Making the first flights in an immaculately restored historic aircraft, notorious for its handling peculiarities | Words: Elliot Marsh – Photos: George Romain

Lined up on Duxford’s runway, it occurred to John Romain that with more than forty years of Bristol Mercury experience behind him, this was a special moment. “It all started here, and the Lysander’s first flight was a culmination of sorts.”

In recounting the maiden flight of the Aircraft Restoration Company’s Westland Lysander MkIIIA V9312 in August 2018, John alludes to the latest chapter of a saga that dates back to 1974 and spans three Blenheim rebuilds, personal involvement with two other airworthy Lysanders, and V9312’s restoration.

The Lysander project surfaced during a visit to Kermit Weeks’ Fantasy of Flight air museum in Florida, early in 2003. There John was introduced to the dilapidated but largely complete V9312, held in Weeks’ storage hangar. “Kermit explained that he had two Lysanders and asked if I would be interested in buying the one he had in storage,” remembers John.

“Interestingly, it was a very rare British-built Lysander, identifiable by its hollow, one-piece undercarriage legs, British electrics and instrumentation, and wood ribbed wings, rather than the more familiar tubular alloy ribs of the Canadian-built Lysanders?perfect for our collection.

“Restoring it would be a real challenge. I thought about it for a few days and we spoke again, agreeing a price. It wasn’t long before we headed back to Florida to break it down for shipment to the UK”.

The Lysander arrived at Duxford in June 2003. Its home, initially at least, was ARCo’s famous Building 66 workshop?the beating heart of the Blenheim rebuild projects?where it was stripped down ahead of a full restoration.

ARCo had Lysander experience, having operated Kermit Weeks’ other example, ex-Brian Woodford V9545, for a short period in the mid-1990s, and the former Strathallan Collection machine that now flies with the Shuttleworth Collection?but they hadn’t yet restored one to flight.

“We knew it would be a huge undertaking. You’re dealing with a massive aeroplane that is classically British in its design, with all the eccentricities of the era. Leading edge slats interconnected with the flaps, huge amounts of woodwork and fabricing, a sensitive engine?all the hallmarks of a pretty monumental undertaking.”

John was intimately involved from the outset: “There are pictures of my son George and me taking the engine out. Dave Ratcliffe used to work on it, as he was just starting out with us. We had a mix of engineers and volunteers getting into it. There was a real team effort up until the Blenheim incident in 2003. The rebuild of that aeroplane then took precedence and the decision to put a MkI nose on it extended its repair. The Lysander continued coming along and was registered [to Propshop Ltd as G-CCOM] in December 2003, but the work didn’t progress at the speed we thought it would as internal resources were being poured into the Blenheim project.

“The emphasis transitioned to that repair work, and big chunks of the Blenheim went into Building 66. Nonetheless, ‘Smudge’ Smith led the work on the wings, establishing how sound the spar booms were and then manufacturing the two-part wooden wing ribs to build up each wing. In the meantime, my involvement was focused on working on the Lysander’s Mercury engine.

“Parts from a Bristol Mercury XX acquired from the Netherlands during the Blenheim project were combined with the engine pulled out of our aeroplane to produce one complete unit. Colin Swann, Smudge, Ian Arnold, Debs Perrot and Mike Terry then stripped and worked on the fuselage and tail. As soon as the Blenheim rebuild finished in 2014, it was Lysander time.

“We effectively split the project between Building 66 and our hangar facility, with 66 doing the cowlings, lift struts and fairings by that point, and the systems fit, fabricing and engine integration taking place down the eastern end. The wings came out of Building 66 for fabricing, which we did in a custom-built tent in the hangar.

“I taught [Aerial Collective’s] Lisa Waterfield about fabricing as we were doing it?she did a great job. Then we were putting in the Perspex windscreens, fitting out the cockpit and conducting a trial fit of the wings whilst the lift struts came together in Building 66. In the latter stage and as everything edged towards conclusion, Billy Kelly and Ian Arnold were heavily involved in finishing the aeroplane?Billy’s a good finisher, and really knows how to wrap up those monumental restorations.

