The later variants of the P-40 were given the company designation Curtiss Model 87 and were named Kittyhawk by the RAF and Commonwealth Air Forces. They saw widespread action in the Middle East and Pacific theatres. Although the Kittyhawk looked generally similar to the earlier Tomahawk, it had a revised engine cowling housing an up-rated Allison and a remodelled canopy. The armament comprised wing-mounted guns only (the nose guns having been deleted). The first twenty-two aircraft delivered to the RAF were fitted with four 0.50-in machine-guns, but all subsequent aircraft were fitted with six 0.50-in guns. The Kittyhawk I was the equivalent of the American P-40D and this was followed by the Kittyhawk IA (P-40E), Kittyhawk III (P-40K and M) and the Kittyhawk IV (P-40N). The aircraft entered USAAF service in July 1941 and deliveries to the RAF commenced later the same year.
The Kittyhawk introduced the ‘F’ model Allison engine, in place of the ‘C’ model as used in the Tomahawk. Although the take-off rating of the ‘F’ was little different at 1150 hp, it could deliver 1470 hp for combat, albeit limited to a maximum of five minutes. The spur reduction gear for the propeller was mounted externally on all P-40s powered by the ‘F’ model Allison (instead of internally as on the Tomahawk). This reduced the length of the aircraft by 6 in and raised the thrust line, allowing the undercarriage to be shortened slightly. The radiator was moved further forward under the nose and the fuselage cross-section was reduced, reprofiling of the rear fuselage allowing better lookout to the rear and below. Provision was also made for a ventral tank to be carried under the fuselage or a 500-lb bomb. Although it offered more power and slightly better altitude performance, this was offset by an increase of approximately 1000 lb in gross weight, with a combat-ready Kittyhawk IA tipping the scales at 8400 lb. As a result its take-off and climb performance was actually worse than that of the Tomahawk, and its manoeuvrability (never a strong point with the P-40) was also inferior.
Boscombe Down received its first Kittyhawks in February 1942 and trials were carried out over the next six months using AK572 and AL229. In contrast with the Tomahawk tested previously, there was much more room in the cockpit, with ample seat adjustment for height. Once again, however, the control column was set too far away from the pilot, which made it difficult to obtain full forward movement. The Sutton harness was satisfactory and a release lever was provided on the left of the seat to permit the pilot to lean forward. The view forward was similar to the Tomahawk and to the rear was good through transparent side panels, but the rear view mirror mounted above the windscreen was of little use.
The Kittyhawk was fitted with trimming controls to the elevator, rudder and ailerons, the trim tab controls for the elevator and rudder being similar to the Tomahawk. The aileron trim tab was fitted to the port wing and was actuated by an electric motor controlled by a switch on the electric control panel. This switch was not particularly easy to operate, but as the changes in lateral trim were so small in flight, it was used very infrequently. Operation of the trimmer was satisfactory with no tendency to slip. However, overall, it was felt that the provision of aileron trim was an unnecessary complication and a fixed trim tab that could be set on the ground would have been sufficient. Engine throttle, mixture and propeller controls were in a box on the pilot’s left. However, with no friction damper, the throttle lever tended to slip back. If adjustments were made on the ground to stop the throttle from slipping when in the air, the propeller control was then almost immovable.
The electric propeller control was that normally fitted on Curtiss aircrews. A master safety switch, with ON and OFF positions, was fitted on the electric control panel. This switch was normally kept in the ON position, except in an emergency. On the right of this switch was a three-position selector for selecting either manual or automatic operation of the propeller. When this switch was moved up to ‘auto’, engine rpm was controlled through a governor unit and a change of rpm was obtained by moving the propeller control lever located beside the throttle lever. When the switch was moved to either of the two down positions, then ‘manual’ operation was obtained. When the switch was moved to the bottom left position, propeller pitch decreased and rpm increased, whilst if it was moved to bottom right there was an increase in pitch and decrease in rpm. The circuits for the ‘auto’ and ‘manual’ positions were independent, and in the event of failure of the governor, the pitch of the propeller could be changed by the ‘manual’ switch.
The fuel cock and flap selector lever were in the same positions as on the Tomahawk. They elicited the same adverse comments as regards their inconvenient location on the left-hand side of the cockpit. In the event of failure of the electric pump the flaps could be raised or lowered by a hand-operated hydraulic pump to the right of the pilot. The undercarriage selector lever was the same as that used on the Tomahawk and the operating procedure was identical. Once again, the emergency hand pump could be used should the electric system fail, the pilot first having to select the required direction of motion of the undercarriage. Should the main hydraulics fail, an emergency system was provided. This consisted of a further hydraulic pump with two changeover cocks on the floor of the cabin. The emergency system operated via a different set of pipelines to those of the main hydraulic system, and would therefore operate when the latter had been punctured. It could only be used for lowering the main wheels; the tailwheel could not be lowered.
The layout of the instruments was generally satisfactory and no undue vibration was noted. The compass was fitted in the centre of the panel and was clearly visible. All the flying instruments were located above and to the left, there being no standard blind flying panel. The Directional Gyro and Artificial Horizon were on opposite sides of the gunsight mounting, which from the pilot’s point of view was not ideal. Engine instruments were located on the right of the panel and were conveniently grouped.
