Being designed in the mid-1930s, the B-17 enjoyed, at least initially, some of the rapid improvements in engine technology that occurred during the decade that led into World War II. Beginning with the B-17A, it was the first Air Corps bomber to employ the turbosupercharger to significantly improved its higher altitude performance. This post will explore some of the details of the engines and superchargers that powered the airplane, and address some of their misconceptions.

What is a Supercharger?

First, to define some terms. In general, any engine supercharger will increase the amount of air available for combustion, thereby increasing the amount of power that can be produced by the engine. In the case of large radial aircraft engines, most superchargers were the centrifugal type, where by an impeller geared to the engine crankshaft compressed either the incoming air before the carburetor or the oxygen-fuel mixture ‘charge’ going into the intake manifold. The superchargers are normally integral to the engine and, for both radial and inline engines, placed between the crankcase and the accessory case on the back of the engine.

Initially, aircraft superchargers were relatively simple single-speed, single-stage devices. Later development produced two-speed, multi-stage superchargers that provided much improved high-altitude performance. The drawback of the supercharger was that, because it was geared to the crankshaft, it added drag to the crankshaft and reduced the net horsepower output of the engine.

What is a Turbosupercharger?

The turbocharger, more commonly referred to as the turbosupercharger back in the day, instead took advantage of engine exhaust to turn a turbine wheel that was connected to an air compressor, the output of which was ducted to the engine intake. The was a very minor power output reduction due to backpressure in the engine exhaust, but it was so minor as to be negligible. The drawback to the turbosupercharger was the required complexity: the bulky device had to be mounted in such a way to have exhaust routed to it, and ducting to feed the air intake and, then, after being compressed, usually in intercooler to cool the compressed air before it went to the carburetor. The intercooler radiator also required ducting to move air from the slipstream, through the intercooler, and then exhausted. Also, the method of controlling the turbosupercharger was usually with a wastegate at the end of the exhaust that controlled how much engine exhaust was directed to the compressor turbine, and how much was bypassed. How that wastegate was controlled added even more complexity to the turbosupercharger.

The other problem with turbosupercharging was that it was great in concept but difficult in execution. The metallurgy required to provide reliable components that could withstand the high-temperatures, high turbine rotation speeds, and contaminated exhaust proved slow to be perfected. It’s no accident that the company that ultimately developed a reliable turbosupercharger, General Electric, was also a pioneer in American jet turbine development as the technology and metallurgy was directly applicable.

Early efforts to develop the turbosupercharger took place in Europe in the first two decades of the 20th Century. In the U.S., however, the early Army Air Service asked General Electric to take on development work on a turbosupercharger during World War I. Initial Army testing began in 1918 at McCook Field in Dayton, Ohio, and further testing was completed at Pike’s Peak in Colorado at an elevation of more than 14,000′. The turbocharger was installed on a Liberty engine for developmental tests that proved the concept but manufacturing defects and challenges plagued the turbosupercharger. General Electric continued development, however, and by the late 1930s had perfected it to where it was operationally feasible for installation. As it turned out, General Electric designed most, if not all, of the turbosuperchargers used on American aircraft during World War II.

Practical Applications in the Late 1930s

So, it can be stated reliably that all U.S. engines installed on combat aircraft, both radial and inline, contained a supercharger of some sort, usually a centrifugal type, that boosted the input air to the engine’s intake manifold. For example, and writing very generally, the big difference between the Allison V-1710 and Rolls Royce V-1650 Merlin inline engines was the multi-speed, multi-stage supercharger/intercooler on the Merlin that provided much better high-altitude performance. It wasn’t that Allison was negligent in design, though, by incorporating just a single-speed, single-stage supercharger. (To quote Craig Ferguson, “I look forward to your letters…”)

The U.S. Army Air Corps was convinced that turbosupercharging was going to be added to all V-1710 aircraft installations, including pursuit (fighter) types. Though the engine went into the P-38, P-39, and P-40, it was only the P-38 that ultimately had a workable turbosupercharging installation. The P-39 and P-40 suffered greatly in high-altitude performance due to the limitations of no turbocharging. The P-51 started with the Allison, but it was quickly replaced by the Merlin engine (and the rest is history, etc.).

