s truly a remarkable aeroplane with some quite exceptional features. Some are exceptionally good, but some are, quite definitely, exceptionally bad. I’ll try to explain.

On reading Pilot’s Notes, for general flying, one is warned that:

“The controls are not well harmonised, the rudder being light, ailerons heavy, and elevator too heavy at or near maximum angle. The elevator trim is powerful, but slow acting and the elevator effectiveness is poor. Therefore, it is essential that the aircraft is trimmed longitudinally though all phases of flight to allow maximum control effectiveness. This is particularly so on landing, when the trim must be set fully back, and during take off, when it must be set to the take off position.”

Additionally, one finds:

“The flying characteristics at very low speeds are such that a foolhardy pilot might be tempted to take liberties to which no aeroplane can with safety be subjected. Particular reference in this respect is made to stalled take offs and climbs, stalled approaches to land, and flying at too low an airspeed. The stall is delayed to an exceptionally large angle of attack, and can seldom be reached. But if this aeroplane does stall, a wing drops very sharply and control is entirely lost until speed is regained after a loss of 1000 feet.”

In the take off section one can read:

“……..The tail should not be lifted”

and:

“A steep angle of climb can be obtained by climbing at 70 mph or even down to a minimum of 60 mph. This is an emergency operation and should only be performed if necessitated by operational considerations. If engine failure occurs while climbing at this speed, the nose must be pushed down instantly, otherwise at least 600 feet will be lost before control is regained”

I know I’ve never read such warnings before in Pilot’s Notes of an aircraft in service – such pros are more at home in a critical report following an unsatisfactory test flight! Nevertheless, the Lysander went into service with such deficiencies and what’s more, made a good account of herself on Special Operations Executive (SOE) clandestine flight operations. It’s a remarkable history, made so by the courage of the aircrew that flew her as well as the aircraft itself.

So what makes the Lysander what it is?

It was designed as an Army Co-operation and Observation aeroplane and with this in mind, lets first look at the good points.

From the cockpit, the field of view is nothing short of superb. The pilot sits well up over the engine with eyes in line with the wing chord and head in line with the wing leading edge. It’s a grandstand position and accordingly, he can see well to the front over the engine and down to the side as well. He also has a good field of view to the rear, both up and down if his head is raised and lowered over and under the wing. The raised position also leads to a good view forward when taxying, so there is not so much of a need to weave the nose from side to side as with other tailwheel types.

Within the cockpit all the controls and switches come well to hand and all the dials except one can be easily seen – the exception is the fuel gauge which is located on the fuel tank behind the pilot’s seat. This places the large 95 gallon internal fuel tank over the centre of gravity which in turn leads to consistent handling qualities as fuel is used.

So far so good: but what about the aerodynamics. The aircraft has been fitted with a remarkable slat and flap system that allows very slow flight indeed down to an equally low stall speed. Full span slats, in two sections on each wing, extend automatically as wing incidence increases. The outer slats move individually, but the inners are inter-connected and are arranged to extend automatically with the flaps. Consequently, as airspeed is reduced, or when more lift is required as the pilot pulls into a turn, the outer slats extend followed, if necessary, by the inner slats and flaps. As speed is increased, or incidence reduced, the slats (and flaps) blow in to provide a clean wing.

As said, it’s a remarkable system. The wing changes shape as required to provide the pilot with the most efficient wing profile for the ambient flight conditions. What’s more, it’s achieved by good aerodynamics and not digits as we have to do today – eat your heart out, Eurofighter! More importantly the pilot can never forget to lower the flaps for landing, nor can he forget to retract them before the limiting speed is reached after take off. It is all done automatically leaving the pilot to look after more important things.

Things are still looking good, as slow flight is also a pre-requisite of the observation roll and the ability to get into short strips is desirable for a communications aircraft. Now lets turn to some of the bad characteristics:

The Pilot’s Notes warned of certain elevator characteristics. The Lysander has been set up in a very pitch-stable condition. In practice this means that a considerable amount of elevator deflection is required to stabilise the aircraft for a relatively small change of speed. Therefore, during an acceleration, say, the pilot rapidly runs out of the forward stick necessary to prevent a pitch up. To counteract this, an adjustable tailplane has been fitted which is controlled by a large wheel on the port side of the cockpit. Full deflection of the tailplane from one extreme to another takes of the order twenty or so hand pushes of the wheel. So as the machine accelerates, the pilot must trim the aircraft nose down by rapidly and progressively moving the tail trim wheel forward. And at the same time as each application of trim is applied, he must move the control column back in sympathy. But, as the aircraft is still accelerating, the control column must be moved forward to compensate when the trim is not being applied. It can easily be appreciated that it takes some considerable pilot skill not to ‘switchback’ after take off.

