Triplane builder Bill Woodall in his own words
Triplane Builder - Articles by Triplane Builder - William (Bill) Woodall
Getting hold of the original plans
The London office of the company for which I worked, was able to provide me with an address for the public relations office of Hawker-Siddeley.(Hawker-Siddeley-[now British Aerospace]--being the continuation of what had been the Sopwith Co.)
I wrote them, and they assured me that the original blueprints for the Sopwith Triplane still existed. They made a quote which included the cost of reproducing, and shipping.
After some soul searching, I mailed an international money voucher, and waited. Eventually, a large carton was waiting at customs at the local airport.
The drawings, dating from early 1916, appeared to cover everything from a stress analysis of the wings, to crating instructions for sending machines off to war. Some of the blueprints were still marked that they were subject to the " War Secrecy Act".
When I compared the type of parts that the airplane would use, to those listed in supply house catalogues used by home-builders, it was obvious that very few "off-the-shelf items" would be available to me. This would truly be a do-it-yourself project.
You could still buy Sitka spruce lumber, and birch plywood. The original aircraft had used lots of that, but beyond a few pulleys, turnbuckles, bolts and nuts, not much else was recognizable.
(If the reader does not have much interest in the nitty-gritty of home-building, he should bail out now, and flip to the back in hopes of finding something he can better relate to.)
The Triplane workshop, skills and tools
One question that is often asked is what kind of a workshop is required to try to build an airplane like this. In my case, a most modestly equipped one.
I had a vise. (Before it was over, I had wrecked several--they aren't designed for metal forming). I had a small jig saw (some call it a scroll saw), saber saw, table saw, and various hand saws. I acquired a band saw before getting very far into the project.(By installing 2 motors, and a belt reduction system, I was able to cut wood and metal). My shop also had the usual hacksaws; coping saws; planes (including a Shinto); Surform tools; a small bending brake; a metal nibbling tool; snips; a hand held spring winder; a router;some ordinary wrenches, "C" clamps, and an electric drill. I had nothing very special.
Although not exactly an aircraft factory set-up,airplanes of that era could almost be considered a cottage industry product.
The fuselage was built as a wooden box girder. That is, wooden longitudinals, verticals, and horizontals, assisted by small metal fittings, lengths of piano wire, and turnbuckles, were kept in compression by tension in the steel members. (A most efficient use of each material)
It was apparent that skills would have to be developed in the bending of sheet metal, and the forming of many gauges of music wire. This instead of the more normal task of learning how to gas weld.
Help from unlikely places
I was fortunate that my employer had amassed a library of donated books that spanned a considerable period of time. One such book was a work by Colvin and Colvin dating from 1919 which detailed the" aeroplane" standards of that time. The book didn't explain "how" you were supposed to manufacture any of these items, but it did show dimensions, and provided drawings depicting "good" and "bad" workmanship.
Another bit of serendipity was my being allowed to make the acquaintenance of a builder in an adjoining state who was constructing a Sopwith "Pup". In addition to introducing me to World War 1 Aeroplanes, he had a list of people, and companies that had bits, and pieces of things that I would eventually need, and got me in touch with a sub-culture of other "nuts" working on things like this.
One of these people was to me a legend. He had produced many museum quality reproductions of airplanes of this period, including a Sopwith Triplane. He was a unique person. He wouldn't communicate much until he was certain that you were serious about your project. If you were just some hangar flier willing to waste his time--forget it! If you were for real, he would answer any question you raised.
I was working as a tire engineer for a major rubber company, and they operated a small machine shop that produced prototype examples of machinery required for the over-all operation of the company. The employees of this lab were talented craftsmen who could offer suggestions as to how certain items might be turned out by an amateur with limited experience and facilities. They also had expediters who were acquainted with the resources available from local supply houses. All of this information, and encouragement was invaluable.
A particular blessing was that three fellow employees were sympathetic. These three were either hobbyist machinists (with equipment at home), or in one case, an expert gas welder with oxy-acetylene tanks in his garage. Additionally, I had the friendship of an experienced cabinet maker from church. Without this kind of back-up, the project might well have foundered.
The Sopwith Triplane build starts
After a period of procrastination, I decided that a good item to test my skills and dedication would be the center section of the top wing. This would give me a start on the 240+ ribs used in the whole airplane (counting stabilizer and landing gear spreader); sheet metal box construction; wing spar shaping/ routing; music wire forming; the first usage of the some 188 turnbuckles ("strainers" to the British) the project would require; and the manufacture of an elusive elliptical steel tube needed for trailing edge.
The center-section would be a microcosm of the entire project. With copious help from the various sources, including lunch hour stints on the company's metal cutting band saw, the parts gradually accumulated, and came to be assembled. It looked good, and I was hooked. I had also learned a valuable lesson. DO SOMETHING EVERYDAY! If you don't, inertia will bring you to a halt, and re-starting will become an increasingly difficult task.
AP121 – Elusive Elliptical Tubing
Regarding the "elusive elliptical tubing". This is specified as AP121" ("Accles and Pollux 121") on the drawings, and is used for trailing edge; wing tips; and fin, rudder and elevator framing.
I had written a string of letters all over the place, in search of the last remaining supplies of this tubing. The final result was that an American company now owned the product line, and if I were willing to pay substantial start-up charges for a small production run, I could get some. This was out of my price range.
Fortunately, one of the people that I had come to know suggested a home-made alternative. This involved purchasing a round, copper finished steel pipe, called Bundy Weld tubing.
After making an adjustable roller die assembly, one could roll this formerly round tube into the correct elliptical shape to match the AP121. It meant multiple passes thru the dies, at gradually tighter settings. This used lots and lots of elbow grease and produced an anything, but straight length of material. It could, however, be returned to straightness by the judicious use of strong hand pressure, with the attendant risk of carpel tunnel syndrome. (I am fortunate that this whole operation was never video taped.)
Building the Sopwith Triplane
As a child, when I had built model airplanes, I tended to begin with the tail assembly. I may have done this to have complete parts to look at and be encouraged to go on to the more complex areas ahead, this strategy was employed on the triplane. The horizontal stabilizer became my next goal, and work began on the metal fittings.
( I had by now bought out the remaining stock of surplus World War 1 turnbuckles from a unique hardware store in an old area of Los Angeles.)
The wooden materials, were ordered, and when they arrived, it was assembled. Next to come would be the elevator, fin and rudder. I experimented with wooden tooling to produce the curved outlines that the bogus AP121 would generate.
A plan view of the curve, sawed from a board, ("Kentucky windage" was used for spring-back), provided a female mould. The edge was then routed with a core box tool appropriate for the radius of the major axis of the ellipse. It was surprising how well this worked. The minor axis of the ellipse projected out from the tooling, and would have to stretch when the tubing was bent around the wood. This it did.
Eventually the familiar curves of the Sopwith fin were accurately reproduced. Both the rudder and elevator assemblies are steel. The joints are brazed together, and some interesting transition pieces are required to blend the different diameters of tubing.
Rudder and Elevator
The rudder and elevator horns were a challenge. I wasn't smart enough to come up with a modern substitution for the two-piece steel pressings that had been used to form the hollow, edge- welded components on the original. A friend produced a wooden reproduction of the shape. The friends in the shop allowed me to use the arbor press.
With a thick rubber pad for backing, the thin sheet steel deformed into the rubber, asthe wooden tooling was pressed against it. This accurately developed first one surface, and then the other. When trimmed, matched and edge welded, the end product looked good. (The same technique would be employed on the top and bottom aileron horns; and the aileron lever on the control stick rocking shaft.)
Building the Wing Ribs
The left top wing panel was the next component to be tackled. The drawings didn't indicate how the original ribs might have been made. They employed spruce cap strips that were grooved on their inner surfaces to accept the 1/8"birch plywood webs. The webs were lightened by removing curved areas behind, between, and in front of the spars.
The drawings called for many tiny flat-head brass wood screws to hold the ribs together while the glue set--(glued and screwed) was the note on the blueprints. The modern price and availability of the correct size screws would have played havoc with the budget and time table. I, therefore, made a jig. I laid out the rib contour on a flat, stabilized, wooden platform. The upper surface could be backed by a curved wooden strip. On the under-side, at the points where the screws would have been used, I located wooden discs with an off-center hole drilled through them. Lag screws held the discs to the jig surface. When the embryo rib was in place, the disc-"cam" could be rotated to put pressure on the caps. Then when all was ready, tightening the lag screws, held the compression. Since I was using Aerolite glue, my rib production could be as slow as one per day.
The wing spars were ordered from an aircraft parts supply house. Since the spars fit within the top, and bottom camber of the wing, they are parallelograms rather than rectangles. I had them shaped by the supplier to the correct profile.
The solid wood compression ribs fit into square steel sockets that are welded to small backing plates that bolt to the spars. The ends of these fittings are bent and drilled to anchor music wire/turnbuckle fittings. The normal drag/anti drag truss is then developed, with the wires passing through the holes in the ribs.
One compression rib is unique. This is the one where the interplane strut attaches.
Here a round steel tube is used. It carries the fittings to locate the strut sheath, and the anchor for the flying wires.
The wing went together satisfactorily, so the port mid-wing was next.
The Port Mid-Wing
The span and chord is identical to the top wing. There is a cut-out near the root. The steel tube compression rib is special. The interplane strut passes completely through it. This results in staggered, streamlined openings in the top and bottom surfaces, sheathed with a thin sheet metal sleeve that passes through the openings. The sheath is continuously welded to the compression rib. Rib, and fitting manufacture continued until enough parts were on hand to make all 4 remaining panels. All that was lacking were the expensive spars.
The Ailerons
This airplane has ailerons on all six wing panels, and I decided to go for some variation by tackling these next. Not the least problem to be faced was the manufacture of 18 special hinges. These required a special rivet for hinge pins, but that is a story in itself. Suffice it to say that I was lucky.
The leading edge of the aileron was made from a thick block of wood that was severely routed between the rib attach points, and then covered by a thin slat of spruce. The spar was a round aluminum tube. Where the aileron horns attached, the rib was unique. It was solid, to accept the through bolts that held the horn assembly. After making 6, I was glad that there were no more.
Some of my associates had begun to needle me because of the lack of any fuselage work. I decided that this area would come next.
The Longerons
The longerons were to be made from 1-1/2" square ash timber that needed to be straight grained and 17 feet long. My son located some suitable material at a specialty lumber yard in Michigan, and had it ripped for me.
I built a plywood jig that developed the plan view of the fuselage. This had wooden blocks strategically placed so they would force the ash to hold the necessary curve. (The 1-1/2" square dimension tapered to 3/4" square at the sternpost).
I bought a length of flexible plastic drain pipe that was somewhat longer than the longerons. This way, by curling the ends up, it could hold hot water. The longerons were inserted,the ends of the pipe fastened up, and then filled with hot water. After a soak, the longerons accepted the bend enforced by the jig, and when dry, retained the curve.
Additionally, the bottom longerons have an upward curve aft of the cockpit, and this was accomplished by jacking up the aft end of the jig while the longerons were drying.
The sternpost is a complex part. It is a combination of steel plate, and square, and round steel tubing of several dimensions. When welded together, it keeps the four longerons in their correct relationships; accepts the triple pitched jackscrew that trims the tailplane; supports the tail skid; anchors the bungee that springs the tail skid; and provides 4 turnbuckle hook-up points that are the terminals for the tension system that is the strength of the fuselage.
Turnbuckles and wire
From the sternpost forward, there are a variety of sheet metal clips at each station where vertical and horizontal spacers occur. These clips fold around the longeron from inside to out. Horizontal bolts are located so that they will provide sockets for the vertical spacers. These bolts do not penetrate the longerons. Encircling each bolt is a folded steel strap. When the strap is slid over its bolt, it will be tangent to the longeron, and the spacer. The strap also provides the anchor point for either a turnbuckle, or the "loop" of a length of music wire. On the inboard surface of each clip, square lightening holes with their sides folded outward, formsockets for the transverse spacers.
