Flight Test: Nanchang CJ-6A
PUBLISHED: 11:56 01 June 2017 | UPDATED: 16:04 01 June 2017
Yak-lookalike it may be, but the Nanchang is a home-grown communist Chinese original with fine handling and excellent aerobatic capabilities
Rolling high above the Berkshire countryside behind a big, round engine may not be the cleanest or greenest way to fly, but it certainly puts a smile on my face! I’d always assumed that the Nanchang CJ-6A was essentially a copy of the Yak-52 from behind the Bamboo Curtain, in much the same way that a Tupolev Tu-4 Bull is basically a reverse-engineered Boeing B-29, and a Shenyang J-5 is a licence-built MiG-17. But you should never assume in aviation, as it often makes ‘an ass of u and me’, and a copy it is not.
Although it doesn’t look dissimilar to Yakovlev’s trainer, and some of the systems are alike (primarily the plethora of pneumatics), it’s really a completely different machine that actually predates the 52. Intended to replace the CJ-5 (which was a copy of the Yak-18A), the Chuji Jiaolianiji-6 (Basic Training Aircraft 6) entered production in 1961, and several thousand were built (the actual number is a little nebulous) at Shenyang in Liaoning Province−the oldest aircraft factory in China.
Even if you didn’t know the CJ-6 was an ex-military machine (if you could ignore the People’s Liberation Army Air Force markings on the side) from any distance it looks exactly what it is – a 1960s air force primary trainer. As well as the PLAAF, it also equipped quite a few foreign air forces, and there is even a COIN (counter insurgency) fighter version armed with twin machine-guns, bombs and rockets.
Walking towards it at White Waltham I was struck by how big it is, particularly as it was standing next to Tim Darrah’s beautiful Beech 18, which we’d just used as the camera ship.
Having greatly enjoyed the resolute roar of the Expeditor’s two P&W Wasp Juniors, I assumed (that word again!) that the ’Chang’s mill was probably somewhat similar. “So what’s under the cowling?” I asked co-owner Lawrence Hawthorn, who explained that the engine hidden behind the neat louvres was indeed an air-cooled radial, but of a relatively modest 285hp. Hmm, that’s a lot of aircraft for only 285hp, I thought to myself.
The prototype was originally powered by a Czech-built flat-eight, but it quickly became apparent that more power was needed, so the Czech engine (curiously known as the Doris-B) was replaced with the Zhuzhou Housai HS-6, which is a licence-built version of the Russian Ivchenko/Vedeneyev AI-14.
This supercharged nine-cylinder air-cooled radial produced 260hp, and the extra power transformed the aircraft. In 1965 the HS-6 engine was upgraded to 285hp, re-designated HS-6A, and all new aircraft fitted with this engine became CJ-6As. I still thought it looked like a lot of aircraft for only 285hp, but a wise old Chinese proverb (attributed−wrongly I believe−to Confucius) states that ‘the proof of the pudding is in the flying!’
Up close, it’s apparent that the CJ-6 is essentially conventional in both design and construction, although what really catches the eye is how well made it is. Flush-riveting is used throughout, and the 51-year-old airframe is in excellent condition, with no corrosion anywhere.
The nine-cylinder radial spins a two-blade J9-G1 hydraulically-actuated, two-blade constant-speed propeller, rotating anti-clockwise (as viewed from the cockpit). The cooling system is very elegant: a ring of 28 gills/shutter slats around the front of the cowling allows airflow through the engine bay to be controlled via a knob in the front cockpit, allowing precise control of the cylinder head temperatures. The generator and air compressor are cooled by their own dedicated air scoops, which project through the shutter.
The starboard scoop directs cooling air to the compressor, while the port one cools the engine-driven generator. An oil radiator is housed in the starboard wing root. The fuel is carried in a pair of 77-litre wing tanks, which flow into an eight-litre header tank in the fuselage and then to the engine-driven fuel pump.
The cantilever wing consists of a centre section and detachable tapered outer wing panels which have a noticeable amount of dihedral. It is the dihedral that readily distinguishes the CJ-6 from the Yak-18, although the squared-off fin is another good clue. The wing uses a NACA 23015 aerofoil for the centre section, which changes to a 4412 section on the tapered outer panels.
It carries an aluminium semi-monocoque fuselage, although what grabbed my attention was the single, split flap. In fact, I’d probably be more minded to describe it as an airbrake, as it goes right across the belly and only has two positions−up or 40° down. Two things are immediately apparent: don’t use it for takeoff, as it clearly creates buckets of drag and very little lift, and in a go-around make sure it’s retracted promptly.