As will be seen, the engine is powerful for an aircraft of this size. This leads to a high-energy prop wash which flows over the tail group and markedly increases the efficiency of the tailplane at high engine power. Consequently, although the aircraft’s speed is approximately the same on take off and landing, the position of the tailplane is fully up (leading edge down) with the engine off (landing) and almost fully down (leading edge up) when the engine is producing high power (take off). If the tailplane is set in the wrong position for take off, even with full forward control column, an uncontrollable pitch up would occur. Given that twenty odd applications of tail trim are required to move the tailplane from one extreme to the other, it is unlikely that the aircraft would survive this mistake.

The breakout forces in pitch and roll are high as are the friction forces in the control circuit. However, as the flight loads on the roll and pitch controls are also high, the static forces do not effect control in flight. The rudder is very light, both statically and in flight.

All in all, we have, as the Pilot’s Notes suggest, an aircraft with controls that are ‘not well harmonised’.

Let’s turn our attentions to the engine? Up front the Lysander has an 860hp Bristol Mercury radial engine. It’s light, air-cooled, economical and powerful. In fact, it produces a higher power than the Collection’s Sea Hurricane which has an 840hp (at sea level) Merlin III fitted. With the 95 gallon tank a flight endurance of up to about 3 hours is possible and at cruise speed, a range of about 350 miles can be achieved. Both are very useful to the design role. However, here comes the first point to ponder, and it’s not in Pilot’s Notes. Although the Mercury was made and successfully used in great numbers during WWII it did have several deficiencies. The carburettor is fitted with a powerful acceleration jet which will quickly lead to a backfire and engine rich cut should the throttle be advanced too quickly – I use a minimum of 4 to 5 seconds when moving throttle from idle to full. Further, the carburettor has a marked propensity to ice up. The symptoms include loss of power, rough running and backfire; the latter two are also symptoms of too much applied carburettor heat. (A cockpit carburettor heat control allows the air supply to the carburettor to be taken from around the engine exhaust, thus allowing hot, rather than cold ambient air to be delivered to the carburettor. Available power is reduced, but carburettor icing can be prevented). Because of the propensity of the carburettor to ice and the associated danger of engine cut, a carburettor temperature gauge is a standard fitted item in the cockpit, so at least, the situation can be monitored. Lastly, if negative G is applied, a lean cut will quickly follow. But the resulting rich cut when positive G is re-applied could take 30 seconds to clear. This is a positive disadvantage for the carefree low-level manoeuvring required for observation duties, and more importantly, it would be an even bigger disadvantage in evasive combat.

So, although the engine is light and efficient, it requires specific handling techniques which must be respected by the pilot at all times.

Finally in the basic aircraft appraisal we have the airframe. It’s certainly large and purposeful and the fuselage is cavernous, but it’s also extremely robust. The undercarriage is based on a very strong ‘wishbone’ shaped arm that passes through the bottom of the fuselage. Tales from the era tell of several pilots who found themselves sitting in the cockpit of a serviceable wingless fuselage, having lost the airfoils in an overly heavy landing…..

We appear, then, to have an ideal design for a battlefield observation and communications aircraft. It has performance to spare in flight, a good take off and landing performance and a reasonably long range and fair endurance. Some of the airframe and engine handling qualities leave a little to be desired, but the aerodynamics are excellent and the cockpit field of view is superb. Coincidentally, these are also ideal requirements for a Special Operations aircraft, with the ability to fly into and out of un-prepared strips during the hours of darkness and daylight.

Now, lets look at the practicalities of flight within the known characteristics of the machine and as they apply to the Shuttleworth aircraft. Walkround, start and taxy provide no specific problems, but the Shuttleworth groundcrew will not remove the chocks for taxy unless the pilot has set the tailplane to the take off position. On the Shuttleworth aircraft, three white marks on the fuselage sides show the two extremes of tailplane movement and the ideal position for take off.

