The Westland Whirlwind: A Catalogue of Advanced Design

The Westland Whirlwind was a brilliantly designed fighter aircraft of World War II. Matt Bearman from The Whirlwind Project talks us through 10 remarkable design innovations of the beautiful and unlucky Whirlwind.

1 . The Tail Acorn

The coke-bottle shape became famous in postwar aviation as a way to avoid wave drag when going transonic – but many years before Westland had used the same shape to smooth out the pressure gradients that make for shocks around the Whirlwind’s T-tail. At the time the flutter that made the Whirlwind vibrate in a dive was understood to be the result of ‘interference drag – and whatever was really causing it, the solution was an acorn and a fairing.

The shape on the prototype’s leading edge grew from a pimple to a bulbous acorn, each iteration proving more effective than the last. The manufacturing drawing for the successful version was sent to the US military at the same time as Whirlwind P6994 was shipped to Wright Field AFB for trials.

This became the only detailed manufacturing drawing – as opposed to General Arrangement drawing – to survive, by ending up in the US National Archive in Maryland. Accompanying it is a letter from the Whirlwind’s designer-in-chief W.E.W. Petter, discussing pressure gradients and suggesting solutions to the high-speed aerodynamic problems of the P-38.

2. Retractable Tailwheel

Often overlooked, a fully retracting tailwheel with doors was not something you would normally find on a fighter in the thirties. Being new technology, it wasn’t perfect. Until modifications made it harder to break, high landing speeds and grass fighter strips meant complete failures of the unit were common – expected even. Never fatal and usually causing no extra damage, this was just an annoyance.

The mechanism was complicated – overly so, as were so many moving parts on Petter’s ‘baby’. The landing light has a complicated retracting mechanism involving hydraulics and joints on multiple axes.

Nothing was kept simple if it could be made complex – and hard to maintain for ground crews, who loved the Whirlwind far less than its pilots did.

3 – Monocoque Fuselage

Truly groundbreaking was the rear fuselage. Not so much in its use of an exotic magnesium alloy instead of the more usual aluminium, but in the fact that it was a pure monocoque.

Belt and braces had been the way until then – while the principle of an aircraft’s skin contributing to its strength was well established, few designers had sufficient faith in the theory to let a tube of thin metal hold it all together without a lot of internal frames, stringers and bracing.

The Spitfire, for example, was a semi-monocoque. While the skin was ‘stressed’ – i.e. it took some of the load – it was rivetted to structural members that were intended to keep things rigid.

The Whirlwind’s skeleton on the other hand was just a series of formers with a couple of bulkheads separating modular, hollow sections. While the cockpit section had some heavier (though sometimes strangely unconnected) members, the rear fuselage really was just a lightweight, empty tube. This was fortunate as it offered smaller groundcrew the opportunity to crawl down inside to fix the frequently broken tailwheel.

4. Fowler Flaps

This was very new in the 1930’s -a flap that didn’t just hinge, it slid backwards and down. Fairly common now but revolutionary at the time, the device gave more chord and more lift – you could use it for take-off in the partially open position, as well as for landing when fully down and ‘draggy’. One odd thing about the Whirlwind’s Fowler flap is that it was just one flap, running across the entire centre-section and incorporating both the curvature of the lower fuselage and the rear of both nacelles into the moving part.

Not only did this introduce an extraordinarily complex geometry to the system of rollers, actuators, jacks and guides, the engineering challenge was compounded by linkage of this system to the cooling gills on the trailing edge.

Using cams, the cooling shutters would be held fully open at half flap for take-off and climb and almost shut, with just a small aperture for exhaust, when the flaps were either fully deployed (as in landing configuration) or fully closed (in the cruise). This was a completely automatic system which made sense in normal operation. There may even have been a slight ‘Meredith Effect’ to the shut condition, further reducing drag. However, pilots weren’t always happy with only indirect control of engine cooling.

5. Slab Sides

Whatever his engineering design quirks, Petter really did seem to know his aerodynamic theory. He may have been informed by aerodynamicist John Carver Meadows Frost, recorded by obituary as the designer of the Whirlwind’s fuselage. Frost went on to design the world’s only viable flying saucer – the Avro Canada Avrocar.

A wing has to decrease in thickness towards the trailing edge if it is to work as a wing. This makes a deeply awkward ‘hollow’ when the resulting shape meets a rounded-section fuselage low down. Worse than the dodgy aesthetics, it produces what was then called ‘interference drag’ – by the bucketload. 

Designers were trying endless variations of fillet shapes to fill the gap, like the Spitfire’s huge fillet with inflections and double curves that needed a craftsman, an English wheel and several days to make.  In 1937 Petter – or perhaps Frost – had a lightbulb moment. This was really all about pressure gradients – just make the fuselage side locally flat, leave the gradients to the wing, and the problem goes away. Frost later used his insight into pressure changes to make the Avrocar fly.

The unorthodoxy of this shape has meant that artists and model kit designers have given the Whirlwind a small fillet ever since. It never had one – it’s just an illusion created by the peculiar transition from flat sides to a round frame aft of the trailing edge.

6. Bubble Canopy

Fighter pilots have always been as concerned about what is behind as in front. Although enclosed cockpits gave the opportunity for at least some comfort, a common complaint was not being able to see an attack shaping up from the rear.

Drawings for Westland Aircraft Co’s proposed P-9 from early 1937 show a perfectly smooth, teardrop-shaped and fully transparent canopy. This was extraordinarily ambitious – the technique and tooling to manufacture of such a large acrylic bubble didn’t exist. It was as though Petter was assuming it would be invented before he had to build a prototype.