“The next thing of real excitement was the first engine run,” Romain remembers. It was early evening on Wednesday, 8 August 2018 when the Lysander was rolled out of ARCo’s hangar, sans cowlings and disrobed of its side panels to expose the fuselage structure and nest of control cables within. With a succession of pops, bangs and an abundance of smoke, the restored 860hp Bristol Mercury XX fired for the first time, turning over in the characteristic low, guttural radial churn.”

Towering above…

Sat on the taxiway, the aeroplane stands taller than most single-engine historic aircraft, at 14ft 6in. A sprawling fifty-foot wing incorporates the aerodynamically-actuated slats and flaps responsible for the Lysander’s extraordinary low-speed handling characteristics.

Independently operating inboard and outboard slats are fitted to the leading edges along the whole length of the main planes, whilst the trailing edge flaps are connected to the inboard slats and deploy in tandem as angle of attack increases.

Access to the lofty cockpit is achieved via a series of footholds on the port undercarriage spat and side of the fuselage which allow the pilot to reach the lift struts and then the cockpit sill. The cockpit itself, John says, is a reasonably spacious and intuitively laid out environment for its era.

The elevator trim wheel?so vital to the Lysander’s operation, as we shall see?sits low to the left of the pilot’s thigh, antilockwise movement of the wheel decreasing the incidence of the tailplane (i.e giving nose-up trim). Throttle and mixture controls are mounted in a quadrant forward of the trim wheel.

The pitch of the two-position de Havilland DH 4/3A propeller is controlled by red knob located on the port side of the instrument panel. The cowling gills are opened and closed by a handle on the starboard side. The six core blind-flying instruments are centrally located in front of the pilot, as was standard for all RAF aeroplanes at the time. Engine instruments are grouped to the right, comprising gauges for the boost, tachometer, cylinder head and oil temperatures, and fuel and oil pressure.

Fuel is carried in a single 95-gallon main tank located between the cockpit and observer/air gunner’s compartment. The tank is pressurised by a handle on the port cockpit coaming. The Kigass primer and priming controls are on the starboard side of the instrument panel, the control lever having three positions?‘Off’, ‘Prime Carburettor’ and ‘Prime Engine’. Finally, the engine starter button sits under a hinged cover beneath the engine instruments.

A cold start requires three shots of primer, injected into the top three cylinders. The ground crew then turn the propeller through five blades before the pilot gives the engine a further shot of primer. Magnetos are switched on and?with sufficient priming?the engine fires as soon as the starter button is activated.

Oil pressure is an immediate concern: the Mercury’s system circulates oil at around 100psi on start-up and if the pressure doesn’t rise within thirty seconds, it’s critical to shut down the engine immediately. “Start-up is quite a visceral experience,” adds John, “and you’re feeling the heat radiating through the instrument panel, smelling the warm oil as the temperature rises”.

As the oil temperature increases through 40°C a valve in the oil system adjusts the flow, the pressure then diminishing and stabilising at 80 to 90psi. The propeller pitch is brought from coarse to fine by pushing in the pitch control knob, oil pressure momentarily dropping as the piston fills with oil and the angle of the propeller blades changes.

This adjustment has a marginal effect on static rpm, though the tachometer only reads from 1,400 rpm and at idle power any rpm changes are noted aurally. Some dissipation in fuel pressure may be symptomatic of the fuel tank’s position when the aeroplane sits in a three-point attitude, as the tank is mounted broadly in line with the engine and the gravity feed through the fuel lines is weak. This can be rectified by directly priming the carburettor chamber.

Says John, “it warms up a lot quicker than the Blenheim, which is disconcerting as the cowling designs are so similar and you would expect the two to behave broadly the same way. On a hot day you need to be conscious of cylinder head temperatures and don’t want to see more than about 180°C on the gauge.

“You need to have the cowling gills wide open on the ground and it’s essential to start into wind, otherwise you’d quickly end up with high temperatures in the cylinders and a low oil temperature.”

