How a Soviet Spy Helped Change the Course of the War

F. W. Meredith was an Irish aeronautical genius working at the Royal Aeronautical Establishment in England. In the 1930s, the RAE was not the fusty organisation we might imagine but a hotbed of radical ideas. Meredith’s fears for Europe, which proved prophetic, led him to spy for the Soviets. Meanwhile, the aerodynamic phenomenon he identified was harnessed to great effect for the P-51 Mustang. And Meredith’s contribution to aviation went even further. He played a key role in creating the first practical drone and even the autopilot. Far from being a thing of the past, his ideas may prove more important than ever for the next generation of electric aircraft.

In his book Spitfire, Mustang and the ‘Meredith Effect’: How a Soviet Spy Helped Change the Course of the War, Peter Spring traces Meredith’s remarkable career in meticulous detail, challenging (and sometimes even reversing) received wisdom about some of the war’s most celebrated aircraft. Which fighter was designed to harness the Meredith effect — the Spitfire or the Mustang? The answer is surprising.

We caught up with Peter Spring for an epic, in-depth, 5,000-word interview about Meredith, his inventions and his enduring influence on aviation. This interview, full of fresh perspectives, is essential reading for anyone interested in the Mustang, Spitfire or Second World War aircraft in general.

Ok, pour yourself a glass of wine, get comfortable and let’s jump in.

“In 1942, the RAE investigated why the Allison-powered Mustang I was so fast at low levels compared to the Spitfire V. US aerodynamicist Edward Warner was sent to England to investigate. He wrote he was told: ‘“Well, it really hasn’t much of a laminar-flow section”; that was modified to a suggestion that the wing design was such that laminar flow could only exist near the tips; and finally, that a skin joint running parallel to the span was sufficient to destroy laminar flow beyond the first 25% of the chord.’ Warren added that what interested the RAE was ‘the position of the radiator’.

Explain the Meredith effect in the simplest possible language?

A radiator placed directly in the airstream produces increasing drag as air speed rises. If, however, the radiator is placed in a carefully designed duct, the otherwise wasted heat from the engine’s coolant fluid will heat the air going through the duct, expanding and expelling it at a higher velocity, creating thrust that will, above a certain aircraft velocity, exceed the drag from the cooling system.

It should be added that Meredith’s seminal 1935 paper, Cooling of Aircraft Engines With special reference to Ethylene Glycol Radiators enclosed in Ducts, was largely concerned with the derivation of the equation which calculates the theoretical speed at which drag from the cooling system is exceeded by the thrust generated in the cooling duct. The key insight—that previously no one had considered worthwhile investigating—was that the tiny conversion of the heat into kinetic energy was worth pursuing. Meredith demonstrated that it was: leading to ducted radiator cooling systems becoming ubiquitous, although their efficiency varied widely; and totally replaced the alternatives of retractable and front shuttered radiators.

Producing the Meredith effect operationally required no particular maintenance. What did was the Mustang’s laminar flow wing. During the Second World War, there was great interest in whether the Mustang was so fast because of its laminar-flow wing or some other reason. For a laminar flow wing to function, its exterior surface must be exceptionally smooth and free of physical defects, contamination, or structural changes that could trigger an early transition from laminar to turbulent airflow.

What kind of speeds do you need to reach for the Meredith effect to be meaningful?

Meredith wrote in his original 1935 theoretical paper ‘It is shown that the power expended on cooling does not increase with speed for a properly designed ducted system but that, owing to recovery of waste heat, a thrust may be derived at speeds of the order of 300mph’. The specific number given by his calculations was 312mph at 14,000ft.

Experiments at the Royal Aircraft Establishment (RAE) in 1936 found that ‘the internal drag becomes negative at speeds over 290 mph ’. In 1941, US NACA researchers tested a propulsive duct. ‘At a Mach number of about 0.5, [360mph at 14,000ft] the propulsive effect had become equal to the internal drag, and beyond this speed substantial net thrust was developed.’ They concluded, “clearly the insignificant Meredith effect had the potential to become a primary jet-propulsion system.”

These experiments were carried on in laboratories using small-scale models. I am not aware of any tests using actual aircraft that established the velocities at which Meredith-effect thrust balanced cooling drag, but in practice such speeds may have been higher.

The question is the definition of ‘meaningful’. In practice, it is doubtful that most ducted radiator systems on Second World War fighters produced thrust that exceeded drag. The real issue was how much they negated cooling drag.

