This is a bit of an old subject here at Aero Vintage Books, but I thought it worth revisiting. The first B-17Es rolled from the production line with a remotely-operated lower turret installed just aft of the radio room on the belly of the B-17. It had a shape and size similar to that of the new top turret installed just aft of the cockpit.
A look at many B-17 information sources, be they magazines or casual books, or even authoritative reference books on the airplane, speak of the first version of this lower (belly) turret as being a “Bendix” turret. Even one of my author/photographer heroes, Peter Bowers, does so in his excellent 1976 book Fortress in the Sky.
But, no. It was clearly a Sperry remote turret installation. Bendix had nothing to do with it. It did not retract and it was operated by a gunner lying prone in the fuselage aft in the turret with a gun sight that protruded into a plexiglass bubble on the belly, with six added viewing windows to help the gunner identify attacking fighters.
As can be seen in paragraph (4)(a) below as extracted from a B-17E technical order, the “Sperry number 645705-D remote sighted twin .50-caliber bottom turret” was installed on the first 112 B-17Es that rolled from the Boeing factory at Seattle beginning in November 1941…serials 41-2393 through 41-2504 inclusive.
Here are two illustrations showing the Sperry remote turret as compared with the Bendix remote turret as installed on the early B-25s.
If that’s all you wanted to know, Bendix or Sperry, the answer is Sperry and you’re done here. Move along now…
However, this minor clarification of the Sperry vs. Bendix turret installation allows the opportunity for a deep-dive into the historical record on getting gun turrets into the B-17 design. My primary reference for what follows is the definitive 1947 AAF Study entitled Development of Aircraft Gun Turrets in the AAF, 1917-1944, by Irving B. Holley, Jr. This 279-page study details, as the title indicates, the development of the gun turrets in general and has specific information on incorporating them on the B-17. That document is available as a download using the link above from the USAF Historical Research Agency site, and is a nearly-primary source on the history of AAF gun turrets.
.
Known Armament Deficiencies of the early B-17 design
To review a small part of early B-17 history, it needs to be remembered that the first production B-17s were the 39 B-17Bs ordered in a necessarily haphazard fashion between August 1937 and November 1938, with deliveries between October 1939 and March 1940. The B-17Bs had numerous improvements over the original Model 299 and the service test Y1B-17s, including the addition of General Electric turbosuperchargers for the engines as developed on the single B-17A.
However, despite the moniker of Flying Fortress applied to the B-17 and the mildly-impressive five 0.30 caliber machine guns sticking out from various fuselage ports, the Air Corps knew the B-17B had serious deficiencies with armament. As discussions were underway about the B-17C series, the Air Corps was working with Sperry and Boeing to get a central fire control system into the B-17C design, with low-profile remote control turrets installed behind the cockpit and on the belly of the bomber. Low profile turrets were desperately needed to reduce the drag of defensive armament and still provide an effective powered defense against the airstream and environment of the high altitude bomber.
Sperry, Bendix, General Electric, and Martin Turrets
From 1938 onwards, Sperry Gyroscope Company was one of several companies the Air Corps Armaments Branch was working with for turret development. Besides Sperry, major players were the Bendix Aviation Corp. at South Bend, Indiana, the General Electric Co. at Schenectady, New York, and the Glenn L. Martin Co. at Baltimore, Maryland.
In a simplified version of the narrative, the Air Corps was not putting all its eggs in one basket and all the companies were moving forward with turret development. Martin, for its part, was developing an upper manned power turret for the new Martin B-26. General Electric was working with Douglas Aircraft on a remote turret central fire control system for the Douglas A-20. And Bendix was developing upper and lower turrets for the North American B-25. In the Bendix case, the upper turret was being developed as a manned or, in the parlance of the day, a “locally controlled” turret. The Bendix lower turret was to be remotely operated.
As it developed, the General Electric remote turrets never made it into the A-20 design, but the turrets and central fire control systems were intrinsic to the later Douglas A-26 Invader, Northrop P-61 Black Widow, and Boeing B-29 Superfortress. The very successful Martin upper turret found widespread use beyond the B-26 and was installed in later versions of the Douglas A-20, Consolidated B-24, and several other types. The Bendix upper turrets were used in the B-25 well as other types. The Bendix lower turret…well, more on that later.
Remotely Controlled Sperry Gun Turrets for the B-17C (September 1939)
Back to Sperry and its turret development for the B-17 design. The B-17C was slated to get a Sperry central fire control system with low profile upper and lower turrets. The B-17C contract for the production of 38 B-17Cs was issued in September 1939 (a month before the first B-17B was delivered) and Boeing was working directly with the Armament Branch and Sperry to incorporate the turrets into the design.
To quote directly from the 1947 AAF Gun Turret Study (page 104):
In May 1940, the Boeing Aircraft Company expected to equip the B-17C with a a twin-.50 caliber upper retractable turret coupled to a remote sighting station containing a computer, and a similar lower remote using a Keuffel and Esser driftmeter noncompensating sight within the structure of the turret itself. When an Air Corps Board rejected the idea of using a modified driftmeter lower sight, Boeing warned the (Air Corps) Material Division that delays involved in replacing the sight “will kill the whole turret program for all B-17C airplanes.” boeing was unwilling to plan on any basis of “mythical anticipated development”.
