My Own Private Wichita

** Melmoth under construction in 1970
beneath a grape arbor in the writer’s
California backyard. It took 6,000 hours to
build, and flew 2,000.**

Reading Ken Scott's very sound advice to would-be homebuilders in Flying's May issue got me thinking about my own experiences, and how things have changed since I got involved in homebuilding.

I started thinking about building an airplane around 1963. At that time there were no prefab kits on the market and only a limited number of plans. Most homebuilts were sport biplanes or small wood or tube-and-fabric cruisers like the Wittman Tailwind. Like most young men I was entranced by speed and fighter-like looks and flying qualities, and so I leaned toward something like the Long/Bushby Midget Mustang, a handsome single-seat taildragger of all-metal construction. But although plans for such airplanes could be bought, I don’t recall ever considering them; I wanted to design an airplane myself. This was a project for which my only qualification — at the time I was on a two-year hiatus from college, where I majored in English — was apparently boundless self-confidence. I knew nothing about engineering, stress analysis, calculus, aerodynamics, materials, construction methods or aircraft systems.

In various college engineering libraries I pored over NACA technical reports and the one or two books on practical construction that existed then. I remember that K. D. Wood's Aircraft Design, in a floppy red binding, was full of fascinating details, many of which were either over my head or irrelevant to the kind of airplane I wanted to build. Despite all the reading, however, I think that much of the real learning consisted not of facts systematically absorbed from study but of insights that came to me from random and often unexpected sources.

For example, the phrase “load path” is often heard in discussions of structures. Forces acting upon a structure can be thought of as three types: tension, compression and shear. The first two correspond to pulling and squeezing; the third results from forces acting in opposite directions on an object, like the blades of a pair of scissors on a sheet of cardboard. If a wing, for example, supports the weight of an engine that is far away from it, there has to be some structure connecting the two, and it should be possible to say what sort of compressive, tensile, and shearing stresses and strains are occurring between one and the other.

This idea seems pretty basic, but my ideas about its practical application remained nebulous. Then I happened to pick up a magazine that contained one of those wonderfully detailed see-through drawings, this one of a Thorp Sky Scooter. The Sky Scooter’s engine rested on a narrow bed mount rather than a weldment of steel tubes. The tension in the upper part of the bed mount had to get into the upper fuselage longerons, which were much farther apart than the bed mount was wide. The load path was through a shelf in the firewall: Tension became shear, shear turned back into tension. This picture brought me my little load-path epiphany: The shelf was analogous to a spar. It is so obvious now that it seems bizarre that it could ever have seemed like a revelation; but that is how learning is done.

Another example: the Scooter’s designer, John Thorp, whom I later got to know and who helped me a great deal, liked to point out the counterintuitive fact that all steel, whether it be a wire hanger as soft as cookie dough or a tempered spring that you can’t scratch with a file, is equally stiff. Now, this is something that engineering students learn in the first week of Materials 101, but I never took Materials 101, and so notions of elasticity, stress and strain had to come to me by other routes.

The concept of stiffness, or more precisely, of “elastic modulus” — how much a material changes size under a given load — is very important in design. Here is why: Suppose you have a 100-pound weight, and you have some 75-pound-test steel wire and some nylon monofilament, also of 75-pound-test, to hang it from. Together the numbers add up to a 50 percent margin of safety, but actually both components will break, because the resistance of each support is only as great as the stretch put into it. The stretchy nylon will provide hardly any support at all while the steel wire reaches its limit and breaks; then all the weight falls upon the nylon, and it breaks as well.

The steel/nylon example is unrealistic, but the idea of relative stiffness turns up all over the place. Suppose you have a wing with two spars of equal strength, but, because of the shape of the airfoil, one spar is deeper in section than the other. It will be stiffer, and therefore will take the lion’s share of the stress whether you want it to or not. The same principle applies all over the airplane; the arrangement of parts, rather than their strength, can determine how they will behave under load.

Common phrases — “shear lag,” “diagonal tension,” “compression crippling,” “stress concentration” — first hung there outside me, elusive and mysterious, before their physical significance eventually sank in and became intuitive. Years passed. By late 1968, after a couple of false starts, I had begun building the first airplane I would actually complete. I started with parts — rudder, ailerons, trim tabs — that I understood and could fabricate with the tools I had available, in the meantime working on the design of more difficult components. I used a slide rule for calculations — a wonderful tool, now seldom seen, that beautifully balances responsibility between itself and its user. I found that many seemingly difficult operations, like calculating the distribution of material in spars and ribs, could be done graphically.

Using a geometrical procedure that dates back to the French philosopher Pascal, I constructed conic sections for fuselage frames on sheets of plywood. These would be sawed into form blocks over which I would hammer out fuselage frames and wing ribs with a leather mallet and a bar of plumber’s solder. Formed in soft aluminum, the parts returned from heat treatment on dry ice, twisted like Pringles. They needed to be hammered onto the form blocks again to relieve stresses, and then allowed to warm and harden there. Today, a kit would provide all such parts ready-made, formed on hydraulic presses, most likely with the rivet holes already punched in them.

Working out of doors, I secured my hand-hammered frames to a wooden spine, wrapped sheet aluminum around them, and drilled hundreds of rivet holes, temporarily holding the skin down with little gadgets called clecos. When every hole had been drilled, the whole assembly would be taken apart, the holes deburred, each hole countersunk — frames and skins separately — and everything acid-etched and primed with zinc chromate. Then I would reassemble it with clecos and drive the rivets that would hold it together permanently. Sheet metal that had been loose and floppy became tight and rigid. At the end the wooden jig was removed, and the subassembly was ready to be connected to another.

In this manner, learning as I went along, I built fuselage, wings, empennage, hydraulically retractable oleopneumatic landing gear, double-slotted Fowler flaps, variable-incidence ailerons, wing-mounted airbrakes, and an IFR panel composed largely of World War II surplus. It was a lot of work — somewhere between four and five years, more or less full time. By the time I finished, I knew — more or less — how to do it.

Comparing that approach with today’s tab-A-in-slot-B kits, I marvel at the elasticity of the FAA’s 51 percent rule. I used to tell people who talked about wanting to build an airplane that they had better like building at least as well as they liked flying, because they were likely to do a good deal more of the former than the latter. That’s no longer true. Today I think you can reasonably expect to put more hours into flying a homebuilt than into building it.

Still, I think the building should not be considered something that just needs to be gotten out of the way. After my first airplane was wrecked on the ground by an errant Cessna in 1982, I spent 20 years building another, this time of composites, also completely from scratch. Clearly I had not learned my lesson. But no, seriously, the real lesson was how much I liked the work. The meditative hours spent in an artisan’s solitary and attentive state are not a painful price paid for the realization of a dream. They are good hours. They are, like the flying that follows, their own reward.

We welcome your comments on In order to maintain a respectful environment, we ask that all comments be on-topic, respectful and spam-free. All comments made here are public and may be republished by Flying.

Peter Garrison taught himself to use a slide rule and tin snips, built an airplane in his backyard, and flew it to Japan. He began contributing to FLYING in 1968, and he continues to share his columns, "Technicalities" and "Aftermath," with FLYING readers.

Your email address will not be published. Required fields are marked *

Subscribe to Our Newsletter

Get the latest FLYING stories delivered directly to your inbox

Subscribe to our newsletter