A Test Pilot's View
Memories of P-39 Airacobra ‘Brooklyn BUM 2nd’

Memories of P-39 Airacobra ‘Brooklyn BUM 2nd’

“I’ve got this mathematical theory that each year seems to go by more quickly because it’s a smaller percentage of your overall life”, says Dave Southwood. “By the time you’re 50, a year is 2 percent of your life. Some things you experienced years ago feel like they happened just yesterday.”

He’s been flying Second World War piston-engine fighters for more than 30 years. The first was Charles Church’s P-51D Mustang Susy, way back in 1988. He’s since flown warbirds for numerous UK-based operators, among them Hurricanes and Spitfires of various marks, P-51C Mustang, Buchón, Bf 109G ‘Black 6’, Hawk 75, P-40B, P-40C, P-40F and P-40M from the Curtiss fighter lineage, FG-1D and F4U-5 Corsairs, and Grummans Wildcat and Hellcat. Above all, The Fighter Collection’s (TFC) Bell P-39Q Airacobra Brooklyn BUM 2nd does, he says, “stand out for being so distinctly different – those particular memories still feel recent”.



The Airacobra arrived at Duxford in summer 2004 and was assembled just in time to participate in the Flying Legends airshow. “When you know you’re going to be flying a new type for the first time it’s always a case of getting into the original manuals”, Southwood says, “but also finding somebody else who has flown the aircraft.” That person was Fighter Rebuilders’ Steve Hinton. He’d overseen the aircraft’s restoration at Chino, CA and had completed its test flight regime before Brooklyn BUM 2nd was crated up and shipped to England to join TFC.

“I hadn’t flown TFC’s two P-63 Kingcobras and so I asked him what the P-39 was like. He gave me this knowing look and said, ‘Oh, you’ll find it interesting when you fly it!’ That’s all he said on the matter, and it felt like a rather loaded statement. I later received a full brief from Pete Kynsey, who gave me similar words of caution. When I came to fly it, it was fascinating to see how Bell had effectively built the aircraft around the 37mm cannon. As a result of that, the P-39 had many peculiarities that set it apart from anything else I’d flown.”

The Airacobra owed its unorthodox features to the US Army Air Corps’ (USAAC) Circular Proposal X-609 specification for a new fighter that could fulfil “the tactical mission of interception and attack of hostile aircraft at high altitude”, incorporating a cannon as its main armament. The 37mm Oldsmobile T9 was the weapon of choice, and to accommodate this, Bell’s chief engineer Robert J. Woods strayed from the typical designs of the time and displaced the aircraft’s engine (a turbosupercharged Allison V-1710 , this version having both an exhaust gas driven turbocharger and a mechanically driven supercharger) amidships, at the centre of gravity, transmitting its power through a ten-foot drive shaft to a propeller gear box in the nose.



The single 37mm cannon was mounted in the nose, firing through the centre of the propeller hub for optimum accuracy and stability. A tricycle undercarriage – the first of its kind used in a fighter – was utilised to distribute weight accordingly. This arrangement also necessitated the displacement of fuel tanks to the wings, limiting the aircraft’s range. In the prototype, the coolant radiator was innovatively buried in the port wing root, being fed by an inlet in the wing leading edge with an outlet in the trailing edge. The oil cooler and intercooler were mounted in the aft fuselage and were fed by large sharp-edged scoops on the right and left side of the fuselage, respectively.

The XP-39 was characterised by these unconventional design features, and it looked fantastic – but there was a penalty for Bell’s ingenuity that limited the aircraft’s combat potential. The prototype, serial 38-326, made its maiden flight from Wright Field on 6 April 1939 with Bell’s test pilot James Taylor at the controls. Though the unarmed prototype reportedly achieved almost 390 mph during testing – albeit it must be acknowledged that extensive research has failed to uncover source material from the time to support this popular claim, and reports indicate that the engine’s performance was also limited by drive shaft vibration issues – it fell significantly short of the 400 mph specification set by the USAAC.

General Henry H. “Hap” Arnold, eager for a new fighter, hoped that the speed could be increased to over 400 mph by cleaning up the considerable drag created by its then-unique design. On 9 June 1939 he formally requested National Advisory Committee for Aeronautics’ (NACA, the forerunner to NASA) approval for immediate testing of the XP-39 in its Full-Scale Tunnel (FST) at Langley. There it was put through one of the most thorough test programmes ever conducted by NACA.