It’s also pretty obvious that its operation would cause significant and possibly rapid changes in pitch trim. Why rapid? Well, like other Communist aircraft of this era that I’ve flown, a lot of the primary systems, including the engine starter, wheel brakes, undercarriage and flap are actuated pneumatically. Lawrence explains that as so many of the primary services are pneumatic a full understanding of how the system works is vital. Basically, an engine-driven compressor keeps two air bottles behind the rear seat permanently topped up.
The main bottle has a capacity of twelve litres and its pressurised air is used to start the engine, raise and lower the flap and undercarriage, and actuate the brakes, via various pressure-reducing and−for the brakes−differential valves. Should the compressor or main system fail there is an emergency system, but this is very much a finite resource. It consists of a three-litre bottle whose primary purpose is to lower the undercarriage, although it can also operate the wheel brakes and flap. If you’ve only ever flown aircraft fitted with electric and hydraulic systems pneumatics may seem somewhat crude, but there are sound reasons why they work, as they are simple both to operate and maintain, and much more tolerant of extreme temperatures.
While the electrically-heated pitot tube in the starboard wing seems unnecessarily large, red-and-yellow poles protruding from the upper surface of each wing provide a neat and reliable indication that the main wheels are down.
The trailing-link undercarriage features relatively large low-pressure tyres and consists of an aft-retracting castoring nosewheel and inward-retracting main legs fitted with drum brakes. Only the mainwheels fully retract in flight and are semi-covered by doors. Down at the tail I note that, like the ailerons, the elevators and rudder are fabric-covered, and that although there is a rudder trim tab it is only ground-adjustable. There is a trim tab in each elevator.
Access to the tandem cockpit is aided by a stirrup under the port wing-root, and as G-BVVG is a syndicate aircraft and some of the vital controls (such as the undercarriage selector and emergency air valve) are deactivated in the rear cockpit, Lawrence must sit in the front. However, before we fly I take the opportunity to sample the front seat. The cockpit is very military, being utilitarian with no apparent concession to comfort, although it is actually quite comfortable. The tall stick carries a lever for the brakes, while the adjustable rudder pedals are mounted on a pivoting bar.
The instrument panel is initially a bit confusing due to the non-standard layout and mostly metric instruments (except the altimeter). However, with Lawrence’s expert guidance, I soon work out which does what, while the extensive use of colour-coding on the gauges (as well as showing max/min limits the fuel system is yellow, the oil brown) is a big help too. Curiously, the various levers are not colour-coded.
The panel is dominated by a large, centrally-mounted attitude indicator. To western pilots its presentation may seem counter-intuitive−at first glance it seems that the sky is brown and the earth blue−but there are sound human factors and engineering reasons why it’s like this.
Left of the AI are the airspeed indicator (in km/h) and manifold pressure gauge (in millimetres of mercury), while on its right are the vertical speed indicator (in m/s) and accelerometer. On the next row is a triple-needle gauge showing fuel pressure, oil pressure and temperature, then the altimeter, turn & slip, a small modern transceiver, a clever little instrument that is halfway between a directional gyro and a horizontal situation indicator, and the fuel gauge.
This only goes up to sixty litres (and remember there are 77 in each wing tank), and a small switch below the two ‘Bingo’ lights (which illuminate if there’s less than ten litres in that tank) selects which tank is being gauged. Curiously, you cannot select which tank feeds−they both flow into the eight-litre header tank in the fuselage, from where fuel is pumped to the engine. More on this later. The final row shows (left to right) prop rpm, CHT, manifold temp and volts.
Below the engine instruments, at the bottom of the panel, is a long row of identical switches for the electrical services, with the undercarriage selector lever on the left. Two switches that look very much as if they should select the magnetos are instead the landing and taxi lights. The actual mags are on a rotary switch that, interestingly, works anti-clockwise! Just below the port side canopy rail is the elevator trim wheel, but no sign of any indicator.
This confused me until Lawrence explained that the small green light just below the ASI lights up when the trim is neutral. Below the trim wheel are two levers that operate the emergency fuel cut-off and alternate air selector, while slightly further aft are the throttle and prop levers, with the mixture just behind and the flap selector by your left hip. The mixture control is worthy of further comment as it works the ‘wrong’ way, i.e. back for rich, and is interconnected with the throttle. The Chipmunk uses a similar arrangement.
On the right side console is the main air system valve and pressure gauge, hand-wheels for the engine shutters and oil cooler door, a lever for the wobble pump and the emergency air valve and pressure gauge.
Appropriately equipped with a parachute (the seats are designed for one) I strap myself down in the back seat and study the panel, which has most of the front cockpit’s instruments in mainly the same places. Lawrence has to start the engine, but talks me through the procedure, and as it’s quite different to the aircraft flown by the average Pilot reader it’s worth explaining. Although its good practice always to pull a radial through to check for hydraulic lock, with the HS-6A you don’t if the CHTs are above 80°C, so Lawrence opens the main air valve and checks we have at least 20kg/cm2.
