Main Menu

Parnall Pipit


The Pipit had its origins in Specification 21/26, issued by the Air Ministry's Directorate of Technical Development (DTD) on September 30, 1926. This called for a "Single Seater Fighter Ship--plane for use from HM ships", fitted with a land undercarriage which could be replaced by floats, and vice versa, within half an hour. The aircraft, which was to have an all-metal structure but could be fabric covered, was to be suitable for launch from a catapult and for taking-off from and alighting on the deck of an aircraft carrier. It was to have a good degree of positive stability about all axes in both configurations, and tail incidence had to be adjustable in flight to enable the aircraft to fly horizontally at all speeds without attention from the pilot.

It was to be "highly controllable" at all speeds, and especially close to stalling speed, with no tendency to 'hunt' in a steep dive. Control had to be adequate to stop an incipient spin when the machine was stalled. A high degree of manoeuvrability in the air and on the ground or water was desired, and it had to respond quickly to the controls while not being tiring to fly. The ailerons were to have the minimum of yawing effect.

As a seaplane, the machine was to have good static stability in the water, and when under tow or running under its own power it was to be stable about all axes at all speeds.

Engines specified were the air-cooled Bristol Mercury radial giving 550hp at 2,000rpm or the water-cooled Rolls-Royce Falcon X giving 480hp at 2,300rpm. The installation had to allow for rapid and easy removal of the engine. The cowling, which also had to be easily removable, had to be finished "to prevent the reflection of light which might betray the presence of the aircraft to the enemy or dazzle the pilot". A metal propeller was specified.

There was to be tankage for 74 gallons of fuel, plus an easily removed 20 gallon auxiliary tank and a gravity tank of sufficient capacity to allow half an hour's flight at full power at ground level. An 11 gallon oil tank was to be provided if the Mercury engine was used, or an 8.5 gallon oil tank and a 2.5gallon reserve water tank for the Falcon.

Alternative exhaust systems for day or night flying were required, and were to be easily changed. The night-flying system had to provide adequate silencing and flame damping, while the daytime system was to be "of minimum weight". Additional equipment to be carried during the acceptance flights weighed 5581b and included a Vickers 0.5in gun and 300 rounds, a Vickers 0.303in gun and 600 rounds, a rocket launching (R/L) tube and six bombs, and flotation gear. A second 0.303in gun with 600 rounds was to be provided for if the 0.5in gun was not available in time.

Minimum performance requirements with this load, using the Mercury, called for a horizontal speed of 132kt (152mph) at 10,000ft and a service ceiling of 23,000ft. With the Falcon X the figures were 127kt (146mph) at 10,000ft and 22,000ft. The length of run to take-off was not to exceed 47ft in a relative wind of 28kt (32mph), and the aircraft was to become airborne at a speed of 55mph when catapulted in still air. The suitability for launching from a catapult or alighting on the deck of an aircraft carrier was "of first importance", and the aircraft had to be capable of taking-off from a turret or cruiser platform.

For fighting, the pilot was to have the best possible view in all directions, and a good view forward and downwards was required for carrier landings. A clear, unobstructed view forward over the machine's centreline was needed to enable him to sight the fixed guns, the installation of which was to dispense with blast tubes. There was also to be provision for the fitting of a G.3 aerial camera as near to the sights as practicable, and standard clips were to be fitted to allow the new “light carrier” to be installed to carry four 20 lb bombs, sufficient clearance being provided to enable the bombs to be released in a very steep dive.

Despite the emphasis placed on the machine's naval use, it was stressed that: "The aircraft is to be designed primarily as a landplane fighter and qualities required for this work are not to be sacrificed in order to improve its characteristics when equipped with the float alighting gear". A padded head support was to be provided to prevent injury to the pilot during catapult launch acceleration.

A limit of 35ft was put on the wing span, the overall length was restricted to 23ft, the height was not to exceed 14ft 9in. Quick and easy removal and erection of the wings was specified, with the ability to remove the wing structure completely in ten minutes and replace it in fifteen minutes.

The contractor was required to provide a full-size mock-up of his proposed aircraft before constructional work was begun, to enable the Director of Technical Development to examine and approve the layout. This mock-up had to include "all parts and components which are likely to interfere with the all-round view of the pilot", and was to show the internal arrangement of the cockpit. Scale model floats for official water-tank tests were also to be provided, along with specimens of ribs, a section of wing, and a length of spar.