Handling trials were carried out with the aircraft at an all-up weight of 8480 lb at CG 26.5 in aft of datum (the CG range due to dissipation of load was from 20.9 in to 26.5 in aft of datum). The ailerons were discovered to be light and quick in response at all speeds up to maximum level speed. They were effective in level flight, and during climbs and glides, but there was deterioration at speeds close to the stall, although control remained satisfactory. Aileron control tended to become heavier with increase in speed, but it was only when diving at speeds above 400 mph IAS that any serious difficulty was encountered. By the time that the limiting dive speed of 460 mph IAS was reached, the ailerons were virtually immovable. At slow speeds there was a slight tendency for the ailerons to snatch. There was little change of lateral trim with speed, engine on or off, and as a result the aileron trimmer was little used.
Elevator control was moderately light and effective throughout the speed range, becoming heavier with increase in speed. Response was quick, although the aircraft’s positive stability at the CG position tested made movement in pitch appear a little heavy and sluggish. The elevator trimmer was effective and gave adequate trim for all conditions of flight.
The rudder was the heaviest of the three controls, particularly with engine on and with increase in speed. With engine off, it was still moderately heavy, but became light at landing speeds and near the stall. Despite this, the rudder was effective under all conditions of flight and the response was quick. Even though the rudder was rather heavy, there was no adverse effect on overall manoeuvrability, owing to the aircraft’s excellent aileron control. One aspect that did cause concern was a fairly large change of directional trim between engine on and off. This was important in view of the rudder’s heaviness because, although the rudder trimmer was effective, the gearing was rather too low and it was not possible to re-trim rapidly to cater for this change of trim.
Directional control of the Kittyhawk was looked at in some detail because of reports received from the USA that the rudder was liable to lock when displaced through more than two-thirds of its range of movement to the left (through about 20 degrees) with the engine on. It was found that rudder locking could occur, but only under abnormal conditions of flight. Most of the tests were made by doing a quarter slow roll to the right and then applying top rudder. When the rudder had been moved through two-thirds of its range, the extreme control heaviness was replaced by extreme lightness, although it did not flick over to the full position but could be moved between the two-thirds and full rudder positions in either direction with very little foot load. Although full rudder was held for a considerable time, there was no tendency for the aircraft to spin, but it tended to go into a sideslip before the nose would drop.
Other methods of inducing rudder lock were tried, and it was noted that it would occur whenever more than two-thirds left rudder was applied with engine on, irrespective of the position of the ailerons and elevator, or of the attitude of the aircraft. The rudder did not lock when the engine was throttled right back, or when the rudder was moved to the right. When locking had occurred, the rudder could be moved back to the two-thirds position quite easily. However, a very large foot load was required to move it beyond this point with engine on. If the engine was throttled back, centralisation was easy and recovery immediate. Although it was proved that rudder locking could occur, it was felt that it would not be encountered often and, even then, was not dangerous provided the method of recovery was known.
At the loadings tested, the Kittyhawk was directionally and longitudinally stable under all flight conditions. With the flaps and undercarriage up, there was little warning of an approaching stall, except for the high position of the nose and a tendency for the aircraft to yaw to the right. At the stall, which occurred with the control column central at a speed of 90 mph IAS, a shuddering was felt and this was followed by a drop of the nose. When the stick was pulled further aft, the left wing dropped sharply, as would occur before entry into a spin. Recovery was effected by moving the control column forward. With the flaps and undercarriage down, there was a tendency for the right wing to go down as speed was decreased below 83 mph IAS, but the aircraft could be kept level by coarse use of aileron at speeds as low as 80 mph IAS. Control forces to stall the aircraft were light. If the control column was pulled back at the stall, there was slight fore-and-aft pitching and the aircraft flicked to the right. Recovery once again was immediate when the elevator force was removed.
High-speed dives were carried out up to the limiting dive speed of 460 mph IAS and the maximum permitted engine speed of 3200 rpm as follows:
In dives 1 and 2, above 400 mph IAS the elevator and rudder forces needed to hold the aircraft in the dive were very heavy. At about 440–450 mph IAS, the pilot was unable to exert enough force on the controls to prevent it from coming out of the dive and yawing to the right. Prior to this, the aircraft was steady in the dive and could be held onto a target. The dives were continued into bumpy air at lower levels, but there was no instability or control surface vibration. Recovery was made by decreasing the forward pressure on the control column, although this had to be maintained to some extent to prevent the recovery becoming too rapid, which may have led to overstressing.
In the third dive, the elevator and rudder trimmers were used to reduce the force needed on the relevant controls as the limiting speed was reached. A small adjustment of the elevator trimmer was sufficient to reduce the force on the control column to a reasonable value, but the rudder had to be trimmed 4 divisions left (of the available 6¾ divisions) at the limiting speed of 460 mph IAS. Even so, a large foot load was required to hold the aircraft straight in the dive. Recovery was accomplished easily, and there was less chance of the aircraft coming out of the dive too quickly.
