By William Pearce
In the early 1940s, Wright Aeronautical decided to utilize their 18-cylinder R-3350 engine as the basis for a new engine to compete with the Pratt & Whitney R-4360. The new engine developed by Wright was the R-4090 Cyclone 22 (Wright model no. 792C22AA). It used 22 R-3350 cylinders arranged in two rows of 11 cylinders. The R-4090 is one of only a few radial engines with 11 cylinders per row. It is also one of only three 22-cylinder aircraft engines ever built.
The 22-Cylinder Wright R-4090 engine of 3,000 hp (2,237 kW). (Aircraft Engine Historical Society image)
The air-cooled Wright R-4090 had a 6.125 in (155.6 mm) bore and 6.3125 in (160.3 mm) stroke. Total displacement was 4,092 cu in (67.05 L) and the engine’s compression ratio was 6.85 to 1. The Cyclone 22 had a two-speed, single-stage supercharger and gave 3,000 hp (2,237 kW) at 2,800 rpm for takeoff. For continuous output, the engine produced 2,400 hp (1,790 kW) at 2,600 rpm. However, increased performance was expected with further engine development. The R-4090 had a diameter of 58 in (1.47 m), was 91 in long (2.31 m), and weighed 3,230 lb (1,465 kg).
The crankcase was a steel forging, following a construction practice pioneered by Wright and used on other Cyclone engines. The three-piece crankshaft was built up through the two one-piece master connecting rods. Ten articulating rods were attached to each master rod. Each cylinder was constructed in typical Wright fashion and had 3,900 sq in (2.52 sq m) of cooling fin area. Each cylinder’s hemispherical combustion chamber had two valves; the exhaust valve was sodium-cooled. It appears that the .333 to 1 propeller gear reduction was provided by Wright’s standard, multi-pinion planetary gear system. The supercharger and accessory drive section was very similar to that used on the R-3350 engine. However, the supercharger had a 14 in (356 mm) impeller and gear ratios of 5 to 1 and 7 to 1.
Front of view of the Cyclone 22 showing the 22 R-3350 cylinders tightly packed around the forged steel crankcase. (Aircraft Engine Historical Society image)
The R-4090 possessed similar power and weight characteristics to early Pratt & Whitney R-4360 engines. While developing the Cyclone 22, Wright was preoccupied with serious developmental issues of the very high priority R-3350 engine and ongoing development of the 42-cylinder R-2160 Tornado; not much time or manpower remained for the R-4090. As a result, only a few examples of the Cyclone 22 were built, and it is doubtful that the engine ever flew. Perhaps three R-4090 engines were completed: two XR-4090-1 engines with a single propeller shaft and one XR-4090-3 engine with a coaxial shaft for contra-rotating propellers. The XR-4090-3 weighed an additional 30 lb (13.6 kg) for a total of 3,260 lb (1,478 kg). In addition, the XR-4090-3 was to have a two-speed nose case to maximize propeller and engine speed efficiency for maximum power and cruise power. Ultimately, the R-4090 Cyclone 22 was abandoned so that more resources could be used for the R-3350 Cyclone 18.
Radial engines with 11-cylinder per row are very rare. With so many cylinders, the engine diameter becomes very large, and the valve train can be crowded and complex. In addition, difficulties can arise with so many power pulses on each crankpin.
The R-4090 was very close to the same power and weight as the Pratt & Whitney R-4360 at this stage of development. (Aircraft Engine Historical Society image)
During World War I, Clerget developed an 11-cylinder rotary engine of 200 hp (149 kW), designated the 11E. Another World War I-era 11-cylinder rotary of 200 hp (149 kW) was developed by Siemens-Halske and designated the Sh.III. The Sh.III was unusual in that its crankshaft rotated one direction within the engine while the crankcase, with propeller attached, rotated in the opposite direction. The result was 1,800 rpm of engine speed with only 900 rpm of propeller speed—an ideal speed in the days of fixed-pitch propellers and no gear reduction. Far removed from aviation, Nordberg Manufacturing Company made a successful 11-cylinder, two-stroke, diesel, stationary, radial engine of 1,655 hp (1,234 kW) at 400 rpm for industrial use.
Other examples of 22-cylinder, twin-row radial engines include the Mitsubishi A21 (Ha-50), with a displacement of 4,033 cu in (66.1 L) and an output of 2,600 hp (1,939 kW) and the Hitachi/Nakajima [Ha-51], with a displacement of 2,673 cu in (43.8 L) and an output of 2,450 hp (1,827 kW). Both of these engines were developed by the Japanese during World War II and, like the Wright R-4090, never entered production. Clerget also studied a 22-cylinder engine between the wars, but it never progressed beyond the design phase.
Rear view of the R-4090 showing the supercharger and accessory drive section which is very similar to that found on the R-3350. (Aircraft Engine Historical Society image)
Sources:
– http://www.enginehistory.org/Piston/Wright/R-4090/Curtiss-WrightR-4090.shtml
– Allied Aircraft Piston Engines of World War II by Graham White (1995)
– R-4360: Pratt & Whitney’s Major Miracle by Graham White (2006)
– http://www.ww2aircraft.net/forum/engines/11-22-cylinders-radials-33342.html
– Model Designation of U.S.A.F. Aircraft Engines (1950)
– The Wright Cyclones by Wright Aeronautical Corporation (1942)
Old Machine Press 
What was planned power of Sh.III contra rotating engine?