The clip is also designed and drilled to provide tabs for the plan view wiring, as well as the diagonal wiring that ties an upper right to a lower left, and vice versa. These clips change in dimension in accordance with the expanding cross-section of the longeron until they reach the maximum at the cockpit.
At the time of the first World War, these parts were standard items, and available in quantity. In my case, it was a one-at-a-time proposition, and very time consuming.
The music wire components have been alluded to, but not described.
First, music wire in whatever gauge, is a very stiff and strong piece of steel, and it is necessary to deform the ends into uniform loops that fold back. The now parallel wire ends are encased in an elliptical spring (called a ferrule).When the ferrule is driven against the aft edge of the "loop" the free end of the music wire is bent back, alongside the ferrule to lock it.
When you consider that the British did not want threads showing when the turnbuckle was tightened and locked, the accuracy needed to make sure these wires were the right length, was a chore. (The ferrules have to be slid onto the wire before the final loop is formed). I made some of the ferrules, but one of my friends made most of them.
It is small wonder that modern fuselages are of welded steel tubing. The wooden box girder is a very labor intensive proposition. (Although hard to beat on a strength to weight basis).
The Cabane Struts
In the Sopwith Triplane, a cabane strut is unlike that used in any other airplane that I know about. It is a forward canted, large piece of wood that has a notched corner to partially sit on the bottom longeron, and a notched channel to accept the top longeron. Inside the fuselage, it is routed extensively for lightness. Above the fuselage, it is streamlined. Within the fuselage each strut provides the front support for the seat. The left strut also carries the engine controls. (Air throttle lever and manette). The right strut mounts the stabilizer control wheel, the fuel sight gauge, and up in the airstream, the Rotherham air pump (for fuel tank pressure). Both strutsprovide attachments for the dash board . In addition, the midwing box rib assemblies encircle these struts (above the top longerons) and at the top, carry the top wing center section. They are well used components.
The Rotary Engine
Since this was going to be as faithful reproduction as I could make, there was really never any thought about using a non-rotary powerplant. If the airplane had been so well liked, how could rotaries be as troublesome as some latter-day writers had made them seem? I wanted to experience flight behind one, first hand.
I've been using the term "rotary engine" rather freely without definition. This is not a rotary combustion chamber engine such as a Mazda. This is a radial engine (nine cylinders in my case), with its crankshaft fixed to the airframe. The entire cylinder assembly rotates around the crank.
The propeller is fastened directly to nose of the engine. It is a four cycle, overhead valve design.
A commutator rotates with the engine, and is energized by brushes fromgear driven magnetos.
The fuel/air mixture enters from the rear of the hollow crankcase--mixes further in the business end of the crankcase, and is drawn and thrown out through a manifold to the intake valves. The exhaust is expelled directly to atmosphere under the ring cowl. The carburetor is rudimentary, and slow in response since it takes a while for the engine to find out that a mixture change may have been made. In some cases it is easer to manage the engine RPM by killing the spark. The engine displacement is large--the RPM low (1200)--the horsepower rating very conservative, and when coupled to the right propeller, the performance is astonishing. There is a noticeable gyroscopic effect. The oil is not recovered. It is slung or burned after having been pumped through the bearings. The horsepower to weight ratio is quite good.
The Engine Plate
Rotary engines are mounted to an airframe in two places. The rear engine mount is a sheet steel box mounted transversely to the fuselage, and provides an anchor for bolting the circular flange that attaches to the crankcase extension. The front engine plate is the forward end of the fuselage. It is a heavy gauge steel plate to which a circular steel plate on the back of the engine is bolted.
At the time these airplanes were manufactured, the front engine plate was a stamped product.
It is extensively lightened by removing steel in 4 radially extending teardrop shaped slots. The perimeters of these slots fold back to form stiffeners. The sides fold back to cradle the foremost vertical struts. The bottom is an arched, folded cut out, to begin the development of a tunnel that will carry away exhaust gasses. -It is drilled or tabbed to accept various wires forming the forward end of the box girder, and anti-drag wires that brace the lower two wings. A cowl-locating ring is welded at the 4 corners of the plate. This is a challenging part to make by hand.Not knowing any better, I cut the outlines with a metal cutting sabersaw. (Many, many blades).
Then, with vise-grips, I incrementally bent the stiffeners by stretching the metal. This produced some horrendous stresses in the plate, which left it badly warped, and my hands somewhat strained. A local company that had an oven, and press, "normalized" the part, and restored it to a flat condition. Too late, I learned that most builders would have welded on the stiffeners, and saved all of the grief that I experienced.
I had always hoped to find a Clerget engine, but when an 80 LeRhone became available, I floated a loan, and bought it. Both front and rear engine mountings were built for the Clerget, but cut, and drilled to take the smaller LeRhone. (A Clerget could still be fitted.)
One enthusiast had produced rough castings for the joy stick spade grip--the manette--the air throttle-- the tampier valve (gasoline regulator)--and the base for the oil pulsator.I bought the set.
The First Lecture
I had been asked to talk about this project to a local group, using slides for visuals. After agreeing, it dawned on me that I=d better have more to talk about. This spurred the assembly of the fuselage in my basement. I also built the sheet metal bucket seat, and got the control stick (with its unique spade grip), the throttle, and manette set up so they could be photographed.
My son, Bob came through again with enough bass wood so that the instrument board could be fabricated.
I discovered that I was missing the blueprint for the forward fuselage formers. A volunteer group in England was building a triplane for the Shuttleworth collection, and they agreed to copy their print and send it along to me at their cost. This was a most generous solution to my problem.
When the talk finally came off, it seemed to be pretty well received, and the fire that it had built under me, to turn out structure, really accelerated the project.
Tyres & Wheels
The tires on the original airplane were 700x75 mm . It worked out that a re-worked 3.00-21 motorcycle size was an almost exact duplication. (Again, friends working on their lunch periods turned out authentic looking tires). Hubs were made by a friend, and another made a punch and die set so that the modern rims with new spoke holes in the bead seats, would allow the unique Palmer wheels, with their cloth covered, asymmetrical appearance, to be duplicated. (The spokes are radial, but the outboard ones form a cone, the inboards do not).
Getting it out of the basement
Eventually, the fuselage was set up with formers and stringers, and the tail assembly lashed in place.
This situation produced a bit of humor; at least to me it was funny. My wife was anything but supportive of this project, so I usually kept the door to my basement workshop closed. One day she strolled in, and was dismayed at the sight of this much structure. She became deeply concerned at what she assumed would be my intention to dismantle the house to get it out. I know that she did not believe me when I told her that it would go out the same way it had come in. That is, up the stairs, or out the window.
Here was one of the few advantages resulting from a box girder fuselage. It could be disassembled, and the longerons, with tagged fittings tied to them, passed out through the small basement window. I could then re-assemble in the garage, and start hunting for a barn. Wing production resumed. I continued down the left stack, and then down the right. As with the fuselage, I was able to trial fit each panel in the basement, and disassemble and do final fit in the garage. The car was destined to spend a winter in the driveway.
Aligning the fuselage
I found fuselage aligning to be a frustrating chore. As with a bicycle wheel, adjustments made in one area, had repercussions in several others. I despaired that the warps would ever be overcome. Finally, I increased the width of the longeron slot in the port cabane strut, by half the thickness of a hand saw blade, and the whole thing snapped into shape.
Fuel and Oil tanks
The fuel and oil tanks presented a challenge of their own. Not the least was obtaining the lead coated "terne plate" in the correct gauge for the project. The oil tank has a curved top that mirrors the top of the fuselage. It sits just below the top shield and just above the fuel tank. Both tanks have baffles, and sumps. After forming, they are riveted (with many, many copper rivets), soldered, then "sloshed" with a sealant. Each tank has a liquid-tight hole extending from top to bottom. These allow the filler pipe for the fuel tank to reach through the top shield; and the supply pipe from the oil tank to be connected to the oil pump.
These were two more examples of "labour intensive" construction.
The fuel tank has the added need to hold air pressure, since a modest pressure may be used to assist in forcing fuel to the carburetor.
Space – the final frontier
One of the men with whom I worked, lived in a small town some distance from my home. He owned an old barn that was used to provide storage space for other people’s automobiles. He offered me the space I would need to set up the airframe so that I could make the measurements needed to place an order for the custom-made external bracing wires. (The British had pioneered the use of streamlined steel rods instead of the stranded cables used by others. This airplane used that type of material).We moved the airplane parts to this location. He flatly refused to charge or accept any rent for his badly needed facility.
Sourcing the propellor
I had read that the famous propeller maker, Mr. Ole Fahlin, although retired, would accept orders for certain of his designs. I wrote Mr. Fahlin, and received a quote for a prop suitable for an 80 LeRhone. His price seemed in line, so he made one for me. Incidentally, the propeller mounting on these old engines is somewhat unique. The prop hub is a machined part that involves a steel cylinder with an integral back plate. The cylinder has a tapered hole running through it, and a key-way. The front of the hole is tapped.
The prop has a center hole that matches the hub. It is drilled for 10 bolts. A matching front plate bolts over the front of the prop. When the prop is aligned on the tapered nose of the engine, and the key has entered its way, a large nut, with a projecting threaded section, is tightened, forcing the tapers into a very tight fit. This large nut is kept from backing off by a 12 point sheet metal nut whose ears align with 2 of the prop bolts. For prop removal, back off the nut; reverse it so that its external threads enter the internal threads of the hub; tighten, and it has become a puller.
Triplane assembly
After carving, and finishing the 2 large interplane struts, I was able to begin assembling the aeroplane. Since the six wing panels had been individually built over some period of time, it was with some anxiety that I installed the struts. As mentioned, each interplane strut passes through the middle wing, and engages fittings on the top and bottom wings. This passage has to be in agreement with the "gap", "stagger", and "incidence" specified. It was a tremendous relief to find that everything fitted. I used hardware store bolts, stranded cable, "U" bolt cable clamps and "C" clamps to hold everything together. Other volunteers gave generously of their time and sweat to help this assembly take place. When finished, it looked like a massive model airplane without a landing gear.
I checked, and re-checked the resulting "pin to pin" dimensions, and sent a rather large international money order to the Bruntons Co. of Scotland.( The same supplier used by Sopwith). These stainless steel "RAF" wires constitute the flying, landing, drag, anti-drag and landing gear cross bracing elements. Each would have right and left hand threads at their extremities, and special fittings and locking nuts.
My machinist friends had been busy. They had supplied the triple pitched jack screw needed for the adjustable stabilizer, and the aluminum pulley that would activate it. Also provided was the machining needed on the pulsator (oil pump indicator); and after I had secured blueprints from Australia, a complete Rotherham air pump. From a loaned original British compass, castings were made, and a convincing duplicate materialized.
One part that promised to require a long lead-time was the aluminum cowl that enclosed the engine, and extended back to the front engine mount.
In connection with my occupation, I had dealt with well known manufacturer. This local company assured me that they could easily spin my part, but they needed for me to have produced an accurate wooden block, against which the spinning would be made. My son signed on for this rather massive piece of tooling. It took a considerable amount of time for fabrication. Additional time passed before the spinning was scheduled. Since set-up time was factored into the price, I elected to order 7 parts. (I sold 5 at cost, and kept 2 for myself).
My daughter-in-law Linda, sewed the cockpit coaming; the windscreen surround; and the wing cover seams. One decision that was taken, was to cover the airplane in a polyester type fabric, rather than Irish linen which had been used by Sopwith. Since the fin carried the hand lettered Sopwith logo, and I was anxious to see if my free hand lettering was going to work, I did this first. It did not turn out too badly.
The Sopwith windscreen frame was supposed to be an aluminum casting. This was beyond my capability, so I made a replica from laminated strips of ash. These were steamed, bent, glued and clamped in a frame that formed the exterior shape. This was then sanded to produce the requisite cross-sectional shapes, and when painted, and covered with the surround padding, looked fairly reasonable.
Using that same technique, and a hub that required four spokes, I also made the pilots hand wheel that controls the horizontal stabilizer trim. Although this looked OK, and I installed it, when a casting became available, much later, I replaced my wooden one with it.