But, to directly address the B-17 and it’s progression through engines…

Boeing Model 299

For the prototype Model 299 that brought forth the B-17, the installed Pratt & Whitney Hornet engine was a civil version of the engine, the model being the S1E-G (military designation was the R-1690). The S1E-G model was rated at 750 hp (35″ MP at 2250 RPM at 7000′ pressure altitude) and rated at 850 hp for take off at 40″ MP and 2400 RPM (limited to one minute at that power setting). The single-speed supercharger had a gear ratio of 12:1. With the engine rated at 7000′ pressure altitude, the available power decreased with an increase of altitude above that. There was no turbosupercharger installation.

The engine had 3:2 reduction gearing to bring the propeller RPM speed down to a more efficient range from the crankshaft RPM.

The empty weight of the airplane was 21,657; the normal maximum gross weight was 32,432 pounds, and absolute maximum gross weight was 38,053 pounds. In flight testing, the Model 299 had a cruise speed of 140-204 miles per hour and a maximum speed of 236 miles per hour at 10,000 feet. The service ceiling was 24,620 feet. Its maximum range was 3,101 miles. Carrying a 2,573 pounds load of bombs, the range was 2,040 miles. By contrast, the competing twin-engine Douglas B-18 had a top speed of 220 mph, but its range was only 1,030 miles with a bomb load of 2,532 pounds.

As is well known, the Boeing Model 299 had a short existence: its first flight was on July 28, 1935, and it crashed at Wright Field on October 30 of the same year with only about 36 hours of flight time.

YB-17s and Switch to Wright R-1820 engine

Those versed in the history of the B-17 know that the Air Corps ordered thirteen service evaluation B-17s, strangely designated as Y1B-17s (instead of YB-17s) due to funding that purchased the airplanes. A major change in the design was the selection of the Wright R-1820-39 Cyclone engine instead of the Pratt & Whitney Hornets. The R-1820-39 was rated at 805 hp at 2,100 RPM at sea level, 850 hp at 2,100 RPM at 5000′, and 930 horsepower at 2,200 RPM for takeoff, thus a modest increase in power over the Model 299. However, with the improvements and changes of the YB-17s over the Model 299, its empty weight increased by 2,808 pounds, with corresponding increases in maximum gross weights. The YB-17s had a top speed of 239 mph, essentially the same as the Model 299.

The R-1820-39 engine had a two speed supercharger with gear ratios of 7.14:1 and 10:1. With the two speed supercharger, the service ceiling increased to 27,000′ over the Model 299’s 24,620′.

B-17A and Turbosuperchargers Added

Things got more interesting with the development of the sole B-17A. Originally part of the YB-17 order as a static-test fourteenth airframe, it was instead set aside for a test turbocharger installation. General Electric had developed its turbosupercharger to the point where it was available for an operational airplane, and it offered a substantial performance improvement.

The engine was changed to the R-1820-51 with a single stage supercharger with a gear ratio of 5.95:1. This engine was rated at 1000 hp for takeoff (at 2,200 RPM) and 800 hp at sea level (at 2,100 RPM). Initially, the new turbosuperchargers were mounted on top of the nacelles aft of the engine, an obviously ungainly installation with the necessary exhaust and intake ducting concealed in a bulky fairing above the engine that also contained the air intake.

This quickly proved aerodynamically unworkable, and it was redesigned to install the turbochargers beneath the nacelles, with exhaust tubing being routed around the landing gear wells on the number 2 and 3 engine. It proved to be a much better design. The engine and intercooler air intakes were moved to the leading edge of the wing inboard of engines 2 and 3, and outboard of engines 1 and 4. Also, the B-17A had the number 2 and 3 exhaust routed to the turbosuperchargers inboard of the nacelles, moved to the outboard sides on subsequent production.

With the turbosupercharger added as a second stage, the B-17A had a service ceiling of 38,000′, a considerable jump over that of the YB-17s. Top speed increased with altitude and was recorded as 271 mph at 25,000′. Cruise speed at 10,000′ was 183 mph. Empty weight was boosted to 31,160 pounds (I guess turbosuperchargers, ducting, and intercoolers are heavy…) but maximum gross weight was also increased to 44,000 pounds. As can be seen, the performance increases offered by the increased in engine output was offset by increased airframe weight.

Final Turbosupercharger Configuration with the B-17B

Whereas the Y1B-17s first flight was in December 1936 (thirteen months after the crash of the Model 299), the B-17A did not fly until April 1938. However, the configuration of the engines and turbosuperchargers was fixed with the B-17A, and carried forth to the contract for the first ten (out of 39) B-17Bs that were ordered in August 1937. When the first B-17B flew in June 1939, it had the engine and turbocharger configuration that largely remained unchanged for the balance of the B-17 production through the B-17G.

