RAF started in the pre-digital age, when the tools of the aeronautical engineer were still slide rule and drafting paper. Around 1980, however, Rutan bought an Apple 2 computer. When the salesman asked him whether he would like a second 160 kilobyte floppy disk in addition to the one supplied with the computer, he declined it, saying that he would certainly never fill even the first disk. Some time later he proudly showed me a Corvus hard drive for which he had paid thousands of dollars. It had the unimaginably vast capacity of 10 megabytes.

I too had acquired a computer, but not a printer. So when I developed a rudimentary program for generating fuselage cross-sections, I delivered the results to Rutan in the form of snapshots of my computer screen, developed and printed in my basement darkroom.

During one of our Wednesday lunches at Reno’s Cafe, Rutan mentioned that some fellow in the Midwest had written to offer to design airfoils for him. At that time, most designers picked their airfoils from catalogs developed by the National Advisory Committee for Aeronautics in the 1930s. I suspect that Rutan may even have designed some of his with a French curve and draftsman’s spline, relying on the time-tested principle that they ought to be round in front and pointy in back. This Midwestern fellow said he could do better, using a computer that he had built and codes he had written. Thus did digital simulation come to RAF.

• READ MORE: Gliders Are Form of Flight That Resonates Above Others

Eventually this Midwestern fellow turned up in Mojave, and it was there that I met him. His name was John Roncz. He was cherubic: plump and pink, with a boyish voice and air of innocent candor. He was, I would soon learn, in addition to his coding and analytical skills, sweet-natured and generous and a formidable classical pianist. His day job, rather incongruous, involved running a metal stamping business he had inherited from his father.

Thereafter our paths would cross from time to time, and at a certain point they converged. A pilot himself—he once glided a Commander 112 to an airport landing after a night engine failure—he had by then left the metal stamping business and become a full-time computational aerodynamicist— that is, someone who uses computer simulations in place of wind tunnels to discover the properties of airplanes and their various parts. He was using a computer program called VSAERO for his analyses of complete aircraft.

It happened that I and a partner, Dave Pinella, were selling a package called PSW that bundled a program I had written for defining airplane geometries with Pinella’s programs for analyzing the digital models and displaying the results. Our analysis code, Cmarc, was descended from VSAERO, and their input “decks” were sufficiently alike that Roncz and I could conveniently collaborate. My main contribution was turning complicated geometries into digital models digestible by VSAERO and Cmarc.

Though we seldom met in person, we corresponded copiously and became good friends. When I needed an answer to some difficult question for an article— such as, how much power would be required to hover a 7,000-pound, four-rotor electric air taxi in ground effect?—I would email Roncz, knowing that a reply would come to me within hours. Roncz’s contributions to aviation were many and significant. He created airfoils for Voyager, the Beech Starship and other airplanes, including mine, and designed several complete airplanes. He also designed the wing sail for a victorious America’s Cup boat and odds and ends such as windmill blades and race car downforce wings.

He had a merry sense of humor and would name his airfoils with funny acronyms. The laminar profiles for my second Melmoth were SODA (Stamp Out Drag Airfoil) and POP (Peter’s Other Profile). When Rutan needed an airfoil with extra trailing-edge thickness for a complex Fowler flap, Roncz produced OSPITE (Olympic Swimming Pool in Trailing Edge). A STOL project yielded GOLA (Gobs of Lift Airfoil). Although his work involved scrutinizing mountains of numbers, Roncz was not an obsessive drudge. He laughed often. Beset by intermittent maladies and amorous tragedies, he had an entirely separate life into which he would disappear from time to time and in which I suspect he found more satisfaction than he did in his fluid-dynamics work.

In this other life, Roncz was a medium. He would regularly visit Arthur Findlay College in England—“the real Hogwarts,” he called it—to practice his spiritualistic skills in sessions in which he would stand before an audience and, as he described it, say whatever came into his mind. Invariably some astounded stranger would confirm that what he had said was true, and what’s more, there was no possible way he could have known it.

Roncz was quite aware of the incongruity between his two lives and said he was just as puzzled as anyone else about how “it” worked—“it” being his weird ability to hit so many invisible nails on the head. He wrote a book, An Engineer’s Guide to the Spirit World: My Journey from Skeptic to Psychic Medium, part autobiography and part case study, in which he matter-of-factly laid out his dealings with the “spirit world” without making any attempt to provide a scientific-sounding explanation for it.