I priced a Vickers machine gun, but the cost of one of these rivaled the cost of the engine. therefore I ordered a scale model plastic kit so I could study details.
Then, from home-made, full-scale drawings, a fairly convincing fake was made from wood and sheet metal. It included a short strip of webbed belting that carried a few dummy rounds, and empty casings. The gun is more than just a visual effect; the windscreen attaches to its rear.
The assembly continues
With the assembled airframe still in the barn, I had the good fortune to be visited by an acquaintance. I had met him only once before. He was the friend of a co-worker, and was doing the restoration of a Cessna 140. Being a retired contractor, he had purchased a rural tract of land, built a house; prepared a grass strip; and put up a hangar. He was concerned, though, about the many leaks in the slate roof of the barn, and the possibility for water damage to my work. He invited me to visit his layout, with the thought in mind that I could move to his place, and eventually fly it off his strip.
Since I was taking up lots of space, and would have to disassemble for the coming vehicle storage season anyway, it sounded like a fine offer. The project was moved again.
With the wings and tail removed, the fuselage protruded some, but not too much, from the back of his Dodge van. The rest of the structure followed in turn. The only drawback was the 40 minute drive from my home to his. On the plus side was the fact that this man could do almost anything; had almost any tool that one could think of, and was generous to a fault.
I really believe that without the help of Carl Fuderer, the completion would have taken much, much longer, if at all. The covering of the wings and ailerons continued at my mother’s house, and I became very familiar with the technique of "rib stitching". For added authenticity I used the frayed tapes that the British had used. This meant that I bought the modern pinked tape--cut off the pinking, and then hand frayed the edges so that individual threads projected from the sides.I had bought a modest sized air compressor, and was able to spray the colored butyrate dope
that would resemble the PC (pigmented cellulose) that had been used originally. With the blue, white and red rudder flashing, and the similarly colored "roundels", my mother's garage began to look quite colorful.
Since these had been British naval airplanes, they had been identified with "N" prefixed serial numbers. The U.S, of course, uses "N" prefixes for our civil license numbers. I contacted the F.A.A. and presented them with a list of Sopwith Triplane "N" numbers, and requested that they assign me one of them for my use. As it happened, they selected N6291, an airplane that had flown with both Naval 8, and Naval 1 squadrons in 1917. It was not a famous number, like used by those in Collishaw's "Black Flight", but it would do.
I had been advised that another of the infamous "AP" numbers, (this time AP15, used for the landing gear), was almost exactly duplicated by the forward lift strut on the post-war Taylorcraft,BC12d. A visit to the nearby T.Craft factory, netted me the remaining stocks of what had now become an obsolete part for them. Since it had served in a tension role on the monoplane, I had round 4130 steel tubes inserted, and tack welded inside the streamlined tubes. This would provide strength for the compression role it would play on the antique.
The landing gear is welded up to form a pair of "vee's". Near the bottom of the "vee", round steel tubes pass through each leg to form an anchor for the bungee cord suspension system. There is also an axle travel limiter, and guide. Being a pretty complex unit, I made wooden jigs to locate everything prior to welding. The steel "vee's" are separated at the bottom by a wooden spreader bar. This bar is shaped like an airfoil, which has a large groove on its upper surface. The groove houses the axle when in flight. The bar is built up like a wing, and is covered in plywood. Since it might see severe buckling during crosswind landings, I used ash instead of spruce, and encased the whole thing in a glass/fiber sheath. Carl helped me install the ash tailskid, and spring it with the bungee lay-up called for in the drawings.
Standing tall
We next got the fuselage up on the gear and installed the streamlined bracing wires that had arrived from Scotland. Besides the wires that brace it from lower right to upper left, and vice versa, there is also a vertical wire from the center of the spreader to the bracket that supports the rudder bar and rear engine mount. This wire resists the downward loads from the inboard ends of the axles. The outboard ends of the axles pass under the bungee suspension, and cause the characteristic front-view Sopwith appearance: the wheels exhibit large amounts of negative camber while on the ground. I now had a very tall structure, over 10 feet from the ground to the top of the top wing center-section.
The elevator, rudder, and the tail-skid control cables, and wires were threaded through the fairleads, and left in the vicinity of the control stick. Covering the fuselage became the next task. I first, did the portion aft of the cockpit. With that done, the stabilizer and the fin could be installed, and secured with the RAF wires.
Wires
The control cables connect to the stick in a rather busy fashion. The two, down-elevator cables connect to a ring at the back of the aluminum stick. The up-elevator cables pass under the floorboards-around two pulleys, and then back to the ring on the front of the stick.
The rudder cables attach to fittings near the outboard ends of the rudder bar, and the tailskid wires anchor to two straps that encircle the bar, closer to the pivot. Cable that operates the stabilizer trim, winds around and through a pulley on the outboard side of the right cabane strut, then around a pulley built into the jack-screw mechanism, then back through a turnbuckle, and on to the control pulley.
Looking back down the fuselage, behind the seat, at the maze of bracing wires and cables that occcupied the space between there and the tail, I was struck with another meaning of the term "fly by wire". Now it was time to install the engine.
Installing the engine
Since the LeRhone is lighter than the Clerget, I decided to set it 6" forward. This meant that a spacer had to be constructed. It would sit between the mounting plate of the engine, and the front engine plate of the fuselage. It was made with parallel plates, spaced apart with large diameter steel rods, located on the centers of the bolt holes in the engine mount. The entire unit was welded to form a unified part. The rods were drilled to accept the much longer bolts that I had obtained. As mentioned, the aft end of a rotary engine is a long cylinder that ends in the flange that carries the carburetor, and attaches to the rear engine mount. It is threaded to take a flange. I had an extension tube made with a threaded collar to account for the front spacer, and the fact that the rear mount was positioned for the longer Clerget.
From the first, I had been a little concerned about what would happen when this rather heavy steel engine was attached to the somewhat fragile-looking fuselage with its slender wooden struts and attendant sheet metal, and wire fittings. We picked up the engine with a sling, hooked to a block and tackle, and swayed the extension tube back through the front engine plate, and then on through the rear engine mount.
When the bolt holes in the front plate lined up with the holes in the spacer, we began installing the long bolts. The rear mount accepted the flange, and when they were bolted together, we released the block and tackle. The fuselage was now carrying the load. It was as if nothing had happened. Seemingly, there was no change in the tension on any of the piano wires, and the aspect of the longerons was un-affected.
Electrics and Carburetor
It was now time to connect the electrical, and carburetor controls. The electrical side of the house is pretty straight-forward. There are only the "P" wires wired parallel to the points in the magneto. These go to ground either through the ignition switches (old fashioned porcelain wall switches), or the normally "open" button switch in the center of the spade grip on the stick. The infamous coupe' or "blip" switch.
Carburetor controls consist of threaded rods that attach the Aball@ type fittings on a bell crank to the air-gate/needle valve on the carburetor, and the large brass throttle on the control quadrant.
A second control is the "manette". Also located on the quadrant, it is a wheel topped lever that controls the tampier valve (A needle valve that regulates the amount of fuel that will get to the 2nd. needle valve that is part of the carburetor). This is also a threaded rod set-up. With this installed, the forward former/stringer assemblies could be attached, and covered, as well as the low wing box ribs which project slightly from the airplane.
Installing the top shields
It was now time to shape, and install the top, and side shields. These are aluminum sheet that produce the round shape that develops in front of the cockpit. The top shield goes under the machine gun and has holes for the oil and gas filler pipes. It is secured by wood screws, and the cable that will hold the nose cowl to the cowl ring. The side shields have doors that give access to the carburetor, and holes for the carburetor air intake pipes. The side shields fit into an aluminum channel at the bottom, slip under the top shield, and pass under a cable that grips from the rear. The front edge rests on the cowl ring, and is under the nose cowl. The bottom shield is held by metal clips that encircle the longerons, and their outboard edges fold to create the channels needed by the side shields. With the nose of the airplane buttoned up, it became possible to go for some engine tests.
Testing the engine
Since the octane rating of the fuel for which the engine was designed was low, un-leaded auto gas is just about right. Instead of castor oil, I opted to use 60 weight racing oil+10% STP. With the tanks containing enough fuel and oil, we pushed the wingless wonder out of the hangar, chained the sternpost to a heavy piece of farm machinery, chocked the wheels with blocks of concrete, and wondered what to do next.
Having gotten some advice from the few people who had any to give, and read everything I could find about the subject, we started out.
First, we hand lubricated the ball bearings and the sliding pins in the rocker arm assemblies. Since the fuel system can be pressurized, we closed the vents, and began using the small hand pump located at the pilot’s right shoulder. With the gasoline tank valve open, and the manette, and air valve full forward, we pumped until
fuel began running out the carburetor overflow pipe protruding from the bottom shield. Then we reduced the manette to about one quarter open, and the throttle to one half open. Priming was accomplished by squirting raw gasoline past the depressed exhaust valves as they appeared beyond an opening in the nose cowl, when the engine was hand rotated in reverse. Since my crewman had never propped an airplane before, he got to sit in the cockpit, as I flailed away on the propeller. There were many silent rotations before it suddenly burst into life with what was to become a characteristic loud, snapping clatter. Attempts to throttle back were greeted with long periods of non-firing, and the sounds of whirling machinery, then movements of the manette would restore the mixture to a runable one, and the "song" would come back.
About all that we learned was that it would run; the dreaded torque would not rotate the wing-less fuselage enough to tip it over; and when one depressed the "blip" switch on the stick, the engine stopped firing, but came back to life when the switch was released. In successive engine runs, it began to dawn on me that at least this rotary engine was very smooth. There was none of the vibration that a normal radial puts into the airframe. It reminded me more of what one might expect from an electric motor, and made me feel strangely disconnected from the source of what was apparently, very powerful thrust.
Moving the project again
Carl had been flying his Cessna for some time, and it was becoming obvious that if the wings were put on the triplane, things would be pretty crowded. In fact, the wheels would have to be removed to get out the hangar door. (We had used a taller, but narrow, back door for the engine runs). As it happened, the folks at a local airport that annually staged a rather ambitious air show, got wind of my project. I was contacted about the possibility of my triplane being exhibited statically at the next edition. To sweeten the offer, they made arrangements for me to rent a hangar with a sufficiently tall door.
Again, Carl made everything possible. He had a trailer that could transport the fuselage with the tail assembly attached. Subsequent trips carried the wings and other sundries that had cluttered his hangar for several years. As with the owner of the slate roofed barn, attempts to establish and pay rent for the space had always been deferred, and never followed up; then finally refused.
Carl either flew, or drove over every day during one of the hottest summers on record, as we worked on final assembly. The fuselage was leveled with a saw-horse and a jack under the tailskid. We built some 2"x2" scaffolding (held together with "C" clamps), to support the low wings.
The steel-spar boxes project slightly from the wing roots, and enter mating spar housings in the fuselage mounted box ribs.(Extensions of the low wing center section). Long steel rods, entering from the trailing edges, pass through the rear and front spars and lock the outer panels in place. Additionally, drag and anti-drag RAF wires attach to fittings on the wings.
They anchor either to the front engine plate, or the fuselage aft of the cockpit, and when tightened, assist in pulling the spar boxes into their housings for the final fit. With the low wings on, we could erect the interplane struts. The middle wings having previously been clamped in place at the struts midpoints. (Additional 2"x2" lumber, clamped to the scaffold, supported the load ) Another set of drag and anti-drag wires pulled the spar boxes into mating housings concealed in the mid wing box ribs on the cabane struts. Steel rods, entering from the leading edges, pass through the nested spars.
Landing wires
We could now install the landing wires. Fittings attached to the compression tube (that is built into the top wing center section), anchor the upper ends. From there, the wires pass through plywood sheathed holes in the middle wing on their way to the fittings installed on the low wing steel compression rib. The single landing wires lie in the plane of the forward staggered cabane, and interplane struts. With 4 of the wings now supported by the fuselage, we could begin to hoist up the top wings, and eventually work them into place.
