Scratchbuilt 1/16 Salmson 2A2 By Jim Landon

Introduction

The Salmson 2A2 was a two-seat biplane used in World War One for photo reconnaissance and artillery spotting. Designed by Emile Salmson in late 1916 to early 1917 for the French Aviation Militaire it went into production in the latter half of 1917. 3,200 were produced, 2,200 by Salmson, the rest by three other manufacturers. After WWI, Japan mass produced a licensed version for reconnaissance called the Otsu-1. Not one Salmson or Otsu remains anywhere in the world, as far as we know, except a replica Otsu in the Kakamigahara Aerospace Museum, and the delapidated hulk of an Otsu fuselage on display in the same museum. There is a scale model of a Salmson 2A2 hanging from the rafters in the WWI section of the USAF Museum.

The American Expeditionary Force purchased 705 Salmson 2A2's and 557 were dispatched to the Zone of Advance, making it the most numerous observation plane in the USAS. Ten Observation Squadrons were equipped with Salmsons. The Salmson 2A2 was a counterpart of the Brequet 14 two-seater. Together the Salmson and the Brequet replaced the A.R.1, A.R.2 and Sopwith 1-1/2 Strutter in the escadrilles. Fifty-five escadrilles were to be equipped with the Salmson 2A2.

The Salmson 2A2 was developed around the newly introduced Salmson Canton-Unne 9Za engine, a nine cylinder water cooled radial rated at 260 hp at 1600 rpm.. For comparison the famous Sopwith Camel only had 130 hp. The Salmson engine and the solid airframe that housed it proved to be of unsurpassed reliability. Several other features are pointed out in the model build story.

Why I Modeled a Salmson 2A2

In April 2000 I was asked by my friends at Wings Over the Rockies Air and Space Museum in Denver Colorado to build a 1:16 scale model of the Salmson 2A2 in which Lt. Francis B. Lowry served. Lowry served as an observer in a Salmson 2A2 in the 91st Aero Squadron, in the American Expeditionary Forces fighting in France. Lowry was killed in action on his 33rd mission on September 26, 1918. He was posthumously promoted and awarded the Distinguised Service Cross, and later the Lowry Air Force Base in Denver was named for him. Lowry AFB was deactivated in 1994 and is now a business and residential community. The Wings Over the Rockies Museum resides in a 1940s era Lowry AFB hangar.

Reference Material

The only reference material I was given initially was two 8-1/2î X 11î photocopied pages with exterior-only views of the plane. I found better copies of the two 8-1/2"X 11" pages in the book 'Scale Aircraft Drawings, Volume I - World War I' by Air Age, Inc. The drawings were done by Richard Anderson in 1961. I enlarged them to 28-15/16" wingspan, to get 1:16 scale templates. I also borrowed and scanned several 8 X 10 b&w photos from Wings Over the Rockies, but it seemed as if nobody else I wrote to had even heard of a Salmson 2A2. I took snapshots of the USAF model from below, but could not see into the cockpits and the model builder did not have his reference material any more. So I assumed that no interior information existed and my initial work started on 4-12-00 with only the exterior in mind, based on the Anderson drawings.

Since then I have collected many other photos and drawings -- thanks in large part to several members of the WWI Modelers web site. By much digging I eventually found a lot of info for the interior of the cockpits, including a priceless photo from Musee de l'Air et de l'Espace, of the interior of an observer's cockpit, and a great French drawing of the interior structure provided by Brian Nicklas at the NASM. I enlarged the French structural drawing to 1:16 scale but of course no two drawings or photographs ever quite agreed. Photocopies of drawings used to build the Salmson replica in Japan were very kindly snail mailed to me by Mr. John O'Reilly in Japan, to whom I am deeply indebted. He also sent me snapshots of the Japanese Otsu replica and snapshots of the Otsu hulk. My friend Mike Fletcher provided a copy of a Salmson replacement parts manual and rigging instructions. A back issue of W-W-1 Aero Journal had photos of the pilot's cockpit interior, so I ended up detailing the cockpits extensively and even made the control surfaces operable from the pilot's cockpit.