Tendering for this demanding specification were Armstrong Whitworth, which offered the AW XVI; Fairey, with the Flycatcher II; Gloster, which tendered the Gnatsnapper; Hawker, which offered the Hoopoe; Vickers, with a modified version of its Type 141 Scout; and George Pamall & Co, which had its head office at the Coliseum Works, Park Row, Bristol, and its aerodrome at Yate, in Gloucestershire.

Parnall's submission to Specification 21/26, designed by chief designer Harold Bolas, was to be powered by a supercharged Rolls-Royce Falcon X. Members of the Air Staff visited the factory at Bristol on February 25, 1927 to inspect the mock-up. They reported that the upper wing was of "rather thick section", and that the pilot's view forward was rather obscured by the wing centre section, despite the fact that its thickness had been reduced to considerably less than half that of the wings themselves. To improve the view, it was arranged that the pilot's seat be raised about two inches, and that the centre section be redesigned and cut away at the trailing edge.
View for deck-landing was described as good, and the wing-root trailing edges of the lower wings had been cut away to give an improved downwards view. There was provision for the R-T equipment in the fuselage, just aft of the cockpit, and it was operated by remote control. Apart from a few alterations, such as a slight rearrangement of the dashboard and adjustment of the position of oxygen connections, the cockpit layout was deemed satisfactory.

The armament of either two 0.303 Vickers guns or one 0.303 Vickers on the starboard side and one 0.5 gun on the port side was I positioned conventionally for manipulation level with the pilot's seat", and there was a good fighting view. The ammunition boxes and feed were satisfactory, as was the release mechanism for the four 20 lb bombs carried under-wing. Some criticism was raised regarding the lack of detail of intended dispositions of equipment on the mock-up.

F/L Kirk likewise reported that the forward and upward view was restricted, though it was good forward and downward. He added that the span and the length of the fuselage meant that there were several blind spots which "are not desirable in a fighting machine-, but that these defects could not be overcome without significant alteration to the design. He approved of the cutaways in the lower trailing edges, and of the position of the instruments as altered during the examination, but found the tail incidence wheel difficult to position. Kirk preferred a quadrant lever with 90' radius, arguing that it would be better for deck landings and faster-acting. This, was not agreed by everyone, but it was decided to fit either a lever or a very high-geared wheel.

Kirk found the size of the cockpit “somewhat small" for access or egress by a pilot wearing a leather coat and equipped with a parachute and a Perrins belt or fife-saving jacket. A fairing on each side of the cockpit and the position of the windscreen were thought to be unsuitable for deck landings, and the screen side pieces were extended two inches downwards and further forward and the fairing was cut away for two inches round the cockpit.

A response to Kirk's criticisms came from S/L Quintin Brand in April. He argued that the defects in the view were inherent in aircraft using water-cooled engines and having somewhat thick wing sections. He opposed the tail incidence lever because it took more space and was a bad obstruction. The wheel control was standard policy and gave greater leverage, and Parnall was providing a high-geared wheel with no more than 180' of movement and a hand grip on the rim.

On March 31, 1927 Parnall was awarded an experimental contract for a Single-Seat Fleet Fighter powered by a Falcon X(S) engine; and work proceeded on detail design. In October and November the duralumin wing ribs and spars for the new fighter, which had by now been named the Pipit, underwent strength tests. At first the rib was below the strength requirements under centre-of-pressure -forward conditions, but modification rectified this. Two aircraft were ordered to contract No. 747093/27 in January 1928.

The Air Ministry issued a corrigendurn to Specification 21/26 on February 21, 1928. The in-fine engine was changed from a Falcon X to a Falcon XI(S) rated at 480hp at 2,25Orpm at 10,000ft, though an unsupercharged engine with an 8:1 compression ratio was to be installed initially. The weights of both the Mercury and Falcon had risen considerably, and the existing tankage for the Falcon was altered to 68 gallons for fuel and five for oil. Military load was changed to 550 lb. The horizontal speed required at 10,000ft was increased to 146kt for both engines, and the service ceiling was increased to 28,500ft. The requirement for an R/L tube and six bombs was withdrawn, but there was now to be provision for identification and navigation lights, instrument lighting and wingtip flares.

To improve the pilot's view for landing, the seat was to be made adjustable over a 4in vertical range. Maximum permissible length of the aircraft was increased by three feet to 26ft.