At 1,800rpm the mag drop should be fifty rpm or so, carb heat causing a twenty rpm drop. Before the first flight full-throttle ground testing gave 4 ¼ lb/sq in boost and static revs of 2,600, which gave a reasonable prediction of the anticipated rpm at maximum boost in flight, which is usually accurate to 100 rpm or so.

However, protracted ground running can be detrimental to bedding in what is effectively a new engine, as running with low loading can cause the oil to form a glaze on the cylinder walls. Should this ‘bore glazing’ occur, the piston rings will not seal properly and combustion gases and oil will get past them.

“We also use straight oil during the initial bedding-in phase, rather than detergent oil,” John explains, “as the latter is self-cleaning and isn’t particularly adhesive, for lack of a better word. Straight oils will stick to the surfaces and stay on the bores. That’s perfect for running the engine in as it ensures proper lubrication. We switch over to detergent oil later on, but having a coating of oil within the engine prevents corrosion and just makes the thing run a lot smoother.

“You do need to get this aeroplane in the air as soon as you can,” stresses John, “as you need to get the engine working at higher boost with the cooling airflow across the cylinders that you just won’t get on the ground. If the cylinder walls get very hot, it compounds the glazing problem and you could end up with an engine that consumes oil heavily and smokes. You’d then be in a position of needing to strip down the engine and remove the cylinders for overhaul”.

Taxi trials were an important step in the aeroplane’s return to flight, as John explains: “You can learn a great deal from taxying an aeroplane around. It’s your first impression of the aeroplane in motion?you’re monitoring the responsiveness of the engine and propeller, the brakes and air pressure, and feeling out the elevator, aileron and rudder to ensure the control runs have all been fitted correctly.

“The Lysander’s brakes aren’t brilliant and you’re instantly aware of that fact you have to anticipate a turn, and that will influence where, and how you taxi the aeroplane around?particularly if the aerodrome is busy. As you apply the brakes you can see the undercarriage leg flexing and moving, and it is initially alarming to see the spat come backwards by about an inch as you brake, before the aeroplane turns.

“You couldn’t instantly stop it with full brake; it slows to a halt fairly sedately by comparison with most aeroplanes and you need to be conscious of that.”

First flight

Lysander V9312’s maiden post-restoration flight came on Tuesday, 28 August 2018. The flight was preceded by a fast taxi run with the tail raised, which confirmed the position of the aeroplane’s centre of gravity.

The air test itself was primarily an assessment of engine performance, control effectiveness, and slat and flap deployment at low airspeed. In all, it lasted just twenty minutes, chock-to-chock. Duxford’s Runway 24?the preferred option for test flying?was used.

“If I’m getting airborne in a ‘new’ aeroplane on Runway 06 and need to abort at a late stage, I’d have to throw it into a high-speed ground loop to avoid the bank of earth at the M11 end,” John explains.

“On 24 I’ll accept going off the runway and taking the grass and the fence, as I’d be hitting them wings-level with an element of control. Ultimately, it’s about whether your intuition is that an issue can be resolved in flight, or that you need to get the aeroplane back on the ground, and that decision-making process has to be second nature. You need to be mentally ahead of the aeroplane.”

Prior to take-off the Lysander is positioned into wind, brakes set, elevator trim to takeoff position, and throttle opened to zero boost for ignition and propeller pitch checks. “You can accept a drop of up to 140 rpm on the magnetos, but it’s rare to get that?normally the Mercury will tell you it’s unhappy on a magneto check by hunting 200 rpm and backfiring, which means the spark plugs are oiled up.”

This is not a trivial problem as the only way to sort it out involves a trip back to the hangar to get the cowlings off, identifying the affected plugs and cleaning them up. If the mag test is acceptable, the propeller is then cycled twice from fine to coarse pitch, moving hot oil into the piston and ensuring that the pitch change mechanism is functional.

Settling onto the grass runway for departure, the aeroplane is held on the brakes for three seconds with the stick aft as the throttle is opened gently to +1 lb/sq in boost. Torque effect is negligible and little rudder input required. As the airspeed increases, the slats and flaps automatically retract.