What was the first aircraft to incorporate the Meredith effect installations?

The Spitfire’s cooling installation was designed to generate the Meredith effect. In 1935, Meredith transferred to the RAE Aerodynamics unit (from its Avionics unit) to tackle the challenge that higher power from newer engines, like the Rolls-Royce Merlin, was being negated by increased cooling drag. That year, ‘Internally placed radiator systems [were rated] Priority A’ by the RAE. In September 1935, Meredith led an RAE delegation to Supermarine to discuss the radiator installation mock-up. E J Davis from Supermarine was present and, in 1986, gave a lecture on The Basic Design of the Prototype Spitfire. Davis wrote that ‘he and his colleagues were of the opinion that the thrust of the warm air from the radiator balanced the drag. It was in effect a rudimentary jet engine’.

Meredith was scathing at the meeting about the Spitfire radiator duct’s design and location, which was too short and would cause significant boundary-layer ingress problems. He was right, and the Mustang’s higher speed at the same engine power was partly due to its superior cooling-system design.

There is a popular idea that the Meredith effect actively contributed to the P-51 Mustang’s thrust; is this true?

It is certainly true that the Meredith effect contributed to the P-51 Mustang’s higher speed. The story, however, of how the Meredith effect came to be incorporated into the Mustang’s design is complicated, and I little understood despite the vast literature on the subject. Indeed,the whole story of the Mustang’s development is shrouded in myth.

The Mustang’s ventral radiator duct allowed a long, deep installation that facilitated the Meredith effect, but that location was not initially selected to produce it. By the late 1930s, US research favoured ‘oversquare’ radiators, that is, wider than they were long. It also favoured burying the radiator. In 1939 North American Aviation (NAA) head of Preliminary Design, Edgar Schmued, commented, when laying out a secret study for an Allison-powered fighter, that “due to the small size of the engine compartment, the oil radiator, coolant tank and cooling radiator will have to be placed to the rear of the pilot”. This had nothing to do with obtaining the Meredith effect, of which Schmued was then unaware.

Incorporating achieving the Meredith effect in the Mustang probably had two key stages:

1) Between February and March 1940 there was a major change in the design of the radiator duct of the P509 which developed into the Mustang. This moved the radiator duct forward and lengthened it. The event that precipitated this change may have been a meeting of NAA engineers in late February, who gathered to break the impasse between the British Purchasing Committee (BPC) and NAA over marketing the P509. At this meeting were CALTEC-trained aerodynamicists Ed Horkey and Irving Ashkenas. They were familiar with the concept of the Meredith effect through NACA publications, which drew heavily on German research that, in turn, developed Meredith’s work. (A 1946 NAA report co-authored by Irving Ashkenas described the “so-called Meredith effect”. This is the second use of the term I found. The first use was in the 1943 report cited above.) The revised duct would have been far more effective in generating the Meredith effect.

2) Between May and October 1940 there were major further changes to the duct design. There was a debate over whether, after purchasing material from Curtiss, including its XP-46 design, NAA incorporated it into the Mustang design. The BPC, which considered the XP-46 to have been designed to incorporate the Meredith effect, insisted that NAA review Curtiss material. NAA vehemently denied taking anything (for legal and reputational reasons) from the Curtiss XP-46. But a careful analysis of the XP-46 and the changes in the NAA design might cast doubt on these denials. I obtained the January 1940 XP-46 report from NASA (then NACA) and the revised design of the NAA aircraft (by then called the X-73) and the similarities are marked: both ducts have similar shapes; both have adjustable entrances; and a round slanted forward radiator like the XP-46 replaces the X-73’s square vertical radiator. I believe that, despite NAA’s denials, the timing and close similarities indicate that NAA did include design material from the XP-46, and that this improved the derivation of the Meredith effect.

It might be added that writings on aerodynamics show that Irving Ashkenas, who redesigned the entire radiator duct installation for the P-51C to use the Packard-built Rolls-Royce Merlin with a two-stage, two-speed supercharger, was very familiar with the Meredith effect.

Paradoxically, the Spitfire was designed from the outset to incorporate the Meredith effect radiator, and it did it badly; in the Mustang design process, its incorporation and enhancement were gradual, but it did it very well.

Which countries, including the Soviets, explored the Meredith effect and achieved it to a meaningful degree?