Ironically, after I had done so much research on my own, a book was published by Flying Machine Press called 'Salmson Aircraft of World War I' It was invaluable for exterior detail but contained no interior drawings and only one worthless interior photo. Another irony was that after I'd finished my model a Datafile was published on the Salmson 2A2 with more extensive drawings than the FMP book, including some interior construction and some cockpit details. Thankfully none of this contradicted or added to my own research.

Construction

The construction of this model took over four years, largely because it was my very first total scratch build. I'd only scratched a few components before for a Guillows balsa and tissue Thomas Morse S4C kit. Much of my time on the Salmson went to trial-and-error and undoing catastrophies. To relate all the steps of construction and methods of building each individual part is impossible here. I did however take hundreds of in-progress photos that can be found on my website, as well as additional finished model portraits. I kept a daily log of the build that fills two 3-ring binders. This article will attempt to provide an overview and hit some things that are not so commonplace in modeling.

The fuselage seemed like the logical place to start. Based on the cross sections A-A through E-E given on the Anderson drawings I made formers from scrap Plexiglas with an aluminum tube down the center for alignment, and piano wire stringers to create the overall exterior shape. After friends started providing drawings and photos of the interior, I started adding fake interior framing where it might be visible to someone peering into the cockpits. This model was now being built from the outside in, backwards of how I would have done it if I'd thought interior info would be found.

The cabane struts had to be strong enough to support the top wing, and yet slender to look like the real thing. I sandwiched stiff piano wire inside layers of styrene sheets. A mixture of oil paints created the appearance of wood. But for the interplane struts I used mahogany instead of styrene for an even more convincing wood appearance. The top wing center section sits on top of the cabane struts and is actually a semi-permanent part of the fuselage. The top wings attach later to this center section. The spars are aluminum tubing and the ribs are styrene. The leading edge and trailing edge are piano wire. Styrene plates were used in lieu of cross-bracing wires. Covering was deferred until later.

I used .015 inch diameter piano wire for the cross-bracing wires in the fuselage, which may be a hair larger than scale, but stays straight during handling and greatly simplified installation. To simulate turnbuckles I first cut and ground the brace wire to the exact length, then slipped telephone wire insulation over the wire and added drops of epoxy to represent the connections. This allowed a rigid one-piece bracing wire with the turnbuckle 'built in'.

The Salmson did not have the usual fixed vertical stabilizer fin with separate hinged rudder, instead the whole vertical fin pivoted on a vertical axle shaft. I wanted my model to have control surfaces operable from the cockpit, so I imitated what I saw on the Anderson drawings. The rudder is made from sheet styrene with brass tubing for the hinge pieces and piano wire for the rudder post or hinge pin. Piano wire simulates the ribs. The Salmson did not have a fixed horizontal stabilizer either - the whole horizontal fin pivoted on a horizontal axle shaft. Construction is similar to the rudder except bearings were needed outboard on the exle shaft for the bracing wires and rods.

Because my previous model was a balsa and tissue kit, I chose Silkspan tissue to cover the elevator and rudder but used epoxy to attach it to the styrene. There probably are as many ideas on how to cover a model as there are model builders out there. I might not even do it this way next time myself.

Throughout the build of this model, I alternated my attention between its various subassemblies, working on the fuselage one minute and the tail pieces or wings the next minute, while epoxy or paint was drying.

Since my previous model was a balsa and tissue kit, I chose to build the wings the same way those kits are made, except with aluminum tubing for spars and styrene for the ribs. This also supported my plan to have cables inside the wings operate the ailerons.

One thing I've discovered about scratch building a model airplane: you spend a large percentage of your time engineering tools and templates and fixtures. You have a bazillion wing ribs to make, and they have to be the correct shape and all identical (in the case of the Salmson) in order for the wings to be uniform and straight. I chose 0.040" thick styrene sheet because it was stiff enough, yet still fairly easy to cut. For the second wing panel I switched to 0.030" styrene sheet because the .040" turned out to be overkill.

