Following its deck landing assessment PS944 was retained by A&AEE for further handling trials which showed that attitudes were beginning to change within the testing establishments. The aircraft was evaluated at a loaded weight of 10,394 lbs and was subject to limiting speeds based upon a Mach number of 0.77. At the beginning of the trial, however, the limit was at the higher figure of 0.85M, but continuing tests using a Spitfire F.22 highlighted a violent and rapid pitching or ‘porpoising’ at a true Mach number of 0.77 (0.82 indicated) and so this lower figure was imposed. At 25–30,000 ft 0.77M represented an indicated airspeed of 345 mph (300 knots) rising to 520 mph (450 knots) at 5–10,000 ft. The reason for the test was to obtain sufficient knowledge of handling characteristics to decide whether the aircraft could be cleared for service use in what was now a post-war environment. The Seafire F.47 was put through a variety of manoeuvres including stalls, dives, baulked landings and a full range of aerobatic manoeuvres including loops, half rolls off the top of loops, rolls and climbing rolls.
Ground handling was straightforward and the aircraft was easily controlled in cross winds of up to 10 knots. The brakes were effective and the engine could be opened up to 2,800 rpm with +8 lbs/sq.in boost without them slipping. This power setting was also the maximum in which the aircraft could be run up against the chocks without the tail lifting. Turns on the ground could be made in calm air without brakes by application of rudder and bursts of power up to zero boost. Take off was easily accomplished using a power setting of 2,750 rpm and +12 lbs/sq.in boost and the rate of climb was good. It was during the climb, however, that the first criticism of the aircraft was made which showed that handling characteristics that had been deemed acceptable during the war would be coming in for much tighter scrutiny in the future.
When trimmed into the climb at 180 mph IAS (156 knots) the aircraft showed neutral longitudinal stability (stick free) and if disturbed it tended to diverge either to a dive or a stall. At the higher speed of 200 mph IAS (174 knots) the stability became positive but the post-war requirements were for an aircraft to have positive stability at all speeds. In all other conditions of flight the Seafire F.47 had excellent longitudinal stability, particularly on the glide with the flaps and undercarriage down. Certain aspects of the aircraft’s directional stability were also remarked upon. Generally the degree of directional stability was good and when the rudder was displaced and then released in any trimmed condition of flight the aircraft quickly returned to straight flight. However, during flat turns that introduced sideslip, the rudder tended to overbalance so that the rudder control force did not increase progressively as the angle of sideslip was increased up to the maximum obtainable. Again this was outside the current requirements and was considered to be unsatisfactory.
Dives were carried out in both ‘out of trim’ and ‘in trim’ conditions. With the aircraft trimmed to fly level at 300 mph IAS (260 knots) with climbing power at 12,000 ft, the speed was increased in a dive to 525 mph IAS (455 knots) at 5,000 ft when a push force of 30 lb on the stick and a left foot force of 15 lb on the rudder pedal was required to maintain the dive and keep the aircraft straight. On release of the stick in this dive the aircraft recovered with an accelerometer reading of 4.1g. The aircraft was also dived in trim from 20,000 ft to 7,000 ft up to 525 mph IAS. The foot force needed on the rudder pedal to generate 5 degrees of sideslip as measured on the directional gyro was measured at 50 lb and a pull force on the stick of 45 lb was needed to recover from the dive with an accelerometer reading of 4g. Stick force per ‘g’ at 20,000 ft was 2½ lbs/g at a speed of 325 mph IAS (282 knots), rising to 15 lbs/g at 430 mph IAS (374 knots). These figures were also not in accordance with the recommended values, being outside the upper and lower limits as laid down in the requirements of the time. As the Seafire F.47 was likely to be a first-line fighter for some time to come and would probably be flown by relatively inexperienced pilots in the event of a national emergency, A&AEE felt that it should be made to comply fully with the requirements, particularly with regard to its deficiencies in relation to longitudinal stability and the inconsistent forces that were encountered during use of the rudder.