The B-17B had the General Electric Type B-2 turbosupercharger installed that matched the final installation details of the B-17A. For the outboard engines (1 and 4), the turbosuperchargers were installed on the bottom of the engine nacelles forward of the front wing spar, while the inboard engines (2 and 3) had them installed aft of the engine nacelle and spar due to the landing gear wells.

The outboard engine intercoolers were installed vertically behind each engine firewall, while the inboard engine intercoolers were installed aft of the nacelle and front spar also due to the landing gear well. Ducting led to the intercoolers from openings in the leading edge of the wing, and the warmed intercooler air was ducted out of the wing from the distinctive slotted openings at mid-wing behind each engine. The carburetor inlet temperature could be controlled by the copilot using a panel installed on the right cockpit wall. The controls adjusted valves in the intercooler exit air. In conditions where carburetor icing was expected, the intake temperature could be increased to prevent its formation.

On the B-17B, the turbosupercharger output was controlled by hydraulically-actuated regulators located at each turbo exhaust that controlled the wastegate position. The pilots controlled the the regulators and thus the output power using four levers located on the top left of the control pedestal, located to the left of the engine mixture controls with both sharing the same control lock mechanism. On the B-17B, the only systems that utilized the aircraft’s hydraulic system were the wheel brakes and the turbocharger regulators.

Of course, there were other major changes to the B-17 design between the YB-17 (and B-17A) and the B-17B, which won’t be detailed here. However, the engine model was unchanged; the B-17B was also powered by the R-1820-51. Empty weight went surprisingly down to 27,652 pounds, with maximum gross weight set at 46,178 pounds. Top speed was similar to that of the B-17A: 268 mph at 25,000′. Fuel capacity was increased to 2,482 gallons giving the B-17B a range of 3,000 miles.

The B-17C and B-17D

There were significant changes that came with the B-17C, which had its first flight on July 21, 1940. However, the changes to the powerplants and turbochargers were minor. The engine was upgraded to the R-1820-65 model. This engine retained the single-speed supercharger but the gearing ratio increased to 7:1 on the new engine and it was rated at 1000-hp at 25000′ (2300 RPM). The empty and maximum gross weights continued to escalate, but the top speed of the B-17C was stated as 300 mph, quite a jump from that of the B-17B. How much of this was the result of the engine/turbo combination being refined is unknown.

There were 38 B-17Cs ordered in September 1939 (the same month that the Germans invaded Poland) with an additional 42 ordered in April 1940. The equipment and armament changes specified in the April order were significant enough to have these 42 aircraft redesignated as B-17Ds. However, there appear to have been no changes to the engine/turbo combination with the B-17D. The only change that might have affected the system was the addition of hydraulically-actuated cowl flaps to the engine cowlings. Two engine driven hydraulic pumps (on engines 1 and 2) plus a manual hand pump provided the hydraulic pressure to operate the only three hydraulic systems on the aircraft (brakes, cowl flaps, turbocharger regulators). The first B-17D was flown on February 3, 1941, with deliveries of the 42 aircraft between February 3 and April 20, 1941.

Comes the B-17E

As is commonly known, the B-17E was considered to be the first combat-worthy version in the series. Changes that brought forth the B-17E were started in 1940 (well before the RAF had combat experiences with its ‘Fortress I’ B-17Cs). The first 277 B-17Es were ordered in late August 1940 and the first B-17E was delivered to the Air Corps just a year later, in September 1941. There were so many changes brought forth in the B-17E that it was essentially a new aircraft. However, its engine/turbo configuration, Wright R-1820-65, the General Electric B-2 turbosupercharger, and associated systems and controls remained unchanged from the B-17D for the first 100 B-17Es.

The B-17E used Eclipse Type A-7 turbocharger regulators. The first 100 aircraft, s/n 41-2393 through 41-2492, used hydraulic pressure to control the regulators. However, beginning with B-17E s/n 41-2493, the airplanes used engine oil pressure to operate the regulators, thus providing each regulator with its own oil source. The cockpit controls remained the same.

The A-7 regulators measured exhaust pressure that was matched with the cockpit-selected turbocharger setting to control the wastegate that controlled the turbocharger output. Reports suggest that it was a labor-intensive effort by the pilots to match manifold pressure on each engine and fine-tune the settings with power changes, a problem that carried forth from the earlier regulators.

An excellent technical description of how this turbosupercharger control system can be found on this page of the Aircraft Engine Historical Society website.