I built a fixture for holding the rib outline template steady while tracing around it. It also ensured that the two spar holes were aligned between the template and the blank. Later I added pins at the leading and trailing edges so that all four holes are aligned and the template doesn't move while tracing around it.

Before starting to assemble the bottom wings, I realized I needed to make sure they would be exactly perpendicular to the fuselage and the finished wingspan needed to be exactly 28-15/16 inches. And each of the four wing panels needed to fit properly onto the attachments on the fuselage. Plus, the wings had to have the correct dihedral. Plus the undercarriage had to be planned so that the axle was positioned correctly, with the right distance down from the fuselage and between the wheels. Plus, the elevator and rudder needed to be properly aligned horizontally and vertically. From past experience I believed the best way to handle all this was with a permanent sturdy assembly fixture.

Another factor is that the top wing spars needed to be directly above the bottom ones. The assembly fixture was first leveled itself, then the fuselage was leveled. Then everything could be aligned and dry fit.

At the top of each cabane strut, brackets were needed for attaching the landing wires and the cross bracing wires from one cabane strut to another. I made templates on my computer and glued them to 0.005" brass sheet with rubber cement. The holes were then center punched with a needle-sharp tool, then drilled out with a #80 drill bit. Manicure scissors were used to cut out the parts, the paper was peeled off and the edges were cleaned up with files and sandpaper. Finally the parts were bent to shape with needle nose pliers. A big hole in the center fits over the 1/16" piano wire that extends out the top of each cabane strut.

I was so anxious to do the wire X-braces between the Cabane struts that I stole ten of the X-braces that had been made to go inside the fuselage! But before installing them I had to figure out where they would pass through holes in the cowling. I used drafting mylar to make trial fit pieces and templates but I liked the mylar so well, I used it for the final cowl panels around the cockpits.

For the wings, a 1:16 scale template was created on my computer, based on photocopies of drawings used to build the Salmson replica in Japan. The American and French 3-view drawings each contained various obvious errors. I had to carefully study many photos to determine which drawing was right.

For the first two wings (bottom wings) I used brass doll house hinges for the ailerons. By the time I built the top wings I had become more sophisticated in my modeling and made my own scale hinges from piano wire and aluminum sheet.

The aileron control horns were fashioned from brass sheet, epoxied to the styrene. The aileron control cables were simulated with Spiderwire, a type of fishing line. Spiderwire is extremely slippery almost like Teflon so even going around a 90 degree turn imparts very little drag. Thanks to Guy Fawcett for telling me about it and providing it. The pulleys for the aileron cables are made from the brass end rings from broken guitar strings. They look exactly like little pulleys. Everybody I showed this wing to asked to pull on the string hanging out from the wing root and see the aileron move ... and the adults were worse than the kids!

Advance planning was needed for the eventual addition of the interplane struts and rigging. I drew rigging brackets on my computer, based in part on photos of the real plane and the Japanese replica.

For the second wing panel I made a few refinements. I switched to 0.030" for the ribs for easier cutting and sanding. The flimsy part of the rib was from the rear spar back, so I epoxied piano wire alongside before cutting out the rib. Trying to install the piano wire for the leading and trailing edges (and keep it there) was a challenge, so I left excess plastic around the hole until after everything was epoxied securely. On the first wing I cut out the aileron after the whole wing was built. This time I had the first wing to use as a guide and built the aileron separately.

Photos of real Salmsons suggested that the wing covering was flat, without the valleys between ribs seem on some planes. So I chose mylar drafting film for the wing surfaces. It is very strong because it was designed to make engineering drawings (before the days of computerized drafting) that would not tear or stretch. I applied epoxy generously to the tops of the ribs and used a dowel to roll the mylar down firmly against the ribs until the epoxy hardened. The mylar works extremely well on flat surfaces, or on uniform curvature like the topside and underside of the wings. But it will not stretch or curve in two directions at once, nor can it be shrunk after installation. So for the wing tips I sliced the mylar outboard of the last rib into narrow strips and epoxied them one at a time to the outer edge of the wingtip former. I pressed the strips concave until the epoxy cured, and then used putty to build up the desired shape. The wing root presented the same challenge as the wing tip.