“That brief transitional period between rotating and settling into the climb gives a good idea of slow-speed handling, which will be important later,” says John. “It’s also the point that any control issues, be it elevator, aileron, rudder, slats or flaps, would become evident.

“I’m looking at acceleration and throttle responses, instantly and subconsciously feeling the controls and anticipating something that indicates a problem. Keep an eye on the air regulator?if the pressure goes off the clock, you’ll need to come on and off the brakes in flight to dissipate the air. If the regulator fails, the air pressure can eventually cause the bottle to rupture. On the other hand, if you lose air pressure, you lose your brakes.”

John also monitors temperatures and pressures constantly and repeatedly. “They’re the life blood of that engine. If you see the worst-case scenario of declining oil pressure and rising oil temperature, you’ll turn downwind and land without hesitation.

If one changes without the other, it could be a gauging problem and you’d keep the aeroplane in an orbit over the aerodrome while you monitor it. Quite often we see gauge issues with these aeroplanes. It’s about reading those dials and understanding the story they tell.

“Smell is another big factor?you’re sensitive to the odours of fuel and oil and the potential leaks that may be associated with them, and you’d then be into looking for visual signs of a leakage. It didn’t happen with the Lysander, but a shiny coating on the tail feathers would be a sure sign of an oil leak.”

Power is maintained at zero boost to achieve 150mph, and as the aeroplane climbs through 500ft the propeller is brought into coarse pitch and the cowling gills closed. Fuel mixture is leaned above 1,500ft, coming back until a slight reduction in rpm is noted.

These settings stabilise the engine temperatures, giving a reasonable airflow whilst not working the Mercury too hard through high boost and rpm. “That said, we didn’t pamper the engine on its first flights,” notes John, “as it needed to run in positive boost for at least 25 to 30 hours to bed it in?a continuation of ensuring the bores didn’t glaze.

“We saw the difference latterly as the oil consumption went from being high initially to stabilising at a lower level, at which point we knew the engine had run in. Thereafter you can handle it like any engine.”

Back to the first flight: “everything is smooth in the ascent, and I’m settling down to what would be a normal cruise power setting of +1 lb/sq in boost and 1,750 rpm, seeing 160 mph on the clock and watching the cylinder head temperature coming down to 150°C. The oil temperature climbs a little and stabilises at 60°C; oil pressure stabilises immediately in the region of 80-90psi.

As soon as the systems and the engine give you a level of comfort that everything is doing what it should, it’s onto general handling and giving the engine a workout whilst getting used to the performance. I fly some gentle pitch ups and pitch downs, and a series of turn reversals and orbits, noting stick loading and climb and turn performance at cruise power.

“Then it’s crucial to bring the power back to assess the low-speed handling, as you never know whether you’ll need to get the aeroplane back on the ground swiftly?the reality is, though, that if there was a major problem you’d get it back on the ground more rapidly than a normal landing.

“I bring the boost back, finding an airspeed of 75 mph that I’m comfortable is satisfactory for landing without edging towards the stall, and monitor the aileron responsiveness, pitch sensitivity and rudder authority at that airspeed and mentally note the aeroplane’s performance in what is effectively its landing attitude.

“Controls are fairly light, but the ailerons feel slow to make any significant rolling motion and the elevator remains sensitive. Stability is good in pitch, a little less so in roll, as is the case across all airspeeds. I note outer slat deployment at 105mph, inner slat and flap deployment at 85mph. Then it’s back to the field for a run through at 700ft and around 180mph and peel into the circuit.”

In the downwind the throttle is brought back to -1 lb/sq in boost/1,200 rpm and the mixture brought back to rich to avoid a lean cut. Incremental nose-up trim is critical to counteract the nose-down pitch characteristic of automatic slat and flap deployment. “You have to just bite the bullet, get the trim in, hold the nose high and match the power to that angle of attack,” explains John.