This was exactly the length of the completed Pipit, which was a well-streamlined biplane of elegant appearance. The choice of the liquid-cooled Falcon in-line engine in preference to the air-cooled Mercury radial had enabled Bolas to give the fuselage exceptionally clean fines. Its rectangular main structure, built of square duralumin struts and stainless steel tubes, was faired to an oval cross-section. The engine cowling and forward fuselage, to a point just aft of the cockpit, consisted of quickly removable aluminium panels, and the fabric panels covering the rear fuselage were also easily detached for access and inspection. The Falcon, which drove a two-bladed metal propeller, was carried on a riveted stainless steel mounting, and was cooled by an underslung radiator which retracted into the fuselage behind the main undercarriage legs and worked in series with an auxiliary skin-type radiator in the top wing centre-section. The two main fuel tanks in the roots of the top wings had a total capacity of 68 gallons, while an auxiliary tank in the fuselage held another 18 gallons. Fuel was gravity-fed to the engine.

Two Vickers guns, synchronised to fire through the propeller arc, were positioned one on either side of the pilot at elbow level, with their barrels protruding into troughs in the fuselage sides which extended right to the nose of the aircraft. The cockpit was located behind and below the cut-out in the upper wing trailing edge, and was "particularly free from draught at all speeds". An adjustable seat and rudder bar were provided for the pilot, and the retractable radiator supplied sufficient heat to rule out the need for electrically heated clothing up to 20,000ft in winter conditions. In hot weather a controllable vent regulated cockpit temperatures.

The pilot's seat, control column, and the handwheel controls for the retractable radiator and for the variable gearing on aileron and elevator were all carried on a 'control unit' supported on a three-point suspension system. A Very pistol, the oxygen apparatus and other Service equipment was conveniently positioned around the cockpit. A faired headrest was provided. Immediately behind the pilot was a compartment housing a short-wave radio working on a fixed aerial. This was accessible by means of a panel in the starboard side, a sliding tray enabling the radio to be pulled out for examination.

A cross-axle vee undercarriage supported the fuselage, the front legs incorporated double-acting oleo shock absorbers. Palmer-type wheel brakes were servo-actuated from the rudder bar, and operated independently on either wheel to facilitate deck landings. A faired, sprung tailskid was provided. The wheeled undercarriage was quickly interchangeable with a seaplane chassis comprising a pair of long single-step duralumin floats which used the same fixing points as the land undercarriage. Although a pair of floats was built, the aircraft was apparently never tested with them.

The Pipit's staggered equal-span single-bay wings spanned 35ft. Each wing panel was built-up on two duralumin spars comprising three drawn sections riveted together. These carried the pressed duralumin ribs, and the structure was braced with stainless steel gap struts and drag struts. Constant-chord ailerons were fitted to the lower wings only, but occupied almost the entire trailing edge, extending from the wing-root cut-out almost to the wingtip. Each aileron was operated by two pushrods. The top centre section was attached to the fuselage by four faired steel struts and bracing wires. The top and bottom wings were braced by faired steel interplane 'N' struts and streamline bracing wires, and the surfaces were fabric covered. While the top planes had marked dihedral, the lower mainplanes had none.

A fin with a curved leading edge carried an unbalanced narrow-chord rudder which had a straight trailing edge, and to which was fitted a tail lamp bracket. The parallel-chord adjustable cantilever tailplane and unbalanced elevators were fabric-covered stainless steel and duralumin structures.

Serials N232 and N233 were allotted to the two Pipits, and the first machine originally appeared with its metal fuselage panels and all struts painted a dark colour, possibly grey, and the remainder of the airframe finished in aluminium dope.

On June 22, 1928 the first Pipit underwent a final Air Staff inspection at the manufacturer's aerodrome at Yate. Its “clean appearance" impressed the examiners, who reported that all views were good and the cockpit was comfortable. The instrument layout was "generally satisfactory". As far as armament was concerned, the fighting view was good and the cockpit was roomy and conveniently arranged, the guns being accessible and easily loaded. The separate ammunition boxes could be withdrawn without difficulty through a removable panel in the bottom of the fuselage. The Aldis sight, positioned centrally on the fuselage top decking, allowed a very clear view, and the ring and bead sight, offset to starboard, was suitably placed. A bomb release for the light bomb carrier beneath the port wing was conveniently situated forward on the port side of the cockpit.