Whenever the wings had wet epoxy or putty, I worked on the fuselage. The first piece of "jewelry" I made for the pilot's cockpit was what I call the elevator transfer axle. It translates joystick forward and backward movement into elevator control cable movement. The design was deduced from the French 3-view drawing. I used brass tubing and sheet and soldered it together, as it would be inaccessible if it broke. The pilot's seat sits above the transfer axle, hiding my little masterpiece.

The second piece of "jewelry" created for the pilot's cockpit was the joystick assembly. It connects to the elevator transfer axle beneath the pilot's seat. The design came from Japanese parts drawings. Again I used soldered brass. After installing the joystick in the pilot's cockpit, I wanted to test its ability to move the elevator properly. The assembly fixture provided a way to hold the elevator and fuselage in correct position for the test. The elevator control cables were temporarily attached to the elevator itself and the joystick operated the elevator as desired, just like the real thing.

The pilot's instrument panel is 0.015 inch thick brass, painted with an oil paint mixture to resemble wood, sealed with sanding sealer and Future. I later dry-brushed black oil paint over it to try to make it look "used" because Lowry was killed on his 33rd mission. The oil pulsator started with brass tubing soldered together, and additional pieces epoxied on. Extra fillets of epoxy were applied to give it the look of a casting, painted brass and dry-brushed with black oil paint. The glass portion was whittled from clear sprue, sanded and coated with future.

Instrument faces and bezels are from a radio control plane aftermarket instrument kit, modified and adapted. The magneto switches and map scroll knobs are made from doll house brass hinge nails. The map was scanned from an issue of Cross & Cockade and shows the town of Crepion France where the plane was shot down. The oil lines in and out of the pulsator are copper wire. The mounting screws for the pulsator and RPM meter are just empty drilled holes, enhanced by the black oil paint treatment. The mounting screws for the air temperature gauge are dots of brass paint done with a toothpick. The four mountings for the height gauge are just painted bits of styrene. The label plate above the wireless switch is a scan from the actual blueprint printed on clear plastic, painted brass on the back side, dry brushed with black. Its mounting screws are dimples made from the back side with a tool similar to an ice pick. The RPM meter bezel needed to be concave, so I installed one of the R/C kit bezels backward and epoxied the dial face on the "wrong" side of the clear plastic.

Japanese drawings of the throttle quadrant parts were scanned and reduced to 1:16 scale, the templates were glued to .005" brass sheet, holes were center punched and drilled. Then manicure scissors were used to cut out each piece, followed by filing and sanding. The quadrants were assembled per japanese drawings, using piano wire for the axles and guide bars.

I constructed the rudder bar and its support structure from brass tubing, brass sheet and piano wire, following the Japanese scale drawings. The cables are routed through the fuselage per drawings and are visible where they pass through the observer's cockpit. The rudder bar operates the rudder just like the real thing. The oil tank formed the right wall of the pilot's cockpit. I used a rectangle of brass, dirtied up with black oil paint. All of the fuel lines are piano wire bent to fit by trial and error. The pilot's chair is built up from styrene sheet per Japanese drawings. Other items are bits of electrical wire insulation, aluminum and brass sheet and dots of epoxy for screw heads.

The left side of the pilot's cockpit featured a hand wheel for elevator trim, and a few unidentified things copied from photos and drawings. The accurately scratch built fuel tank with its support structure and tie-down straps is out of sight behind the pilot's chair but I know it's all there. Behind the pilot's headrest is an enclosure covering the fuel fill pipe and numerous fuel lines and vent lines. No drawings or photos show seat belts so I didn't invent any.