“You’re still putting in nose-up trim at that point and if you let go of the stick, or if you overcompensate on the power, the slats will retract, and you’d end up with pitch oscillations. You should aim to be over the threshold somewhere between 75 and 90mph, holding off near full nose-up trim with forward stick, which feels counter-intuitive but helps with the round-out after touchdown as you’re coming from a positive push forward to a fairly relaxed aft pull to pin the aeroplane on the ground.

“That trim issue is perhaps the most critical handling trait of the Lysander,” John muses, “and the RAF Pilot’s Notes don’t quite reflect the reality of what can happen to you if you get it wrong?I don’t share their optimism! Ultimately you’ve got to have a go yourself and make your own determination.”

That first test flight logged twelve minutes of airtime and gave John preliminary notes on all aspects of the Lysander’s engine, systems and performance to feed back to the engineers for fine-tuning ahead of the second flight.

Those notes, he says, are made on practically any suitable surface: “I’ll write on anything I can?usually my gloves, which come back covered in times, temperatures and pressures and the like. I have a kneepad I can use as well. I’ve even stuck masking tape to my thigh and written on that, then peeled it off back in the hangar and handed it to the engineers. Anything that won’t get in the way if I need to get out of the cockpit quickly.

“I landed after the first flight and everybody was jubilant,” he remembers. Those who stood by as the Lysander ran down on the taxiway had been intrinsically linked to the aeroplane for years: Romain’s eldest son, George, who had accompanied his father to Florida to recover the aeroplane in 2003, and younger son Alex, now one of ARCo’s engineers; ‘Smudge’ Smith, Colin Swann, Ian Arnold, Billy Kelly and Col Pope, whose irreplaceable expertise had been so instrumental in completing the Lysander’s restoration; and Lisa Waterfield, who had worked on fabricing the wings. For them as much as John, the maiden flight signified the completion of a significant chapter.

However, John had been wrong-footed by the experience. “I had felt incredibly exposed in flight,” he reveals with great candour. “Sad to say, I couldn’t share their enthusiasm. I just didn’t want to be up there any longer. It seemed uncharacteristically loud?far more so than I remember from the other two Lysanders I’d flown, to the extent that it was uncomfortable.

“The lift struts played on my mind for most of the flight, almost like a fixation. The originals were badly damaged, and we’d gone through an authorised repair on them, including X-raying. I knew they were sound but couldn’t shake that feeling of apprehension… I started thinking about the struts failing, the wing folding and it being curtains for me and the aeroplane.

“I think I ended up with a bit of vertigo. I’d taken off with the hood open and the side screens down, completely open from the waist upwards. Most aeroplanes wrap around you, even open cockpit types, but the Lysander doesn’t – you’re perched up there, feeling like you’re sat on a stack of crates, with the wing at eye level behind you and out of sight.

“I remember climbing out of Duxford, seeing the farms and fields getting smaller below and just having this fear. That was a new and unsettling experience for me, and at the time I couldn’t get my head around why.”

High speed, low speed

The Lysander’s second test flight came the following day, This was a longer 35-minute flight from Duxford, in which Romain assessed the aeroplane’s climb performance and performed a dive to Vne.

“I closed the hood and the side screens on takeoff, and it was a fundamentally different experience?a world apart from the first flight,” he says. The noise issue that plagued the maiden sortie, it transpired, was down to the new helmet John was wearing. The mic was later found to be transmitting background noise through the earpieces at higher volume than was necessary?exacerbated by the aerodynamic din created by the open cockpit.

“The concerns I had first time round faded away and it was incredible how my mindset changed in an instant. Everything rationalised, I was able to delve into the testing whilst appreciating the experience.”

Once north-west of Cambridge and within sight of the diversionary airfield at Bourn, a timed climb to height was flown at +2¼ lb/sq in boost with the propeller in coarse pitch, cowling gills a quarter open, two divisions of nose-up trim and an airspeed of 120 mph with full fuel and a gross take-off weight of 5,568 lb.