Being on the knuckle-dragging level of spiritual evolution myself, I could never view Roncz’s mediumistic side other than with a skeptic’s condescending amusement. But he didn’t mind. For him, it was a lived experience, and my doubts were like those of a shut-in who questions a visitor’s assertion that outside the sky is blue.

Roncz died in September of cancer at the age of 75. I talked with him a little before his death and suggested that he come knock on the wall of my house once he was comfortably ensconced in the spirit world. He laughed and said he hoped he would.

I have not yet heard the knock. But his airfoils keep me aloft, and whenever the raccoons are dancing on the roof, I think of him.

This column first appeared in the January-February 2024/Issue 945 of FLYING’s print edition.

The post Recalling a Good Pilot Friend and One Curious Character appeared first on FLYING Magazine.

Recent years have seen a general drift toward longer spans and higher aspect ratios. The Beechcraft Bonanza has a span of 33.5 feet and an aspect ratio of 6.2; the Cirrus SR22, which might be seen as today’s Bonanza, has a span of more than 38 feet and an aspect ratio of 10.1. The trend is generally toward greater aerodynamic efficiency, partly in response to fuel costs and partly because the increasing use of turbocharging leads to higher cruising altitudes, where longer wings are more at home.

The two airplanes I’ve built, Melmoth and Melmoth 2, are (or were—the first Melmoth was destroyed in an accident long ago) broadly similar, with low wings, T-tails, bubble canopies, retractable gear, and the same 200 hp Continental 360 engine and Hartzell constant-speed prop. The first Melmoth was aluminum, with 2+1 seating; the second is composite and seats four. Both were built with long-distance travel in mind and have lots of internal tankage: Melmoth’s wing and tip tanks held 155 gallons; Melmoth 2’s completely wet wings hold 142 gallons. The two Melmoths, with the same engine, propeller, empty weight, and cabin cross-section, differ significantly in one aspect: wingspan. The first began life with a wingspan of 23 feet and went through 21-foot and 28-foot iterations before its eventual demise. Melmoth 2 has a wingspan of 36 feet but only a little more wing area—106 square feet to the first Melmoth’s 93. (For comparison, the wing areas of most commercial four-seaters range from 145 to 180 square feet.) The first Melmoth’s aspect ratio was 5.75; Melmoth 2’s is 12.6.

Span and area are entangled with one another in the sense that structural strength and stiffness (not to mention space for retracting landing gear) require a certain wing thickness, and that in turn implies a minimum chord (the distance from leading to trailing edge), because airfoils shouldn’t be too thick. So you can’t just increase span willy-nilly without at some point having to increase chord and area as well. However, increasing the wing area, which was originally selected to permit a certain landing speed at a certain weight, adds drag and makes the airplane heavier.

Increased wingspan—other things remaining the same—rewards you with better efficiency and climb rate, and improved high-altitude performance. The first Melmoth had a maximum lift-drag, or L/D, ratio of about 11.8 and a “Breguet range”—a fictional, greatly exaggerated number that ignores takeoff, climb, and varying engine efficiency and assumes that you always fly at a low and ever-decreasing ideal speed—of 3,000 nm. Melmoth 2, with half again the span, has an L/D ratio of 17 and a Breguet range of 3,600 nm, despite carrying 8 percent less fuel. Rate of climb is less strongly influenced by span than L/D and range are, but Melmoth 2, climbing at 1,800 fpm at full power and a typical weight of 2,200 pounds, betters the original Melmoth by about 20 percent.

Note that I said “half again the span” and added nothing about aspect ratio. That is because, contrary to widespread belief, aspect ratio actually does not enter into it. Aspect ratio is generally thought of as the quintessential measure of efficiency, but if you could double an airplane’s wing area (thereby halving the aspect ratio) without increasing its parasite drag, the L/D ratio and Breguet range would remain the same. But you can’t increase wing area without increasing drag and weight, and that’s why aspect ratio becomes important: It’s a measure of how little wing area you can have with a given span.

Curiously, and I think unexpectedly for most pilots, altitude also does not enter into it. You might intuitively suppose that thinner air would make the airplane more efficient, but in fact neither the maximum L/D ratio nor the maximum range is affected by altitude.

You will object that at 8,000 feet you will go faster, with the same fuel flow, than at 2,000 feet. True. But that is because your indicated airspeed is lower. If you flew at the same indicated airspeed and fuel-air ratio at both altitudes, you would find your fuel flow is greater at the high altitude. The reason is that drag at a given indicated airspeed is the same at all altitudes, but the power required to overcome it is proportional to the square of the true airspeed, not the indicated airspeed. At the bestrange speed, the miles per gallon is at a maximum, however, and is unaffected by altitude except to the extent the engine’s efficiency might vary at different settings of manifold pressure and rpm.