The problems that the later versions of the Spitfire/Seafire had encountered during testing at high Mach numbers caused considerable concern at A&AEE and the rapid imposition of a flight restriction limiting high speed flight to a True Mach Number (TMN) of 0.77. High speed dives in a Spitfire F.22 had resulted in an extremely uncomfortable pitching motion setting in at 0.77 TMN (0.82 IMN or 420–430 mph IAS at 20,000 ft). Supermarine produced a modification (Seafire mod 971) to increase the critical Mach number and this took the form of 9 in of angle strip which were mounted above and below the trailing edge of each side of the elevator. The total depth of the angle was about 0.75 in. PS944 was tested at Boscombe Down without modification to ascertain the limiting Mach number and it was then sent to the manufacturers to have the strips fitted.
In its pre-modification state, PS944 showed a slight fore-and-aft oscillation at a TMN of 0.78-0.79 (0.83-0.83 IMN), however, this became violent with just a slight increase of speed, the top of the control column then moving 3–4 in forwards and backwards. After returning from Supermarine PS944 was tested with the angle strips in place at a normal service take off load, with full internal fuel tanks and wing combat tanks. The rear fuselage tank was used in the climb so that at the commencement of the dive, CG was approximately 7.5 in aft of datum (this represented the forward CG limit of the Seafire F.47). During the dive an auto observer recorded height and airspeed.
In the second set of dives with the modified elevators the aircraft behaved normally up to a TMN of 0.82 at 25,000 ft (0.87 IMN), the only trim change being a slight nose-down tendency which was easily trimmed out. With a slight increase in Mach number the ailerons snatched sharply and the aircraft started to roll to port. This was restrained by the ailerons which were effective, but without any appreciable increase in Mach number the aircraft started a pitching motion that rapidly became violent. A brief handling check was made to ascertain the modified aircraft’s handling at 5,000 ft with a full service load and the CG at 9.4 in aft of datum (the aftmost acceptable CG for the type). A qualitative assessment at moderate Mach numbers and at low speed showed the aircraft to have handling that was virtually unchanged from the unmodified case with no loss of stability. As the modification raised the maximum Mach number without any apparent deterioration in general handling characteristics, A&AEE recommended that mod 971 be applied retrospectively to all Seafire F.47 aircraft and that for modified aircraft the limiting speeds should be raised and based on a true Mach number of 0.81.
Following the criticism of the Seafire F.47 as regards a) its neutral longitudinal stability at the best climb speed, b) the variable foot loads experienced during sideslips and c) unacceptable stick force per ‘g’, Supermarine attempted to improve the situation by fitting a larger inertia weight of 15½ lbs to the elevator circuit of PS952. This produced a moment of +143 lb/ins about the elevator hinge (the corresponding figures for the standard weight were 6½ lbs and +66 lb/ins). In addition a spring tab was fitted to the starboard side of the elevator immediately outboard of the trimmer tab. These two modifications were intended to improve items (a) and (c) above, however, despite attempts to remedy item (b), it was eventually decided that the only way this aspect of directional control could be improved was for the fin to be re-modelled to include a dorsal fillet. As this would have required extensive modification work, the Seafire F.47’s inadequacy as regards rudder control had to be accepted, albeit reluctantly. PS952 was delivered to Boscombe Down in September 1948 but unfortunately it had not been tested prior to modification to compare it with PS944. Because of this the results of the subsequent trial were somewhat speculative in certain areas as it was quite possible that PS952 in its unmodified state was slightly less stable than PS944 had been [throughout its service life, the Seafire F.47 was inevitably made to carry more weight which tended to reduce longitudinal stability].
The outcome of the trial was extremely disappointing as the larger inertia weight and modified elevator gave no appreciable improvement in longitudinal stability on the climb at the best climbing speed compared with PS944, indeed there appeared to be some deterioration in longitudinal handling in various other respects. Even when using fuel from the rear tank first (as was normal practice) no improvement was apparent and the aircraft was difficult to trim acurately during climbs at any altitude. There was some improvement, however, in the stick force per ‘g’ characteristics, but at high altitudes and at aft CG, stick force per ‘g’ was still too low at low speeds and too high at high speeds. The eventual conclusion from the trial was that the advantages of the modifications, such as they were, did not warrant action being taken, particularly since production of the type was drawing to a close. [In view of the recent modification to the elevators to improve the Seafire F.47’s limiting Mach number, several dives were carried out in PS952 to evaluate its handling at high IMN. Despite the fact that mod 971 had been incorporated on the aircraft, the porpoising started at only 0.82 IMN (0.78 TMN) or slightly earlier than on PS944 before it was modified].
