Though at the time of Pearl Harbor, on 7 December 1941, the Europeans had vastly greater experience of air warfare than the Americans, the close collaboration between Britain and the United States from mid-1940 had gone far to put the United States well in the picture. With characteristic energy the US Army and Navy re-thought their day-fighter procurement, buying aircraft with the much more powerful R-2800 engine and much heavier firepower. Planned mainly for night use, the Army P-70 and P-61 saw limited action in the Pacific war against Japan. But the most immediate burden of the Japanese war fell upon the US Navy and Marine Corps. Both deployed large and important forces of combat aircraft, the most basic difference between the two services being that the Marine Corps squadrons were nominally all land-based and trained to operate chiefly in support of amphibious assaults on land targets.
Both the Army and the Navy took a keen interest in Britain’s struggle in the winter of 1940–1 to defeat the night bomber. It is worth emphasizing that neither adopted an isolationist policy, nor said that it could see no threat to the United States. Though it did not at first collaborate with the British with quite the same affinity as the Army, the US Navy – which handled procurement and technical development for the Marine Corps also – gradually came to recognize that the need to be able to intercept enemy aircraft at any time of day or night was a pressing one. In collaboration with the Army Air Corps (Army Air Force from March 1942) it organized the nationwide Aircraft Warning Service, with equal attention paid to both the Atlantic and Pacific coastal areas. The defence problem was alleviated by the geographical remoteness from any enemy, other than carrier-based forces such as had crippled the Pacific Fleet at Pearl Harbor; but it was exacerbated by the sheer size of the territory to be defended, and the absence of any US night fighters. Fortunately the development by the Radiation Laboratory of advanced and powerful microwave surveillance radars had been started in 1940, and small numbers of pre-microwave sets had been used by the Army Signal Corps as early as 1938. By December 1940 the Navy was working with its own team at the Radiation Laboratory on large shipboard radars, not only for air and surface warning but also for GCI by the fleet’s own fighters. By February 1941 American microwave ASV radar was being designed, and a different team then began the challenging task of developing an AI radar that would fit into a single-seat carrier-based fighter.
Though the British had been unable to build much research hardware on this wavelength, they had conducted limited studies into 3 cm radar from the spring of 1940, and were able to give the Radiation Laboratory a lot of useful information. On virtually no evidence, they had believed that this extremely short wavelength was a prerequisite to radar being satisfactory in a single-seat fighter. The US Navy was persuaded to the same conclusion, and in January 1941 it requested a 3-cm AI radar that was light, compact, simple, difficult to jam, operable by the pilot, and useful in night air interception with secondary capability in attack on surface vessels. By August 1941 sufficient success had been achieved with 3 cm components for design to begin on a complete radar, called AI-3 (SCR-537). The complete installation was proposed to the Navy on 2 November 1941. Six days later the Navy requested from the Chance Vought company a study for a night-fighter version of the Corsair, the XF4U-2. In December Sperry received a contract for AI-3 development and production, with the Radiation Lab serving as consultant. On 28 January 1942 the mock-up review board accepted the proposal for the XF4U-2. The radar was contained in the fuselage behind the pilot’s seat, with a waveguide running out along the right wing to a scanner in a fairing near the tip. Spiral scan was used for search (120° cone at fifteen scans per minute) and a 10° conical scan for the second phase, then called ‘sight’, in which accurate aiming was needed for a blind firing attack. The single CRT display in front of the pilot showed a Type G (Az-El) presentation with ‘wings’ growing on the target blip in inverse ratio to range.
Flight trials began in a JRB (Twin Beech) in April 1942, with the scanner in the nose. Soon the pointing error had been refined to less than 0.25°, though problems were met with sea return, with humidity and with performance above 20,000 feet. In June the set was taken to Marine Corps Air Station Quonset Point, Rhode Island, for tests and training, while a second set flew in an F4U. This was delivered as an operable night fighter in October 1942, but by this time extreme pressure on the F4U-1 programme had unfortunately caused Chance Vought to drop work on the F4U-2. By late 1942 the first Sperry set, the AIA, was in production for the Grumman F6F Hellcat. A total of 604 were delivered, and installed in F6F-3E and -5N Hellcats, as described later, and the AI Mk XIVA (ASH) in British NF Fireflies and TF Sea Mosquitos. In 1943 this radar was given the new Army/Navy designation of APS-4.
In January 1943 work began on an improved radar, with a new Dalmo Victor scanner, Westinghouse pressurized RF head (this did away with the long waveguide from the fuselage to the wing pod) and Stromberg Mk II modulator. This radar was completed in October 1943, and flight-tested in an SNB-1 (Beech Kansan). Westinghouse began to deliver production sets, designated APS-6, in April 1944. A few months earlier deliveries had begun of the APS-6A in which, because of difficulties with the new RF head, the receiver was the Philco APS-4 type. Many modifications were made, including changing the polarization of the aerial feed through 90° to improve beacon reception, before production APS-6 sets were cleared for use. Sea clutter remained a major problem, but on the whole this challenging miniaturized set worked well even at heights above 30,000 feet and in combat manoeuvres. It could paint coastlines at forty miles, and dependable search range was rarely less than five miles. Minimum range was a mere 360 feet. In the search mode the well-balanced in scanner rotated at no less than 1,200 rpm in sweeping through a 60° spiral scan. No tremors were felt in the cockpit, and the fighter’s stability and manoeuvrability were unchanged, though in a sideslip the pod falsely increased the airspeed reading. The whole installation weighed 242 lb, and was priced at $10,936. Altogether the enormous total of 5,260 were delivered, 2,161 of them by 1 August 1945.
