One Hundred Years of U.S. Navy Air Power (38 page)

NAVAL CONTROL VERSUS AIR UNIFICATION

The U.S. Navy watched the creation of the Royal Air Force (RAF) and drew the lesson that it did not want to suffer a similar disaster. Although in theory the RAF was expected to provide air services for the Royal Navy and for the army, in practice it disdained both. For example, it became interested in large flying boats as a substitute for the fleet, a consequence being that the Royal Navy needed long-range carrier reconnaissance aircraft. Because carrier aircraft capacity was so limited, the British became interested in using catapults aboard battleships and cruisers to launch combat aircraft. At first that must have seemed an excellent idea, but as aircraft performance improved, the requirement that, for example, torpedo bombers must be suited to such operation dramatically limited their performance—particularly because they also had to function as long-range fleet scouts. The U.S. Navy managed to retain control of its aircraft, including flying boats. It came to see big sea-based airplanes (most famously the PBY Catalina) as a way around the limits imposed by the Washington Treaty—at the least, as a vital fleet scouting force.
¹⁶

Naval control of naval aviation had other consequences. The U.S. Navy had in-house technical aviation expertise, which could respond to its own needs; the Royal Navy depended on the separate Air Ministry, whose experts might not understand naval warfare at all. Perhaps most important of all, by keeping naval aviation within the U.S. Navy, the United States maintained an organic relationship between flyers and the rest of the Navy. To further encourage air-mindedness in the fleet, a law passed in 1926 extended the aviator-in-command requirement to carriers and seaplane tenders.

Again, the Royal Navy was very different, because after 1918 all pilots held RAF ranks. They could rise to flag rank within the RAF, but they could not achieve high naval rank. Naval officers learned what existing naval aircraft and carriers could do for the fleet, and they certainly became acutely aware of the need for fleet air defense (probably more than in other navies of this period), but without air experience of their own they probably could not imagine more remote possibilities. Without
in-house technical expertise, it was difficult for the Royal Navy to frame requirements effectively. For example, based on its own experience that British naval aircraft had limited performance, the Royal Navy seems to have concluded that all naval aircraft would be limited in performance—with unfortunate consequences when it faced the Imperial Japanese Navy in 1941–1942.

Given Reeves' innovation and the understanding that numbers were vital, the U.S. Navy bought large numbers of aircraft during the interwar period. Numbers gave the Navy's Bureau of Aeronautics the clout to demand high performance (roughly like that of contemporary land-based aircraft) and also the rationale to finance the necessary engine development. During the interwar period, the U.S. Navy bought about as many aircraft as the U.S. Army Air Corps, in distinct contrast to the British experience. Moreover, because the Navy incorporated the full range of technical expertise, it could understand the potential of new aviation technology far better than could the contemporary British Admiralty.

TREATY CARRIERS

The same Naval War College exercises that motivated Reeves to recast carrier operating practices led U.S. naval thinkers to try to juggle the tonnage left over from the two ex–battle cruisers to get the greatest possible number of aircraft.
¹⁷
In March 1926, in connection with an abortive proposal for a light carrier (below 10,000 tons, hence not limited by the Washington Treaty), BuAer pointed out that carrier operating capacity was not merely a matter of hangar and flight-deck size but also of the rates at which airplanes could be assembled, brought up to the flight deck, and launched and also the rate at which they could be recovered and rearmed for a second strike. A larger number of small carriers might also be less vulnerable than a smaller number of larger ones. The main threat to a carrier lay in the gasoline spread among her aircraft and in fuel tanks and lines. Given this explosive load, the same weapon could destroy a carrier of any size, since all it had to do was touch off the gasoline and its vapor. Size bought little in the way of protection, it seemed. So what was the minimum acceptable size of a carrier? Surely it was set by the needs of the aircraft. Hangar headroom decided what sort of aircraft the ship could operate; BuAer considered 17 feet a minimum for the largest ones (the torpedo bombers). Hangar width was also important; the bureau considered 68 feet optimum. The ship should have the largest possible hangar, which would necessarily extend to the sides of the ship.
¹⁸
The desirable minimum deck width was 80 feet; cutting to 70 would be possible, but would dramatically reduce the area available for aircraft warming up for takeoff (by an entire fore-and-aft row). That is, the strength of the strike the carrier could launch would depend on how much parking space was available aft in which to warm up airplanes. The island particularly blocked the flight deck; without it, aircraft could
be placed outboard until their wheels were in the scuppers at the edges of the flight deck. Thus the island denied the parking space extending abaft it, and even somewhat to the side. The success of the flush-decked
Langley
suggested that it might not really be so difficult to vent smoke over the side of the ship rather than through conventional uptakes. BuAer had C&R sketch three carriers with its new deck and island arrangement, all capable of 32.5 knots: 23,000 tons (sixty-four aircraft in a mixed air group), 13,800 tons (forty-four aircraft), and 10,000 tons (twenty-four aircraft). Another set of studies showed that if the speed of the 13,800-tonner were reduced to twenty-five knots, aircraft capacity would increase to fifty-six.
¹⁹
The studies showed that hangar deck area (hence aircraft capacity) shrank more slowly than tonnage, so on a fixed total treaty tonnage it would pay to buy more, smaller carriers rather than a few large ones.