Work sometimes alternated between the pilot's cockpit and the observer's cockpit. Because Lt. Lowry was an observer it was essential that I show what his cockpit would have been like. It is based on the photo I bought from Musee de l'Air et de l'Espace and some Japanese drawings. An important part of the observer's equipment is the wireless telegraphy system. I made a scale telegraph key from bits of styrene and brass. I almost made it operable but decided that was too insane so I epoxied it together.

Other equipment included a reel of antenna wire (the observer had to reel out a long wire behind the plane for the antenna), other wireless transmitter equipment, a storage rack for photographic plates or maps, and an ammunition magazine rack.

Perhaps the most important part of the observer's equipment was the camera used to take photos of enemy activity. The body is sheet styrene, the handles are piano wire. The site with the cross hairs was printed by computer onto overhead projection film.The cross bracing wires in the observer's cockpit were arranged so that the crossover point was well below center, so that observer had access to the camera and to the spare ammunition magazines.

Other equipment included a spring-loaded fold-up seat that would retract out of his way if the observer had to stop his recon work and man his guns. There was also what appears to be an electric lamp with a pistol grip, possibly for sending morse code. People at the museum, my family and others thought it was a crime to cover the model, hiding all the neat detail. I explained that the neat detail was only intended to be seen through the cockpit opening, and would still be visible. They didn't buy it. I plan to build uncovered models in the future.

The compound curve portions of the fuselage framework were covered with tissue paper and shrunk with water, just like the old fashioned balsa and tissue models I grew up with in the '50s.

Despite efforts to apply the tissue carefully, some areas had imperfections requiring putty - much like doing body and fender work on a car. Then I discovered that my stringers were not quite evenly spaced, so I used 1/64 inch wide Chartpak tape on top of the tissue to create fake stringers. Chartpak tape is self-adhesive and can be pulled up and repositioned repeatedly - which was a life saver. A thick liquid filler was applied alongside the Chartpak tape to eliminate the square edge of the tape. The filler was then sanded to create a radius alongside the tape to immitate the appearance of fabric over stringers. A great deal of time was spent sanding, applying more filler, sanding again, applying sanding sealer and sanding. It needed to look as if linen was simply applied over correctly spaced stringers.

The front cylindrical cowling and access panels are made from an empty Slim-Fast can. These steel cans are stronger than aluminum soft drink cans, and the model can be handled without deforming. Surface details like louvers can be glued on and removed if wrong, without ever loosing the perfect cylinder. Knowing that the rounded nose of the fuselage was going to be a problem for me I started saving empty plastic bottles early in the project. Many round plastic bottles were tried, but only a mustard squeeze bottle proved to be exactly the correct shape. Of course it was not nice material to work with but I had to have that shape

The shutters in the front end of the fuselage controlled air flow to the radiator, which was immediately behind the shutters. This regulated engine temperature and prevented the radiator from freezing. The shutters were controlled by a lever in the pilot's cockpit. I used thin styrene for the blades and set them at half open with flat black paint behind them.

Ever since I started this project, people had asked how I planned to do the louvers, and suggested ways. I didn't know, I'm no tool & die maker. First I used a computer drafting program to lay out paper templates to get equal spacing and alignment for the louvers, rocker arm covers and other items on the engine cowling. I had been thinking that I would make glued-on louvers out of pop can aluminum or thin brass sheet, etc., but I didn't have much luck, so I decided to go with styrene quarter-round and use flat black paint on the rear face to simulate the "opening". I filed the rear faces slightly concave to aid the illusion.

I bought three radio control airplane propellers in a hobby shop, thinking one of them might be close enough to modify for my Salmson. When I got them home and compared them to photos I was dumbfounded to discover the Salmson prop turns backwards from all the props in the hobby shop - and in fact backwards from most WWI props. So I proceeded to make my own from six layers of walnut laminated and carved. The reason I made the prop now, when the rest of the model was nowhere near done, was because I needed some fun. I was sick and tired of puttying and sanding.

As I understand it, the plumbing on the top of the nose of the Salmson would be analogous to the top radiator hose from your car's engine cylinder head to the top inlet of the radiator. My model is based on a photocopy of a Japanese drawing, plus photos.