This is very important; otherwise we may have enough air to start the engine, but not enough to work the brakes! Another important check is to note the manifold pressure gauge. The engine-driven supercharger is not designed to increase performance at altitude, but simply to increase power overall, and although the engine’s modest compressions mean it’s very difficult to over-boost it, good practice is to use no more than ambient plus 80mmHg.
With engine primed, wobble pump wobbled and the gill shutters and oil cooler door opened (if the engine were cold the oil cooler door would be closed) the rest of the checks are fairly standard and the start button is pressed. An electric solenoid energises the starting coil, creating a shower of sparks at the plugs and opens a valve, allowing high-pressure air into the manifold distributor. A poppet valve opens on the cylinders that are just past top dead centre and the engine turns over. It starts with a hiss, then a roar−and although it sounds a little agricultural at first soon settles down to a grumpy growl. The temperatures and pressures soon come up (babying a round engine is so important) and Lawrence encourages me to start taxying.
This is not easy. I always like to taxi a new type quite slowly but, as I’ve found with other aircraft with castoring nosewheels, this is actually counter-productive. A little faster is better, along with judicious squeezes of the brake lever and plenty of rudder. The nosewheel castors up to 52° either side of neutral but the differential braking system doesn’t kick in until you have more than 15° of rudder on. White Waltham’s soft, damp grass isn’t helping either, and I’m sure it’s a lot easier on a smooth tarmac taxiway. So initially we proceed in a rather erratic fashion with the pneumatics hissing like an angry anaconda.
The ride quality seems slightly ‘soft’ and it almost feels as if the wing dips into the direction of turn. In fact, this is perfectly normal; the trailing link undercarriage is cleverly designed so that although it seems quite pliant it also has excellent energy-absorbing characteristics. During the run-up the checks are pretty standard for a single engine piston aeroplane, notable differences being setting the engine ‘gills’ and oil cooler door, and checking that both the main and emergency systems have at least 30 and 50kg/cm2 respectively.
With all checks complete we slide our respective canopies closed and I taxi out onto Runway 29 and slowly open the throttle. It’s not a great takeoff. I can’t really see over the nose and the grass runway means there aren’t clearly defined edges. I don’t apply anywhere near enough left rudder and the combination of a lack of peripheral cues, poor technique, a bumpy runway and general incompetence result in the takeoff being a rather untidy affair.
The rudder starts to come alive at about the same time as the ASI; I ease the stick back at around 100km/h (55kt) and the Nanchang flies at about 130 (70kt). As I’d suspected, the ground roll and initial climb are slightly underwhelming, as the Nanchang doesn’t so much accelerate as ‘gather speed’.
However, with the undercarriage retracted, it transforms from a slightly clumsy land vehicle into an elegant flying machine−and continues to accelerate. At 500ft I pull power and prop back to the max continuous settings of 710mmHg and 2,250rpm but it continues to accelerate. As both Lawrence and Tim had intimated during the pre-flight briefing, this really is a surprisingly clean aerodynamic machine, and the faster it goes, the better it gets.
The initial rate of climb is around 1,000fpm and about ninety seconds later we’re passing through 1,500ft so I make a further power reduction to the cruise setting of 680/2,000 and head off to the west to find somewhere to play while still climbing at around 500fpm.
Straight and level at 4,000ft, 680mmHg and 2,000rpm, the Nanchang lazily lopes along at 150kt while burning roughly 70 lit/hr, and I must say I’m certainly beginning to sense its charm and character. The trimmer is nicely geared, and I notice that some left rudder is nearly always needed to keep the ball centred (unless you’re turning right).
A few turns of varying steepness confirm my initial impression of taut, crisp handling, well-coordinated controls and surprisingly light ailerons. I’ll freely admit to being a frustrated fighter pilot, and looking through the windscreen at that big round cowling while feeling the pulse of the powerful engine through the throttle is a very visceral experience. It may have been designed as a trainer but it’s also a warbird, and it’s easy to see how the arrival of a pair of CJ-6B COIN fighters armed with bombs, rockets and 7.62mm machine-guns could tip the balance in a border skirmish.
It’s also slightly surreal to see the big red star on the wing, as aircraft in communist Chinese markings are very rare in the skies over Britain. Nevertheless a litre a minute is still a litre a minute and Lawrence disturbs my reverie by reminding me that this is supposed to be a flight test and I need to get back to work.
I want to see what the slow speed handling is like and gently pull some power off. The almost-infinitely-variable gills are very useful here, as the CHT needle hardly seems to move. The clean stall is preceded by a reasonable amount of buffet before the wing finally gives up at around 60kt.