High up on step ladders, and without the help of any drag/anti drag bracing, to assist the panels into place, getting the spar boxes to nest and accept the securing rods, is an acrobatic exercise. As with the mid wings, the rods enter from the front. These rods are pointed on their entering ends, and flattened on the rears. Single wood screws penetrate the flattened portions, and fasten to wooden structure behind.
One problem in setting up this aeroplane, is the fact that there are no adjustments to the incidence angles for the wings. Once bolted in place, whatever angle selected, is it. As a result, Carl and I took considerable care, even though hardware store bolts and nuts were being employed, to get things right. (The deadline for displaying at the air show made it impossible to wait for aviation hardware from a supplier.) The original drawings did not call for the wings to be washed for propeller, or engine effects.
I, in my infinite wisdom, elected to reduce the tip incidence on the right middle wing, and increase it on the left side. We successfully finished in time for the air show, and a friend who lived in Pennsylvania, flew his Fokker triplane over, and we parked side by side. Our two airplanes attracted considerable attention, and made all of the work seem worthwhile.
With the rush to exhibit out of the way, aircraft hardware could be ordered, and gradually installed. One obvious feature on the fuselage, is a series of eyelets that extend along the right top longeron between the sternpost and the cockpit. Criss-crossing those eyelets are short lengths of twine, tied frequently with knots. This is an inspection panel. Untie the knots; remove the strings; and the panel flops down to provide access to the guts of the fuselage.
Before calling for an FAA examination of the project, I wanted to re-check the interior of the fuselage for the presence of cotter pins; tension of piano wire etc., but I did not want to unlace that panel, and risk that the appearance of the fabric would suffer.
I, therefore, elected to crawl back down into the fuselage. (After first having removed the seat). This meant that the vertical cross-bracing wires would have to be opened up and re-set as I worked from back to front. The heat of summer continued, and this would be no place for someone with claustrophobic problems. I was alone in that thing for hours, and if I had gotten stuck, no one would be coming to help me out.
I decided that I would never do that job this way again. The old guys must have known what they were doing when they provided that big opening.
As the construction had progressed, I decided to find whether I remembered anything about flying..
Back to flying
The first thing that came up was the scarcity of tail dragger opportunities. The airplanes themselves have become rare enough, let alone instructors who were current in them.
The first episode involved a former instructor who had a PA18, and would let me fly from the right seat. This would preserve the hand relationships regarding throttle and stick.
As it happened, I picked a windy, snowy day. I also uncovered a problem with the steerable tailwheel, and with no brakes on my side to help correct for it, the take-off produced a yell for help.
As we flew around the pattern, I remember that I questioned myself about ever having felt comfortable doing this in the past. I decided to hook up with a pilot from work who was currently instructing, and take some formal training from him in his PA28 Archer. I did shake off some rust, but I began acquiring bad habits for future tail dragging. I was encouraged to forget that I had feet, especially during landing roll outs.
The FAA Inspection
Engine tests, culminating in a taxi test took place with the triplane. The steerable tailskid actually was of some use, and the plow on the skid did provide some braking.
Prior to asking for an inspection by the FAA, it was necessary to do a weight and balance. Having that information was well and good, but there is no C/G information on the blueprints. Fortunately, the group in England had gotten an aerodynamicist to study the problem. He reduced the Sopwith Triplane to a theoretical monoplane, and arrived at the fore, and aft limits.
They mailed me a copy of his presentation, and I was greatly relieved to find that my airplane balanced within the rather tight limits he had calculated.
I suppose every home-builder has trepidations about the FAA inspection. I know that I did.
It was a bitterly cold morning when I met the two men from the regional office. The hangar had been specially heated, and I brought everything that I thought would be required.
Even though this was probably the first time either one of them had ever seen an airplane held together the way this one was, they gave it pretty fair inspection. They discovered a couple of missing cotter pins, and suggested that some of the music wire ends could cause cuts to unwary flesh, but not much else. We then had to open the door; (letting out all of the heat) for neither had ever seen a rotary run before. Luckily, it fired on the first swing, and I let it rev as long as my un-helmeted/goggled parts could take the wintry blast and then shut down.
The sign-off was almost anti-climactic.
Learning to fly again
Meanwhile, the adventures of the would-be (again) tail-dragger pilot continued to stagger on. I was finally allowed to take some dual instruction in an Aeronca Champ. Here I really re-discovered my feet, and began to take charge of the ground antics of the airplane on take-off and landing. If the nose was going to swing, it was because I caused it! Not because I was chasing the airplane. The final confidence booster came in a Stearman owned by a friend.
After a bouncy landing, I took off my glasses; pulled the goggles over my eyes, and shot three acceptable landings. One of them might have even qualified as a "grease job". My glasses were making it appear that I was closer to the ground than I really was. I passed my next third class physical without glasses.
Time to fly the Sopwith Triplane
By virtue of a fortunate friendship, I was able to solo a J3 Cub, and build some time in it. This built my peace of mind as well, and finally led up to deciding that it was time to go "Sopwithing". The October morning that I had selected was clear, frosty, and calm.
I drew a larger crowd than I would have liked. I didn't want to possibly embarrass myself in front of a lot of people. (I planned to stay in "ground effect", and if things didn't feel right, I could set back down on the remaining grass).
The engine was balky. At first it would only run out the prime. Finally it accepted fuel from the tank, and accelerated to 1100 rpm. By carefully playing with the mixture, I idled down to 550 rpm.
While waiting for an airplane in the pattern to land, I signaled for a mechanic to check the security of the spring loaded latches on the side shield inspection doors. Everything was OK, so when the traffic cleared, I waved away the chocks, and advanced the throttle and manette.
The airplane began to roll instantly. A little forward stick, and the acceleration was surprising.
When I applied a little back pressure, the feel of the tires went away at once, so I eased off and glanced at the airspeed indicator. It was reading zero. Feeling that I was still pretty low, I looked overboard, and was shocked to see that I was about 100 feet up, and still climbing. With the "ground effect" plan out the window, I went to plan "B" which was "fly the airplane".
I climbed to pattern altitude, throttled back a little, and tried to set up a pattern that would get me back to the grass. It was then that I realized that I was flying cross controlled to keep the ball centered. It was taking lots of left stick, and right rudder, I judged that my wash-in/wash-out decision had been a bad one.
On my first attempt at a pattern, throttling back, and "blipping", like I knew what I was doing, I turned base. As it happened, I had been too close on downwind, and not wanting to do any steep banks to correct it, I just went on around.
The second pattern was better, and I carried on until I was fairly low, but again I was going to have to "ess" turn to reach my wind line, so I fired up and climbed again. This time, there was an audible howl from something in the slipstream. Just before turning crosswind, The right side inspection door popped into my line of vision. It seemed to be standing out from the shield, and then it was gone.
I throttled back and considered landing in a cornfield that was dead ahead. I was too high to just glide in. It would take some side-slipping. Since I knew nothing of the stall characteristics, and needed cross controlling just to fly, I decided to try to get around one more time.
A thunderous racket began, and this was accompanied by a frightening vibration.
I needed to apply power just to hold my well-below-pattern altitude. It was a relief to find that when I added power, the noise and vibration reduced markedly, and raised my hopes that there was nothing wrong with the engine. On downwind, just before turning final, there was sudden blast of cold air on my right leg, and daylight illuminating the interior. Then there was a jolt, and the airplane skidded left. I corrected, and Ablipping@ regularly, and praying constantly, I brought it around and reached the grass area that I was supposed to be using. Relieved, I relaxed my grip on the coupe= and the airplane ballooned slightly . Thinking again, I squeezed tightly, deciding that we were going to land now! I kept increasing back pressure, and the airplane began to settle at a rather fast rate. The main gear hit, and I thought I was into a 10 feet bounce, but it only seemed that way. When the tail skid got down, the airplane hardly rolled one fuselage length.
With the engine shut off, I was joined by a host of people.After climbing out, I was shocked at what I saw.
The right side shield was gone, and the nose cowl was sagging into the trailing edge of the propeller. Cavities had been grooved in the prop, and there were several slits in the fuselage fabric that skipped almost to the stabilizer. The missing door and side shield were recovered, so we pushed the airplane back to its hangar.
During the car ride home, I began to experience an adrenalin let down, and realized that I was very tired. I also realized that the people on the ground had been quite frightened for me. I had been very excited, but I don't think that I had time to get scared. That afternoon, I returned to the hangar, and with the help of an experienced pilot/mechanic went over the mornings events. We traced the path of the departing side shield.
It had struck the lower part of the fuselage as it tumbled aft, and slammed into the right front flying wire on the stabilizer. The force was great enough to damage the fitting that held it to the stabilizer spar, and almost tore it loose. This had allowed the right side of the stab. to assume a higher incidence, and accounted for the skidding I had experienced.
If that fitting had failed totally, someone else would be writing this.
Our analysis of the incident went something like this:
1. The spring loaded latches on the right inspection door had failed to stay latched.
2. The door swung out, and acting like a sail, pulled the right side shield back from under the cable that held down the nose cowl.
3. The loosened nose cowl, still anchored at the bottom, tilted forward, and came down on the whirling rocker arms of the engine, machining the trailing edge of the propeller in the process.
4. The side shield, held only by its rear cable, blew away, and tumbled back, striking the stab. bracing wire before falling clear.
5. At higher rpms, the air whirling out from the engine floated the cowl away from the rockers, and reduced the vibration and noise.
Almost another year was to pass before the triplane was again ready to fly.
Repairs to get flying again
The nose cowl was replaced, and it and the inspection doors were held by screws as well as by the traditional fastenings. The propeller was repaired. The stabilizer had been removed, repaired internally, recovered, and installed; and the fuselage tears repaired. The airspeed indicator plumbing was redone and tested. This time, the take-off was no surprise, and the airspeed indicator worked.
As I sailed around the pattern, I began to perform some tests. First, at about 80 knots, I let go of the stick and rudder. The nose didn’t rise or fall, but the airplane rolled right, and yawed left.
It required full left stick, and full right rudder to bring the ball back to the middle. I reduced power and explored the realm of stabilizer trim.
( I forgot that I had lately received missing drawings involving a clutch, built and installed it).
I shoved the control wheel rather hard. Nothing happened. I made a mental note that I would have to look into this.
A couple of minutes later, there was an audible "bang" in the tail, and the airplane dived sharply.
My seat belt/shoulder harness set-up allowed for considerable vertical movement on my part.
I used it all. I was shot up out of the cockpit. My head was well above the windscreen, and my goggles blew over my head.
Thinking that something had failed, I managed to get back on the seat, and get my feet on the rudder bar. (I was not wearing a parachute and with the drag wires surrounding the cockpit I'm not sure one could bail out anyway])
I pulled back the power and considered an off-field landing.As I glided lower, the airplane seemed to be responding OK, so I re-applied power and climbed to pattern altitude, and went around and set up an approach. Harking back to the first flight, I expected the airplane to float some on final, so I let it get pretty low. It didn't float at all, and the wheels touched down in a fairly decent wheel landing.Since I was not yet to the grass where the taxi lights had been removed, I applied power and went around. This time, I made a 180 degree side approach, and hit the grass that I meant to use.
The landing was more tail high than I intended, but the gear took the shock--the tailskid planted itself, and with very little roll-out, the airplane stopped.
The next day I checked the incidence angle of the stabilizer. It was well above the 1 degree that was there when I had started the flight.
I deduced that the strain that I had put on the cables had eventually over-powered the clutch and caused the jackscrew to rotate. (The clutch is set up to require an inboard pull on the wheel to overcome a spring load, followed by wheel rotation, then release of pull, for re-engagement of the clutch).
That stabilizer set-up provides for one powerful control system. (In theory, it mimics the F100 flying tail).
I now believe the words of a WW1 pilot who stated that with the engine two thirds open, rolling back the trim control could initiate hands-off loops that continued until the fuel ran out.
I have leisurely re-rigged the airplane for no wash in/ wash out on the middle wings, and am ready for a flight that hopefully will leave my knuckles pink this time.
When I view the airplane again after an absence, I need to re-assure myself that I actually caused this all to happen. To me, it is still an attractive airplane, but beauty truly is in the eye of the beholder.