C&R (presumably based on BuAer comments) argued that the rate at which strike aircraft could be rearmed for another strike was the key design consideration. The lengths required to land (about 250 to 400 feet) and to take off (70 feet for fighters, 250 for torpedo bombers) were relatively fixed. The space between the two determined how quickly aircraft could be rearmed, and even whether they could be parked at all after landing. In other ships, length was chosen to achieve the desired speed for a given power plant. For a carrier, length came first, and an unusually long hull might achieve high speed even with limited power. Minimum length was clearly 600 to 650 feet, but anything longer would be highly desirable. Conversely, anything that made for slow landings would ultimately limit carrier operation. A 10,000-ton carrier could not have a deck more than about 600 feet long, which ruled it out from this point of view. It seemed that, beyond a point, any carrier would launch aircraft at about the same rate. On this basis a larger number of smaller carriers would put aircraft into the air, for a maximum air effort, more quickly than a smaller number of larger carriers.

For the first post–battle cruiser carrier, then, the General Board chose the smallest acceptable ship (13,800 tons), because it could get five such ships (the maximum number of aircraft) on the available tonnage: USS
Ranger
(CV-4).
²⁰
One advantage of minimum size was that it would be associated with the minimum power plant, whose smoke might be easiest to vent through the desired side pipes. BuAer's arguments were taken so seriously that many in the Navy considered the two ex–battle cruisers white elephants, before either had been completed. In line with the BuAer arguments,
Ranger
was designed with a flush deck and no island at all, venting her smoke from deck-edge pipes, which could fold down for air operations. Her flight deck was built as a light superstructure atop the hull. The large openings in the sides of the hangar were closed by sliding doors. Aircraft of this era had to warm up their engines before flight. Those stowed in an open hangar could do so before being lifted up to the flight deck, so a second strike could be launched soon after the
aircraft spotted aft had taken off. This arrangement also made it relatively easy to cut large holes in the flight deck for elevators. The big open hangar accommodated growth in the aircraft themselves. The ship was designed with two catapults but completed with none.
²¹
Speed was cut to 29.5 knots and the number of aircraft carried was increased (108 aircraft as compared to 91 that would normally be carried on a 32.5-knot designated ship). The lower speed of 29.5 knots simplified the smoke problem. The demand for a flush deck precluded anything like the 8-inch battery of the big carriers, so
Ranger
was assigned a combination of their twelve 5-inch antiaircraft guns and machine guns, all mounted alongside her flight deck (she had eight 5-inch as completed). While
Ranger
was being built, experience with the big carriers showed just how valuable an island could be, and she was redesigned while under construction. It was too late to change the run of her smoke pipes. They gave considerable trouble in service.

As
Ranger
was being designed, in 1926, U.S. naval aviators first demonstrated dive-bombing. This revolutionary means of attack made it possible to hit maneuvering ships, and it was far more difficult to defend against. In his FY28 report (September 1928), the fleet commander recommended abandoning torpedo bombers (which were large and heavy and had limited performance) in favor of the much more agile dive-bombers, which could be made so fast that fighters could effectively escort them. On the other hand, bombs from dive-bombers could not accelerate sufficiently to penetrate heavy armored decks; without torpedo bombers U.S. carriers would pose threats to enemy carriers and cruisers, but not to enemy battleships. As a member of the General Board later put it, the question was not whether carriers would ultimately supersede battleships, but when. When that would happen depended on when carriers could wield a battleship-killing weapon, a torpedo bomber with sufficient performance to survive in the face of enemy fighters and anti-aircraft guns. Given very limited funds,
Ranger
was not ordered until 1930 and not completed until 1934.

