Comet 4C

Way back in 2000 I attended the Woodvale Rally for the first time, flying the then new, ducted fan Gloster Javelin that I’d just completed. This was in fact the first large show that I’d flown at, and I freely admit to being slightly nervous! Would the engines start on request? Would any of the unforeseen problems so many people suffer from catch me out? Fortunately the model performed flawlessly, despite the heat and the pressure!

Travelling home that night my thoughts started to wander onto what could be built next, and the usual sort of things passed through my mind – something bigger, more complex and more graceful. Commercial airliners are so rarely modelled, and with my interests lying in aircraft from around the ‘40s and ‘50s my next project could only be one thing – the de Havilland DH106 Comet. The disasters that befell the Comet 1, with its famed square windows, are well documented. No, my aircraft had to be one of the later variants, indeed I rather preferred the long, sleek nose of the Comet 4 to the short, fat nose of the Comet 1.

With budget constraints a constant issue, the power would have to be four O.S. 91s and Ramtech fans, similar to the Javelin. With approx 36 lb of thrust to play with and aiming at a 1:2 power-to-weight ratio, the AUW could be around 72 lb (32.7kg). But how big can a model be and still weigh in at under 72 lb? That, I still can’t really answer. We all know that models end up weighing far more than originally expected, and this was always a big concern during the Comet’s design.

I started to draw bits and pieces on scrap paper during spring 2001, but the sums didn’t really add up. I had great trouble trying to get the retract units to fit around the outer nacelles, and I was also having problems sorting out the scale intake and outlet diameters. As I enjoy scale models so much I couldn’t just cheat, so the idea was shelved whilst I concentrated on more pressing matters – i.e. work!
Of course, it wasn’t long before my thoughts started to return to the project. Towards the end of summer 2001 I’d formed a plan that would solve all my problems, the key being to replace the four ducted fans with a pair of gas turbines! It was the only way forward: two engines would give nearly 52 lb of thrust, weigh less than the four ducted fan units and would resolve the space issue around the outer nacelles, which could now be dedicated to a strong and robust retract mount. Requiring less volume of air, turbines would also solve the intake and outlet problems. With the engine situation resolved it was decided that to help things along, my father would join in and build the fuselage while I concentrated on the wings. With 52 lb of thrust in the offing, the size of the model crept up a bit – and so did the predicted workload!

Construction began in November 2001, and progress was initially fairly swift. The basic fuselage and wing structure was in place by the spring but, as is often the way with such a project, exterior forces caused the progress to slow. My father had been offered a job in Kuala Lumpur for the year, and after Sept 11, my job within the airline industry was under threat. In fact come April 2002 my father was abroad and I was redundant, concentrating my efforts on getting back into employment. Once more, the Comet was on hold.
I managed to secure a new position in November, and dad returned from abroad a few months later to enjoy full-time retirement – something that seems to be far more hectic than any employment! All this meant that the Comet project was back on… again.

I’d originally intended the model to be transported in a small caravan, but it soon became apparent that this wasn’t going to be practical; fortunately, though, the sizes of the main components could stay the same and still fit into a medium-size van. The fuselage comprises three sections, each of about 70” in length; the wing is also in three sections, and there are two tailplane halves. In total the Comet comprises seven pieces, the centre fuselage and wing sections being permanently joined. All the important electronics and systems are kept within this centre-section to avoid problems with continually connecting and disconnecting linkages.

Being mainly of balsa and ply construction the main structure grew in size very rapidly, and took around 6 months to complete. The fuselage is basically a tubular structure of 1/8” liteply formers on a 1/4” balsa crutch, with a few cyparis stringers to help keep things straight; the whole lot is sheeted using 1/8” balsa. What took time, as with any project, were the little fiddly jobs – the cockpit windows being a particular challenge. Looking visually simple on the finished article, integrating a single curvature into the double curvature of the nose wasn’t easy, the main problem being in finding a reference point from which to take all the other measurements. A simple 1/6” ply frame holds the acetate windows in place, and with a bit of trial and error the correct position was finally found.

Wing construction follows the same principles as with any built-up mainplane, it’s just a bit bigger! 1/8” balsa ribs and 21/2” x 1/4” spruce main spars form the foundation for the outer panels, whilst 1/8” ply ribs and two 1/4” sheet full depth spars sandwiched together do the same for the centre-section. In fact there’s only one rib in the centre-section, just outboard of the engines. 11/4” aluminium tubes make up the removable outer wing panel fixings, and just two bolts stop the wings from sliding out of place.