None of which provides any real answer to WHY! Maybe none is needed.
Getting hold of the original plans
The London office of the company for which I worked, was able to provide me with an address for the public relations office of Hawker-Siddeley.(Hawker-Siddeley-[now British Aerospace]--being the continuation of what had been the Sopwith Co.)
I wrote them, and they assured me that the original blueprints for the Sopwith Triplane still existed. They made a quote which included the cost of reproducing, and shipping.
After some soul searching, I mailed an international money voucher, and waited. Eventually, a large carton was waiting at customs at the local airport.
The drawings, dating from early 1916, appeared to cover everything from a stress analysis of the wings, to crating instructions for sending machines off to war. Some of the blueprints were still marked that they were subject to the " War Secrecy Act".
When I compared the type of parts that the airplane would use, to those listed in supply house catalogues used by home-builders, it was obvious that very few "off-the-shelf items" would be available to me. This would truly be a do-it-yourself project.
You could still buy Sitka spruce lumber, and birch plywood. The original aircraft had used lots of that, but beyond a few pulleys, turnbuckles, bolts and nuts, not much else was recognizable.
(If the reader does not have much interest in the nitty-gritty of home-building, he should bail out now, and flip to the back in hopes of finding something he can better relate to.)
The Triplane workshop, skills and tools
One question that is often asked is what kind of a workshop is required to try to build an airplane like this. In my case, a most modestly equipped one.
I had a vise. (Before it was over, I had wrecked several--they aren't designed for metal forming). I had a small jig saw (some call it a scroll saw), saber saw, table saw, and various hand saws. I acquired a band saw before getting very far into the project.(By installing 2 motors, and a belt reduction system, I was able to cut wood and metal). My shop also had the usual hacksaws; coping saws; planes (including a Shinto); Surform tools; a small bending brake; a metal nibbling tool; snips; a hand held spring winder; a router;some ordinary wrenches, "C" clamps, and an electric drill. I had nothing very special.
Although not exactly an aircraft factory set-up,airplanes of that era could almost be considered a cottage industry product.
The fuselage was built as a wooden box girder. That is, wooden longitudinals, verticals, and horizontals, assisted by small metal fittings, lengths of piano wire, and turnbuckles, were kept in compression by tension in the steel members. (A most efficient use of each material)
It was apparent that skills would have to be developed in the bending of sheet metal, and the forming of many gauges of music wire. This instead of the more normal task of learning how to gas weld.
Help from unlikely places
I was fortunate that my employer had amassed a library of donated books that spanned a considerable period of time. One such book was a work by Colvin and Colvin dating from 1919 which detailed the" aeroplane" standards of that time. The book didn't explain "how" you were supposed to manufacture any of these items, but it did show dimensions, and provided drawings depicting "good" and "bad" workmanship.
Another bit of serendipity was my being allowed to make the acquaintenance of a builder in an adjoining state who was constructing a Sopwith "Pup". In addition to introducing me to World War 1 Aeroplanes, he had a list of people, and companies that had bits, and pieces of things that I would eventually need, and got me in touch with a sub-culture of other "nuts" working on things like this.
One of these people was to me a legend. He had produced many museum quality reproductions of airplanes of this period, including a Sopwith Triplane. He was a unique person. He wouldn't communicate much until he was certain that you were serious about your project. If you were just some hangar flier willing to waste his time--forget it! If you were for real, he would answer any question you raised.
I was working as a tire engineer for a major rubber company, and they operated a small machine shop that produced prototype examples of machinery required for the over-all operation of the company. The employees of this lab were talented craftsmen who could offer suggestions as to how certain items might be turned out by an amateur with limited experience and facilities. They also had expediters who were acquainted with the resources available from local supply houses. All of this information, and encouragement was invaluable.
A particular blessing was that three fellow employees were sympathetic. These three were either hobbyist machinists (with equipment at home), or in one case, an expert gas welder with oxy-acetylene tanks in his garage. Additionally, I had the friendship of an experienced cabinet maker from church. Without this kind of back-up, the project might well have foundered.
The Sopwith Triplane build starts
After a period of procrastination, I decided that a good item to test my skills and dedication would be the center section of the top wing. This would give me a start on the 240+ ribs used in the whole airplane (counting stabilizer and landing gear spreader); sheet metal box construction; wing spar shaping/ routing; music wire forming; the first usage of the some 188 turnbuckles ("strainers" to the British) the project would require; and the manufacture of an elusive elliptical steel tube needed for trailing edge.
The center-section would be a microcosm of the entire project. With copious help from the various sources, including lunch hour stints on the company's metal cutting band saw, the parts gradually accumulated, and came to be assembled. It looked good, and I was hooked. I had also learned a valuable lesson. DO SOMETHING EVERYDAY! If you don't, inertia will bring you to a halt, and re-starting will become an increasingly difficult task.
AP121 – Elusive Elliptical Tubing
Regarding the "elusive elliptical tubing". This is specified as AP121" ("Accles and Pollux 121") on the drawings, and is used for trailing edge; wing tips; and fin, rudder and elevator framing.
I had written a string of letters all over the place, in search of the last remaining supplies of this tubing. The final result was that an American company now owned the product line, and if I were willing to pay substantial start-up charges for a small production run, I could get some. This was out of my price range.
Fortunately, one of the people that I had come to know suggested a home-made alternative. This involved purchasing a round, copper finished steel pipe, called Bundy Weld tubing.
After making an adjustable roller die assembly, one could roll this formerly round tube into the correct elliptical shape to match the AP121. It meant multiple passes thru the dies, at gradually tighter settings. This used lots and lots of elbow grease and produced an anything, but straight length of material. It could, however, be returned to straightness by the judicious use of strong hand pressure, with the attendant risk of carpel tunnel syndrome. (I am fortunate that this whole operation was never video taped.)
Building the Sopwith Triplane
As a child, when I had built model airplanes, I tended to begin with the tail assembly. I may have done this to have complete parts to look at and be encouraged to go on to the more complex areas ahead, this strategy was employed on the triplane. The horizontal stabilizer became my next goal, and work began on the metal fittings.
( I had by now bought out the remaining stock of surplus World War 1 turnbuckles from a unique hardware store in an old area of Los Angeles.)
The wooden materials, were ordered, and when they arrived, it was assembled. Next to come would be the elevator, fin and rudder. I experimented with wooden tooling to produce the curved outlines that the bogus AP121 would generate.
A plan view of the curve, sawed from a board, ("Kentucky windage" was used for spring-back), provided a female mould. The edge was then routed with a core box tool appropriate for the radius of the major axis of the ellipse. It was surprising how well this worked. The minor axis of the ellipse projected out from the tooling, and would have to stretch when the tubing was bent around the wood. This it did.
Eventually the familiar curves of the Sopwith fin were accurately reproduced. Both the rudder and elevator assemblies are steel. The joints are brazed together, and some interesting transition pieces are required to blend the different diameters of tubing.
Rudder and Elevator
The rudder and elevator horns were a challenge. I wasn't smart enough to come up with a modern substitution for the two-piece steel pressings that had been used to form the hollow, edge- welded components on the original. A friend produced a wooden reproduction of the shape. The friends in the shop allowed me to use the arbor press.
With a thick rubber pad for backing, the thin sheet steel deformed into the rubber, asthe wooden tooling was pressed against it. This accurately developed first one surface, and then the other. When trimmed, matched and edge welded, the end product looked good. (The same technique would be employed on the top and bottom aileron horns; and the aileron lever on the control stick rocking shaft.)
Building the Wing Ribs
The left top wing panel was the next component to be tackled. The drawings didn't indicate how the original ribs might have been made. They employed spruce cap strips that were grooved on their inner surfaces to accept the 1/8"birch plywood webs. The webs were lightened by removing curved areas behind, between, and in front of the spars.
The drawings called for many tiny flat-head brass wood screws to hold the ribs together while the glue set--(glued and screwed) was the note on the blueprints. The modern price and availability of the correct size screws would have played havoc with the budget and time table. I, therefore, made a jig. I laid out the rib contour on a flat, stabilized, wooden platform. The upper surface could be backed by a curved wooden strip. On the under-side, at the points where the screws would have been used, I located wooden discs with an off-center hole drilled through them. Lag screws held the discs to the jig surface. When the embryo rib was in place, the disc-"cam" could be rotated to put pressure on the caps. Then when all was ready, tightening the lag screws, held the compression. Since I was using Aerolite glue, my rib production could be as slow as one per day.
The wing spars were ordered from an aircraft parts supply house. Since the spars fit within the top, and bottom camber of the wing, they are parallelograms rather than rectangles. I had them shaped by the supplier to the correct profile.
The solid wood compression ribs fit into square steel sockets that are welded to small backing plates that bolt to the spars. The ends of these fittings are bent and drilled to anchor music wire/turnbuckle fittings. The normal drag/anti drag truss is then developed, with the wires passing through the holes in the ribs.
One compression rib is unique. This is the one where the interplane strut attaches.
Here a round steel tube is used. It carries the fittings to locate the strut sheath, and the anchor for the flying wires.
The wing went together satisfactorily, so the port mid-wing was next.
The Port Mid-Wing
The span and chord is identical to the top wing. There is a cut-out near the root. The steel tube compression rib is special. The interplane strut passes completely through it. This results in staggered, streamlined openings in the top and bottom surfaces, sheathed with a thin sheet metal sleeve that passes through the openings. The sheath is continuously welded to the compression rib. Rib, and fitting manufacture continued until enough parts were on hand to make all 4 remaining panels. All that was lacking were the expensive spars.
The Ailerons
This airplane has ailerons on all six wing panels, and I decided to go for some variation by tackling these next. Not the least problem to be faced was the manufacture of 18 special hinges. These required a special rivet for hinge pins, but that is a story in itself. Suffice it to say that I was lucky.
The leading edge of the aileron was made from a thick block of wood that was severely routed between the rib attach points, and then covered by a thin slat of spruce. The spar was a round aluminum tube. Where the aileron horns attached, the rib was unique. It was solid, to accept the through bolts that held the horn assembly. After making 6, I was glad that there were no more.
Some of my associates had begun to needle me because of the lack of any fuselage work. I decided that this area would come next.
The Longerons
The longerons were to be made from 1-1/2" square ash timber that needed to be straight grained and 17 feet long. My son located some suitable material at a specialty lumber yard in Michigan, and had it ripped for me.
I built a plywood jig that developed the plan view of the fuselage. This had wooden blocks strategically placed so they would force the ash to hold the necessary curve. (The 1-1/2" square dimension tapered to 3/4" square at the sternpost).
I bought a length of flexible plastic drain pipe that was somewhat longer than the longerons. This way, by curling the ends up, it could hold hot water. The longerons were inserted,the ends of the pipe fastened up, and then filled with hot water. After a soak, the longerons accepted the bend enforced by the jig, and when dry, retained the curve.
Additionally, the bottom longerons have an upward curve aft of the cockpit, and this was accomplished by jacking up the aft end of the jig while the longerons were drying.
The sternpost is a complex part. It is a combination of steel plate, and square, and round steel tubing of several dimensions. When welded together, it keeps the four longerons in their correct relationships; accepts the triple pitched jackscrew that trims the tailplane; supports the tail skid; anchors the bungee that springs the tail skid; and provides 4 turnbuckle hook-up points that are the terminals for the tension system that is the strength of the fuselage.
Turnbuckles and wire
From the sternpost forward, there are a variety of sheet metal clips at each station where vertical and horizontal spacers occur. These clips fold around the longeron from inside to out. Horizontal bolts are located so that they will provide sockets for the vertical spacers. These bolts do not penetrate the longerons. Encircling each bolt is a folded steel strap. When the strap is slid over its bolt, it will be tangent to the longeron, and the spacer. The strap also provides the anchor point for either a turnbuckle, or the "loop" of a length of music wire. On the inboard surface of each clip, square lightening holes with their sides folded outward, formsockets for the transverse spacers.
The clip is also designed and drilled to provide tabs for the plan view wiring, as well as the diagonal wiring that ties an upper right to a lower left, and vice versa. These clips change in dimension in accordance with the expanding cross-section of the longeron until they reach the maximum at the cockpit.