Whatever inside information Petter possessed, as it was also in 1937 that Mouldrite, a division of ICI, began producing beads of their new ‘Perspex’ material that could be placed in a large mould to produce a complete rounded canopy.

7. Slats

Sitting on the leading edge of each outer wing, these spring-loaded slats were meant to ‘pop out’ at low speed and delay the stall. It was a piece of advanced aerodynamics that was generally sound but the automated deployment presented problems. Being independent, if one wing was closer to the stall than the other – in a tight turn for example – only one slat would deploy.

The asymmetry was often too much for the aerodynamic balance of the fighter and two Whirlwinds came down as a result before the slats were simply wired shut in 1942.

8. Leading edge intakes

This was a real shot in the dark by Petter. There had been no research anywhere in the world into what effect taking a large chunk of the leading edge out of a wing would have on lift and drag at the time of the design. It was a suck-it and see design that more or less worked. The duct itself, containing radiators and oil coolers, was designed – sculpted even – according to the work of Fred Meredith of the Royal Aircraft Establishment, whose work was later ‘borrowed’ by the North American Aviation team to produce the P-51’s drag-reducing belly scoop.

It is a tribute to Petter’s instinct that the wing performed as well with the large aperture in the leading edge as without, whilst having no external protuberances for cooling did make the Whirlwind slipperier than most. Retracted flaps narrowing the exit aperture and capturing the ‘Meredith Effect’ helped.

However, the complete lack of data on how to avoid flow breakdown inside the duct meant that very often not all the air that should have got to the radiators did, and overheating did happen. How much of that was due to the shockwaves coming off the propeller blades (spoiler alert) and how much was down to turbulence generated by the lips and walls of the duct will probably never be known – but it added to the myth of ‘unreliable’ engines.

9. Cannon

Nothing so subtle about these. While many of us now admire the Whirlwind as a flying machine, the four prominent 20mm cannon are impossible to ignore – The Whirlwind was after all a weapon of war and it functioned as a piece of flying artillery. Developed from the Oerlikon FF S, the 20mm Hispanos were aimed by simply looking down the nose. The concentrated fire needed no ‘harmonisation’ – they all pointed the same way and delivered rounds within inches of each other regardless.

Though the Whirlwind was mooted as a ‘Bomber Destroyer’ – essentially an interceptor – it was rarely used for this. Even during the Battle of Britain, it was recognised that the Whirlwind was better suited to other tasks. Should German armour have attempted to cross the channel, the barges would have had little chance on encountering a Whirlwind. Any vehicles that made it across would have been intensely vulnerable to the cannon. All the Whirlwinds delivered by the Autumn of 1940 were kept back in Scotland by Dowding for this possible use.

From 1941 until 1943 the relatively few Whirlwinds in service wreaked havoc on ground targets in occupied France using these weapons, as well as the 500lb bombs they were later modified to carry.

10 . The Culprit – DH Constant Speed Propeller

 The big one. It was discovered in 2018, by this author, that there was in fact nothing wrong with the Rolls Royce Peregrine engine, the one famously blamed for the Whirlwind’s cancellation. What looked like an inexplicable fall-off in engine power with altitude was in fact an entirely explicable fall-off in the propeller blades’ ability to turn power into thrust. To explain requires a series of concepts of varying degrees of difficulty to be ‘lined up’ and born in mind simultaneously.

The first is that a propeller blade is an aerofoil, working perpendicular to line of flight. All aerofoils accelerate air over their surfaces. Propeller blades are moving considerably faster than the aeroplane they are attached to, their velocity a combination of rotation and the aircraft’s forward movement. With all this it is inevitable that locally over the blade’s foil some air will become supersonic, especially further out where geometry means the blade section is moving faster for any given rpm.

That was the easy bit.

The fatter an aerofoil, the greater the local acceleration. Modern thin propeller blades won’t go ‘locally supersonic’ right up to the point where they are actually hitting the speed of sound – and even then the resultant shock, and drag, will be less. Fat aerofoils hit big problems much earlier. The Whirlwind’s propeller blades would start to create shockwaves at a combined blade velocity of 0.7 Mach.

Now bear in mind that the speed of sound decreases with altitude. So at combined speed of, say, 575 mph a fat bade section would produce no draggy shockwaves at 1,000 feet and yet be an aerodynamic mess at 30,000ft.

It all comes together when one considers the final innovation of the Whirlwind, the De Haviland Constant Speed Propeller. The idea is simple enough – all reciprocating engines have speeds at which they are most powerful or most efficient. Cars use gearboxes to keep close to a desired rpm whatever the speed. The constant speed system – of which the Whirlwind was a pioneer – works a little like an automatic gearbox. When it senses deceleration, it lowers the angle of attack of the propeller blade aerofoils. Drag on rotation decreases, and rpm comes back up. Too fast, and the blades go ‘coarse’ again, to slow it down.

As the Whirlwind climbed, it started to experience shockwaves on the blades. The higher it went, the draggier things got for the propeller. The constant speed unit, oblivious to the real cause of the drag, simply fined the blades in response, keeping up the revs. It wouldn’t stop until it had passed through a negative incidence, producing no thrust whatsoever. The Whirlwind appeared to lose power at height and as almost no-one had knowledge of transonic aerodynamics, nearly everyone blamed the engines, including Petter himself. Famously, this got the program cancelled.

You can support the project here. Buy The Hush-Kit Book of Warplanes here, and support Volume 2 here.