Unfortunately, technology just wasn’t where the Air Corps needed it to be in 1940, particularly with turret drives and the precise control needed for accurate defensive fire. The central systems lagged far behind the B-17C schedule. The first B-17C flew on July 21, 1940, and was delivered to the Air Corps in September 1940, unfortunately without any turrets. The B-17C armament as delivered consisted of single 0.50 caliber installations on top of the fuselage (radio room), two waist guns, and the bathtub gun on the belly. A lone 0.30 caliber gun was provided in the nose but the ante was upped because now there were three available gun sockets to hold that gun in the nose section. Both Boeing and the Air Corps knew that the airplane remained woefully deficient in defensive armament, but the Air Corps desperately needed B-17s and could not wait for turret development.
B-17D as a Continuation of the B-17C Contract (April 1940)
The B-17D was technically just a change to the B-17C contract that ordered 42 additional aircraft in a change dated April 17, 1940. However, their were enough refinements of the B-17C design that the airplanes ordered were redesignated as B-17Ds. The defensive armament change was that the upper and lower gun installations were modified to hold two 0.,50 caliber guns. Turrets were to wait for the coming B-17E.
B-17E Development
The B-17E was the long-awaited redesign of the B-17 that would incorporate what the Air Corps and Boeing both recognized as essential to make the bomber an effective weapon. Tail guns were essential, so the process started with a redesign of the aft fuselage. Powered gun turrets were also essential, and the intent remained for the Sperry low-profile upper and lower gun turrets to be installed.
Initial discussion for the redesign were held in May 1940, just as Germany was invading France and two months before the first B-17C flew. At first, the Air Corps just added 150 airplanes to the B-17C contract with the improvements still to be determined. The Boeing engineers got to work on the changes. There was a flurry of activity leading up to July 1940 when budgets and funds got involved, and Boeing was forced to cancel some subcontractors they were retaining. On July 12, 1940, the Air Corps advised Boeing to plan on a purchase of 277 of the improved B-17Es, but a contract was not actually put in place until late August 1940, and that was for the whole lot of 277 aircraft, which were assigned serials of 41-2393 through 41-2669, inclusive. Less than three weeks later, a contract change added another 235 B-17Es, these receiving serials of 41-9011 through 41-9245, inclusive, for a total of 512 B-17Es now on order.
Royal Myth: RAF Experience With the B-17C Led to the B-17E
As an aside, it is common wisdom (commonly incorrect) in the history of the Flying Fortress that the B-17E was the design response to the Royal Air Force’s experience with the B-17C in combat with the Luftwaffe in 1941. As can be seen, the Air Corps long knew of the deficiencies in the B-17’s (and other types) defensive armament and was working diligently to get powered turrets, preferably remote control turrets, into the bomber. The Royal Air Force did not fly the B-17C in combat until a year after the design changes that brought forth the B-17E had been laid down on paper.
Sperry Turret Development
Since 1939, Sperry had been closely working with the Air Corps Armament Branch to develop a complete Central Fire Control (CFC) system. Initially, Sperry wanted to let the airframe manufacturer, Boeing in this case, build the turrets and the Sperry computing sights and controls would be incorporated. However, by April 1940 the Air Corps wanted Sperry to develop a complete CFC system. As noted earlier, by May 1940, Boeing was expecting Sperry to provide remote turrets for the B-17C. As late as the following month, though, the Air Corps and Sperry were still floating the expectation that Boeing would build the turret dome, the ring gear mounting, gun mounts, and gear trains; Sperry would provide the sight, computers, flexible shafts, power units, and controls. As seen above, Boeing rejected that concept and proceeded forward with the B-17C without turrets.
Requirements and contracts were constantly changing. Sperry hedged its bets a bit and designed components that were adaptable to the changing requirements and interchangeable between proposed turrets and systems. To the AAF, they proposed an experimental, remote, retractable turret mounting twin 0.50 caliber guns and a computing sight.
By the following month, July 1940, as the design details for the B-17E were coming together on Boeing drawing boards, the Air Corps Armament Branch abruptly changed direction. The 1947 AAF study noted that:
“…the plan for a complete fire control system for the B-17 was abandoned in favor of a more immediate solution to the armament problem, so that the threatened delays in armament development would not endanger B-17 production. At a conference of Armament Branch and Sperry representatives in July 1940, it was decided that Sperry would build a locally controlled upper turret mounting twin .50’s, each with 400 rounds of ammunition.”
This change of requirements suited Sperry’s approach to designing the turret as many of the components developed for the remote upper turret were able to be quickly incorporated into developing the manned turret. The remote lower turret development moved forward.
In September 1940, in conjunction with an overall review of lower turrets, the Armament Branch decided to consider the development of a spherical, non-retractable lower turret. Thus, the development of the ball turret was started. Notable, the 1947 AAF Study documents that “…the spherical lower or ball turret idea proved so satisfactory that the (Air Corps) Material Division sent rough drawings of the installation for Sperry to elaborate on, and authorized production of the ball turret for use on the B-17E in lieu of the originally contemplated lower remote.” Thus, even as Boeing was still doing the initial design work for the B-17E, the Sperry ball turret was being developed.