FST tests on the ‘as received’ XP-39 and then on a completely faired and sealed airframe revealed that 43 percent of the airframe drag resulted from the engine installation. NACA’s analysis showed that some 14 percent of the aircraft’s drag originated from the oil cooler alone, owing to its large rectangular scoop and oversized inlet duct. Similarly, the intercooler intake was also sharp-edged and oversized. The turbocharger, in a protruding installation in the belly of the aircraft, produced considerable drag. Lesser percentages of the overall drag component were found in the exhaust piping, carburettor inlet, and the radiator and intercooler designs. NACA also determined that an uprated 1,350 horsepower supercharged V-1710 engine could yield an airspeed as high as 429 mph at 20,000ft without the need for the turbocharger.

Subsequent redevelopments suggested by NACA led to the removal of the turbocharger and redesign of the engine installation in production P-39s. Notably, the oil cooler intake was relocated to the wing root, mirroring the leading-edge inlet feed of the radiator, with the oil cooler itself being mounted in the belly, underneath the engine and flanking the radiator. The carburettor air intake was displaced to the fuselage centreline aft of the canopy, while internal cooling ducts underwent a redesign. Modifications also provided for full nose-gear retraction and allowed the main gear to be fully submerged in coverless wheel wells in-flight. Reducing the height of the cockpit enclosure and fitment of a new windscreen and canopy tail section further reduced the drag coefficient.

The reality was that even after the drag reduction modifications, production model Airacobras carrying full armament never achieved 400 mph. The decision to omit the turbocharger and the unavailability of the uprated Allison engine recommended by NACA greatly hindered its high-altitude performance, limiting its career as an ‘interceptor’ with the western air forces. The engine placement also had its drawbacks; as mentioned, the absence of fuselage fuel tanks limited the aircraft’s range, particularly as compared to its US Army Air Force (USAAF) successors, though the addition of a belly tank sought to mitigate that issue. Though production P-39s served with the USAAF, Royal Air Force and Soviet Air Force amongst others, it was flying against the Luftwaffe over Russia that it was most successful.



For all the ingenuity of its design and in spite of a fine combat record, history has done the Airacobra a disservice. The type’s extensive development history could fill several weighty tomes, and that isn’t the intention of this article. Just three remain airworthy. The subject of this article has been operated by Lewis Air Legends in San Antonio, Texas since 2010, while the others – P-39Q 42-19597 with the Commemorative Air Force and P-39F 41-2175 with Jerry Yeagen’s Military Aviation Museum – are based in Texas and Virginia respectively.

P-39Q-6-BE serial number 42-19993 was manufactured at Bell’s factory in Buffalo, New York and delivered to the USAAF on 8 June 1943. It was shipped to the Pacific theatre on 16 August 1943 and assigned on arrival to the 87th Tactical Reconnaissance Squadron (TRS) of the 71st Tactical Reconnaissance Group, 5th Air Force. After assembly and a quick check flight, 42-19993 was flown almost coast-to-coast from Port Moresby to the front line airbase at Dobodura, crossing the Owen Stanley Mountain Range mid-country. The P-39 was assigned to Lt Peter A. McDermott, whose recollections of air combat in New Guinea and opinions of the Airacobra were published in Jim Busha and John Dibbs’ book The High Battleground in 2006. “I enjoyed flying almost every airplane I had been in”, he said, “but the P-39 was a real lousy airplane; simply put, it was a dog. But this dog could take a beating and still bring me home, minus some important pieces!

“I thought I finally made it to where the action was and now hopefully I would be given a real fighter plane to help win this war. When I saw the P-39s on the flight line at Port Moresby, I could have killed myself. I wanted to be in fighters, not the ‘dopey’ P-39. Adding insult to injury, New Guinea was covered by jungle and foliage. This wasn’t the barren wasteland of North Africa that we were trained to take pictures of.” McDermott named his P-39, a modified Q-6 variant fitted with a fuselage-mounted camera for tactical reconnaissance, Brooklyn BUM 2nd as an ode to his childhood favourite Brooklyn ‘Bum’ Dodgers baseball team. The names of his crew – Squadron Engineering Officer Capt James Parks, Crew Chief Sgt James W. O’Mara, and Armourer Sgt John Conway – were painted onto the side of the fuselage. Painted onto the doors by the squadron’s artist was a nod to McDermott’s Irish heritage: A large insignia comprising a green shamrock, a horse shoe and boxing gloves within a white circle. 



