Most of Dan Ford's books are available in digital editions; click
 here for more

HOME > BUFFALO > CYCLONE ENGINE

About that Wright Cyclone engine . . .

Charles Lindbergh chose a Wright Whirlwind to power his Atlantic passage in 1927--a double compliment, since Spirit of St. Louis was a single-engined plane, with no backup if the Whirlwind faltered in mid-ocean.

The Wright Cyclone was a straightforward development of that engine--nine large cylinders around the propeller shaft, like the spokes of a wheel. As a piston moved toward the hub, the cylinder filled with 201 cubic inches of air mixed with gasoline, to be fired by two sparkplugs; the heat was carried away by cooling fins like those on a lawnmower. The Cyclone was designated R- 1820, for radial engine, displacing 1820 ci. The early models developed 575 horsepower at takeoff, less at higher altitudes.

The Cyclone powered most U.S. Army bombers of the 1930s, including the immortal B-17 Flying Fortress. Donald Douglas picked it for a new passenger plane that the airlines thought would be a trimotor. Douglas gave it two, and when the prototype took off and crossed the Continental Divide on only one engine, the airlines never again questioned the reliability of the Cyclone or the plane that became the DC-3, arguably the greatest aircraft of all time.

Other Cyclone-powered planes included the Douglas SBD dive bomber and the Lockheed Hudson light bomber--but not one important fighter. The engine wasn't suited to the high-g stress of fighter combat, probably because of the way lubricating oil reached the cylinders in the early models. Oil leaks and oil starvation are a constant of the Brewster Buffalo story, even in the B-239 that did so well in Finnish service.

There was one exception to the general rule. When the Navy wanted a fast-climbing fighter for convoy duty, it chose a Wildcat variant built by General Motors and powered by an R-1820- 56 Cyclone. The valves, cylinder heads, transmission gears, lubricating system, and supercharger had all been improved. The FM-2 still couldn't match the high-altitude performance of its Grumman cousins with their Pratt & Whitney engines--but then it was designed for low-level work, finding and strafing submarines.

In the end, the Wright Cyclone had a production run of 25 years, and it flew more miles than any other piston-driven aircraft engine ever built. - Dan Ford (with help from Ben Schapiro)

And commentary . . .

From Gorka Martinez Mezo in Spain

I can`t share this opinion: on its Soviet form (the M25 produced by the Shvetsov OKB), the Cyclone powered thousands of I-15, I-152 and I-16 fighters, used in combat successfully in Spain, China, and WWII. In many cases, against Buffaloes flown by the Finns! At least in Spain, the M25 was considered a good, reliable engine and a copy (based in blueprints sent during the war) built in the 1940s and used for many years.

Besides the M25, the Republican Air Force used genuine, US made, Cyclones. The CC&F G-23 Delfin (licence built Grumman FF-1) fighter bomber used US made Cyclones (Wright R-1820F-52 according to Spanish sources), while a batch of 775hp Cyclone R-1820F-54s obtained subrepticiously was fitted to a number of I-16 Type 10 fighters to allow them to operate at high altitudes, them the realm of German Bf-109s and off limits to standard I-16s fitted with the 750hp M25B. They formed an elite unit known as the "pacifier Scuadron" as the hasty fitted oxygen breathing apparatus looked like a baby`s pacifier.

The Republicans built more than two hundred I-15s under license, but they were unable to build the engines and there was a shortage of both engines and weapons for these. M25/Cyclone spares were built locally and many engines though too worn out for use were succesfully overhauled and put back into operations.

According to David Gesali`s DELFIN book on the CC&F G-23 career in Spain and based on original documents and many interviews with crewmen and support personnel survivors, US made Cyclones had a life of 400h under combat condtions, while Soviet built M25s were good only for some 100h.

After the war M25 blueprints were used to create a local clone, known as the Elizalde/ENMASA Beta 9C and used until the 1970s. First engines were delivered in the 1940s and some I-15s were reengined with the Spanish engine replacing the original Soviet M25 in the mid 1940s. Unfortunately, these engines proved rather unreliable and usually underpowered due to material quality issues, as the Spanish industry was ravaged during the war and substandard components were used as nothing better was available due to industrialization problems and the international isolation of the Francoist regime at the time.

From Panu Kolju in Finland

The oil circulation with one or two of the pistons in the Finnish Brewster fighters was a problem which shortened the life of the engine and rised the temperature in the engine, which you also mentioned in your story. I recall that one of the piston oil rings was turned upside down to "push the oil" to the piston chamber, which in turn lowered thw temperature and provided better oil surface for the piston. This job was done to all Brewster engines as soon as the problem was discovered. I was surpriced that it was also problem with other Brewster types, like you said in your story, and wonder if this is the problem behind all those heat problems you described in the article.

From Ben Schapiro in the U.S.

The idea of a world market impacting the way you designed your engines really didn't sink in until the middle of WWII. Most aircraft engines did not have any type of filtering at all for the intake air. (The exceptions that come readily to mind are the Spit and Hurricane with the stokes filters and the tropical versions of the Bf 109 and FW190 etc.) Dust, which wasn't much of a problem on fields with grass or paved surfaces, could halve an engine's time between overhaul in Egypt or dusty dirt fields in SE Asia. Some of the grit will get past the piston rings to mix with the oil (exhaust gases can blow past the rings and contaminate the oil) and wear at the crankshaft bearings, valves and rocker arms too. Eventually you get a wide tolerance engine and low power output.