A further report was submitted by S/L H R Malet of RAF Base Gosport to the Officer Commanding the station. He stated that the Pipit's narrow fuselage would allow a clear view of a carrier deck when landing, but thought that more than the two seat adjustments were needed. He found the throttle position "uncomfortable" because the levers were too close to the fuselage and the fitting was too far from the pilot, which would cause his left arm to become tired if he was to fly in formation for any length of time. This was being altered. The control column was satisfactory and the gun triggers were easy to operate, the tailplane incidence wheel was "OK”, and the radiator control and variable aileron and elevator gearing wheel were within easy reach for the average pilot. All of the instruments were clearly visible, and the Very pistol stowage was good. Malet found the radio stowage extremely convenient, and thought it a great improvement on existing methods.

Although he found the quick-removal fuselage panelling "very satisfactory", Malet felt that there was a slight danger that the locking devices could catch in the clothing of airmen handling the aircraft's tail, and suggested that they be doped over. He had been informed that the Pipit's rudder was rather small, and that a larger one was to be fitted.

The date of the Pipit's first flight is unknown, but it had apparently taken to the air either before or shortly after the final inspection, for it was reported in an Air Ministry minute dated July 23, 1928 that a larger rudder had been fitted to N232, and had proved more effective in flight. Photographs show that this rudder was of significantly greater chord than the original and slightly taller.

At Yate on September 6, de Havilland chief test pilot Hubert Broad, who freelanced his services to other manufacturers, made his first test flight in the Pipit, this one lasting one hour 30 minutes. It seems to have followed further and more extensive modifications to N232's tail surfaces which are visible in a series of official photographs. The unbalanced rudder was replaced by one of similar shape but having a large horn balance, and the top of the fin was removed to accommodate this. Likewise, the plain, unbalanced elevators were replaced by new elevators having square balance areas at their tips, the tailplane being cut away to allow for them.

At the same time the aircraft was fitted with a wooden propeller and the forward fuselage was either stripped to bare metal or repainted to match the rest of the airframe, save for a dark anti-glare panel on top of the nose forward of the cockpit. In addition, wooden vibration damping rods were fitted at the junctions of the flying and landing wires in the wing bays.

Unfortunately there is no record of the exact nature of Broad's tests or of his opinion of the aircraft, merely a bland record in his logbook that he flew it. On the 7th, after a further 20 minute test flight, Broad flew N232 to the Aeroplane and Armament Experimental Establishment (A&AEE) at Martlesham Heath for its official tests.
A&A.EE trials revealed a number of shortcomings in the Pipit's handling qualities. Its rudder control was "rather weak", and while the aileron control was described as “fairly good in general", use of the control at very high speeds resulted in such flexing of the wings that the vibration damping rods connecting the lift and anti-lift bracing wires were broken in flight. Elevator control was very heavy, and the machine was longitudinally unstable, a large force being necessary to pull it out of a dive. It was also reported that the aircraft had “a strange tendency to drop its right wing” which increased with higher forward speed.

Some of the foregoing information was obtained during a flight on September 20, 1928, when Pipit N232 was completely written-off in a crash while being flown by the accomplished Service test pilot S/L Jack Noakes. The circumstances of this accident are controversial, and as the official report by the Accidents Investigation Branch, No E.3, has not been traced, one can only quote from the contradictory available evidence.

A preliminary A&AEE Report dated September 22, 1928 states that Noakes was carrying-out a moderate dive at about 140mph and at approximately 2,000ft, while investigating the aircraft's tendency to drop its starboard wing at high speeds. At about 1,500 to 1,200ft he heard a loud bang, and, looking back, saw that the port tailplane had apparently folded up and backwards. He did not see the starboard side. Noakes unfastened his seat belt to bale out, but found that the Pipit had settled into a slow glide, and that he could "just keep it in control and prevent it stalling" by pressing the control column hard against the dashboard. There also appeared to be "a certain amount of rudder control".

According to the report, Noakes thought that the machine's trim would enable him to flatten it out, so he decided to attempt to land on the rough ground of Martlesham Heath's speed course, which was straight ahead. On approaching the ground and releasing his pressure on the stick, however, he found that it moved "as if disconnected" and had no effect at all. Consequently the Pipit struck the ground heavily at a flat angle and, after sweeping a large quantity of heather and bracken before it, turned over; apparently making 1½  somersaults before coming to rest on its back. Noakes was thrown out in the first somersault and suffered a fractured neck, and was laid up for some time.

The accident prevented Noakes from continuing the testing of the Beardmore Inflexible all-metal three-engined monoplane, and nullified his selection as the pilot to fly the Fairey Long Range Monoplane from Cranwell to the Cape in an attempt to gain Britain the record for the longest-duration unrefuelled flight.