An additional 235 B-17Es were ordered on September 16, 1940, bringing the total B-17E production to 512 aircraft, all from the Boeing plant at Boeing Field in Seattle. On June 2, 1941, the Air Corps ordered 300 more B-17Es but, by the time these were in production, they were transformed into the refined B-17F.

B-17F: Minor Changes to Engine/Turbosuperchargers

The B-17F was the first to be built by three manufacturers: Boeing, Douglas, and Lockheed. The engine model (and thus the supercharger) and the turbosupercharger installation remained essentially the same as on the B-17E. The turbosupercharger regulators were upgraded to the Eclipse A-11 with no identified changes to the system. The big change to the powerplants was the use of the wide Hamilton Standard ‘paddle-blade’ propellers to increase performance.

The top speed of the B-17F was stated as 325 mph at 25000′ and service ceiling was 37,500′.

As noted above, the first 300 B-17Fs produced on the Boeing contract were actually bult on an earlier B-17E contract, and the AAF ordered another 3,435 from Boeing. (The last 1,435 from this order were actually delivered as B-17Gs.) Douglas built 605 B-17Fs and Lockheed built another 500.

Final Configuration: B-17G

The production shift from the B-17F to the B-17G was gradual. Some late production B-17Fs had the Bendix chin turrets installed at the factory. Other changes were gradually added to the production line in a continual process.

The engine version that powered the B-17G was the R-1820-97, which had the same power rating as the R-1820-65 and just incorporated some minor changes from the earlier engine. The single-speed supercharger gear ratio (7:1) and the propeller reduction gear of 1.78:1 remained the same. Similarly, the engine was rated at 1200-hp for takeoff, 1380-hp for ‘war emergency’, and otherwise, 1000-hp at 25,000′.

Not surprisingly, hanging a big turret on the nose of the airplane had an effect on the top speed of the B-17G, which was stated as 302 mph at 25,000′. Maximum gross weight was now 48,726 pounds and it could carry up to 3,600 gallons of fuel. The first B-17G was delivered on September 4, 1943, and the last on August 28, 1945. Depending on how one counts the B-17F/B-17G transition deliveries, about 8,680 B-17Gs were built.

On earlier B-17s, controlling the turbochargers to match the power requirements had proven to be cumbersome with the older regulators that used a measurement of exhaust pressure to set the wastegate position. There were too many variables that could affect the exhaust pressure to provide the desired precise control and balancing of the engine power. To address this problem, the Minneapolis-Honeywell Company developed an electronic turbocharger control system that utilized a pressure reading on the intake ducting to the carburetor which, combined with several other sensors and electronic controls, proved to be much more suitable. The wastegate was now controlled by an electric motor. The new system was dependent upon AC power, though, so a DC-AC inverter was required for operation.

Instead of turbocharger control levers on the throttle pedestal in the cockpit, an electronic turbo boost selector was installed at the same location.

A detailed explanation of how this electronic system worked can be found on this page of the excellent Aircraft Engine Historical Society website.

For Boeing-built B-17Gs, the new electronic turbo boost selector system was installed beginning with the B-17G-10-BO block. Lockheed-built B-17Gs had the new controller with the B-17G-5-VE block and, for Douglas, with the B-17G-15-DL block.

Also beginning in early B-17G production was the replacement of the original General Electric B-2 turbosupercharger with the upgraded B-22 turbosupercharger. The designed RPM of the B-2 was 21,300 vs. 24,000 on the B-22, presumably accounting for the increase of the rated altitude from 25,000′ to 30,000 for the B-22. The upgraded B-22 was incorporated in production at and after the B-17G-35-BO block, the B-17G-25-VE block, and the B-17G-40-DL block. In practice, the B-2 and B-22 turbochargers were interchangeable as the mounts and controls were identical. Some loss of performance would theoretically, at least, be expected but there were no limitations in maintenance instructions to address such a change.

The Wright R-1820-97 single-speed supercharged engine, General Electric B-22 turbosupercharger, and the Minneapolis-Honeywell Electronic turbocharger control system carried forth as the standard engine/turbocharger installation for the balance of B-17 production.

References: This material was mostly derived from the various technical manuals for the B-17 and R-1820 available on the excellent Air Corps Library site. If you are reading this to the end, and don’t have a subscription, you should. Another excellent website for such information is the Aircraft Engine Historical Society that contains a wealth of information about old engines and systems.

Now, I’m confident that most of this material is correct but the nit noids no doubt crept through. I welcome any comments and corrections to make this posting more accurate.