Stop-Gap Sperry Remote Turret into the First 112 B-17Es
In December 1940, the Materiel Division set a contract with Sperry for production of 540 upper manned turrets ($7,185 each) and 113 lower remote turrets ($11,275 each). Thus, at this point, it was envisioned that the lower remote turrets would only go into the first 112 B-17Es produced (with only one spare turret purchased??) as the Sperry ball turret was being finalized and set up for production. The obvious need was not to delay B-17E production to allow for the installation of the Sperry ball turrets; B-17 deliveries had the highest priority and the less-than-ideal remote turrets would be a stop-gap. And, it was planned that those initial 112 B-17Es delivered were designed to accept the Sperry ball turret as a bolt-in replacement; the converse was not true (the latter B-17Es were not built to accommodate the remote turret installation). The 112 B-17Es delivered with the Sperry remote turrets carried the serials of 41-2393 through 41-2504, inclusive, and were delivered between late September 1941 and mid-January 1942. The initial production of the B-17E lagged due to multiple factors, including turret deliveries, and the first article was about six months behind schedule. In actuality, Sperry subcontracted the production of the 540 upper turrets to the Steel Products Engineering Co. of Springfield, Ohio.
By the way, there is only one more or less in intact B-17E that originally had the Sperry remote turret installed, that being B-17E 41-2446 (Swamp Ghost) on static display at the Pacific Aviation Museum at Ford Island in Hawaii. Rare for many reasons, it also was recovered from Papua New Guinea in 2010. The vestiges of the early B-17E remote turret sighting windows are quite evident on the lower fuselage also.
The detailed history of this B-17E can be found, of course, in Final Cut: The Post-War B-17 Flying Fortress and Survivors available right here on this website.
The Sperry Ball Turret
As noted earlier, the Sperry ball turret seemed to be the answer to the lower turret problem for the B-17. Though there was hedging by the Air Corps about switching from the remote to the ball design, the prototype ball turret was ready for testing in early June 1941. Sperry designed the turret but sub-contractor Briggs Manufacturing Co. of Detroit, Michigan, took on more of a contractor role as the program progressed. In later June 1941, even before firing tests had been accomplished, the Air Corps Materiel Division ordered 250 ball turrets from Briggs as a starting point to get production moving.
As per the 1947 AAF Study (pages 204-205), “…the ball was reported to ‘compare most favorably in control with any turret which has been demonstrated in this county‘” and “…’was unusually correct in conception and design for a first experimental article.'” Problems of production of the ball turrets were eased when Sperry released manufacturing rights to the Air Corps and both Briggs and Emerson Electric Manufacturing Co. of St. Louis were soon building enough ball turrets to meet demands. Variations of the turret also went into the B-24. There were the normal teething problems with the ball turret, but the basic design proved remarkably successful. B-17E s/n 41-2505 rolled from the Boeing factory with a Sperry ball turret on January 14, 1942, and every subsequent B-17 built through the end of production in August 1945 had that ball turret installed.
Those Bendix Turrets
Back to those Bendix remote lower turrets for a bit. How they got associated with the B-17 lower turret is unknown but the proclamation that the B-17E remote lower turret as Bendix is long-lived and far-spread. Be that as it may, the story of the Bendix remote turret reveals major shortcomings in design and delivery of the turret. Development paralleled that of the Sperry effort at remote turrets, the Bendix experience with the lower turret was even worse than that of Sperry. The resulting Bendix Model K turret was installed in production B-25B and B-25Cs and B-24Ds. They were dropped from installation during subsequent production runs due to the turret basically being unusable in combat. On the B-24s, the Sperry ball turret replaced the Bendix remote turrets. Subsequent B-25 production just deleted a lower turret.
The Bendix remote lower turrets dropped from B-25 production enjoyed a brief if also unsuccessful resurgence. The 1947 AAF Study notes that, in April 1943, the AAF directed that the surplus Bendix lower turrets be utilized as chin turrets on the B-17F until the newly designed Bendix chin turrets could be put into production. That lasted until July 1943 when the AAF Materiel Center directed units to dismantle all such Bendix turrets installed in B-17F airplanes and “hand-held guns were continued until a production-version direct-sight chin turret could be procured in quantity.” (I’ve searched far and wide for a photo of a Bendix Model K turret on a B-17F as a chin turret but have not found one yet.)
Of course, that redesigned Bendix remote direct-sight chin turret proved to be a great success as later installed on late-series B-17F and all B-17G production. That turret had many similarities to the earlier remote-sighted Bendix turret but as developed initially for the XB-40 and YB-40 (B-17 gunship program) with the direct-sight installed at the bombardier’s station, it was much more effective than any other effort to provide nose armament for the B-17.
Final Answer
Back to the question: it was a Sperry remote lower turret on the first B-17Es and not a Bendix remote lower turret.