At the time of the first World War, these parts were standard items, and available in quantity. In my case, it was a one-at-a-time proposition, and very time consuming.
The music wire components have been alluded to, but not described.
First, music wire in whatever gauge, is a very stiff and strong piece of steel, and it is necessary to deform the ends into uniform loops that fold back. The now parallel wire ends are encased in an elliptical spring (called a ferrule).When the ferrule is driven against the aft edge of the "loop" the free end of the music wire is bent back, alongside the ferrule to lock it.
When you consider that the British did not want threads showing when the turnbuckle was tightened and locked, the accuracy needed to make sure these wires were the right length, was a chore. (The ferrules have to be slid onto the wire before the final loop is formed). I made some of the ferrules, but one of my friends made most of them.
It is small wonder that modern fuselages are of welded steel tubing. The wooden box girder is a very labor intensive proposition. (Although hard to beat on a strength to weight basis).
The Cabane Struts
In the Sopwith Triplane, a cabane strut is unlike that used in any other airplane that I know about. It is a forward canted, large piece of wood that has a notched corner to partially sit on the bottom longeron, and a notched channel to accept the top longeron. Inside the fuselage, it is routed extensively for lightness. Above the fuselage, it is streamlined. Within the fuselage each strut provides the front support for the seat. The left strut also carries the engine controls. (Air throttle lever and manette). The right strut mounts the stabilizer control wheel, the fuel sight gauge, and up in the airstream, the Rotherham air pump (for fuel tank pressure). Both strutsprovide attachments for the dash board . In addition, the midwing box rib assemblies encircle these struts (above the top longerons) and at the top, carry the top wing center section. They are well used components.
The Rotary Engine
Since this was going to be as faithful reproduction as I could make, there was really never any thought about using a non-rotary powerplant. If the airplane had been so well liked, how could rotaries be as troublesome as some latter-day writers had made them seem? I wanted to experience flight behind one, first hand.
I've been using the term "rotary engine" rather freely without definition. This is not a rotary combustion chamber engine such as a Mazda. This is a radial engine (nine cylinders in my case), with its crankshaft fixed to the airframe. The entire cylinder assembly rotates around the crank.
The propeller is fastened directly to nose of the engine. It is a four cycle, overhead valve design.
A commutator rotates with the engine, and is energized by brushes fromgear driven magnetos.
The fuel/air mixture enters from the rear of the hollow crankcase--mixes further in the business end of the crankcase, and is drawn and thrown out through a manifold to the intake valves. The exhaust is expelled directly to atmosphere under the ring cowl. The carburetor is rudimentary, and slow in response since it takes a while for the engine to find out that a mixture change may have been made. In some cases it is easer to manage the engine RPM by killing the spark. The engine displacement is large--the RPM low (1200)--the horsepower rating very conservative, and when coupled to the right propeller, the performance is astonishing. There is a noticeable gyroscopic effect. The oil is not recovered. It is slung or burned after having been pumped through the bearings. The horsepower to weight ratio is quite good.
The Engine Plate
Rotary engines are mounted to an airframe in two places. The rear engine mount is a sheet steel box mounted transversely to the fuselage, and provides an anchor for bolting the circular flange that attaches to the crankcase extension. The front engine plate is the forward end of the fuselage. It is a heavy gauge steel plate to which a circular steel plate on the back of the engine is bolted.
At the time these airplanes were manufactured, the front engine plate was a stamped product.
It is extensively lightened by removing steel in 4 radially extending teardrop shaped slots. The perimeters of these slots fold back to form stiffeners. The sides fold back to cradle the foremost vertical struts. The bottom is an arched, folded cut out, to begin the development of a tunnel that will carry away exhaust gasses. -It is drilled or tabbed to accept various wires forming the forward end of the box girder, and anti-drag wires that brace the lower two wings. A cowl-locating ring is welded at the 4 corners of the plate. This is a challenging part to make by hand.Not knowing any better, I cut the outlines with a metal cutting sabersaw. (Many, many blades).
Then, with vise-grips, I incrementally bent the stiffeners by stretching the metal. This produced some horrendous stresses in the plate, which left it badly warped, and my hands somewhat strained. A local company that had an oven, and press, "normalized" the part, and restored it to a flat condition. Too late, I learned that most builders would have welded on the stiffeners, and saved all of the grief that I experienced.
I had always hoped to find a Clerget engine, but when an 80 LeRhone became available, I floated a loan, and bought it. Both front and rear engine mountings were built for the Clerget, but cut, and drilled to take the smaller LeRhone. (A Clerget could still be fitted.)
One enthusiast had produced rough castings for the joy stick spade grip--the manette--the air throttle-- the tampier valve (gasoline regulator)--and the base for the oil pulsator.I bought the set.
The First Lecture
I had been asked to talk about this project to a local group, using slides for visuals. After agreeing, it dawned on me that I=d better have more to talk about. This spurred the assembly of the fuselage in my basement. I also built the sheet metal bucket seat, and got the control stick (with its unique spade grip), the throttle, and manette set up so they could be photographed.
My son, Bob came through again with enough bass wood so that the instrument board could be fabricated.
I discovered that I was missing the blueprint for the forward fuselage formers. A volunteer group in England was building a triplane for the Shuttleworth collection, and they agreed to copy their print and send it along to me at their cost. This was a most generous solution to my problem.
When the talk finally came off, it seemed to be pretty well received, and the fire that it had built under me, to turn out structure, really accelerated the project.
Tyres & Wheels
The tires on the original airplane were 700x75 mm . It worked out that a re-worked 3.00-21 motorcycle size was an almost exact duplication. (Again, friends working on their lunch periods turned out authentic looking tires). Hubs were made by a friend, and another made a punch and die set so that the modern rims with new spoke holes in the bead seats, would allow the unique Palmer wheels, with their cloth covered, asymmetrical appearance, to be duplicated. (The spokes are radial, but the outboard ones form a cone, the inboards do not).
Getting it out of the basement
Eventually, the fuselage was set up with formers and stringers, and the tail assembly lashed in place.
This situation produced a bit of humor; at least to me it was funny. My wife was anything but supportive of this project, so I usually kept the door to my basement workshop closed. One day she strolled in, and was dismayed at the sight of this much structure. She became deeply concerned at what she assumed would be my intention to dismantle the house to get it out. I know that she did not believe me when I told her that it would go out the same way it had come in. That is, up the stairs, or out the window.
Here was one of the few advantages resulting from a box girder fuselage. It could be disassembled, and the longerons, with tagged fittings tied to them, passed out through the small basement window. I could then re-assemble in the garage, and start hunting for a barn. Wing production resumed. I continued down the left stack, and then down the right. As with the fuselage, I was able to trial fit each panel in the basement, and disassemble and do final fit in the garage. The car was destined to spend a winter in the driveway.
Aligning the fuselage
I found fuselage aligning to be a frustrating chore. As with a bicycle wheel, adjustments made in one area, had repercussions in several others. I despaired that the warps would ever be overcome. Finally, I increased the width of the longeron slot in the port cabane strut, by half the thickness of a hand saw blade, and the whole thing snapped into shape.
Fuel and Oil tanks
The fuel and oil tanks presented a challenge of their own. Not the least was obtaining the lead coated "terne plate" in the correct gauge for the project. The oil tank has a curved top that mirrors the top of the fuselage. It sits just below the top shield and just above the fuel tank. Both tanks have baffles, and sumps. After forming, they are riveted (with many, many copper rivets), soldered, then "sloshed" with a sealant. Each tank has a liquid-tight hole extending from top to bottom. These allow the filler pipe for the fuel tank to reach through the top shield; and the supply pipe from the oil tank to be connected to the oil pump.
These were two more examples of "labour intensive" construction.
The fuel tank has the added need to hold air pressure, since a modest pressure may be used to assist in forcing fuel to the carburetor.
Space – the final frontier
One of the men with whom I worked, lived in a small town some distance from my home. He owned an old barn that was used to provide storage space for other people’s automobiles. He offered me the space I would need to set up the airframe so that I could make the measurements needed to place an order for the custom-made external bracing wires. (The British had pioneered the use of streamlined steel rods instead of the stranded cables used by others. This airplane used that type of material).We moved the airplane parts to this location. He flatly refused to charge or accept any rent for his badly needed facility.
Sourcing the propellor
I had read that the famous propeller maker, Mr. Ole Fahlin, although retired, would accept orders for certain of his designs. I wrote Mr. Fahlin, and received a quote for a prop suitable for an 80 LeRhone. His price seemed in line, so he made one for me. Incidentally, the propeller mounting on these old engines is somewhat unique. The prop hub is a machined part that involves a steel cylinder with an integral back plate. The cylinder has a tapered hole running through it, and a key-way. The front of the hole is tapped.
The prop has a center hole that matches the hub. It is drilled for 10 bolts. A matching front plate bolts over the front of the prop. When the prop is aligned on the tapered nose of the engine, and the key has entered its way, a large nut, with a projecting threaded section, is tightened, forcing the tapers into a very tight fit. This large nut is kept from backing off by a 12 point sheet metal nut whose ears align with 2 of the prop bolts. For prop removal, back off the nut; reverse it so that its external threads enter the internal threads of the hub; tighten, and it has become a puller.
Triplane assembly
After carving, and finishing the 2 large interplane struts, I was able to begin assembling the aeroplane. Since the six wing panels had been individually built over some period of time, it was with some anxiety that I installed the struts. As mentioned, each interplane strut passes through the middle wing, and engages fittings on the top and bottom wings. This passage has to be in agreement with the "gap", "stagger", and "incidence" specified. It was a tremendous relief to find that everything fitted. I used hardware store bolts, stranded cable, "U" bolt cable clamps and "C" clamps to hold everything together. Other volunteers gave generously of their time and sweat to help this assembly take place. When finished, it looked like a massive model airplane without a landing gear.
I checked, and re-checked the resulting "pin to pin" dimensions, and sent a rather large international money order to the Bruntons Co. of Scotland.( The same supplier used by Sopwith). These stainless steel "RAF" wires constitute the flying, landing, drag, anti-drag and landing gear cross bracing elements. Each would have right and left hand threads at their extremities, and special fittings and locking nuts.
My machinist friends had been busy. They had supplied the triple pitched jack screw needed for the adjustable stabilizer, and the aluminum pulley that would activate it. Also provided was the machining needed on the pulsator (oil pump indicator); and after I had secured blueprints from Australia, a complete Rotherham air pump. From a loaned original British compass, castings were made, and a convincing duplicate materialized.
One part that promised to require a long lead-time was the aluminum cowl that enclosed the engine, and extended back to the front engine mount.
In connection with my occupation, I had dealt with well known manufacturer. This local company assured me that they could easily spin my part, but they needed for me to have produced an accurate wooden block, against which the spinning would be made. My son signed on for this rather massive piece of tooling. It took a considerable amount of time for fabrication. Additional time passed before the spinning was scheduled. Since set-up time was factored into the price, I elected to order 7 parts. (I sold 5 at cost, and kept 2 for myself).
My daughter-in-law Linda, sewed the cockpit coaming; the windscreen surround; and the wing cover seams. One decision that was taken, was to cover the airplane in a polyester type fabric, rather than Irish linen which had been used by Sopwith. Since the fin carried the hand lettered Sopwith logo, and I was anxious to see if my free hand lettering was going to work, I did this first. It did not turn out too badly.
The Sopwith windscreen frame was supposed to be an aluminum casting. This was beyond my capability, so I made a replica from laminated strips of ash. These were steamed, bent, glued and clamped in a frame that formed the exterior shape. This was then sanded to produce the requisite cross-sectional shapes, and when painted, and covered with the surround padding, looked fairly reasonable.
Using that same technique, and a hub that required four spokes, I also made the pilots hand wheel that controls the horizontal stabilizer trim. Although this looked OK, and I installed it, when a casting became available, much later, I replaced my wooden one with it.
I priced a Vickers machine gun, but the cost of one of these rivaled the cost of the engine. therefore I ordered a scale model plastic kit so I could study details.
