Five Years With Melmoth 2

_Melmoth_2 was not really finished in 2002, but what original-design homebuilt ever is? It was ready to fly. I have continued to work on it since then at a rate of 10 hours or so a week. I've made many changes, added a lot of weight-185 pounds and counting-and somewhat improved its performance. But there's still plenty to do.

Little has changed in the overall configuration. Other than a couple of air inlets and some minor fairings over the flap tracks and exhaust pipe-the latter intended less to reduce drag than to keep exhaust stains off the fuselage-the only change to the external aerodynamics has been the addition of upturned wingtips. I had built the wing without tips hoping to be able, if necessary, to tailor the dihedral effect-the ability to pick up a wing with rudder-after flight testing. As it turned out, an adjustment was necessary. The tips I added, with 45 degrees of sweep and 30 degrees of dihedral, brought the roll-yaw coupling into the acceptable range. These aren't winglets, by the way. It took me a while to get used to their look of having been added merely to comply with the dictates of fashion; I prefer the appearance of the more austere, clip-tipped wings of the 1950s.

Not surprisingly, the wingtips had no discernible effect on speed or rate of climb. The only change that has yielded a distinct speed increase so far was the addition of nosewheel doors. I had always intended to have them, but had been stalled by my inability to come up with a way to operate them with a mechanical linkage-bellcranks and pushrods being, obviously, morally purer than springs and cams. The mains are fully enclosed, and their doors open and shut under the control of a complicated but aesthetically rather satisfying (to me, at least) mechanical linkage. The problem with nosewheel doors in general is that they have to open very quickly when the gear is coming down and stay open very late when it's coming up, and wait around the rest of the time. It's hard to make a purely mechanical system that does nothing for a long time and then suddenly does a lot. Many nonhydraulic systems solve this problem by spring-loading the doors open and having the gear engage some sort of lever or stirrup on the way up, pulling the doors closed behind it. I finally gave up trying to figure out a prettier way and followed the crowd.

I had not considered the open well a big source of drag because I supposed that since it was a sealed cavity in a flat surface, air would just blow past it. Wrong: Closing the well reduced the total drag of the airplane by nearly 10 percent.

In the process of adding the wingtips I had to decide whether to provide access to the attachment of the pitot head, which was mounted on the end of the then-tipless left wing. I decided not to; why would I ever need to get at the pitot head?

Then last year, on the occasion of my biennial static/transponder check, the mechanic announced that he had discovered a leak in the pitot plumbing. I was surprised, since the plumbing consists almost entirely of lengths of aluminum tubing with epoxied aluminum sleeves at the joints. What could leak? Bowing, however, to authority, I went through the system, starting at the panel and working outward. I found nothing. Having finally arrived at the pitot head, I concluded that the leak must be there, securely buried inside the wing. I excised a hunk of the trailing edge, burrowed forward through the foam core, and finally extricated the pitot-only to find that it too was leak-free.

The leak, to make a long story short, turned out to be in the shop's testing equipment. But now I had cut up the pristine left wingtip, and so-in for a dime, in for a dollar-I sawed off its trailing edge, hinged it, and put a little servomotor into the cavity where I had excavated the foam. Thanks to the shop's leaky test equipment, I now have roll trim. It's not very powerful, but a single-engine plane shouldn't need powerful roll trim, and few have it at all. It's sufficient for trimming out fuel imbalances early in a flight; when the airplane is in balance an electronic timer, which switches tanks at seven-minute intervals, takes over.

Engine cooling gave me some trouble. At first, the oil temperature was high; flushing out the cooler, which, like the engine, had been sitting around for 20 years collecting varnish, fixed that. Cylinder head temperatures were uneven, with the right rear and left front cylinders running considerably hotter than the others. The engine, a 200 hp Continental TSIO-360, is obviously designed for downdraft cooling, and I was using updraft; but that did not explain the strange disparity in cylinder temperatures. For a long time I could not understand what it was about those two cylinders, located diagonally opposite one another, that made them hotter than the others.

