HMM Crossfire
Kit Review by Trevor Hewson
(With sub headings inspired by Alice's Restaurant, written and sung by Arlo Guthrie)

Selection
My P38 Lightning has attracted its fair share of comments and, whilst the complimentary remarks are always appreciated, they are not necessarily the most memorable. After what I thought was a pretty good flight at Kings Park one day, Dave remarked "There's only one thing wrong with that model." "Ye-es," says I, wondering what was coming next. "The basic concept," says Dave!

Having got my attention, Dave proceeded to remind me that, early on in the P38 project, I had expressed the hope that, as well as being an interesting scale model, the Lightning would also serve as a replacement for my late lamented Aerojet. He was, of course, quite right - that was indeed a fundamentally flawed concept.

So, how to fill the Aerojet hole? After considering a range of options from fun fly models at one extreme to high power glider types at the other, I returned to the good old compromise of a sleek but aerobatic sport model. The Crossfire seemed to fit the bill because, as well as promising to do everything the Aerojet did, it also had a rudder, was a little larger and had room for eight cells. This last point was quite important because I have been looking after Peter's 8-cell Dash 7 power packs for him and they could do with a bit of exercise!

Collection
Howard Metcalfe lives in Durley, near Southampton, so I opted to collect the kit in person. There was a slight delay because Howard had been unwell but, true to his word, he duly called to tell me the kit was ready. It was helpful to discuss with him various aspects of the design and to see what he thought of my ideas, as well as to examine the prototype which he still has in flyable condition.

Inspection
Designed for mail order, the Crossfire comes in a plain cardboard box, some 39" x 4" x 5", which contains mostly uncut sheet and strip balsa, a bag of hardware, a rolled plan and stapled instruction sheets. Closer inspection reveals that the ribs are die cut as are a few ply components. For the remainder of the shaped balsa parts though, peel and stick templates are provided, some of them already stuck onto sheets of the appropriate thickness.

At first I felt a little cheated by this, I did after all buy a kit not a plan and a pack of balsa. On reflection though, the tedious part of building from a plan is not the cutting out, but rather the process of transferring the shapes from the plan to the wood. This system avoids all of this and, since the majority of the parts have straight edges, the cutting out doesn't take too long.

More of a nuisance though is the fact that, whilst the templates are clearly labelled, there are no clues as to how to arrange them on the relevant sheets of wood to avoid ending up needing to cut one long thin part out of a short, wide piece of balsa. It took me most of an evening to identify which pieces of wood were for which application. However, by the time I had done so, I felt I understood the construction much better, and the instructions always tell you to study the plan thoroughly before starting - and we all do, don't we?! I was later to discover that the wood quantities were sufficiently generous that I needn't have been concerned to optimise the layout.

As for quality, well, fairly good really. If I was hand selecting the wood, I would have chosen lighter grades for many applications, but it was all firm and usable.

Direction
The instructions are quite detailed and take you step by step through the building process. There are no really tricky steps and most of the head scratching related, as always, to the modifications I had in mind. These stemmed from the fact that I was using a geared speed 600 rather than the direct drive arrangement shown on the plan. This meant that the second former had to have a large cut-out for the rear of the motor to poke through. The motor access hatch on the top now seemed superfluous and was dispensed with, the bottom hatch behind former two serving both to insert and remove the motor and also the battery, which sits, Aerojet style, beneath the wing.

A plastic motor mount is provided which acts as the front bulkhead but, since the bearing boss on my gearbox was longer than that on a standard motor, the mount had to be moved rearwards by a quarter of an inch or so. Once I had worked out how these changes impacted on the construction sequence, I was ready to go.

The fuz builds up quite quickly. Note I said 'builds up', not 'takes shape'. After the experience of planking the P38, where the shape emerges progressively with each plank, building the Crossfire was quite a contrast. Chunky blocks of balsa are glued on at odd angles, resulting in a structure that looks as if it has been designed by whoever builds cartoon dogs' kennels. It is all quite straightforward, but care is needed to chamfer the edges of the blocks so that they fit well. If they only fit where you can see, gaps will appear at the next stage.

The next stage is to transform the ugly duckling into the proverbial swan. Now the graceful shape emerges. To get the smoothest shape without overdoing it anywhere needs some care and also some courage - there is a lot of wood to come off. However the final shape is well worth it.

The wings are a conventional built up structure. I was a little concerned that the spars are all balsa, but resisted the temptation to beef them up. However, since I was putting in a heavier power plant than used in the prototype, I did extend the dihedral brace out one more bay and doubled up on the balsa shear webs in the next bay. Other than that, the wing went together without significant problem, the spar slots in the wing ribs being very accurate, and the jigging strip positions on the plan also working out exactly right.

Tail surfaces are simple built-up frameworks, which look fairly frail but should do the job. The last task, which requires a bit of care, is to seat the wing on the fuz and fit the all balsa canopy. The canopy has to mate up with the wing faring, the wing itself, the extension of the wing seat and the front of the balsa turtle decking. Not only are you looking for a gap-free fit, but the contours of the canopy need to blend smoothly in with the rest of the fuz, and the retaining mechanism (a cut down ball link) needs to be positioned precisely - and is invisible. It's not really as difficult as it sounds, but is not a job to be rushed.

Projection
I generally don't like to put stickers on models saying what they are. This is partly because I feel that, if I'm going to fly an advertising hoarding then I should be paid for it, but also because it deprives people of the fun of guessing what it is. I do though like to give a clue, which is why my thermal soarer has some strange characters on it. (I was somewhat taken aback one day at the slopes when a woman pointed to the model and exclaimed "that's Greek!". It turned out that she was too. At least she confirmed that my spelling of 'algebra' was correct, but apparently there should be a couple of funny accents somewhere.