Noakes' own recollections of the crash, which he recorded in a letter in 1990 (without having seen the A&A-EE report), differ in several details. He wrote:
“After preliminaries I took the machine to the air in the normal way and did gentle manoeuvres - nothing severe - and found it was acting all right when suddenly there was a slight bang and the machine stalled. Knowing that something had gone wrong and that I was momentarily out of control, I released my safety belt anticipating the need to jump. However, believing that I still had control I proceeded to land and thought there would be no difficulty in flattening out on reaching the ground, but on doing the landing manoeuvre, by pulling the stick back slightly, nothing happened and the machine went into the ground at a normal landing angle. She did a somersault and catapulted me out”.

After reading the A&A-EE report, of which he had no prior knowledge, Noakes disputed the fact that he was flying at some 140mph, and said that he was "quite unable in the circumstances to see the rear of the plane following the bang". He also said that he had full control of the rudder, rather than a “certain amount", and that when landing he did not fly "straight ahead", but had to manoeuvre a half turn. He had never heard of the Pipit dropping its right wing when flown at high speeds.

Noakes said that the whole machine was transported to the Inspection Department, and that during the inspection it was found that rivets used in the construction of the tailplane had sheared completely in half. "The face of these broken rivets", he stated, were freshly painted. Thus, the shearing effect had occurred in the workshops of the Parnall firm by being put in a vice and twisted, causing this failure in the leading edge of the (tailplane) main spar which had broken downwards. Therefore, the tailplane became twisted prior to assembly in an effort to remedy this fault. In a later letter Noakes said that, if he remembered correctly, “...the Parnall company failed to meet the delivery date for testing. They were given an extended time and the whole thing became a rushed job, which might have contributed to the failure".

In the absence of the official report giving the findings of the Accident Investigation Branch, it is impossible to assess these remarks or comment upon the events which might have led to the accident. However, the Air Ministry asked Parnall to submit a tailplane rib for strength tests by the Royal Aircraft Establishment (RAE) at Farnborough, and this was tested on September 27 and found satisfactory. On October 17 the Air Ministry instructed the RAE to conduct wind tunnel tests on a 1/8 full scale model of the Pipit tailplane supplied by the manufacturer, and these took place in the seven-foot tunnel during the week ending November 3. The tailplane tested was of the balanced type latterly fitted to N232, and its lift and elevator hinge moments were measured over an incidence range of 0 deg to 6.6 deg and an elevator range of about ± 20 deg.

Meanwhile, the Pipit's designer, Harold Bolas, was conducting his own mathematical investigation into the accident. His study, entitled Tail Flutter - a New Theory, was published in the first issue of the new journal Aircraft Engineering, dated March 1929. Bolas opened his analysis with the comment: "The stresses existing in the structure of a tailplane, in varying conditions of flight, are possibly more a matter of doubt than obtains for any other part of an aeroplane structure". "It therefore seems desirable", wrote Bolas, "in view of the greatly increased speed capabilities of modern aircraft, to attempt to extend our knowledge of the subject. The above considerations, and the fact that there have recently occurred more than one case of tailplane failure at high speed, have prompted the writer to investigate the matter from a somewhat different angle".

Bolas argued that the majority of the load on a tailplane of symmetrical section was carried by the front spar, and that there was therefore a tendency for the aerodynamic load to deflect the front spar to a greater extent than the back spar. This produced increased incidence and, in turn, still greater loading. These deformations would cause a change of trim which the pilot could correct either by the application of elevator or by adjusting the tailplane incidence, though he would remain unaware of any distortion. However, if the rate of increase of aerodynamic load with deflection exceeded the rate of increase of elastic force in the spars, the tailplane would continue to distort until it eventually collapsed.

After presenting 5½ pages of mathematics, Bolas concluded that any normal tail became structurally unstable at some critical speed, although no danger would arise if that speed was beyond the ultimate diving capabilities of the aircraft. In general, however, fracture would take place at a speed less than the critical speed, because the front spar would reach its stress limits first. Any oscillations set up by the slipstream or from other causes would reduce the critical speed still further. Finally, Bolas reasoned that the critical speeds for a tail provided with stays were "considerably higher" than those for a similar cantilever tail.

Bolas' paper was studied by the staff of the RAES Airworthiness Department before publication. The Materials Testing Department had carried out a number of ad hoc tests on the Pipit tailplane, including the measurement of deflections under loads along the span and variously distributed between the spars. In general the RAE agreed with Bolas' figures, though it was thought that: "Some of the underlying assumptions need further consideration".