In this aircraft McDermott flew the vast majority of his 139 combat missions, including dive-bombing, ground attack and photo reconnaissance. Some of his recollections were published in the Flying Legends airshow programme, penned by Michael O’Leary: “I never saw any Japanese airplanes in the sky, only crashed ones in the jungles below. Our missions became so routine, shooting up the same targets and using the same tactics, that at times it became boring for me. On one mission, I had a little fun with my wingman. My wingman pulled into a Lufberry circle and I was right behind him as he set himself up for his gun run. When he made his turn in, I cut right under him and my prop was mere feet away from the belly of his P-39. I waited for him to fire and when I saw his tracers going off, I fired all my guns at once. My bullets whizzed underneath and out in front of him as he pulled straight up and began screaming over the R/T ‘Ack Ack, Ack Ack’. I couldn’t stop laughing at the shrill in his voice. When we landed, he damn near punched me out! Though McDermott was no advocate of the P-39, he conceded that the aircraft was well-suited to the air-to-ground combat role it adopted in the Pacific, and praised the devastating toll wrought by its armament. As he commented postwar, it served him well and always brought him home.

In September 1944, the 82nd TFS’s P-39s were replaced by P-40s, and its contingent of Airacobras were flown to Tadji where they were wheeled into the long grass of the airfield’s adjacent boneyard and cannibalised to keep the resident 110th TRS’s aircraft airworthy. There Brooklyn BUM 2nd languished until it was struck off charge on 21 May 1945. McDermott’s P-39 suffered an undignified end, having been stripped of its propeller, armament, radios, maintenance panels and other parts. The elements had also taken a heavy toll over the years; the airframe was severely corroded, and the fabric covering the control surfaces had long since rotted away.

It was in this dilapidated state that Charles Darby and Monty Armstrong discovered Brooklyn BUM 2nd‘s wreck in 1974. After recovery to New Zealand, it passed through several private owners in the southern hemisphere and USA before acquisition by TFC in 1994. Over the course of a complex ten-year restoration, Fighter Rebuilders tore down the heavily corroded structure to the smallest components, restoring what they could and replacing other parts where necessary (including parts from other P-39s TFC had recovered from the former Soviet Union), and returned the cockpit to stock condition. The completed aeroplane was painted in its original 87th TRS markings and flew for the first time on the evening of 17 June 2004.



Dave Southwood was one of several TFC pilots who flew Brooklyn BUM 2nd after it arrived at Duxford. The insights in this article are drawn from his recollections of flying the aeroplane over a brief period from 2007 to 2008, and with reference to his corresponding pilot’s notes. “There are people who have flown other P-39s far more extensively than me”, he says, “and they could surely write a thesis on flying it. The intent here isn’t to do that; it’s to share an overview of my experiences flying this particular P-39 for a brief period more than a decade ago – experiences I deem to be an incredible privilege courtesy of Stephen Grey’s generosity.

“First impressions were that the armament came at the expense of everything else. That really became apparent when you looked at the manuals and saw where the various systems were displaced to in relation to the armament and engine.” The cockpit’s position in the fuselage centre section made for an unusual method of access, and entrance was through an automobile type door on the right side of the cabin. “I’m 6’3” and you needed to be almost my height to easily clamber up onto the wing, but then I was almost too tall to actually get into the cockpit! There’s very little headroom, and I needed to take the cushion off the parachute and couldn’t wear a hard helmet in it, so it was very limited for back length. In and out through the right door, which would be jettisoned in the event of an emergency – the left door was for emergencies only.”

A précis of the cockpit layout identifies instrumentation split between an upper panel containing flight instruments, and a lower panel comprising engine instruments and other ancillary gauges. The battery, generator and magnetos were clustered to the left of the main instruments. Flying controls were of a conventional type; the stick was originally equipped with a squeeze-type trigger for fuselage and wing machine guns, and a push button for firing the cannon, while the rudder pedals were a standard arrangement with toe brakes. Aileron, elevator and rudder trim tab controls sat in a unit to the left of the pilot. An electric toggle switch on the left of the instrument panel controlled the landing gear.