The complex double curvature of the engine nacelles was, to say the least, a complete nightmare! I never did manage to draw up a proper plan; I simply cut wood until it looked about right. I was toying with the idea of using foam and carving out the complex shape, but carving has never been particularly fun for me, so I stuck with the 1/8” planking method. To tell the truth, this part of the build probably took longer than the rest of the wing put together!

We decided to keep the radio gear as standard as possible, and as the model was clearly over the 20kg limit it had to conform to the LMA guidelines and have a double Rx system, each with two main battery packs to supply the servos and a smaller battery to supply the receiver itself. The idea here is that, no matter what current draw the servos require, the Rx always has a constant and uninterrupted power supply. Simple, really! On top of this, due to the long servo cables involved, an opto-isolator was installed after each Rx; this basically stops the servo cables either acting as aerials or creating their own radio waves that interfere with the primary signal. I don’t claim to know the system inside out, in fact there’s probably only one person in the UK that does; Terry of SM Services, the main supplier of systems for large aircraft in the UK, has a near perfect record. I wouldn’t consider any other.

As you can imagine, the complex double curve of the engine naschelles was a complete nightmare.

The engines we finally settled on were Jetcat P120’s. These German-manufactured units give about 26 lb of thrust each (roughly equivalent to a 75 – 100cc petrol engine) and burn about 3 litres of fuel in 10 minutes – considerably more than a petrol engine! To give a good reserve of fuel I elected to fit a total of 4 litres per engine – weight wise, a not inconsiderable 14 lb! A German friend of mine advised me that European Coke bottles are far better than any custom-made fuel tank, and being made of a far thicker plastic than UK bottles, they prove to be virtually indestructible. I’ve suffered the misfortune of crashing a jet model fitted with these tanks and not a drop was spilt, certainly saving the wreckage from fire. Clearly, a worthwhile safety feature!

The basic airframe of the Comet approached completion at the end of August 2002, and was due for inspection by an LMA examiner before the final sheeting was put in place, permanently sealing the construction method. Mike Jackson had been allocated this task, and it was arranged that we should meet up at Woodvale for the viewing to take place. There’s always a slightly tense atmosphere when someone’s poring over your handiwork, trying to find mistakes. It was instantly apparent that this wasn’t Mike’s intention, he was simply there to offer his thoughts on areas that he thought could be improved and any details that he would have done differently had he been building the model. Following Mike’s inspection we went away with some good ideas to add to the model, and the peace of mind that all had been double-checked.

Many aircraft rely heavily on the covering material for strength, but as the Comet is fully sheeted from 1/8” balsa, only a light skin was needed to produce a smooth surface to paint on. We decided that glass cloth was still the sensible way to go, as it really can’t be beaten for lightness and is the best finishing method to represent the stressed skin surface of the full-size. Using 0.66oz / sq. dm glass cloth and Fibretech resin, the airframe was fast approaching a stage whereby it could be flown. I prefer to test all models before the final paint scheme is applied, for the obvious reason that should anything go wrong, repairs can easily be hidden.
Prior to flight-testing, a final inspection must be carried out before applying to the CAA for a Permit to Test. This permit doesn’t allow the model to be displayed at shows, just simply to run through a flight test schedule that proves both model and pilot. Covering a minimum of 1 hour (soon to be increased for models similar to the Comet), the flight test requires the pilot to demonstrate all the ‘stock’ manoeuvres one would expect: take-off, circuits, landing, stall, crosswind landing and any other manoeuvre that would be performed in front of the public during an air show. Again, this must be supervised by an LMA appointed examiner, which in our case was none other than LMA Chairman Dave Johnson. Well known for his Vulcan bomber and fleet of giant aircraft, I knew that he wasn’t going to sign his name against a borderline model!

The first flight was initially planned to take place at Hucknall, just north of Nottingham but, upon arrival, the wind was blustery and 45° across the runway – not ideal by any standards. After much deliberation we decided not to chance it, and drove home again.

Lettering and motifs are all sprayed on using CNC-cut paint masks.

Whilst the following day was forecast much better Dave was due to fly at an event on Bruntingthorpe airfield. However, he kindly made arrangements for us to use the runway after the event. This turned out to be perfect. With virtually no wind, a 2-mile long runway that was as wide as a B-52 and a perfectly clear sky, I clearly had no excuses.
Both engines ignited on cue and the Comet slowly taxied toward the take-off point. With three video cameras and two still cameras firmly trained on it, I advanced the thrust. She accelerated along the runway with authority and she began to rotate with the slightest breath of elevator, needing a slight amount of correction to stop her becoming airborne too soon. The climb-out was slow, smooth and very predictable. A few notches of down trim were required, but nothing major, indicating that the C of G was perhaps a shade too far back. I climbed to height to give a little margin should anything untoward happen, and tried to relax. The Comet weights about four times more than anything else I’ve ever flown, but to be honest this just seems to make the model even easier to fly. The feedback from all that weight was certainly noticeable, though it simply reminded me to be careful and take it steady.