All the first deliveries were installed in Corsairs and Hellcats. The very first Navy night fighters were 12 F4U-1 Corsairs converted by MCAS Quonset Point and the Philadelphia Navy Yard. The ‘hand-built’ conversion installed the APS-4, usually removed the outermost of the three 0.50 in guns in the right wing, and rearranged the cockpit to accommodate pilot radar controls, a viewing scope and an autopilot. These dozen aircraft saw a fantastic amount of action. Six went to VF(N)-75 at Munda, New Georgia, where Lieutenant-Commander Widhelm and his experienced pilots – all said to have logged over 2,000 hours – slaughtered the Japanese bombers that had previously come over every night mainly to deny the exhausted Americans any sleep. Their first combat mission, and first kill, was on 31 October 1943. Thereafter every ‘washing-machine Charlie’ that approached New Georgia was destroyed. Indeed, Lieutenant O’Neill (so the story goes, and the Navy confirms it) shot down a Japanese bomber that unexpectedly got in the way as he tested his guns one night over Bougainville! The other six aircraft served with VF(N)-101 under Lieutenant-Commander R.E. Harmer aboard Essex, Hornet and Intrepid, and then were assigned to Marine Corps VMF(N)-532. So single-seaters were too difficult to fly at night? It is doubly significant, because until a short time earlier the Navy had insisted that the Corsair could not be operated from carriers even in the daytime.
These arduous missions from pitching carrier decks and island airstrips pioneered a big US Navy and Marine Corps night-fighter effort. Numerically the most important night fighter in the Pacific was the Grumman Hellcat, designed and developed with amazing rapidity long after the start of the Corsair programme, and produced to the tune of 12,274 aircraft in just over two years! The two fighters were both powered by the mighty R-2800 Double Wasp engine, both were large and strong, and both were extremely capable and popular. The main snag with the Hellcat was night vision, and the first NF version, the F6F-3E, was a conversion of the ordinary -3 with APS-4, red cockpit lighting and no curved Plexiglas windscreen in front of the bullet-proof front panel. These eighteen conversions were followed by 149 F6F-3Ns with APS-6, a radio altimeter and IFF, and finally by 1,434 F6F-5Ns (one of which had a scanner pod on both wings) plus eighty designated Hellcat NF.II for the British Fleet Air Arm. It was an officer in the Royal Marines, Major ‘Skeets’ Harris, who almost singlehandedly set up a naval FIU, equipped with Fulmars, Fireflies and Hellcats. In the Pacific, partly because of their faster climb, a Navy F6F-5N squadron replaced an AAF P-61 squadron in the defence of Leyte in early 1945. The NF Hellcats and Corsairs were usually painted midnight blue or jet black, and operated by day as well as by night in defence of the fleet and shore bases, and in escorting Avengers and other attack aircraft. At the end of the war Vought produced the more powerful F4U-4E (APS-4) and -4N (APS-6), which served in Korea together with the final NF Corsair, the -5N, which had, among other things, a rearranged cockpit. Total F4U production (which did not end until 1952) was 12,571.
Creating an all-weather day and night air defence for the fleet was a task comparable in magnitude to that on land. Thanks to British help, the US Navy immediately had available VHF radio and IFF, and the Radiation Laboratory developed a superb GCI radar for Fleet carriers, and installed the prototype in the Lexington, which perpetuated the name of the earlier carrier lost a year previously in the Coral Sea. Designated SM, this radar had a large dish that scanned conically to give the azimuth, elevation, altitude and range of multiple targets. Production SM radars were made by General Electric, the first coming off the line at Syracuse in August 1943. But there was a serious delay in the 3-cm airborne radar, leading to a variety of temporary expedients. These included close GCI control of day fighters used at night, use of day fighters in conjunction with radar-directed searchlights, and even brief trials (in the United States) of searchlight-equipped night fighters reminiscent of the RAF’s Turbinlites. The only scheme that was widely used in combat service was to operate as a two- to four-aircraft team, with from one to three fighters co-operating with a larger aircraft fitted with radar. The usual sensor aircraft was either the Grumman TBF Avenger with ASV or the Douglas SBD Dauntless with Westinghouse ASB. The team would hold tight formation, if possible without showing any lights, while the big surveillance radar(s) of the fleet scanned the sky from horizon to horizon. Upon a hostile force being spotted, the GCI controller would vector the team towards it until the airborne radar, used in the air-to-air search mode, had gained a good contact. The observer watching his scope would then guide the formation until the Japanese aircraft were close enough to be seen visually. After combat, the aircraft returned to the carrier individually.
Though seemingly cumbersome, the system worked quite well, and was perpetuated for many years in an even more complex form in the hunter/killer teams used in anti-submarine warfare. American losses were minimal, though one early casualty was ‘Butch’ O’Hare, after whom the world’s busiest airport (Chicago) is named. Tommy Blackburn, one-time commander of the famed VF-17, the first operational Corsair squadron and later Head of Fighter Branch in BuAer in Washington, reached the opinion that this form of CAP (Combat Air Patrol) might even have been extended to include larger formations. Right up to the present day the ‘lead ship’ concept has remained competitive in attacks on surface targets, and in the Pacific war at night there was often the need to intercept Japanese night attacks by as many as thirty or forty aircraft. In the earliest such engagements the use of IFF had not yet been introduced into the Navy and Marine Corps. This made it imperative to recognize the target before opening fire, whereas after mid-1944 it was by no means unusual for pilots of AI-equipped Hellcats to open fire on a hostile aircraft they had not even seen, and to hit it first time. This was an authorized practice, whereas in the European conflict night-fighter crews were required to identify their targets visually – or at least to do their best to do so.