Hello Aurum,
The Siemens-Halske Sh.III bi-directional rotary went into production during World War I. The engine produced 160 hp (119 kW) at 900 rpm. Keep in mind that this was really 1,800 rpm because the crankshaft and crankcase were both rotating at 900 rpm in opposite directions. The Sh.III could produce up to 210 hp (157 kW) by increasing the engine speed to 970 rpm (1,940 rpm relative) for short periods. The Sh.IIIa had a higher compression ratio and could produce 240 hp (179 kW) at 970 rpm for short periods.
OOh, ooh, I have a question!
“With so many cylinders, the engine diameter becomes very large, and the valve train can be crowded and complex. In addition, difficulties can arise with so many power pulses on each crankpin.”
So, in order:
“With so many cylinders, the engine diameter becomes very large.” The engine is based on the R1820, R-2600 and R-3350 and uses the same bore. The stroke is the same as the latter two engines. Are the cylinders the same? Finning, valve train? The diameter of the first three is 54, 55 and 55 inches, more or less. The R4090 is 58″. I assume the increase in diameter is because of the need for clearance between pistons, so the connecting rods are larger and the master rod to link rod (center to center) is increased. Do you know?
“Valve train can be crowded and complex” So, the pictures of the R3350 and R4090 seem to show the same heads. Can you say more about crowded and complex? I would imagine the cams themselves are the same general layout with the same gear reduction drive.
“…difficulties can arise with so many power pulses on each crankpin.” Is this about crankshaft resonance or lubrication of the master rod or something else entirely? With a power pulse every 33 degrees, it would seem slightly harder to lube than a 9 cylinder master rod. I remember there being issues with resonance increasing with the number of cylinders or rows of cylinders… is that to which you are referring?
“The crankcase was a steel forging, following a construction practice pioneered by Wright and used on other Cyclone engines”. What part of the crankcase was burning on the R-3350 that doomed so many B-29 crews to an early death? Which of the other engines were steel crankcases?
I would love to read a Cyclone engine family development article tracing the designs, their successes and attributes. I don’t remember Pratt & Whitney using the same bore so often, but maybe I am missing something. Bristol did with the Perseus and Hercules and with an increase in stroke, for the Centaurus. I noticed in Wikipedia the Centaurus has the same diameter as the Perseus despite an increase in stroke from 165 to 177mm.
Thank you for your scholarship. Is there a good book to read on how a radial was designed, as in what were the primary criteria and the issues that had to be kept in mind while deciding?
thank you!
Hello Jiro,
I will do my best to answer your questions, but I am no expert.
Are the cylinders the same? Finning, valve train?
– There were many different revisions and refinements that went into the Wright cylinders. The R-4090 cylinder appear to be pretty typical of Wright engines of the era. I cannot say for certain that they were identical to R-3350 cylinders.
The R4090 is 58″. I assume the increase in diameter is because of the need for clearance between pistons, so the connecting rods are larger and the master rod to link rod (center to center) is increased. Do you know?
– Yes, the crankcase is larger in diameter to allow clearance between the cylinders. With the stroke unchanged, the center-to-center length of the connecting rods would need to increase due to the larger diameter crankcase. That small 3.0 in increase in diameter is an 11% increase in frontal area.
Can you say more about crowded and complex?
– Crowded in the sense that you want to keep the dimeter to a minimum, so everything is moved as close together as possible. Complex in the sense that with everything as close together as possible, you still need to have space for cooling air and induction and exhaust manifolds. If the cylinders in each row are right up against each other, there is no room for air to get to the rear row. Likewise, there is no room between the rear row for the intake and exhaust manifolds to reach the front row. You can move ports to the top of the cylinder, but this would increase the engine’s diameter with the manifolding.
“…difficulties can arise with so many power pulses on each crankpin.” Is this about crankshaft resonance or lubrication of the master rod or something else entirely?
– Just a general statement. V-16s had a lot of twist on their long crankshaft. V-12 had a shorter crankshaft, and did not have the issues V-16s had. However, X-24s had a lot or stress on their relatively long crankshaft, with each crankpin absorbing basically twice the power of a V-12’s crankpin. Radials did not have this issue because of the relatively short crankshaft. But, if you put enough cylinders acting on a single crankpin, eventually some limit will be reached. Perhaps 11-cylinders was not that bad for each crankpin of the R-4090’s crankshaft, but it is a lot of power.
What part of the crankcase was burning on the R-3350 that doomed so many B-29 crews to an early death?
– The gear reduction case and supercharger housing were made of magnesium. The early engines leaked oil and had poor fuel distribution. Fuel would pool in the intake manifolds during startup attempts and ignite from a backfire. The fuel fire would then ignite oil on and around the engine. If bad enough, the fire would eventually ignite the magnesium housings. The engines were also quite typically run beyond their high-temperature limit, so any leaking flammable liquid was that much more likely to ignite.
Which of the other engines were steel crankcases?
– R-1820 G-100-sereis (1937 and later) and R-2600 C-series (1943 and later).
Is there a good book to read on how a radial was designed, as in what were the primary criteria and the issues that had to be kept in mind while deciding?
– I don’t know of anything that focuses on radial engines. “Aircraft Engine Design” by Joseph Liston (1942) might be a good place to start. But again, I am not an expert.