• READ MORE: Do Flaps Make Lift?

“Best range” and “best efficiency” are not seen in normal flying. Under actual cruising conditions, Melmoth 2 is not that different from the original Melmoth. The reason is that maximum L/D and the Breguet range assume speeds that are quite low—around 40 percent above the clean stalling speed—and remote from those we actually use. At real-world speeds, 65 percent or 75 percent power, the differences shrink. Melmoth 2 will cruise at 170 knots at 12,500 feet using about 8.5 gallons an hour—about 60 percent of rated power; the first Melmoth would burn about 9.6, around 70 percent power, at the same weight and altitude. So you see that despite a 50 percent improvement in best L/D, the practical benefit of the longer wing is much smaller.

When I designed the first Melmoth, I was strongly influenced by John Thorp and his T-18 homebuilt, whose wing I copied almost exactly. Thorp, who also designed the original rectangular-wing Piper Cherokee, used to say that low aspect ratio wings perform better than theory would lead you to expect, and he was adamant there was no reason to taper the wings of any airplane weighing less than 12,000 pounds. When I designed Melmoth 2, however, I was more influenced by Burt Rutan’s derisive observation that if I intended to fly long distances, I had certainly chosen the wrong wing to do it. Aesthetics, too—hence the long, slender, tapered wing of Melmoth 2, of which Thorp might have disapproved.

For efficiency—the least fuel burned for the most work done—a large wingspan is necessary. But Melmoth 2’s long wing cost it the rollicking roll rate I enjoyed so much in the first Melmoth. Melmoth 2 rolls more like an Airbus. Sometimes, I think I would pay for the extra fuel just to have the rolls back.

This column first appeared in the November 2023/Issue 943 of FLYING’s print edition.

The post The Importance of Wingspan Can’t Be Underestimated appeared first on FLYING Magazine.

It was a clear, calm day. We climbed out of Whiteman Airport (KWHP) and through the Newhall Pass, turning northwestward toward Gorman. I leveled out at 10,500 feet. Mount Pinos crept past our left wing and the vast flatness of the San Joaquin Valley lay fading into the haze before us.

How do you know something is wrong? Sometimes it’s obvious: a big noise, shaking, smoke, pieces flying off. But sometimes it’s so subtle that you ask yourself if you’re imagining things. There was something different about the hum of the engine. Crossing oceans, out of sight of land for 10 hours, I had become intimately acquainted with automatic rough. I knew how an anxious mind could tease ominous signals out of the engine’s chaotic noise. But now the engine was not rough—it was just different.

I clicked through the EGTs. Nothing there. I fiddled with the mixture. Nothing. But whatever it was, it was getting worse. “There’s something funny about the engine,” I said to Bob. He reacted with a characteristic face that meant, as far as I could tell, “Hmm.”

The change grew more insistent, harsher, impossible to ignore. I checked the mags. Roughness on the right—that must be the problem. I decided to land at the earliest convenience, which in this case meant Porterville.

Once on the ground, we opened up the cowling, unscrewed the inspection plug in the right mag, and peered inside. Well, that was easy. Some teeth were stripped from the plastic timing gear.

I’ve forgotten now—I don’t possess the faculty of total recall apparently enjoyed by other autobiographers—what became of Bob, but obviously, he jumped ship and found a car to Bakersfield and a flight to Reno. I waited while the local A&P pulled the mag, rebuilt it, and put it back onto the engine. I did a runup; all OK. I paid the bill and taxied to the runway.

Full throttle, rotate, gear up. Only when I was in the air did I realize that I was not seeing the expected 2,800 rpm, just 2,600, and the engine still felt funny. I went around the pattern at 500 feet, landed, taxied in, and parked the airplane.

It was evening now, and I was stranded. I called a friend, Mike Melvill, chief test pilot at Burt Rutan’s operation at Mojave. Could he come rescue me? Sure. It was night when he landed in his VariViggen, a queer-looking, tandem-seat, wooden delta-wing canard that was Burt’s first offering to the homebuilding world. Mike, a machinist by trade, had built the airplane while living in Indiana with his wife, Sally (they had been teenage runaways in South Africa, but that’s another story). He had brought it out to show to Burt, who promptly offered him a job with the nascent Rutan Aircraft Factory and an opportunity to live in unappetizing Mojave. In a particularly lucky gamble, Mike took the job and ended up becoming the world’s first civilian astronaut.