One A&AEE trial that did proceed more or less to plan was an investigation into the spinning characteristics of the Seafire F.47 at take off weights up to 10,885 lbs. In this condition wing combat tanks were carried (during tests at two lower weights they were not) and fuel in the rear tanks was used up during the climb to height to maintain CG as far as possible at 9.4 in aft of datum. As the trial took place in February 1948, mod 971 had been incorporated in the trials aircraft (PS944). In each spin the aircraft was allowed to make two complete turns (this was visually assessed by the pilot) before standard recovery action was taken consisting of the application of full opposite rudder, a pause, and then forward stick.
In general, the Seafire F.47 was reluctant to enter a spin, full pro-spin controls being required both before and during the spin, indeed pilots gained the impression that even a relaxation of the controls would have been sufficient encouragement for the aircraft to begin its recovery. There was no tendency for the aircraft to depart into a spin in the other direction after recovery and when it was flick rolled into a spin from a stalled turn entry there was no sign at any stage of an inverted spin being the result. Once the Seafire F.47 had settled down into the spin, the rate of rotation was fast with a steep nose-down attitude, although the longitudinal pitching that had been apparent on previous Seafire variants was still present. There was also a certain amount of lateral wallowing and both of these characteristics were accentuated with altitude. Spins to the right and left were similar, except that the aircraft was prone to dropping its port wing at the stall entry which meant that a spin to the left was entered rather more readily than one to the right.
With the wing combat tanks fitted two spins were made in each direction from 30,000 ft following straight stalls (engine off) and from stalls off turns. The entries into the turns were made at approximately 208 mph IAS (180 knots) and the stick was eased back until the aircraft stalled, the throttle then being closed. Entry to the spin was by a flick roll which was followed by a steep fast spin with considerable longitudinal pitching and lateral wallowing, with buffet being felt from the tail surfaces. From the straight stall the entry was slow and flat but after about half a turn the spin got faster and steeper, again with pitching and wallowing. The amount of height lost from the commencement of the spin to the achievement of full recovery was of the order of 6,000 ft [recovery from spins at reduced take off weight from 15,000 ft was in the order of 3–4,000 ft].
Towards the end of its development life the later Spitfire/Seafire variants carried significantly more fuel than the early aircraft and the Seafire F.47 was no exception. The main internal fuel tanks mounted forward of the cockpit held a total of 85 gallons which were supplemented by a rear fuselage tank of 33 gallons and two leading edge tanks with a combined total of 36 gallons. In addition, two underwing combat tanks could be carried, each holding 22½ gallons, and the aircraft was also capable of accommodating a 90-gallon overload tank on the fuselage centreline. Total fuel capacity was thus 289 gallons, sufficient for a range of around 1,000 miles. With full fuel and a full warload the Seafire F.47 weighed a prodigious 12,450 lbs, although it was capable of even more than this as it was cleared for gentle flying when overloaded to 12,900 lbs. At the latter weight wing loading was 52.95 lbs/sq.ft which compared with a wing loading of only 24.5 lbs/sq.ft on Spitfire I K9793 which had been the subject of trials at Boscombe Down in 1939. With such significant weight growth it was necessary to re-test the Seafire F.47 and in March 1949 VP463 was delivered to A&AEE for handling trials with and without the 90-gallon ‘airship type’ drop tank.