If you think about something long enough, however, a light usually goes on. I still have not unraveled the mystery of existence-why is there something, when there could be nothing?-but I did at last notice that what those two cylinders had in common was that the unfinned surfaces of their exhaust ports were not adjacent to another cylinder. On all the other cylinders, the neighbor formed a converging channel that accelerated a stream of air over the hot exhaust-port surface. I added baffles to duplicate that effect on the two corner cylinders, and their temperatures dropped by 40° F.

But still there were cooling issues. I could not maintain a long climb on a 100° day without temperatures getting near the top of the green, and, conversely, on below-standard days I could barely keep temperatures from dropping out of the bottom of the green at economy cruise or when descending. An additional problem turned up during long periods of taxiing or idling: Fuel would begin to boil in the injector distributor, which sat on the top of the engine, making for an extremely rough idle.

I fixed the fuel problem by moving the distributor to a position below and in front of the engine, in the single cooling air inlet under the spinner. Now almost all of the fuel plumbing is on the cold side of the engine.

I experimented a good deal with cowl flaps. The air outlets are in an unusual location, in the cowling top just behind the spinner. I tried chutes with a simple flush exit ramp, and then with a guide vane in the middle of the chute. The guide vane approach looked promising, and I tried it first with the trailing edge of the vane flush with the top of the cowl and then with a raised trailing edge, so that the vane formed a sort of hood over the outgoing air. Each modification brought maximum temperatures down further while of course making the overcooling problem more severe. Interestingly-since cooling drag is a matter of great concern to experimenters-none of these changes affected speed in a measurable way.

Eventually, I mounted the vanes on pivots with an electric motor to operate them. Cooling on hot days is no longer a concern, but this is not yet a perfect solution. Temperatures are still too low at economy cruise, even with the air outlets choked down to a total of 18 square inches (the inlet area is 55), and while the engine runs too cool, the oil temperature still hovers some 30° F above the recommended 170°.

I usually cruise at around 11,000 or 12,000 feet, 165 ktas, eight gallons an hour. As an adherent of the Church of Alcor and GAMI, I lean 50 to 75 degrees to the lean side of peak EGT. The tuned GAMI injectors keep the EGT peaks of the different cylinders sufficiently close to one another that such a lean mixture does not cause roughness (which is usually due to different cylinders running at different mixtures, and therefore developing different amounts of power). I have not been able to make use of very high altitudes, however, because I start to get misses above around 16,000 feet. This was a problem on the first Melmoth as well, until I installed pressurized Slick mags; I am currently using my original unpressurized Bendixes again. I'll get around to putting the Slicks back on one day, but for the most part I prefer to cruise below oxygen altitudes anyway.

I still have a long list of future projects. One is to move the aileron hinge lines aft in order to reduce the stick forces in roll. But the principal and current one is to get the flaps into working order. The airplane has big Fowler flaps that travel all the way to the trailing edge of the wing before deflecting 30 degrees; but it has no actuation system for them. Actually, I've never missed them; I usually operate on the light side, the airbrake-two square feet of board that pops out of the belly-slows the airplane down and keeps it from floating, and the wheel brakes are powerful. But it's not about practical necessity; it's about making a certain style of airplane-a rather complicated one, full of confusing gadgets.

What has held the flaps up for a long time has been the design and fabrication of a system for synchronizing them. In retrospect I think it would probably have been simpler to do this with screw jacks and torque tubes, as Boeing does, but, for whatever now-forgotten reason, I decided to actuate them hydraulically. Since hydraulic fluid neither apportions itself equally to the two sides of an airplane, nor cooperatively moves the inboard end of a tapered flap faster and farther than the outboard end, I had to come up with a system to make it do those things in spite of itself. By the time you read this, the results of my efforts will probably have found their way onto the website where I sporadically narrate the airplane's progress, melmoth2.com.

By the end of its second year, the first Melmoth had flown to Europe, and to Japan by the end of the third. Melmoth 2 is a more backward child; despite its 3,000-mile range, it has not been more than 2,000 miles from home. Ah, for the days of 50-cent avgas again!

Also read these related stories:

Melmoth Flies… Again!

Melmoth 2: A Personal Airplane

Cleaning Up Melmoth