The optimum approach speed with the flaps and undercarriage down was the same as the Tomahawk – 100 mph IAS. The aircraft could be sideslipped but pilots found it difficult to maintain a steady rate of slip, as the nose tended to drop and speed increase. The landing was straightforward, with touchdown occurring at around 75 mph IAS. The handling characteristics in the case of a baulked landing were the same as for the Tomahawk.
Tests were also made with an overload tank fitted, which increased the weight to 8840 lb, but there was little difference in the way the aircraft handled. On take-off there was slightly more bucketing on rough ground and in the climb there was a slight tendency to wander in yaw, although directional stability was maintained. Stalling speeds with the tank were marginally higher at 92 mph IAS with the flaps and undercarriage up, and 82 mph IAS with the flaps and undercarriage down. Loops, slow rolls, climbing rolls and rolls off the top of loops in each direction were performed, the behaviour being similar to the clean configuration. The aircraft was dived to its limiting speed with the tank fitted (280 mph IAS) with trim set for full throttle level flight. The behaviour on recovery was similar to that at the lighter load at the same speed. The best approach speed with the tank fitted was higher at 110 mph IAS, the rate of sink becoming too rapid if a slower speed was attempted. Apart from this, the landing characteristics were the same as those at the lower weight.
Climb tests were made using AK572, the second aircraft in the first batch of Kittyhawk Is. As a very early four-gun model, it was not representative of aircraft that would be used operationally. It did not have an aerial mast or aerials fitted, nor did it have a rack for a bomb or overload tank. Its exhausts were individual stubs as distinct from the multi-fishtail ejectors that were fitted subsequently. The propeller was a Curtiss Electric of 11 ft diameter. Climbs were made using an initial speed of 145 mph IAS, which was the recommended best climb speed from testing carried out in the USA. The trials showed that there had been deterioration in climb performance compared with tests carried out on the earlier Tomahawk, due to the higher all-up weight, which was recorded at 8480 lb (1180 lb more than the Tomahawk).
The full throttle height was thus 11,400 ft (compared with 13,500 ft for the Tomahawk) and the greatest height recorded during the trials was 28,500 ft. It was estimated that the absolute ceiling would have been 29,900 ft.
Level-speed performance was measured from trials with several machines, including Kittyhawk IA ET573 in May 1943. By now, the take-off weight had crept up to 8650 lb, but compared with the Tomahawk, the maximum speed was improved slightly to 344 mph TAS at a full throttle height of 13,800 ft. All speed runs were made with the radiator cooling gills in the neutral position. The full test results were as follows:
The inadequacies of the Allison engine at altitude were recognised by Curtiss at an early stage in the development of the P-40. Shortly before the end of 1940, a Rolls-Royce Merlin 28 was fitted to the second production P-40D, which became known as the XP-40F. This led to the production of 1311 P-40Fs powered by a Packard-built Merlin V-1650-1 rated at 1300 hp for take-off and 1120 hp at 18,000 ft. In comparison with Allison-powered P-40s, the ‘F’ model could be easily recognised by the lack of a carburettor air intake on top of the engine cowling. A number of P-40Fs were supplied to the UK under the Lend-Lease scheme and were designated Kittyhawk II. One of these was FL220, which was tested at A&AEE in August 1942. The aircraft was armed with six 0.5-in machine-guns, the muzzles of which protruded about 3 in from the leading edge and were tape-bound during the course of the trials. Aerials were stretched from the fin to the wing tips and to the rear of the cockpit, but there was no aerial mast. IFF aerials were fitted, as was an external rear-view mirror and fittings for an overload tank. The take-off weight was 8910 lb.
Climbs were made at 160 mph IAS to 20,000 ft, reducing the speed by 2 mph per 1000 ft thereafter. The change from MS to FS supercharger gear was made at 13,000 ft and the engine speed was increased from 2850 rpm to 3000 rpm at 20,000 ft. The cooling gills were left fully open during all climbs. Testing showed a considerable improvement as the following results show (* denotes full throttle height in MS gear and ** in FS gear).
The service ceiling was considered to be 34,300 ft and the absolute ceiling 35,400 ft. Further trials showed a more moderate advance in level top speed.
Although the Kittyhawk II possessed superior performance to the Allison-powered variants, it did not see widespread use with the RAF and only 330 examples of the P-40F were delivered. This was largely due to the fact that the Kittyhawk was employed principally as a fighter-bomber and, as the difference in performance between the Allison and Merlin below 10,000 ft was only marginal, from an operational point of view there was little to be gained by any major changeover. To further reinforce this situation, the Kittyhawk III, deliveries of which commenced in mid 1942, was powered by the uprated Allison V-1710–81, which offered 1200 hp on take-off and 1125 hp at 17,300 ft.
Having given valiant service as a close air support fighter, the Kittyhawk was largely obsolescent by mid 1944, and former Desert Air Force aircraft operating in Italy were gradually replaced by the North American Mustang. It was to remain in first-line service with the Royal Australian Air Force until the end of the war. Indeed, the last aircraft to be lost by the RAAF in the Second World War was Kittyhawk A29-1161 of No. 80 Squadron, which was shot down by ground fire on 9 August 1945 at Samarinda in Borneo.