There is very little wing-drop, and having recovered by simply unloading I gain a little speed, add some power and ask Lawrence to drop the undercarriage and then the flap. Undercarriage extension is quite subtle, but the flap comes down pretty quickly and there’s a definite change in pitch, which is easily trimmed out. As I thought, the flap is primarily a drag-producer as the stall speed only reduces by a couple of knots. There’s a bit more wing-drop at the stall with flap, although this may have been my fault.
The CJ-6 is very much a rudder aeroplane, and as well as good hands it also requires sure feet. Overall though, the slow-speed handling and stall characteristics are very benign, so I ask Lawrence to clean up the airframe and gently bring the power back in. Once we’ve got good speed Lawrence takes control and demonstrates a loop, roll, wing-over and half-Cuban, which I then do my best to replicate, with varying degrees of success.
The airframe is stressed to +6/-3g, but the syndicate tends to stick to positive g, and rarely above four. The Vne and Vfe have also been reduced (by the CAA) to 180 and 87kt respectively, leaving very good safety margins.
It’s interesting how little energy the CJ-6 loses on the up-line; once you get it going it just keeps going, and it really is tremendous fun. I don’t really care for the ‘yank & bank’ style of many modern aerobatic flyers, and the Nanchang is ideally suited for what I term ‘gentleman’s aeros’. Tracing an elegant sequence of long flowing lines across the sky really is great sport, and it occurs to me that a pair of CJ-6s fitted with laser designators and receptors would make a fabulous aerial version of paintballing. It wouldn’t be cheap, but it would be fun.
All good things must end, however, and after a spirited ‘missed approach’ (ahem) at a friend of Lawrence’s strip we rumble sedately back to White Waltham. En route Lawrence mentions that a quirk of the fuel system is that sometimes you end up with one tank a lot emptier than the other.
It’s not an issue (as the engine is fed from the header tank) but if a significant imbalance occurs the solution is to fly with the wing containing the fuller tank raised for a minute. Once fuel starts to flow it tends to keep flowing until the imbalance has been rectified.
One aspect that has really impressed me is how precisely the cylinder head and oil temperatures can be controlled. All big round engines appreciate TLC, and whoever designed the Nanchang’s cooling system really put some thought into it. Turning downwind at White Waltham, I wriggle my parachute into a more comfortable position and sit up a bit straighter.
Having made a bit of a dog’s dinner of the takeoff I’m determined to rectify matters with a good landing. The field of view is excellent and it’s nicely speed-stable so, with the wheels down (and having checked both “three greens” and the poles protruding from the wings), I turn onto a curved base at 80kt (which is also best glide), wait until it really does look as if we’re too high, then ask for the flap. My experiments at altitude have convinced me that with the flap down (again; it really is more of an airbrake) it’s probably quite hard to overshoot, but equally easy to undershoot!
On short final I start bleeding the speed back to seventy and, as we sail over the fence, ease the nose up into the flare. The main wheels touch gently, and mindful of the castoring nosewheel’s shimmy damper I do my best to hold the nose up. Overall, I’d say the landing is the diametric opposite of the takeoff, and Lawrence indicates he’s more than happy.
In fact, he insists that I try another takeoff, and this is a vast improvement on my initial attempt. The trick seems to be to use the rudder as your intention and not your reaction. Taxying back in with canopy slid open, my arm resting on the rail and the brakes hissing like a peevish python is very satisfying, particularly as I seem to be able to make it go in a straight line this time. As is so often the case, confidence is the key.
To conclude; it’s great! I was impressed by the CJ-6, and would love to own one. I’d like a little more grunt for takeoff (who doesn’t) and believe that in America the 360hp M-14 is a popular and easy retrofit. In fact, the CJ-6 is quite a common ‘warbird’ in America (there are over 200 on the FAA’s register) and it’s not hard to understand why. Not only is it not too thirsty for fuel (and oil) but with over 2,000 built, production only ending relatively recently, and many still in service, it is well supported.
It’s certainly significantly more affordable−and tractable−than its capitalist counterpart, the T-6. And while I know that the T-34 is probably a more accurate comparison, being nearer to the Nanchang chronologically and with regard to horsepower and undercarriage configuration, it doesn’t have that lovely round engine!
Specification – Nanchang CJ-6
Wing Area: 17.07sq m
Weights and Loadings:
Empty weight: 1,095kg
Max AUW: 1,400kg
Useful load: 305kg
Wing loading: 82kg/sq m
Power loading: 6.57kg/kW
Fuel capacity: 154 lit
Baggage capacity: Very little!
Climb rate: 1,150ft/min
Service ceiling: 16,500ft
Engine: Zhuzhou Housai HS-6A supercharged nine-cylinder air-cooled radial, producing 285hp (213kW) at 2,350rpm driving a J9-G1 two-blade constant-speed propeller
Manufacturer: Nanchang Aircraft Manufacturing Co.