VP463 was flown at various loadings as follows–Loading 1–10,825 lbs (no external stores but combat tanks fitted); Loading 2–11,590 lbs (as loading 1 plus 90-gallon tank); Loading 3–12,645 lbs (as loading 1 plus 90-gallon tank and two 500-lb smoke bombs); Loading 4–12,035 lbs (as loading 1 plus 90-gallon tank and eight 60-lb rocket projectiles, but less combat tanks). The controls of the aircraft were standard Seafire F.47 and included 9.25 in of angle above and below the trailing edge of each half of the elevator. The inertia weight in the elevator control circuit was of 6½ lbs. Preliminary trials were carried out with no external stores (except for the combat tanks which were a standard fit) but these showed that the aircraft’s handling characteristics were very much worse in several respects than those of PS944 which had been used to clear the Seafire F.47 for service use. The trials were continued, however, so that any serious adverse features resulting from the carriage of the 90-gallon drop tank might be brought to light at an early stage.
Without external stores (loading 1) VP463 was climbed between 3,000 ft and 10,000 ft at the recommended climbing speed of 172 mph IAS (150 knots) at the maximum permissible climb power of 2,600 rpm and +9 lbs/sq.in boost. Accurate trimming was found to be very difficult but after doing so, when the stick was released, the aircraft diverged rapidly in either direction. Other speeds were then tried and the aircraft continued to be statically unstable (stick free) up to nearly 230 mph IAS (200 knots). After trimming in level flight at 270 mph IAS (235 knots) at 10,000 ft, normal turns were made with the speed constant at 252 mph IAS (220 knots). During turns in either direction, a very small pull force (estimated at 2 lb) was required for turns of 2g, while zero stick force was required for 3g turns.
From trimmed level flight at 20,000 ft with maximum climb power set, the aircraft was dived to a speed of 515 mph IAS (450 knots) at 5,000 ft. A push force of about 30 lb was required to maintain this speed and on release of the stick the aircraft recovered at 4–5g. In similar dives to 460 mph IAS the aircraft had to be restrained at 5½g after the stick had been released. Approximate measurements of the stick forces in pullouts at 5,000 ft from trimmed dives gave very low figures up to about 460 mph IAS (3–4 lb for a 3g accelerometer reading). At higher speeds the forces required in similar pullouts were rather higher. At speeds up to about 345 mph IAS (300 knots) the aircraft tightened into the pullout above approximately 3g. Overall, pilot’s impressions at this loading were that the aircraft was most unpleasant to fly, especially in bumpy weather, and that constant concentration would be needed, particularly when flying in cloud or at night.
At loading 2 with the 90-gallon tank in place VP463 exhibited longitudinal characteristics in the climb similar to the clean configuration case. The divergence in pitch was most marked after speed had been decreased by 23 mph IAS (20 knots) by pulling the control column sharply back and releasing. In this case the aircraft stalled after just two seconds. Flat turns were made to port and starboard at the best climbing speed by applying rudder and keeping the wings level with aileron. In each direction the foot force increased progressively to about three-quarter travel when it was rated as being heavy. At greater travel, however, the force suddenly lightened and the rudder overbalanced, a heavy foot load being required to unlock the rudder [rudder locking in flat turns was a known feature of the Seafire F.47; Supermarine had tried several modifications, but without success]. Both out-of-trim and in-trim dives were similar to loading 1 but on one occasion during an out-of-trim dive the acceleration built up so quickly on pulling out from the dive that 5½g was recorded before the pilot could apply a restraining force.
The aircraft was also tested in the glide at 110 mph IAS (95 knots) with undercarriage and flaps down. Accurate trimming was found to be relatively easy, the elevator trim setting being 1.3 divisions nose up. Speed was adjusted by approximately 17 mph IAS (15 knots) above and below the trimmed speed and in both situations the aircraft returned slowly to its orginal speed. In this condition, however, flat turns did not lead to rudder overbalance, neither was there any tendency for foot loads to vary. During stalls with undercarriage and flaps up, slight elevator buffet began at 110 mph IAS and increased down to the stall. Slight aileron twitching was also felt just before and at the stall, which occurred at 98 mph IAS (85 knots) and on occasions the ailerons could also snatch vigorously. Stalling characteristics were similar with undercarriage and flaps down, the stall occurring at 86 mph IAS (75 knots). Just before the stall there was marked snatching of the ailerons after which the port wing dropped followed by the nose. In both cases recovery was straightforward on easing the control column forward and applying opposite rudder when stalling in the landing configuration. Paradoxically, when the same tests as noted above were carried out at loadings 3 and 4, the aircraft’s handling characteristics actually showed a slight improvement, although stability on the climb was still negative.