The fuel system was controlled via the fuel booster pump switch on the left side of the instrument panel and a primer pump mounted to a unit at the base of the panel, while the fuel tanks were selected via the fuel selector valve on the floor to the left of the pilot. Engine controls consisted of the coolant shutter to the right of the pilot’s seat, and the oil shutter control handle mounted on the fuselage turn-over beam at the pilot’s right hand. A conventional quadrant to the pilot’s left featured the throttle, three-blade Aeroproducts propeller pitch control and mixture control levers.

“For starting”, Dave says, “the battery and generator were turned on, the throttle was cracked a quarter of an inch and the propeller lever set fully forwards, with the mixture lever in Idle-Cut-Off. The fuel pump was turned on, and the engine was primed with three sets of three second bursts from the priming switch. Magnetos were switched on, starter switch pressed and held and as the engine fired, the mixture was moved forward from Idle Cut-off to Auto-Rich – it stayed in that position throughout ground handling, take-off, climb, aerobatics and landing, only being brought into Auto-Lean in the cruise. The engine burbled away at 1,000 rpm before the throttle was opened a little to settle at 1,200 rpm.

“Allison engines are prone to overheating on the ground in a heartbeat, and that was one of the big things with the Airacobra’s V-1710. There were no external radiators on it to reduce drag and enhance performance, but the penalty was that the P-39 was fearsome for overheating. Keeping the coolant temperature below the 120°C redline during ground handling and the engine run-up was critical. It’s not dissimilar to how you’d operate an early Spitfire or the Bf 109G, but the Airacobra was the most critical of all.



“As soon as the engine was running smoothly, you needed to taxy as you had a matter of maybe six to ten minutes before the coolant boiled off depending on ambient temperature and whether the aeroplane was starting ‘cold’. A lot of people assume that the tricycle undercarriage improved forward visibility, but the view out of the front wasn’t good – it was almost like a tailwheel aeroplane and sat nose-high. It has a castoring nosewheel, and the steering on it was by differential toe brakes on the rudder pedals. Ground handling was very straightforward despite the unusual gear arrangement.

“The big challenge with the Airacobra was that if you didn’t get it right, the coolant temperature was threatening to redline before the oil temperature was high enough to run up the engine. Most aeroplanes of the era need 15 to 20°C oil temperature before you can do your pre-flight power checks, but the Airacobra’s threshold was higher – 30°C – and the oil and coolant temperatures certainly didn’t rise at the same rate. A bit of misjudgement and you might see 120°C coolant temp’ and 15°C oil temp’, so it was a balance between getting the oil warm enough to do the engine run-up without boiling the coolant.”

To augment cooling, Fighter Rebuilders retrofit Brooklyn BUM 2nd with a water tank and pump in the nose that the pilot could activate on the ground or in-flight to deliver water to the radiator via a spray bar. “You know you broadly have ten minutes of water to use”, adds Southwood, “but there’s no contents gauge for the tank. You don’t want to be caught short, so it’s all about judgement. The spray rig is turned on as soon as you start to move, but the needle’s still climbing immediately and it’s pushing 100°C before you know it.



“The engine power checks were conventional for Allison engines. A magneto cut check preceded a run-up to 2,300 rpm and three consecutive propeller cycles from fine to coarse, then the magnetos were checked, and the maximum drop off on a single magneto was 50 rpm. In the pre-take off check, you’d close the oil cooler door and make sure the radiator was open the whole time. That would remain the case throughout the flight, and you’d only open the oil cooler door on landing if you were running hot.”

Take-off saw the power fed in slowly to a conservative 44 in. Hg, not exceeding maximum take-off power of 52 in. and 3,000 rpm, with full right rudder to counteract the propwash and other effects from the propeller. Neutral aileron was maintained on take-off. When operating from a sufficiently long runway – Duxford’s hard being just shy of 5,000ft – no flap was required. The aeroplane flew off the ground at 110mph and the brakes applied once positively airborne before retracting the gear. The throttle and propeller were brought back to maximum continuous power of 39 in. and 2,600 rpm in the climb, and the coolant temperature typically dropped to a less critical level immediately after take-off.