Initial experience with the two big new carriers,
Lexington
and
Saratoga
, completed in 1927, was a surprise. Each could accommodate about a hundred aircraft. War games (with the ships rather than on a game floor) showed that they derived considerable advantages from their speed coupled with their large air groups. It seemed clear that
Ranger
had been a mistake, but she could not be cancelled and reordered. To hold down her size, she had been made much slower than the two big carriers, and she lacked their protection, particularly against underwater hits. One factor in the preference for large carriers was the conclusion, based on operations, that carriers would have to operate individually. Ironically, the Japanese operated their carriers together at Pearl Harbor and Midway after having seen a film of four U.S. carriers proceeding together (on their return from a fleet exercise, not during it).

With the two ex–battle cruisers and
Ranger
built, around 1930 the U.S. Navy had about 55,000 tons of carriers left to build (not counting the obsolete
Langley
, which could be replaced at any time). The choice lay between three 18,400-ton carriers or two 20,000-ton and one 15,200-ton carriers. Given experience with the big, fast ex–battle cruisers (34 knots), speed was set at 32.5 knots, a figure that became standard.
²²
According to a paper justifying the new design,
Ranger
corresponded to theoretical analysis of carrier requirements made before it was clear just how versatile such ships would be—and how often they might have to operate with few or no escorts. High speed was essential. After considering a 15,200-ton design, about October 1931 the General Board chose the 20,000-tonner, which could be armored to some extent, provided with underwater protection, and also be faster than
Ranger
. It seemed that no carrier short of 18,000 tons could have sufficient speed and protection, and also that no power plant sufficient for 32.5 knots could be compatible with a flush-deck design (i.e., could rely on side exhausts). Experience with the two big carriers showed that the island seemed not to be any problem in air operations, and it certainly simplified ship and anti-aircraft fire control and also provided a valuable flight-deck control position. There was also a suspicion, which turned out to be justified, that too much freeboard had been sacrificed to make the
Ranger
design practicable; she was too wet. Elevators were again entirely rearranged, with one in the bow, one in the stern, and one amidships, corresponding to the evolving practice of spotting all aircraft either forward or aft. The midships elevator could serve aircraft massed either in the bow or the stern. Both the 15,000- and the 20,000-ton designs showed large islands to serve their considerable power plants. General Board characteristics required provision for hangar deck catapults, to be installed if the
Ranger
installation proved successful, but the sketch design showed no catapults at all. The ships were to carry 90 aircraft, 40 percent of them of the largest type (i.e., torpedo bombers), compared to 110 for
Saratoga
and 70 for the projected 15,200-tonner.

BuAer justified the
Ranger
design (perhaps retroactively) on the grounds that the U.S. Navy needed at least seven carriers (two
Lexington
s and five
Ranger
s). The flight-deck cruiser idea offered a real possibility that the U.S. Navy could gain up to eight cruiser-carriers, in which case (according to BuAer) the seven-carrier requirement could be foregone, and the 20,000-tonner adopted (which, according to a 12 November 1931 BuAer letter, had been shown some years earlier to be the ideal carrier size). BuAer's position did not, however, change with the demise of the flight-deck cruiser shortly afterward. Meanwhile war planning studies suggested that at least fourteen carriers would be needed to fight Japan. Few existing American merchant ships were suitable for conversion, and that would take nearly a year in any case. The U.S. Navy would have to fight with what it had, and each carrier would have to provide maximum air effort. These arguments were used to defeat a suggestion to build a pair of 25,000-tonners armed with centerline 8-inch guns (at the
expense of flight-deck length), with about a third fewer aircraft than the 20,000-ton “pure” carrier (the Japanese tried much the same idea at about this time). BuAer did point out that with five
Ranger
s it could operate 635 aircraft, but with the planned two 20,000-tonners and the 15,000-tonner that would be reduced to 553; if only two 25,000-tonners were built, the total would fall once more, to 429, and there would be only five carriers instead of six or seven.