While these investigations were underway, Parnall had been completing the second Pipit, N233, modifying the design in the light of lessons learnt from the first aircraft. It was previously believed that the second aircraft was given the same serial as the first. It now seems that this was not the case, and that it was correctly marked as N233 from the outset.

Pipit N233, which was powered by the 520hp Rolls-Royce F.XIIS engine, had Frise-type ailerons on both the top and bottom wings, to give a greater surface area, and they were apparently linked by struts. Its tailplane was braced by two struts on each side, and the elevators were larger and had a shielded balance coupled with a static balance. Although the fin was the same as that of N232 after the horn-balanced rudder had been fitted, the rudder was larger and more curvaceous, without the straight trailing edge, though it retained the horn balance. The centre of its area was raised appreciably. Palmer-type wheel brakes with "a type of servo actuation" were fitted, and both anti-lift wires in the wing bracing were brought to the front spar of the planes, in line with the front interplane strut.

Once again, Hubert Broad was called in to undertake test flights. His logbook records that he made his first flight in what he calls the "Parnall Pipit II” on January 17, 1929, though it is not known if this was N233's first flight. This 35 minute flight was followed by one lasting an hour on the 19th, and one of 11/2 hours the next day. Perhaps some minor modifications were required, for Broad did not fly N233 again until February 3, when he made a 45 minute test flight. On February 4 he made another flight lasting 30 minutes, and this was his last flight in the aircraft. Once again, we have no further details of the nature of the tests or of Broad's opinion of the Pipit.

On February 6, a pilot from the A&AEE tested N233 for its handling qualities at Parnall's Yate aerodrome, making two flights. The first of these lasted 45 minutes, and the pilot's first impression was of "increased and better feel of control on the ailerons and elevator, but that the rudder was heavy, ineffective and out of harmony with the other controls". These remarks suggest that this pilot had also flown N232 during its brief spell at Martlesham, as he was obviously able to compare the handling of the two aircraft.

He found the aileron control light throughout the speed range. There was increased control, and at the same time it was much lighter in action and was "positive and smooth". In fact, with the variable control in the low gear position it was "almost too fight", and there was a proportionate loss of control. With the gear set about halfway the control was "very satisfactory for all-round manoeuvres including aerobatics". However, there was still some flexing of the mainplanes if the ailerons were moved rather quickly when flying at over 130kt (150mph), and there was evidence of some slackness in the anti-lift wires as the control was moved over, though this disappeared as the control was returned to normal.

The new elevators with both static and aerodynamic balance were quite fight and positive in action. Although they did not stiffen up abnormally when pulling out of a steep dive at 180kt, they became very heavy approaching the top of a loop, and the pilot had the impression of "having to pull the machine over against its own fore and aft stability". There was a "vast improvement", but the elevators were not quite in harmony with the ailerons in their "feel". At speeds near the stall there was a lack of elevator control, and this was particularly noticeable when landing. After the aircraft had been flattened out, elevator control was lost at the last moment, making it difficult to get the tail right down.

The pilot's first impression of the rudder was that it was "heavy and sluggish in action compared with the other controls", and "insufficient in action'.' at high speeds. After a further flight he realised that the control stiffened up so much as the speed increased that, at 140kt (160mph), the rudder could be moved through only about 15O of its travel, allowing only a very small yawing movement to be given to the machine. At 80kt (92mph) full travel was obtained and the aircraft was "fairly responsive", but the rudder was decidedly heavier than the other controls. Although there was no sense of 'hunting', the rudder control was jerky owing to the high foot pressure required to move it, even at low speeds.

For his second flight, which lasted 20 minutes, the pilot moved the elevators to the second of the three available incidence positions allowing different degrees of aerodynamic balance. Although they felt "a little fighter", the feel on the top of a loop was much the same, and while they seemed to harmonize more with the ailerons, there was still an impression of a lack of control at low speeds. All normal aerobatics, including loops, slow rolls, etc, were "quite easily carried out", but there was a different "feel" in the control on the change over from the ailerons to the elevators in a slow roll, and the rudder was quite out of balance the whole time. At high speeds the pilot found it "almost impossible to move it at all".

In summary, the controls were "fairly satisfactory". Aileron control was good; the elevators appeared to require a little more surface and needed to be lightened, though they were quite light and positive and it was now quite easy to pull the aircraft out of a steep dive; but the rudder was much too heavy and slow in action and out of harmony with the other controls. It was suggested that a rudder with a shielded balance might overcome the difficulty.