The conclusions to the Boscombe Down report on VP463 highlighted the deterioration in longitudinal stability compared to that of PS944. At a CG only 0.1 in forward of the aft limit, PS944 had approximately neutral static stability (stick free) on the climb at normal climbing speed, whereas at a more forward CG VP463 became neutrally stable only at a speed 46 mph IAS (40 knots) higher. It was also emphasised that the longitudinal characteristics of PS944 at the time of its test were regarded by A&AEE as marginal. In comparing the two aircraft it was noted that the static friction of the elevator circuit of VP463 was approximately double that of PS944 (5–6 lbs against 3 lbs) and it was considered that this would lead to a reduction of stick free stability. Because of the basic defects in handling characteristics it was difficult to assess accurately the effect of the 90-gallon tank on handling. The main deficiencies of VP463 were summarised as follows
a. The aircraft was statically unstable stick free on the climb at the normal climbing speed; pilots considered the behaviour to be unpleasant and potentially dangerous.
b. Zero stick force was required to maintain 3g turns at 10,000 ft.
c. High normal accelerations were liable to develop on release of the stick in out-of-trim dives.
d. The aircraft tightened into the pullout beyond 3g accelerometer reading at speeds up to about 345 mph IAS (300 knots) at 10,000 ft (stick force per g values, based on tests at 3g, were low up to 460 mph IAS).
A&AEE were aware, via Supermarine, that other production Seafire F.47s were showing similar unsatisfactory characteristics so an improvement was clearly essential for all service aircraft. VP463 was later tested with a larger inertia weight in the elevator circuit and although this appeared to bring its longitudinal handling characteristics into line with those of PS944, as already noted, the experience with PS952 showed that this was not a complete cure for the aircraft’s instability in pitch.
In Fleet Air Arm service most Seafire F.47s were converted to the fighter-reconnaissance role as the FR.47. Two electrically-heated F.24 cameras (one vertical and one oblique) were mounted in the rear fuselage. The aperture for the vertical camera had a spring-loaded flap for protection against debris being thrown up when on the ground. Total production of the Seafire F.47 amounted to 89 which were built at South Marston. Serial batches were PS944-PS957, VP427-VP465, VP471-495, VR961-VR971.
While the Seafire F.47 continued to cause problems at Boscombe Down, it was taken on charge by No.804 Squadron at Ford in February 1948 and later saw active service in Malaya and Korea with No.800 Squadron aboard HMS Triumph. Following operations against communist terrorist targets in Malaya from October 1949 until February 1950, HMS Triumph was in Japanese waters when the communist forces of North Korea crossed over the 38th parallel on 25 June 1950, an action which led to conflict in the area for the next three years. Together with No.827 Squadron (Firefly I), the Seafire F.47s of No.800 Squadron carried out the first naval air strikes of the war on 3 July when rocket attacks were made on the airfield at Haeju. Further strikes were launched over the coming weeks together with Combat Air Patrols and other patrols on the lookout for enemy submarines. One aircraft (VP473) was lost on 28 July when it was shot down in error by a USAF B-29 Superfortress, however, the pilot baled out and was picked up by a US destroyer.
Further strikes saw the Seafires sink a number of small vessels during operations to blockade ports on the west coast of Korea, together with attacks on railway targets near Mokpo using rockets and 20 mm cannon. By mid September the Seafires were beginning to show signs of wear and tear with several aircraft having to be taken off operations due to damage to the external skin caused by over-stressing. With more modern equipment being sent to Korea in the shape of the Hawker Sea Furies of No.807 Squadron, the decision was taken to retire the Seafire F.47 from active service. The type continued to be used by No.1833 Squadron at Bramcote before being supplanted by the Sea Fury in 1954.