The Airacobra comfortably cruised with a fuel burn of 28 US gallons per hour with the power brought back to 23 in. and 1,950 rpm and the mixture set to Auto-Rich, while Auto-Lean mixture coupled with 31 in. and 2,300 rpm settled into a fuel burn of 56 US gallons/hour. Some right rudder input was needed to counteract power increases, but in that respect, Dave describes the Airacobra as “benign compared to some of the other aircraft of the era – it’s more akin to the Spitfire than the Mustang in that respect.



“At maximum continuous power of 39 in. and 2,600 rpm, I’d want a minimum of 2,500ft cloud base and airspeed in excess of 300 mph for the first figure, which in my display profile is a reverse half-Cuban. I’d pull up to around 60° nose-up at the end of my first pass, and I’d start the roll once I’d gone through 1,500ft; I make the judgement call on when to complete the half-roll based on the altimeter, not the ASI, as I know from experience that an entry airspeed of more than 260 mph will give me the penetration I need.

“The next figure I fly is a loop, and in the P-39 I wanted a minimum entry airspeed of 260 mph for that. The ‘gotcha’ with the Airacobra is that it’s fast in a straight line, but it’s quite a low aspect ratio wing with a small wing area and short wingspan. Aerodynamically, there was good lift from the wings, and you could always climb well in the vertical, but the wing profile meant that there was a lot of lift dependent drag. You’d lose a lot of energy in looping manoeuvres, and after the reverse half-Cuban and loop, the airspeed would be well below 260 mph and insufficient to fly a third vertical figure.

“It’d then be a case of flying some slack manoeuvres to build the energy and airspeed back up. Seeing that sort of energy loss in a fighter of that era is unusual and struck me as being unique of all the World War Two fighters I’d flown. It made it a bit of a challenge as an air display aircraft in that you had to be extremely conscious of the ASI reading and ensuring you had 260 mph indicated for looping manoeuvres. The stick force per g was light enough that you could fly it quite comfortably with one hand throughout aerobatic figures and though it was light in pitch you’d need both hands on the stick above 200 mph to get full aileron deflection, with some rudder coordination to keep the aircraft in balance. With the aft CoG, there was very little elevator trimming in pitch throughout the speed range, and little directional trim change with speed or power either.



“One winter shortly after the P-39 arrived, TFC reset the tabs on the ailerons, and it flew completely differently in the roll. With the latter set up, if you were at 140 mph and put the stick all the way over and let go it would just about stay there. As it got faster towards VNE of 360 mph the stick would come back to the middle, so the stick floating characteristics changed across the airspeed regime, which is different to its contemporaries.

“Very little trimming was required once airborne, unlike some aircraft of the era – the starkest contrast is with the P-51, where you would constantly be trimming to different airspeeds and needed both hands on the stick during aerobatics. Flying a P-51 in the vertical, particularly if you were flying formation aerobatics, you might set some nose-up trim at the pull-up so you’re holding a push force, and as you climb, that push force reduces to zero and you have a manageable pull over the top of the loop. The P-39 was not like that, and you could loop it with one hand on the stick, with no pitch trim, and progressively feed in right rudder as the airspeed reduced until you gently held it over the top of a loop. It was very pleasant in that respect.

“Stall characteristics were benign in clean and landing configurations, with no pitching or rolling on departure. The stall was arrested by releasing back stick pressure and applying power incrementally. In a clean stall, buffet onset at 105 mph. With full flap and gear down, the buffet was at 87 mph. The most unusual stall characteristic was the aileron snatching in a high-speed power-on stall. I always take a new type to the stall in a high-speed turn, as I want to know how it would perform if I needed to make a sudden pitch change, say when recovering from a vertical manoeuvre.



“In a left turn, the stick would float to the left before the aircraft departed to the right. In a right turn, the stick would snatch very rapidly to the left, which could be quite alarming, and the aircraft would roll off to the right. That said, the departure was certainly nowhere near as bad as in the P-51, where the stall and flick come rapidly without any warning. In the Airacobra, you get plenty of warning with the buffet, and the accelerated stall came with a marked aileron snatch, pitch up and wing drop to the right. It talks to you more, as does the Spitfire. The Mustang just decides it’s had enough and throws you on your back without warning!