It was planned to carry out further tests after alterations had been made to the elevators and rudder, but an Air Ministry signal then put a stop to all flying with the Rolls-Royce F.XII engine. It has not been possible to ascertain the reason for this ban, but it seems to have been fairly short-lived, for A&AEE test pilot S/L S L G Poppy Pope went to Yate on Sunday February 24, 1929 to test N233 following the fitting of new control surfaces. Unfortunately, the precise nature of the modifications is unknown, but they presumably concerned the rudder and elevators.

Once again the official Accidents Investigation Branch report, No E.4, cannot be traced, but luckily we have Pope's own first-hand account of what ensued, written only a few days after the event. This takes the form of a letter to Irving Air Chute of Great Britain Ltd, applying for membership of the Caterpillar Club, which was exclusive to those who had saved their lives in a genuine emergency descent, using an Irvin-type parachute "in circumstances where they had no intention of jumping at the commencement of the flight". Pope's account, on RAF Martlesham Heath headed paper, reads as follows:

“On Sunday February 24, 1929, about midday, I ascended on a new type high-performance single-seater fighter to test some new control surfaces. As this type of machine had suffered a previous structural failure of the tail in flight, I got as high as I could before I started to put a strain on the controls. Owing to clouds, however, I could not get above 3,900ft. Having carried out trials, at the end of which I was usually about 1, 000ft, I decided to land as the rudder was not satisfactory, and in a final dive when the speed reached registered 170 knots on the ASI, the rudder fluttered and fin and rudder blew off”.
“On looking over my shoulder I saw these in the air as they blew away and, realising that I was certainly under 1,000ft, I knew I should have to act quickly to even get out at all. I had enough control to pull back the elevator and put the machine in a climbing angle, but with the engine off. I had difficulty in getting my Sutton Harness straps undone. This was mostly my own fault. On looking for the pin to release these straps, I found I could not get my head forward enough to see it, as it was buried in my chest. I looked for the string holding the pin to the harness. I pulled at, this string hard and for what seemed a long time with no result, until at last I found I had pulled off the writing pad which was strapped to my knee (and on which I made notes of the tests) and to which was tied a pencil. I had been pulling the wrong string.”
“I remember now realising that I was on my back (the machine had done a half roll), with the nose of the machine coming down. I at last got the pin out, but had difficulty with the straps, which, being bran' new, were stiff on the stub pin and did not wish to come apart. Having picked them off, I half pushed and fell out, as the machine had now nearly completed its roll to a normal position nose down. I can distinctly remember seeing the ground looking very close, and thinking: 'Well, it's too late now even for an Irving (sic) Parachute to save me, but it may 'deaden the bump'. As I fell out and realised I was clear, I fear I did not wait to count one, two, three or four, but looked at once for the ring. This, being the old type ring, had in my hurried fumbles been knocked out of its sheath and it was hanging by my leg. Got it and pulled hard and fast, and it seemed to me that, before I had pulled it right out, there was a bump. I thought 'The machine has hit me', and looked up, but, lo and behold, there was the parachute open, and I saw the machine diving away.”

The wreck of the Pipit was found on a railway embankment at Westerleigh, half a mile away.

An extract from the official accident report was later published in ARC Report and Memoranda No 1247 by W J Duncan and A R Collar, entitled Tail Flutter of a Particular Aeroplane, which was entirely devoted to an analysis of N233's accident. It says:

The pilot .. put the aeroplane into a dive but this time he held the rudder bar stationary. Up to a speed of 140-150 knots the machine behaved normally, but almost as soon as the pilot opened up the engine a little more, in order to gain further speed, the rudder started to flutter very violently.

At a meeting of the ARC's Accident Investigation Sub-Committee at the Air Ministry on April 4, 1929, it was pointed out that the Pipit's rudder post was a long, unsupported tube. The pilot had reported that the rudder was very sensitive to small movements but quite ineffective for large movements. During the flight on which the accident had occurred, Pope had reached 4,000ft and had proceeded to test the rudder, finding it unsatisfactory. There were three marked engine vibration speeds at 80-90, 110-120 and 140kt (90-100, 120-130 and 160mph).
During a test of the rudder for 'hunting', a slow oscillation was observed, and this was followed on a second trial by a sudden development of violent flutter, when the rudder broke off sideways. It was thought that the method of taking the control wire to the rudder might permit considerable slackness, and it was suggested that the rudder's ineffectiveness at large angles might have been due to flexure of the body as a whole. The pilot had been unable to keep up the nose of the aircraft in steeply-banked turns. As it was believed that a lack of torsional stiffness in the fuselage could have been a contributory factor, the committee agreed that the fuselage aft of the pilot's scat and the tail unit should be reconstructed and tested at the RAE. The remaining tail portions of both Pipits were sent to the RAE, but there was not enough of either fuselage available (the last two bays only) to enable torsional stiffness to be measured.