But that was later. Now, we made the starlit flight over the mountains back to Whiteman, where he dropped me off, waved away my thanks, and disappeared into the darkness.

It took me a couple of days to rustle up a big trailer from John Thorp and a suitable tow vehicle, a pale blue and cream ’57 Ford pickup from Homer Knapp, a legendary North Hollywood motorcycle machinist who would let me make parts in his shop. Early one morning, Homer and I drove to Porterville.

Getting Melmoth, whose gear track was 11 feet, onto the 8-foot-wide trailer was not easy. Homer, a resourceful man, figured it out using some irrigation pipe and two-by-eights he picked up somewhere. We weren’t street legal, but we figured we looked strange enough that other cars would give us a wide berth. It was night when we drove back over the mile-high pass that separates the San Joaquin Valley from Los Angeles. We left the trailer and its cargo in front of my hangar.

When I returned in the morning, one of the trailer’s tires was flat. It had timed its failure considerately.

I took the engine off the airplane and delivered it to Mike Attolico at Western Cylinder, who had overhauled it five years earlier. It wasn’t until well into a full teardown, when they detached the No. 1 connecting rod from the crankshaft, that they phoned me to come have a look. 

The crankshaft was broken.

The rear cheek of the No. 1 throw had parted clean through. Because the break was at the back end of the crankshaft, the engine had continued to develop power. The two pieces of the broken cheek had remained interlocked by the angle of the fracture and had continued to drive the accessories and the mags. The small, rapidly oscillating slippage between the two parts of the crank had broken the teeth on the right mag’s timing gear and was also the cause of the unfamiliar vibration I first felt.

So, now, I could boast the twin distinctions of having, first, experienced a broken crankshaft, and, second, taken off and flown around the pattern with one.

I sent the smaller piece of the crank to Glenn Dail, an investigator at the National Transportation Safety Board. “This did create quite a bit of interest among the metallurgists at NARF [Naval Air Rework Facility] Norfolk,” he wrote back in December. “Classic fatigue, it is.” It seems that most cranks that fail—not many do—fail near the propeller end. In this case, a fatigue crack had begun just below the surface at an “inclusion,” a tiny pellet of non-metallic material, and had gradually propagated outward, weakening the crank until it broke. All steel contains inclusions—this one just happened to be close to the surface, in the radius at the highly stressed boundary between the cheek and the journal.

“It could have begun somewhere just south of Shemya,” Glenn cheerfully wrote, referring to the lonely island at the tip of the Aleutian chain that Nancy and I had passed when flying in Melmoth to Japan three years earlier.

There was something to think about. How long, really, had that crack been propagating? No way to know. But the seed of it had been with us from the start, biding its time, nibbling at the heart of the engine, a secret saboteur who knew no difference between Porterville, California, and the middle of the North Pacific Ocean. But oh, the difference to us!

This column was originally published in the December 2022/January 2023 Issue 933 of FLYING.

The post The Propagation of a Broken Crankshaft appeared first on FLYING Magazine.

My friend Longbridge has been working for years—these things always take far longer than you think they will—on a Lancair 320 with a lot of airframe mods, the most conspicuous of which are a double-slotted Fowler flap, enlarged empennage surfaces, and leading-edge cuffs on the outer panels of the wings. And then there are the powerplant things—some engine and cooling mods, electronic ignition and so on.

He finally got it finished a few months ago. Then he hired another pilot to do the first flights.

I was surprised. First of all, Longbridge is an experienced pilot himself. He has modified and tested a series of airplanes, all of which he got to go faster and stall slower. So it’s not as if he were a stranger to flight testing. But, to my mind at least, there was the jus primae noctis aspect of the thing. Don’t you want to be the first?

He said three things had entered into his decision. He had no stick time to speak of in Lancairs, the airplane was heavily modified, and he had not flown in the past 15 months.

This got me thinking about first flights.

I did the first flights in both of my homebuilts, which were original designs and therefore of unknown quality. I never even gave a thought to having someone else do them. I suppose the same rashness that made me think I could create an airplane made me imagine I was equal to whatever surprises it might spring on me. (A seven-minute video of the first flight of my second airplane can be found on my YouTube channel—my first upload. It’s rather boring but gets livelier at the end. This wasn’t strictly a first flight, though; I had made three long hops on what is now the Mojave Air and Space Port’s 10,000-foot runway the day before.)