Then, from home-made, full-scale drawings, a fairly convincing fake was made from wood and sheet metal. It included a short strip of webbed belting that carried a few dummy rounds, and empty casings. The gun is more than just a visual effect; the windscreen attaches to its rear.
The assembly continues
With the assembled airframe still in the barn, I had the good fortune to be visited by an acquaintance. I had met him only once before. He was the friend of a co-worker, and was doing the restoration of a Cessna 140. Being a retired contractor, he had purchased a rural tract of land, built a house; prepared a grass strip; and put up a hangar. He was concerned, though, about the many leaks in the slate roof of the barn, and the possibility for water damage to my work. He invited me to visit his layout, with the thought in mind that I could move to his place, and eventually fly it off his strip.
Since I was taking up lots of space, and would have to disassemble for the coming vehicle storage season anyway, it sounded like a fine offer. The project was moved again.
With the wings and tail removed, the fuselage protruded some, but not too much, from the back of his Dodge van. The rest of the structure followed in turn. The only drawback was the 40 minute drive from my home to his. On the plus side was the fact that this man could do almost anything; had almost any tool that one could think of, and was generous to a fault.
I really believe that without the help of Carl Fuderer, the completion would have taken much, much longer, if at all. The covering of the wings and ailerons continued at my mother’s house, and I became very familiar with the technique of "rib stitching". For added authenticity I used the frayed tapes that the British had used. This meant that I bought the modern pinked tape--cut off the pinking, and then hand frayed the edges so that individual threads projected from the sides.I had bought a modest sized air compressor, and was able to spray the colored butyrate dope
that would resemble the PC (pigmented cellulose) that had been used originally. With the blue, white and red rudder flashing, and the similarly colored "roundels", my mother's garage began to look quite colorful.
Since these had been British naval airplanes, they had been identified with "N" prefixed serial numbers. The U.S, of course, uses "N" prefixes for our civil license numbers. I contacted the F.A.A. and presented them with a list of Sopwith Triplane "N" numbers, and requested that they assign me one of them for my use. As it happened, they selected N6291, an airplane that had flown with both Naval 8, and Naval 1 squadrons in 1917. It was not a famous number, like used by those in Collishaw's "Black Flight", but it would do.
I had been advised that another of the infamous "AP" numbers, (this time AP15, used for the landing gear), was almost exactly duplicated by the forward lift strut on the post-war Taylorcraft,BC12d. A visit to the nearby T.Craft factory, netted me the remaining stocks of what had now become an obsolete part for them. Since it had served in a tension role on the monoplane, I had round 4130 steel tubes inserted, and tack welded inside the streamlined tubes. This would provide strength for the compression role it would play on the antique.
The landing gear is welded up to form a pair of "vee's". Near the bottom of the "vee", round steel tubes pass through each leg to form an anchor for the bungee cord suspension system. There is also an axle travel limiter, and guide. Being a pretty complex unit, I made wooden jigs to locate everything prior to welding. The steel "vee's" are separated at the bottom by a wooden spreader bar. This bar is shaped like an airfoil, which has a large groove on its upper surface. The groove houses the axle when in flight. The bar is built up like a wing, and is covered in plywood. Since it might see severe buckling during crosswind landings, I used ash instead of spruce, and encased the whole thing in a glass/fiber sheath. Carl helped me install the ash tailskid, and spring it with the bungee lay-up called for in the drawings.
Standing tall
We next got the fuselage up on the gear and installed the streamlined bracing wires that had arrived from Scotland. Besides the wires that brace it from lower right to upper left, and vice versa, there is also a vertical wire from the center of the spreader to the bracket that supports the rudder bar and rear engine mount. This wire resists the downward loads from the inboard ends of the axles. The outboard ends of the axles pass under the bungee suspension, and cause the characteristic front-view Sopwith appearance: the wheels exhibit large amounts of negative camber while on the ground. I now had a very tall structure, over 10 feet from the ground to the top of the top wing center-section.
The elevator, rudder, and the tail-skid control cables, and wires were threaded through the fairleads, and left in the vicinity of the control stick. Covering the fuselage became the next task. I first, did the portion aft of the cockpit. With that done, the stabilizer and the fin could be installed, and secured with the RAF wires.
Wires
The control cables connect to the stick in a rather busy fashion. The two, down-elevator cables connect to a ring at the back of the aluminum stick. The up-elevator cables pass under the floorboards-around two pulleys, and then back to the ring on the front of the stick.
The rudder cables attach to fittings near the outboard ends of the rudder bar, and the tailskid wires anchor to two straps that encircle the bar, closer to the pivot. Cable that operates the stabilizer trim, winds around and through a pulley on the outboard side of the right cabane strut, then around a pulley built into the jack-screw mechanism, then back through a turnbuckle, and on to the control pulley.
Looking back down the fuselage, behind the seat, at the maze of bracing wires and cables that occcupied the space between there and the tail, I was struck with another meaning of the term "fly by wire". Now it was time to install the engine.
Installing the engine
Since the LeRhone is lighter than the Clerget, I decided to set it 6" forward. This meant that a spacer had to be constructed. It would sit between the mounting plate of the engine, and the front engine plate of the fuselage. It was made with parallel plates, spaced apart with large diameter steel rods, located on the centers of the bolt holes in the engine mount. The entire unit was welded to form a unified part. The rods were drilled to accept the much longer bolts that I had obtained. As mentioned, the aft end of a rotary engine is a long cylinder that ends in the flange that carries the carburetor, and attaches to the rear engine mount. It is threaded to take a flange. I had an extension tube made with a threaded collar to account for the front spacer, and the fact that the rear mount was positioned for the longer Clerget.
From the first, I had been a little concerned about what would happen when this rather heavy steel engine was attached to the somewhat fragile-looking fuselage with its slender wooden struts and attendant sheet metal, and wire fittings. We picked up the engine with a sling, hooked to a block and tackle, and swayed the extension tube back through the front engine plate, and then on through the rear engine mount.
When the bolt holes in the front plate lined up with the holes in the spacer, we began installing the long bolts. The rear mount accepted the flange, and when they were bolted together, we released the block and tackle. The fuselage was now carrying the load. It was as if nothing had happened. Seemingly, there was no change in the tension on any of the piano wires, and the aspect of the longerons was un-affected.
Electrics and Carburetor
It was now time to connect the electrical, and carburetor controls. The electrical side of the house is pretty straight-forward. There are only the "P" wires wired parallel to the points in the magneto. These go to ground either through the ignition switches (old fashioned porcelain wall switches), or the normally "open" button switch in the center of the spade grip on the stick. The infamous coupe' or "blip" switch.
Carburetor controls consist of threaded rods that attach the Aball@ type fittings on a bell crank to the air-gate/needle valve on the carburetor, and the large brass throttle on the control quadrant.
A second control is the "manette". Also located on the quadrant, it is a wheel topped lever that controls the tampier valve (A needle valve that regulates the amount of fuel that will get to the 2nd. needle valve that is part of the carburetor). This is also a threaded rod set-up. With this installed, the forward former/stringer assemblies could be attached, and covered, as well as the low wing box ribs which project slightly from the airplane.
Installing the top shields
It was now time to shape, and install the top, and side shields. These are aluminum sheet that produce the round shape that develops in front of the cockpit. The top shield goes under the machine gun and has holes for the oil and gas filler pipes. It is secured by wood screws, and the cable that will hold the nose cowl to the cowl ring. The side shields have doors that give access to the carburetor, and holes for the carburetor air intake pipes. The side shields fit into an aluminum channel at the bottom, slip under the top shield, and pass under a cable that grips from the rear. The front edge rests on the cowl ring, and is under the nose cowl. The bottom shield is held by metal clips that encircle the longerons, and their outboard edges fold to create the channels needed by the side shields. With the nose of the airplane buttoned up, it became possible to go for some engine tests.
Testing the engine
Since the octane rating of the fuel for which the engine was designed was low, un-leaded auto gas is just about right. Instead of castor oil, I opted to use 60 weight racing oil+10% STP. With the tanks containing enough fuel and oil, we pushed the wingless wonder out of the hangar, chained the sternpost to a heavy piece of farm machinery, chocked the wheels with blocks of concrete, and wondered what to do next.
Having gotten some advice from the few people who had any to give, and read everything I could find about the subject, we started out.
First, we hand lubricated the ball bearings and the sliding pins in the rocker arm assemblies. Since the fuel system can be pressurized, we closed the vents, and began using the small hand pump located at the pilot’s right shoulder. With the gasoline tank valve open, and the manette, and air valve full forward, we pumped until
fuel began running out the carburetor overflow pipe protruding from the bottom shield. Then we reduced the manette to about one quarter open, and the throttle to one half open. Priming was accomplished by squirting raw gasoline past the depressed exhaust valves as they appeared beyond an opening in the nose cowl, when the engine was hand rotated in reverse. Since my crewman had never propped an airplane before, he got to sit in the cockpit, as I flailed away on the propeller. There were many silent rotations before it suddenly burst into life with what was to become a characteristic loud, snapping clatter. Attempts to throttle back were greeted with long periods of non-firing, and the sounds of whirling machinery, then movements of the manette would restore the mixture to a runable one, and the "song" would come back.
About all that we learned was that it would run; the dreaded torque would not rotate the wing-less fuselage enough to tip it over; and when one depressed the "blip" switch on the stick, the engine stopped firing, but came back to life when the switch was released. In successive engine runs, it began to dawn on me that at least this rotary engine was very smooth. There was none of the vibration that a normal radial puts into the airframe. It reminded me more of what one might expect from an electric motor, and made me feel strangely disconnected from the source of what was apparently, very powerful thrust.
Moving the project again
Carl had been flying his Cessna for some time, and it was becoming obvious that if the wings were put on the triplane, things would be pretty crowded. In fact, the wheels would have to be removed to get out the hangar door. (We had used a taller, but narrow, back door for the engine runs). As it happened, the folks at a local airport that annually staged a rather ambitious air show, got wind of my project. I was contacted about the possibility of my triplane being exhibited statically at the next edition. To sweeten the offer, they made arrangements for me to rent a hangar with a sufficiently tall door.
Again, Carl made everything possible. He had a trailer that could transport the fuselage with the tail assembly attached. Subsequent trips carried the wings and other sundries that had cluttered his hangar for several years. As with the owner of the slate roofed barn, attempts to establish and pay rent for the space had always been deferred, and never followed up; then finally refused.
Carl either flew, or drove over every day during one of the hottest summers on record, as we worked on final assembly. The fuselage was leveled with a saw-horse and a jack under the tailskid. We built some 2"x2" scaffolding (held together with "C" clamps), to support the low wings.
The steel-spar boxes project slightly from the wing roots, and enter mating spar housings in the fuselage mounted box ribs.(Extensions of the low wing center section). Long steel rods, entering from the trailing edges, pass through the rear and front spars and lock the outer panels in place. Additionally, drag and anti-drag RAF wires attach to fittings on the wings.
They anchor either to the front engine plate, or the fuselage aft of the cockpit, and when tightened, assist in pulling the spar boxes into their housings for the final fit. With the low wings on, we could erect the interplane struts. The middle wings having previously been clamped in place at the struts midpoints. (Additional 2"x2" lumber, clamped to the scaffold, supported the load ) Another set of drag and anti-drag wires pulled the spar boxes into mating housings concealed in the mid wing box ribs on the cabane struts. Steel rods, entering from the leading edges, pass through the nested spars.
Landing wires
We could now install the landing wires. Fittings attached to the compression tube (that is built into the top wing center section), anchor the upper ends. From there, the wires pass through plywood sheathed holes in the middle wing on their way to the fittings installed on the low wing steel compression rib. The single landing wires lie in the plane of the forward staggered cabane, and interplane struts. With 4 of the wings now supported by the fuselage, we could begin to hoist up the top wings, and eventually work them into place.
High up on step ladders, and without the help of any drag/anti drag bracing, to assist the panels into place, getting the spar boxes to nest and accept the securing rods, is an acrobatic exercise. As with the mid wings, the rods enter from the front. These rods are pointed on their entering ends, and flattened on the rears. Single wood screws penetrate the flattened portions, and fasten to wooden structure behind.