Now, because of the aircraft’s sensitivity to longitudinal trim, we learned very quickly that putting a passenger in the rear seat caused a pitch up on take off and a rather steep climbout with the trim set to the normal take off position. Imagine the hapless pilot, as the nose of the aircraft continued to pitch up, rapidly applying down trim with his left hand whilst simultaneously holding the control column firmly on the instrument panel with his right.

By trial, we discovered that three bites of down trim from the normal take off position were enough to have the machine in trim at lift off. I remember on one particular flight, a transit with a passenger in the back, that I’d set the corrected trim position on chocks, but as briefed, the groundcrew would not remove the chocks with the tailplane in that position. No amount of gesticulation would convince him otherwise, so there was nothing for it but to set the normal position of the tailplane, and reset the new correct position during taxy. It was my fault, I’d not briefed him of the change, but I was most pleased with his diligence and I praised him accordingly after the flight.

Engine run up includes a check of the magnetos, the effectiveness of the carburettor heat (most important) and the pitch change on the two-pitch propeller. Pilot’s Notes require a large drop in rpm from 1800 to confirm a successful change of pitch from fine to course, but in practice, it takes up to 30 seconds for any change of rpm to be noted and a further 30 to 40 seconds to achieve a full change. With a heavy elevator, which must be held fully back during this process, believe me, the effort required does not equate to the satisfaction achieved.

The take off is the first major difference to normal aircraft operation. The stick is held fully back and the take off is carried out tail down. If the tail were allowed to rise, the slats and flaps would ‘blow’ in, thus reducing the lift available and significantly increasing the take off run. As full power is applied, the swing is light and easily contained with the rudder. Lift off occurs very quickly after a run of about 200 yards or so and at a speed of about 65 mph. At lift off, the nose immediately pitches up. Forward elevator is applied and trim rapidly pushed forward as the aircraft accelerates. As far as the pilot is able, the nose is held down to gain speed as quickly as possible so that full control could be achieved in the event of engine failure – see the Pilot’s Notes warning on slow speed engine failure after take off. At about 100 ft above ground and 110 mph, the propeller pitch is set to course, the engine throttled back slightly and the aircraft set into a steady climb.

The Shuttleworth aircraft has been modified into the SOE fit of external ladder and 150 gallon external fuel tank. The fuel tank cannot be used but the ladder is fully serviceable. Both combine to reduce lateral stability to the extent that the aircraft wanders excessively in yaw. However, the maximum speed achievable in level flight is not much different to that without the tank owing to the high engine power and (I suspect) the relatively high drag of the rest of the aircraft. In pitch and roll, the aircraft is heavy and sluggish. Pitch stability is high, roll stability is lower, but acceptable. Yaw stability is low with the tank and good without it. The rudder sensitivity is high and the forces are light, thus there is a propensity to over-control in yaw when the tank is fitted.

Notwithstanding the above, with practice, the aircraft becomes a delight to fly in a display. The automatic slats and flaps provide a good indication of available aerodynamic performance, as the movement of the inner slat can be easily perceived at the extremity of the pilot’s peripheral vision. The engine power is high and the only requirements are to bring the carburettor temperature gauge into the normal engine health scan and, of course, avoid negative G and throttle slams. Speeds achieved during a display tend to be between 100 and 200 mph, the latter being the never exceed with the tank fitted. The aircraft accelerates quickly to high speed and the throttle must be retarded to prevent engine overspeed, even with course propeller pitch selected.

We have never stalled the aircraft; quite rightly, stalls are prohibited on the Permit to Fly. But it is possible to essay slow flight. With the engine at idle, the trim set fully back and at full back stick, the aircraft can be flown down to about 65 mph. At this speed, the controls are still effective, and their use can be relatively aggressive. If power were to be applied at this point, thus energising the tailplane and increasing its effectiveness, a slower speed could be achieved. Theoretically, this would lead to a full stall at somewhere between 35 and 45 mph. At the stall, we would be below minimum power off control speed by at least 20 mph, and should the machine fall into a spin, or the engine fail, the results would be as the Pilot’s Notes suggest. There is no current plan at the Collection to discover the stalling characteristics of this aircraft.

The approach and landing provide another significant difference from the conventional aircraft. Lets start a description of landing the Lysander from the end of the downwind leg. Conventional pre-landing checks will have been carried out, the propeller would be in fine pitch, the airspeed about 100 mph and the height around 1000 feet. As the aircraft had decelerated through about 115 mph, the outer slats would have moved off their stops, the inners following at around 105 mph. The carburettor heat would be fully out and the throttle almost closed.