In this configuration, the Lysander climbed to 3,800ft in two minutes and reached 4,520ft by two and a half minutes, at which point the climb was stopped on account of engine considerations. Subsequent climbs at different power settings provided additional engine data. “The concern here is running those cylinders too hot and deviating from the comfortable range of 140 to 160°C. Up at 180 you’re approaching the critical zone.

“It’s also essential not to temperature-shock the engine, and when flying above a couple of thousand feet on a cool day with moisture in the air, carb heat should be left on and maintained during the descent. During that second test flight I established that -2 lb/sq in boost with carb heat on in an orbital descent away from the airfield was an effective means of maintaining sympathetic temperatures.

“The Mercury is all about doing things slowly. It doesn’t respond well to rapid advancements or reductions in throttle and you’ll get a rich cut if you slam the throttle open?recovering from that safely might not be possible at low altitude, as it could take fifteen to twenty seconds to pick back up.

“Mid-range boost is preferable, between +/- 2 1b/sq in, where the engine is running smoothly, the temperatures should be sat in a comfortable region and the propeller isn’t driving the engine. Those techniques were learned on the Blenheim and are an essential part of Mercury engine management.”

For the Vne dive, John climbed to 5,000ft and applied two divisions of nose-down trim. With the nose pitched down approximately 40°, the airspeed slowly and progressively increased to 280mph at +2 ¼ lb/sq in boost, John noting 2,500rpm on the tachometer.

“It was really hammering down at that point! A visceral experience, for sure, with hot air rushing around the cockpit, the engine roaring and the airflow creating a terrific noise.

“I’m holding off the nose-up pitch with forward stick and working the rudder pedals to remain in balance, but the aeroplane feels solid and stable. Recovering from that at 280mph by slowly retarding the throttle, the stick forces are heavy, but the aeroplane remains responsive”.

The third and final thirty-minute test flight was flown later on the same day, focusing predominantly on general handling at high and low airspeeds. The Lysander’s control harmonisation isn’t particularly good?“Inputs are light across all axes,” says John “but the elevator isn’t effective, the ailerons heavy and the rudder light. Full aileron deflection is possible, and the Lysander starts rolling instantly, but the roll rate isn’t quick.”

So much so that John is convinced that there wouldn’t be the aileron responsiveness to recover safely from any roll excursion caused by turbulence at low level. “There’s no propensity to yaw, but it does have a very small amount of pitch instability?too much negative pitch, say on a gusty day, and you might get a stumble out of the engine due to the float carburettor.”

Slow-speed handling at altitude sees the Lysander brought down to 65 mph?still, John notes, not slow enough to stall the aeroplane, owing to that highly effective slat and flap combination, and power-off it descends with a rapid sink rate at a high angle of attack.

“You could turn final at 55 mph with full nose-up trim, slats and flaps down, holding it in the air with power on. Develop a big sink rate in that position?maybe through a gust of wind or an engine problem?and you’re hitting the ground hard. The elevator trim just won’t allow you to escape in the way a conventional aeroplane would.”

It is in the go-around the Lysander’s worst traits come to the fore. Aborting a landing requires consideration of the limitations of both the engine and elevator trim. Progressively advancing of the throttle (avoiding the rich-cut ) combined with forward stick should maintain longitudinal stability.

However, such is the change in trim that, as the throttle reaches approximately half of its travel, the stick will hit its forward stop. At this point nose-down trim must be wound on rapidly to contain pitch-up, the throttle being opened incrementally until the aeroplane is climbing away from the runway.

“You do not want to open the throttle without retrimming, otherwise the Lysander will pitch nose-high, roll over and crash. There was a lovely story told by Ken Wallis, who flew Lysanders in the Second World War. He had an American pilot who did just that and opened the throttle without taking off the nose-up trim.

“He instantly pitched but immediately rolled it on its side and allowed the pitch to put him in an orbit while he figured out what to do. As Ken said; a brilliant way of overcoming the problem!” (As he recalled in his autobiography ‘Adventure with Fate’, Westland test pilot Harald Penrose thought the Lysander should have been fitted with an automatic trim sy

Image(s) provided by:

George Romain