“There was no sense of the tumbling characteristics that have long been associated with the P-39”, he adds. “Certainly, during what you’d describe as standard aerobatic figures and power-on stalls in testing, there was never any suggestion of aggressive departure characteristics – but obviously we were flying it very conservatively and without war weight.” The Airacobra hadn’t long entered service when pilots began to report that ‘during flights of the P-39 in certain manoeuvres, it tumbled end over end’. It was a vice that Pete McDermott commented on postwar, where he referenced the P-39’s poor stall characteristics and propensity to flat spin and snap out of high-G manoeuvres.

The reports were serious enough to prompt Bell to investigate – during individual efforts to replicate the tumbling characteristics, its pilots were unable to spin the aircraft in the way described. NACA conducted further flat and steep spin trials in 1942, reporting that “all recoveries from the steep spins by full rudder reversal will be rapid” and that “in the normal flying condition, for normal spinning control configuration (rudder full with, elevator full up, and ailerons neutral), recovery from the flatter of the two types of spins will be fairly satisfactory if the rudder is reversed fully and rapidly’.



Yet, the stories persisted. The USAAF even released an informational video in 1943 dedicated to spinning and tumbling the P-39, which disputed claims of the aeroplane’s volatile stall. It wasn’t until NASA conducted an informal study in the 1970s that the P-39’s spinning characteristics were put under the microscope once more. Using the original P-39 spin test model, NASA replicated earlier tests in its 20-foot Free-Spinning Tunnel with consistent results. It was only when the model was re-ballasted to simulate no ammunition load, which moved the aircraft’s centre of gravity aft, that the model was found to tumble. It remains a point of contention hotly debated amongst pilots and enthusiasts alike.

In the circuit, Southwood says, “you’d bring the power smoothly back to idle at 23 inches, bleeding the airspeed to below 200 mph, which was the gear lowering speed. It had electric landing gear retraction and extension, but it had one red light for when all three legs are up and locked and one green light for when all three legs are down and locked. It took about 15 seconds for the gear to cycle each way, so there were a few moments of contemplation!”

In the downwind leg several methods were used to mitigate against the coolant reaching a critical temperature. “You flew a steep approach on idle power, avoiding a shallow approach as you’d then need to open the throttle to make the field, which increases the coolant temperature very quickly. At max. continuous power or in the cruise, the airflow through the radiators was sufficient to keep coolant temp below the 120°C redline – but it would only be sat maybe 5 or 10°C below that for most of the flight. For that reason, it was necessary to turn on the water pump downwind, one minute before landing, just to try and keep the temperature down. We didn’t take full flap until we rolled out on short final because there was a lot of drag from them and again, you risked needing a burst of power.



“Flap limiting speed was 148 mph, and normal approach speed was 110 mph. There was no instability at lower airspeed, and no untoward tendencies. Visibility was akin to most other fighters of the era; a common misconception is that the tricycle undercarriage improves visibility, but the aircraft is still performing the same as any other in terms of its angle of attack at low speed. It was a fairly conventional flare on landing. One curious thing unique to the P-39 was that when landing at certain angles into-sun on the grass, the quarter panels turned completely green, which was disconcerting!

“It pitched forward harshly when touching down on the main wheels, and there was a tendency for directional oscillation when braking at lower speed where there was insufficient rudder authority. The brakes weren’t particularly effective and it was a long landing roll of about 3,000ft, considerably longer than many warbirds, though it was never a limiting factor as the aeroplane rarely flew away from Duxford. However, the critical thing with the P-39 was getting it back, chocked and shut down before the coolant boiled!

“I have always enjoyed comparing the performance and handling qualities of different contemporary fighters”, says Southwood, “and flying something like the P-39, where there are so few of them left in airworthy condition, was a tremendous privilege. History hasn’t been kind to the Airacobra, but there’s a lot to enjoy when flying it in the present day. I remember after my first trip, Pete Kynsey asked me what I thought of it. I told him it’s certainly different, but not unenjoyable.



“You can fly any warbird and it’ll have its individual quirks and vices, but the P-39 distinguishes itself from everything else because of its peculiarities. That’s part of the pleasure of flying these aeroplanes. Of all the Second World War fighter types I’ve flown across my career, the Airacobra stands out as being unique compared to anything else.”