Anyway, a name like 'Crossfire' cries out for a colour scheme with intersecting, flaming arrows or something of the sort. All my design attempts along these lines though turned out rather naff and, by the time I got around to visiting the local model shop, I was contemplating a relatively simple red and yellow design, suggesting 'fire', with no real suggestion of the 'cross' bit except that the fire did run across the wing. After a good twenty minutes of indecision in the shop, even this feeble idea went out of the window and I eventually emerged with blue and yellow Profilm having decided to copy the colour scheme of the ex-Aerojet. I even relented and used the Crossfire sticker (actually a rather racy X-fire). So much for creative intentions!

Deflections
Control surfaces were set up as instructed and, with everything on board the weight came out at 2lb 14oz, with a cg about 1/4" ahead of the plan position. Whilst this is well over the official upper weight limit of 40oz, the wing loading is still a fairly modest 17.2 oz/sq ft. The airframe in fact accounts for only a little over 12oz; the receiver, three servos and linkages push this up to just over the one pound mark, and the rest is accounted for by propeller, motor, gearbox and speed controller - and of course, 16oz of battery (8 x 1700mah cells).

To give me plenty of scope for experiment, I fitted a 40a speed controller (Microstar 40 BEC). The motor is a Graupner Speed 600 Race, driving a Graupner 10x6 folding prop through a Simprop 2.65:1 gearbox. This set up draws a modest 15amps on the ground, producing a theoretical thrust of 27oz - more than enough for a safe test flight.

Dejection
Saturday morning dawned wet. So wet in fact that, while waiting for the rain to stop, I fitted the floats to the Lazy Bee. Arriving at the park at the appointed hour, the rain promptly stopped and we squelched out onto the grass under brightening skies. A quick warm-up whirl with the Lazy Bee, including a successful landing in the disappointingly tiny puddle, and it was maiden flight time. The range check was interference-free, but Dave did comment that, although the control surfaces were steady during the power-on check, the tailplane did flap about a bit in the prop wash. Laughing this off I gave Jim the go ahead to launch and the Crossfire climbed away into a steady left hand circuit.

All seemed well so, once trimmed out, I eased the stick forward to pick up speed for a loop... "brrrrrrrrr...!" - serious vibration! All calmed down again when slowed up, but a cautious repeat showed that any attempt to build up speed immediately gave the same alarming result. After a brief discussion about motor resonance, prop balance, prop wash, etc., I opted for one more test, this time with the motor switched off. Brrrrrrrrrr... - just the same! This time though, as I levelled out, something flew off. For a heart stopping moment I feared it was half a tailplane but, as the model came by, it was clear that it was minus the canopy. Once my nerves settled, it was time to land - and think.

The discussion that followed focussed initially on the tailplane, which is indeed quite flexible. After a while though, suspicion shifted to the ailerons, which are driven by torque rods at the inboard ends and are long, thin and flexible. As an aid to diagnosis, I opted for a second flight, with the outboard ends of the ailerons immobilised with tape. The aileron servo had no problems moving the inboard ends - as I said, the ailerons are very flexible!

After the climb out, in which I was relieved to find that the aileron response was still okay, a gentle dive was entered - no flutter. After regaining height I tried a longer dive and pulled a loop - no problems. Half roll and pull through? - fine. Whilst I didn't try any really fast stuff, the point was proved: the ailerons had been tried and found guilty.

The second landing, like the first, was smooth enough, but a long way away. At this weight, the Crossfire lands just like an Aerojet - it cruises in low over the trees on the glide, slides by at knee height then disappears off into the distance.

Reflection

Suggested treatments for the offending ailerons included replacing them with a stiffer grade of balsa, stripping and stiffening the originals with various combinations of tissue, dope, epoxy and glass, or introducing some kind of damping into the hinges, perhaps using silicone. The last of these was my suggestion. Frankly the only thing to commend it was that it didn't involve stripping the covering or replacing the ailerons. On the downside was the minor point that I didn't know how to set about it, let alone whether it would do the trick.

Back home and a little more reflection led me to wondering whether the vibration was induced by the tip vortex and, if so, whether the fact that the ailerons went all the way to the tip was contributing to the problem. On the basis of trying the least drastic remedy first, I chopped the last two inches off the ailerons and epoxied this portion to the tip, in the hope that this would shed the vortex cleanly. I had to fit an additional mylar hinge towards the end of the shortened aileron but, by the next weekend, I was ready to try again.

Correction
Well, I was ready, but this time the weather had other ideas. However, eventually, on a beautiful Saturday in mid December, the Crossfire was once more launched into the wide blue yonder. Climb out, level pass at full throttle - no flutter. climb a little higher, gentle dive and loop - no flutter. Steadily, the speed was increased and everything was fine. The landing circuit was flown more slowly this time and a more local landing was achieved - with 10mins flight time on the clock. Terrific!

Perfection?
The Crossfire does what the advert says. It is vice-free, even when over the design weight, and its aerobatic repertoire exceeds mine so I look forward to improving my flying skills as I get to know it better. It won't climb vertically away from your hand, but it does show that you don't need 27 cells and carbon fibre reinforced wings to enjoy smooth aerobatics.

With a 10min flight time, the average current draw in the air can be no more than 10 amps, yet this model will loop from level flight and, even with this set up, can just about manage a vertical roll. The secret is in the clean lines, which allow the model to get up to speed without too much effort, and the modest pitch on the propeller which unloads as the speed builds, so reducing the current consumed. Once the motor and gearbox are fully run in, I intend to try a larger prop - if it goes this well at 10amps, it should provide a fantastic five minutes at 20amps!