One problem in setting up this aeroplane, is the fact that there are no adjustments to the incidence angles for the wings. Once bolted in place, whatever angle selected, is it. As a result, Carl and I took considerable care, even though hardware store bolts and nuts were being employed, to get things right. (The deadline for displaying at the air show made it impossible to wait for aviation hardware from a supplier.) The original drawings did not call for the wings to be washed for propeller, or engine effects.
I, in my infinite wisdom, elected to reduce the tip incidence on the right middle wing, and increase it on the left side. We successfully finished in time for the air show, and a friend who lived in Pennsylvania, flew his Fokker triplane over, and we parked side by side. Our two airplanes attracted considerable attention, and made all of the work seem worthwhile.
With the rush to exhibit out of the way, aircraft hardware could be ordered, and gradually installed. One obvious feature on the fuselage, is a series of eyelets that extend along the right top longeron between the sternpost and the cockpit. Criss-crossing those eyelets are short lengths of twine, tied frequently with knots. This is an inspection panel. Untie the knots; remove the strings; and the panel flops down to provide access to the guts of the fuselage.
Before calling for an FAA examination of the project, I wanted to re-check the interior of the fuselage for the presence of cotter pins; tension of piano wire etc., but I did not want to unlace that panel, and risk that the appearance of the fabric would suffer.
I, therefore, elected to crawl back down into the fuselage. (After first having removed the seat). This meant that the vertical cross-bracing wires would have to be opened up and re-set as I worked from back to front. The heat of summer continued, and this would be no place for someone with claustrophobic problems. I was alone in that thing for hours, and if I had gotten stuck, no one would be coming to help me out.
I decided that I would never do that job this way again. The old guys must have known what they were doing when they provided that big opening.
As the construction had progressed, I decided to find whether I remembered anything about flying..
Back to flying
The first thing that came up was the scarcity of tail dragger opportunities. The airplanes themselves have become rare enough, let alone instructors who were current in them.
The first episode involved a former instructor who had a PA18, and would let me fly from the right seat. This would preserve the hand relationships regarding throttle and stick.
As it happened, I picked a windy, snowy day. I also uncovered a problem with the steerable tailwheel, and with no brakes on my side to help correct for it, the take-off produced a yell for help.
As we flew around the pattern, I remember that I questioned myself about ever having felt comfortable doing this in the past. I decided to hook up with a pilot from work who was currently instructing, and take some formal training from him in his PA28 Archer. I did shake off some rust, but I began acquiring bad habits for future tail dragging. I was encouraged to forget that I had feet, especially during landing roll outs.
The FAA Inspection
Engine tests, culminating in a taxi test took place with the triplane. The steerable tailskid actually was of some use, and the plow on the skid did provide some braking.
Prior to asking for an inspection by the FAA, it was necessary to do a weight and balance. Having that information was well and good, but there is no C/G information on the blueprints. Fortunately, the group in England had gotten an aerodynamicist to study the problem. He reduced the Sopwith Triplane to a theoretical monoplane, and arrived at the fore, and aft limits.
They mailed me a copy of his presentation, and I was greatly relieved to find that my airplane balanced within the rather tight limits he had calculated.
I suppose every home-builder has trepidations about the FAA inspection. I know that I did.
It was a bitterly cold morning when I met the two men from the regional office. The hangar had been specially heated, and I brought everything that I thought would be required.
Even though this was probably the first time either one of them had ever seen an airplane held together the way this one was, they gave it pretty fair inspection. They discovered a couple of missing cotter pins, and suggested that some of the music wire ends could cause cuts to unwary flesh, but not much else. We then had to open the door; (letting out all of the heat) for neither had ever seen a rotary run before. Luckily, it fired on the first swing, and I let it rev as long as my un-helmeted/goggled parts could take the wintry blast and then shut down.
The sign-off was almost anti-climactic.
Learning to fly again
Meanwhile, the adventures of the would-be (again) tail-dragger pilot continued to stagger on. I was finally allowed to take some dual instruction in an Aeronca Champ. Here I really re-discovered my feet, and began to take charge of the ground antics of the airplane on take-off and landing. If the nose was going to swing, it was because I caused it! Not because I was chasing the airplane. The final confidence booster came in a Stearman owned by a friend.
After a bouncy landing, I took off my glasses; pulled the goggles over my eyes, and shot three acceptable landings. One of them might have even qualified as a "grease job". My glasses were making it appear that I was closer to the ground than I really was. I passed my next third class physical without glasses.
Time to fly the Sopwith Triplane
By virtue of a fortunate friendship, I was able to solo a J3 Cub, and build some time in it. This built my peace of mind as well, and finally led up to deciding that it was time to go "Sopwithing". The October morning that I had selected was clear, frosty, and calm.
I drew a larger crowd than I would have liked. I didn't want to possibly embarrass myself in front of a lot of people. (I planned to stay in "ground effect", and if things didn't feel right, I could set back down on the remaining grass).
The engine was balky. At first it would only run out the prime. Finally it accepted fuel from the tank, and accelerated to 1100 rpm. By carefully playing with the mixture, I idled down to 550 rpm.
While waiting for an airplane in the pattern to land, I signaled for a mechanic to check the security of the spring loaded latches on the side shield inspection doors. Everything was OK, so when the traffic cleared, I waved away the chocks, and advanced the throttle and manette.
The airplane began to roll instantly. A little forward stick, and the acceleration was surprising.
When I applied a little back pressure, the feel of the tires went away at once, so I eased off and glanced at the airspeed indicator. It was reading zero. Feeling that I was still pretty low, I looked overboard, and was shocked to see that I was about 100 feet up, and still climbing. With the "ground effect" plan out the window, I went to plan "B" which was "fly the airplane".
I climbed to pattern altitude, throttled back a little, and tried to set up a pattern that would get me back to the grass. It was then that I realized that I was flying cross controlled to keep the ball centered. It was taking lots of left stick, and right rudder, I judged that my wash-in/wash-out decision had been a bad one.
On my first attempt at a pattern, throttling back, and "blipping", like I knew what I was doing, I turned base. As it happened, I had been too close on downwind, and not wanting to do any steep banks to correct it, I just went on around.
The second pattern was better, and I carried on until I was fairly low, but again I was going to have to "ess" turn to reach my wind line, so I fired up and climbed again. This time, there was an audible howl from something in the slipstream. Just before turning crosswind, The right side inspection door popped into my line of vision. It seemed to be standing out from the shield, and then it was gone.
I throttled back and considered landing in a cornfield that was dead ahead. I was too high to just glide in. It would take some side-slipping. Since I knew nothing of the stall characteristics, and needed cross controlling just to fly, I decided to try to get around one more time.
A thunderous racket began, and this was accompanied by a frightening vibration.
I needed to apply power just to hold my well-below-pattern altitude. It was a relief to find that when I added power, the noise and vibration reduced markedly, and raised my hopes that there was nothing wrong with the engine. On downwind, just before turning final, there was sudden blast of cold air on my right leg, and daylight illuminating the interior. Then there was a jolt, and the airplane skidded left. I corrected, and Ablipping@ regularly, and praying constantly, I brought it around and reached the grass area that I was supposed to be using. Relieved, I relaxed my grip on the coupe= and the airplane ballooned slightly . Thinking again, I squeezed tightly, deciding that we were going to land now! I kept increasing back pressure, and the airplane began to settle at a rather fast rate. The main gear hit, and I thought I was into a 10 feet bounce, but it only seemed that way. When the tail skid got down, the airplane hardly rolled one fuselage length.
With the engine shut off, I was joined by a host of people.After climbing out, I was shocked at what I saw.
The right side shield was gone, and the nose cowl was sagging into the trailing edge of the propeller. Cavities had been grooved in the prop, and there were several slits in the fuselage fabric that skipped almost to the stabilizer. The missing door and side shield were recovered, so we pushed the airplane back to its hangar.
During the car ride home, I began to experience an adrenalin let down, and realized that I was very tired. I also realized that the people on the ground had been quite frightened for me. I had been very excited, but I don't think that I had time to get scared. That afternoon, I returned to the hangar, and with the help of an experienced pilot/mechanic went over the mornings events. We traced the path of the departing side shield.
It had struck the lower part of the fuselage as it tumbled aft, and slammed into the right front flying wire on the stabilizer. The force was great enough to damage the fitting that held it to the stabilizer spar, and almost tore it loose. This had allowed the right side of the stab. to assume a higher incidence, and accounted for the skidding I had experienced.
If that fitting had failed totally, someone else would be writing this.
Our analysis of the incident went something like this:
1. The spring loaded latches on the right inspection door had failed to stay latched.
2. The door swung out, and acting like a sail, pulled the right side shield back from under the cable that held down the nose cowl.
3. The loosened nose cowl, still anchored at the bottom, tilted forward, and came down on the whirling rocker arms of the engine, machining the trailing edge of the propeller in the process.
4. The side shield, held only by its rear cable, blew away, and tumbled back, striking the stab. bracing wire before falling clear.
5. At higher rpms, the air whirling out from the engine floated the cowl away from the rockers, and reduced the vibration and noise.
Almost another year was to pass before the triplane was again ready to fly.
Repairs to get flying again
The nose cowl was replaced, and it and the inspection doors were held by screws as well as by the traditional fastenings. The propeller was repaired. The stabilizer had been removed, repaired internally, recovered, and installed; and the fuselage tears repaired. The airspeed indicator plumbing was redone and tested. This time, the take-off was no surprise, and the airspeed indicator worked.
As I sailed around the pattern, I began to perform some tests. First, at about 80 knots, I let go of the stick and rudder. The nose didn’t rise or fall, but the airplane rolled right, and yawed left.
It required full left stick, and full right rudder to bring the ball back to the middle. I reduced power and explored the realm of stabilizer trim.
( I forgot that I had lately received missing drawings involving a clutch, built and installed it).
I shoved the control wheel rather hard. Nothing happened. I made a mental note that I would have to look into this.
A couple of minutes later, there was an audible "bang" in the tail, and the airplane dived sharply.
My seat belt/shoulder harness set-up allowed for considerable vertical movement on my part.
I used it all. I was shot up out of the cockpit. My head was well above the windscreen, and my goggles blew over my head.
Thinking that something had failed, I managed to get back on the seat, and get my feet on the rudder bar. (I was not wearing a parachute and with the drag wires surrounding the cockpit I'm not sure one could bail out anyway])
I pulled back the power and considered an off-field landing.As I glided lower, the airplane seemed to be responding OK, so I re-applied power and climbed to pattern altitude, and went around and set up an approach. Harking back to the first flight, I expected the airplane to float some on final, so I let it get pretty low. It didn't float at all, and the wheels touched down in a fairly decent wheel landing.Since I was not yet to the grass where the taxi lights had been removed, I applied power and went around. This time, I made a 180 degree side approach, and hit the grass that I meant to use.
The landing was more tail high than I intended, but the gear took the shock--the tailskid planted itself, and with very little roll-out, the airplane stopped.
The next day I checked the incidence angle of the stabilizer. It was well above the 1 degree that was there when I had started the flight.
I deduced that the strain that I had put on the cables had eventually over-powered the clutch and caused the jackscrew to rotate. (The clutch is set up to require an inboard pull on the wheel to overcome a spring load, followed by wheel rotation, then release of pull, for re-engagement of the clutch).
That stabilizer set-up provides for one powerful control system. (In theory, it mimics the F100 flying tail).
I now believe the words of a WW1 pilot who stated that with the engine two thirds open, rolling back the trim control could initiate hands-off loops that continued until the fuel ran out.
I have leisurely re-rigged the airplane for no wash in/ wash out on the middle wings, and am ready for a flight that hopefully will leave my knuckles pink this time.
When I view the airplane again after an absence, I need to re-assure myself that I actually caused this all to happen. To me, it is still an attractive airplane, but beauty truly is in the eye of the beholder.
None of which provides any real answer to WHY! Maybe none is needed.