A normal delay at the end of the downwind leg before making a 180 degree turn onto finals should allow the air speed to further reduce to about 90 mph. As the speed falls through 85 mph, a descent to maintain the speed and the final turn are commenced. A small increase of power may be required, but it must be used with care. The engine is sensitive to power changes at low speed, both in the sense of rich cut and also power change with throttle position. Further, any increase in speed will allow the slats and flaps to retract, thus lowering drag and further increasing speed.

If a satisfactory final turn has been achieved, the aircraft should now be on a straight in approach at about 400 feet and about 85 mph. Ideally, the throttle setting should allow the airspeed to bleed of slowly as the threshold approaches, to be 70 – 75 mph over the airfield boundary. However, we do not live in an ideal world. The increase in drag from the automatic slats and flaps as speed reduces demands an increase in power to stabilise the aircraft’s speed. If this power increase is not controlled, the aircraft speed will increase, drag will reduce as the flaps etc retract and again, the airspeed will end up far too high.

Similarly, when on an approach at, lets say, 50 feet too high, if the aircraft nose is lowered to reduce height, the increase in speed will cause the drag to reduce and the airspeed to markedly increase. Then, when the aircraft nose is raised to reduce to the normal approach speed, a gain of more than the original 50 feet can be expected. Thus, the result of lowering the nose to loose height has resulted in a height increase.

But, the Lysander is not a conventional aircraft. If, when, 50ft too high, the nose of the aircraft is raised, the slats and flaps will extend, the drag will increase and the aircraft will descend. Then, at the correct height, if the nose is lowered to increase speed, equilibrium will result and the correct flight path will have been achieved. Intuitively, this technique is wrong. During training, all pilots are taught that it is dangerous to reduce speed to loose height on the approach. This is because a normal aircraft approaches at a relatively small margin over the stall speed and a stall at such a low altitude could be catastrophic. As we have seen, this is not so in the Lysander. A normal Lysander approach is carried out at twice the stall speed, so there is more than adequate safety margin to carry out such a manoeuvre. Indeed, it is the only way to adjust height effectively on an approach.

It’s worth noting that owing to the sensitivity of the throttle at low engine speeds and the problems associated with a rich acceleration cut of the engine, the pilot must avoid any tendency to ‘jockey’ the throttle on the approach to try to stabilise airspeed.

As the field boundary approaches, the pilot must ensure that full up trim has been applied, even if it means that the control column now approaches the full forward position. As the throttle is closed before landing and as tailplane effectiveness reduces, all available back stick will be required to arrest the descent rate to achieve a three-point landing.

Given the excess longitudinal stability of the Lysander and the propensity for the engine to stop on harsh acceleration demand, it is best not to attempt a practice go-around. However, actual go-arounds are necessary from time to time and knowledge of how to carry them out safely is essential to all Lysander pilots.

The throttle must be advanced slowly and progressively in sympathy with forward stick to stabilise the machine. At about one third to one half throttle, the control column will be close to the forward stop and the tailplane must now be moved. The left hand is taken off the throttle and as much trim as is necessary should be applied to bring the control column close to the aft stop. Owing to the high power output of the engine, the aircraft should be in a shallow accelerating climb at this stage, even with only one third, to one half throttle set. After trimming, the throttle is then advanced to the required position, or until forward stick limits further throttle movement. Then, the trimming process is repeated. In practice, adequate power for all but the most steep of go-arounds can be carried out with only one application of trim in between two applications of throttle.

Under normal conditions, with a medium to light weight aircraft, the landing run rarely exceeds 300 yards. As with the take off, it is possible to land a lot slower, but this would require high power on the approach to energise the airflow over the tailplane sufficiently to allow adequate elevator control. An engine failure in such a configuration would lead to an un-containable pitch down with the obvious result – see Pilot’s Notes at the beginning of this article. Slow approaches are not carried out on the Old Warden Lysander, but they were, by necessity, carried out in the field by the SOE.

Given the above characteristics is easy to see why the aircraft was well liked by the SOE pilots and agents alike. However, the machine does have some challenging characteristics that must be taken into account if the pilot is to survive. In the hands of such pilots, the Lysander became a formidable SOE weapon.

It was, and still is, a truly exceptional aeroplane.

© Andrew J Sephton

ANDY SEPHTON