Wright worked constantly at changing the cyclinder head and fin design of the Cyclone. I recall metallurgic changes and a 'W' fin design to improve heat flow. All internal combustion engines depend on the oil system to carry away a large part of the heat.... Bad oil cooling either due to design or operating conditions (climb, high altitude, high ambient temps) will raise the cyclinder temps and lead to increased engine wear.

The Buff's problems where in the the way the engine's oil system was sealed. The Cyclone had valves and rocker arms at the top of the cylinder head. Yup it was a pushrod engine with the cam mounted concentric to the crank. The rocker reversed the pushrod motion to operate the valves. All this stuff needs oil and the rocker/valve stem clearance needs regular adjustments. On the Cyclone there is a stamped metal cover, one to each side of the cyclinder seated on a gasket to allow access to the rocker and valve stem. [Some of] the RAF buffs in Burma had rebuilt engines off DC-3s, and I doubt Brewster paid for top of the line rebuilds. Once this stuff got good and warm and the wind got to blowing the covers leaked and the slipstream carried the oil mist out the cowl flap openings to the windscreen. It doesn't take much junk on the windscreen to make your eyes want to focus on it and not the outside world. [Also see what Vic Bargh said on this subject.]

Good maintenance, clean oil, lower ambient temps and correctly rebuilt engines would have probably prevented this problem and may explain why the Finish Buffs didn't have the oil leaks. I will hasten to point out I am NOT an airframe and powerplant mechanic.

[As for the structural difference between the Wright Cyclone on the Buffalo and the Pratt & Whitney Twin Wasp on the F4F Wildcat, Ben adds:]

The significant difference between those two engines is not the 10 cubic inches, but the bore and stroke of the cylinders. The Twin Wasp had 14 cylinders to the Cyclone's 9 and almost the same displacement. Less displacement per cylinder in the R-1830 means a smaller bore and stroke. Short stroke engines tend to rev higher and develop peak power at higher rpm vs long stroke engines which tend to develop max tourque at lower rpm. Back before Detroit put the same blocks in everything truck engines were all long strokers and sports cars had high reving short strokers. The anology may not be completely accurate in aircraft, but a look at the spec for the R-1830 show peak power at 2,100 rpm and for the R-1820 peak power at 1,500 rpm.

From Birgir Thorisson in Iceland

[First Birgir wrote:] There is something fishy in whole Cyclone story. The cylinder head overheating I would think had something to do with the air-cooling, not the lubricating. [To which he now adds the following:] From the engine section of "Jane4s Fighting Aircraft of World War II" comes this information on the cyclone.

The G-5 model used by the Finns was rated at 1000 h.p. for take off, 850 h.p. at 6000 ft, and 750 h.p. at 15.200 ft. (2100 rpm).

The G-105A used by the 339 export models was rated at 1100 for take-off, (2.350 r.p.m.) 900 at 6700 ft. (2300), and 800 h.p. at 2350 rpm at 17,200 ft.

the G-205 used by the F2A2 and A3 and the 439 was rated at 1200 h.p. at 2,500 rpm at take off, 1000 h.p. at 2300 rpm at 6700 ft.; and 1000 hp at 2500 rpm at 14,200 ft.

The increase in power at altitude is small for the 339, compared to weight increases, but the US examples seem to have a better power loading, and even the fully loaded A3 compares well with other allied fighters.

I have no figure for the normal loading of the Buffalo I, but guessing at a figure of 6400 lbs. the wing loading is approximately the same as the Hurricane Mk IIB trop, (30) and the power loading would be 8 at 17,200 ft, compared to 6.76. At a normal loaded weight of 6350 lbs, the A3 would have had a power loading of 6.35.

The F4F4 had the parameters, w.l. 28.5, and p.l. of 7.4 at 19,000 ft.

The P-40E had a w.l. of 35 and p.l. of 7.36 at 15.000 ft.

Spitfire Vb trop; w.l. 27.5, p.l. 5.6 at 16.000 ft.

At 5800 lbs, the 239 would have had a w.l. of 27.75, and a p.l. of 7.73 at 14.200 ft. At 5200 lbs, the parameters would have been 24.8 and 6.9. Which figure is accarute for Finnish Brewsters, I do not know, (the higher is from Eric Brown).

The fall off in power of large bore american and british radials at altitude is notable. The Sakae 21 had a take off rating of 950 hp at 2600 rpm, 1020 at 6,400 ft, and 885 at 15,700 ft. The Zero had a wing loading of 22 and power loading at altitude of 6.

From Jukka in Finland

In DF's Brewster Buffalo page there is some discussion on the Wright Cyclone and its "bomber" engine heritage. First I'd like to mention that calling an engine a "bomber" engine or "fighter" engine is terribly misleading (it is quite safe to say that no aero engine ever was designed to strictly one role). And it is an absolute fact that the Twin Wasp (which is called by some a "fighter" engine) was designed to similar purposes as the Cyclone was. Anyway, the early Cyclone had its problems. There was a mention how Finnish mechanics inverted the lowest oil scraper ring in FAF Buffaloes. It seems that the problem was not unknown to the Wright people. At least up the G100 series the Cyclone had 5 piston rings (3 compression, 2 oil scraper) per piston. In the H series (e.g. in FM-2) a 6th oil ring was added below existing ones. And it was inverted (scraped oil back to cylinder)! An excellent description of the engine can be found Jane's Fighting Aircraft of WW Two (Studio Editions reprint). In fact, in many basic and crucial areas the H series was pretty much a new engine compared to the old G series. And it should be noted that Wright was able to refine the R-1820 to give over 1500 hp (on 115/145 fuel) P&W had to enlarge the R-1830 to R-2000 to give such ratings.