A couple of circuits later I was keen to get her back on the ground. I’d decided before the flight that only the primary flight controls would be used at first, so no retracts, flaps etc. to complicate matters at this important time. I set her up for a wide, flat circuit that provided nice, long, gentle finals to help me stabilise the approach. Very little flying was required from me from here on. The model simply sat firmly on the approach with just a trickle of thrust to keep her coming forwards, and she descended earthwards with grace. I meddled a bit too much with the flare and overdid it, but by just relaxing the up elevator all ended smoothly. Back on the deck, safe and sound!

With the weather so perfect we decided to have one more flight before packing up, this time trying the retracts, flaps and a gentle stall. Having climbed to a safe height and throttled back to idle, I carefully maintained the height with gentle up elevator, and as the speed bled off and the nose began to rise I reached full up with nothing untoward happening. A few more seconds, a little less speed and the stall was nothing more than gentle. A slight tendency to break away to the left, but as soon as the up elevator was released and power reapplied, the Comet flew cleanly out.
I then put her through a few lower passes for the growing camera crowd prior to another easy landing, this time trying not to overdo the flare. Certainly, half flap at this point helped stop the ballooning tendency. The only effect the flaps had upon being lowered was to increase the nose-down attitude (due to the extra lift produced).
With the car repacked and the required approval signatures from Dave Johnson we drove home elated, full of enthusiasm to finish the paint job and get cracking with the final flight-testing.

Half flap is very useful for landing, the only effect being a slight lowering of the nose.

Using mostly Japlac enamel paint (from B&Q) on an MDF primer base, the painting was more an exercise in masking than anything else. The lettering and motifs are all sprayed on, using CNC-cut paint masks from, all having being drawn digitally via a PC graphics program.
Most airliners give that ‘new and shiny’ appearance from a distance, but once you’re up close, most show the amount of work they do in their lives! Hence the Comet sports some simple weathering, including dirty streaks under the windows from being left out in the rain, and discoloured metal panels from the heat of the engines. All this was accomplished using a cheap airbrush and some soft cloths. The common mistake made when weathering is to put too much on, using black paint. I tend to dull down weathering by using browns and greys, heavily thinned and applied very lightly. It’s much easier to add more weathering than take some off!

Finally, one coat of Aerokote fuel proofer sealed the surface to stop fingerprints from permanently marking the model.

All in all a lot of work has gone into the Comet, though I don’t think we ever regretted starting the project. I freely admit that due to work pressures and workshop space, my father completed the majority of the finishing and painting – I was just left in charge of engines and avionics! With about 30 flights now under her belt, the model is still going strong. There have been a couple of minor glitches (one partial ‘gear up’ landing at Cosford, and one double engine failure during testing are all we’ve encountered), though I still can’t get my dad to fly her. He says he’s happy just being ground crew. In 2006 we’ll be at the Scottish Airshow and the LMA Cosford event, but, due to my getting married that’s possibly all we’ll manage this year. Such sacrifice!

The Comet weighs four times more than anything I've ever owned but this seems to make it easier to fly!

Name: de Havilland Comet 4C
Model type: Large scale jet airliner   
Designed by: John and Steve Rickett
Wingspan: 15' (4.6m)
Fuselage length: 151/2' (4.7m)
Fin height: 41/2' (1.4m)
All-up weight: 120 lb (54.5kg) wet, 105 lb (48kg) dry
Power: 2 x Jetcat P120 turbines

Radio: 2 x Futaba 149DP PCM Rx
2 x 700mAh 4.8V Rx batteries
4 x 2100mAh main flight 6V batteries
2 x 1500mAh 7.2V turbine batteries
1 x 3000mAh 12V landing light battery
2 x SM Services opto-isolators
2 x SM Services battery backers
Servos: 9 x Futaba 3303 sail winch servos for: flap (x4); aileron (x2); elevator (x2); rudder (x1)
4 x Futaba S136G retract servos (operating undercarriage doors), 1 x Futaba 148 (nose wheel steering)

  • This article was first published in 2006 – the model still makes regular show appearances.