In August 1929 the battered rudder of N233 was straightened and repaired at the RAE, special care being taken to avoid any alteration in the disposition of weight of the material in the rudder. The Pipit rudder and tailplane were then used in experiments to investigate the flutter of tail units, using a light rectangular frame balancing on knife-edges. When the rudder was received it had a tail lamp bracket of "considerable weight" attached to its trailing edge, and the RAE decided to test the rudder with and without the bracket and also with a standard tail lamp attached to the bracket. It was found that the addition of the lamp and bracket more than doubled the moment of inertia of the rudder about an axis parallel to the hinge line.

In addition to these tests, the RAE also conducted extensive tests with a wooden 1/8 model of N233 in the National Physical Laboratory's 4' windtunnel (these formed the subject of R&M 1247). It was concluded that the accident was due to tail flutter , essentially of the rudder-fuselage type", in which fuselage twist was accompanied by rudder oscillation. This could be avoided by having the rudder symmetrically positioned about the torsional axis of the fuselage, and it is noteworthy that the Pipit's rudder was entirely above this axis. It was also stated that flutter would not have occurred had the rudder been dynamically balanced, and that, in the case of the Pipit, it would have been avoided if the tail lamp and bracket had not been fitted to the rudder.
Because of the Pipit's susceptibility to tail flutter, both prototypes had been lost and two courageous test pilots had barely escaped with their lives. The aircraft had been designed at a time when officialdom and manufacturers were turning from traditional wooden airframes to all-metal structures, and aircraft performance, was being pushed to new extremes. Flutter had always been a problem, but with aircraft being built using new structural techniques and flying at speeds which imposed greater stresses than ever before, it posed major problems in aeroplane design.

Alas, it was all too late for the Pipit, which was taken off the Secret List and dismissed as "quite unsuitable as a Fleet Fighter". Apart from the serious flutter problems, the machine's poor handling qualities had not been completely resolved at the time of the second accident, and any hopes Parnall may have had of winning a production contract and becoming a front-rank manufacturer of naval aeroplanes crashed with the Pipit.

Engine: Rolls-Royce F.XI, 495 hp
Wingspan: 35ft 0in
Chord: 6ft 0in
Wing area: 361 sq ft
Length: (landplane) 26ft 0in, (seaplane) 28ft 6in
Height: (landplane) 10ft 5.5in, (seaplane) 12ft 0in
Empty wt: 3,050 lb
Loaded wt: 3,980 lb
Wing loading: 11 lb/sq ft
Power loading: 8 lb/hp
Maximum speed, ground level: 168mph, (With F.XIIS 188mph)
Maximum speed at 3,000ft: 173mph
Maximum speed at 10,000ft: 168mph. (Cruising speed with F.XIIS 168mph)
Landing speed: 55mph
Rate of climb at 3,000ft: 1,600ft/min
Time to 10,000ft: 7.5min
Armament: Two fixed forward-firing 0.303in Vickers machine guns, each with 600 rounds
Bomb load: four 201b Cooper bombs beneath the port wing.

Engine: Rolls-Royce F.XIIS, 520 hp
Wingspan: 35ft 0in
Chord: 6ft 0in
Wing area: 361 sq ft
Length: (landplane) 26ft 0in, (seaplane) 28ft 6in
Height: (landplane) 10ft 5.5in, (seaplane) 12ft 0in
Empty wt: 3,050 lb
Loaded wt: 3,980 lb
Wing loading: 11 lb/sq ft
Power loading: 8 lb/hp
Maximum speed, ground level: 188mph
Maximum speed at 3,000ft: 173mph
Maximum speed at 10,000ft: 168mph
Landing speed: 55mph
Rate of climb at 3,000ft: 1,600ft/min
Time to 10,000ft: 7.5min
Armament: Two fixed forward-firing 0.303in Vickers machine guns, each with 600 rounds
Bomb load: four 201b Cooper bombs beneath the port wing.


Parnall Pipit



Copyright © 2021 all-aero. All Rights Reserved.
Joomla! is Free Software released under the GNU General Public License.