I know that a disproportionate number of accidents involving amateur-built airplanes occur on first flights or within the first few hours of testing, and that not everyone is going to be as lucky as I was. In an excellent article on this subject in Kitplanes magazine, Ron Wanttaja reported that during a 10-year period in which more than 10,000 new homebuilts took to the air, about one in 130 first flights ended in a reportable accident or incident. He noted that while this rate is similar to the annual rate for all homebuilts, it represents a single hour of flying rather than a whole year. (He might have added that the annual rate would be lower if the first-flight mishaps were omitted from the denominator.)

The causes of first-flight mishaps break down very roughly into three main categories: builder error, powerplant problems and, finally, our old nemesis pilot error. A catchall fourth category, “other,” accounts for about a tenth of mishaps.

Builder errors can be quite serious and particularly difficult to cope with in flight. Elevators or ailerons rigged backward are almost always fatal—trim not so often, but still a hazard. Missing fasteners, incorrect fastener types (hardware store bolts instead of aircraft ones, for instance), faulty adhesive bonding, loose nuts, missing cotter keys or safety wire—any of these can be hard to spot and can result in a structural failure or loss of control.

The risk of assembly error increases when, as is usually the case, an airplane has been built at home, disassembled, trailered to the airport and reassembled there. I flew my first airplane for 50 hours before I noticed that a bolt connecting a pushrod to a bell crank in the aileron linkage had no nut on it at all. Luckily, gravity had kept it in place.

Read More from Peter Garrison: Technicalities

Most such problems can be prevented by careful preflight inspection using multiple sets of eyes. Just as authors are bad proofreaders of their own galleys, builders are bad inspectors of their own airplanes; other experienced builders or A&Ps should be involved—the more nitpicking and spiteful, the better.

Powerplant issues often involve fuel- system faults and cooling problems. An engine that rapidly overheats forces the pilot to return to land, in a heightened state of anxiety, before having gotten much of a feeling for the airplane. Fuel-feed problems can usually be detected by high-power runs on the ground; the tail should be tied down in order to hold the airplane in a climbing attitude. Some, however, are caused by construction debris in fuel tanks, and these may not turn up until a second or third flight. Fine sanding dust and fibers adhering to the walls of a tank may elude a visual inspection. There have been many instances of loose connections in fuel systems, perhaps because fuel-system plumbing is often assembled and dismantled several times during the course of construction.

Pilot error takes the usual forms: getting too slow in the pattern, overcontrol, loss of directional control during takeoff or landing, ground loops, and so on. A low-altitude stall is the mistake most likely to cause grave harm.

Whether a first flight should be a runway hop or an up-and-away flight is controversial. Neither method is guaranteed. As innocuous as a runway hop seems, it can go wrong. Engines and propellers surge, elevators turn out to be more sensitive than expected (or pilots less sensitive), and the airplane gets too high to get back down within what’s left of the runway. On the other hand, a well-controlled runway hop, just a couple of feet above the ground, gives the pilot a chance to detect nose or tail heaviness, misrig or mistrim, engine feel, and braking effectiveness. The benefits outweigh the risks, in my opinion, but I would take the trouble to find a reasonably long runway, even if it meant some inconvenience.

It’s a mistake to invite friends and family to witness a first flight. The pressure to hurry, to satisfy expectations, to proceed even when feeling misgivings is multiplied by each set of eager eyes. Even factories perform a secret first flight before the public one. It’s better to have a few knowledgeable people in attendance—not as cheerleaders, but as careful observers and critics.

It used to be that only the pilot was allowed to be aboard during the testing phase of a homebuilt airplane’s life. A few years ago, the FAA—reasoning that some mishaps might be prevented if a second pilot, presumably more competent than the builder, were present—has changed that rule. If a more competent pilot is available, you might ask, why not just have the more competent pilot do the first flight? Because it’s a learning experience for the owner.

And this brings us back to my friend Longbridge and his Lancair 320. He had owned and modified an MFI-9 and an RV-7A earlier, so he knows about light stick forces. (He also had a Cardinal, which he cleaned up considerably, but it probably didn’t teach him much that applies to a Lancair.) His engine had already flown 130 hours, so there were no break-in requirements. His aerodynamic modifications were aimed to make the airplane more docile, not less. So why not just go ahead and fly it?

An excess of caution, I guess. Maybe when you’ve been married a couple of times already, the right to be the first doesn’t seem so important any more.

This story appeared in the November 2020, Buyers Guide issue of Flying Magazine

The post Maiden Flights in Homebuilt Aircraft appeared first on FLYING Magazine.