I always figured the pilot's Vickers machine gun would be glued onto the exterior of the finished model at the last minute, but then it dawned on me that holes had to be drilled, sawed and filed in my nice steel cowling for the gun and the spent casing chute. So I had to build the gun now. Fortunately the Scale Aircraft drawings book had nice Vickers drawings. The gun body was built up with three layers of styrene to get the three dimensional receiver detail, the mounting ears and the top cover hinge points. The cooling shroud around the barrel was made from a plastic drinking straw saved from my Coke in a restaurant. I attached styrene strips to create the flutes, then applied liquid putty and sanded to form the desired flute contour. A larger diameter straw made the end rings. Muzzle pieces were fashioned from tubing, wood and styrene. The louvers on the cooling shroud were postponed as I was sick of louvers. I did a final fit check on the Vickers mounting before putting the unfinished gun away and going back to work on the fuselage and wings.

On my first W.W.I biplane model I used 1/16" lengths of thread glued to the wings to represent stitches. I had a better idea for the Salmson, based on a method I had been taught for making simulated rivet heads. If J-B Weld epoxy thinned with denatured alcohol makes nice rivet heads, maybe it would make little raised lines to mimic rib stitches. I tried it on the underside of the elevator where my mistakes would be less obvious. It worked well so I used the same trick on the undersides of the wings. I found that once the templates were all in place, I could mix up the epoxy and "paint" the 254 stitches on one side of one wing in only 50 minutes.

Rib tapes over the stitches were created by masking off the desired width strip, then brushing on paint. When the tape is removed it leaves a raised strip of paint that looks like rib tape. Ditto for the reinforcing tape around the perimeter of the wings and tail pieces.

One of the bigger problems with scratch building in 1:16 scale is that there are no decals. As time wore on I found it very tedious to lay out the paint boundary design based on photos (can't trust drawings), then make Frisket masks, apply the paint in what you believe to be the smartest sequence, paint each color, sand away the paint defects and ridges, then apply overall clear coats to further eliminate the paint defects and ridges.

Another challenge in scratch building is there are no instructions regarding build sequence.You often realize in hindsight that you should have done something else first, especially if it's your first scratchbuild, as this was for me. And you don't necessarily want to do things in the same order that you would on a full size plane. For example on my first attempt on the top surfaces of the bottom wings I applied the fake stitches and rib tapes on the unpainted wings like you would on a full size plane. But the stitches and rib tapes then got buried in paint and damaged by sandpaper as I struggled to paint the camouflage color scheme. The old stitches and rib tapes had to be sanded off and new stitches and tape had to be applied. But there is a 'Catch 22' here too: the gray epoxy stitches have to be touched up to match their surroundings. The perimeter reinforcing tape is easier because it can be simulated by applying clear varnish. It produces a paint ridge like the simulated rib tapes, but shows the color beneath.

The 91st Aero Squadron insignia was crucial to making the model represent Lowry's plane. After experimenting with homemade decals, images printed on adhesive backed clear film, and considering having custom decals made, I finally decided (right or wrong) to go with hand painted markings. I did it on the rudder first, and proved to myself that I could do it. I made stencils from mylar drafting film, traced the patterns onto the fuselage with a ball point pen, then hand painted each portion. Additional templates were used to locate the shield, the knight's leg, the saddle, etc.

The number eleven and the squadron stripes on the wings were also hand painted, using Tamiya masking tape. Elsewhere on the plane you will see 'TSF', hand-painted the same way. TSF stands for "Telegraphie Sans Fil" which is French for wireless telegraph. It indicated the plane was an observation craft, equipped with a wireless for sending observations to headquarters on the ground.

To build the undercarriage, without actual plans, I used my assembly fixture upside down. This seemed to me to be the best way to determine the precise length needed for each of the six legs and end up with the axle the correct distance from the fuselage, the wheels the right distance apart and everything square and true. I first built the soldered brass spreader bar and axle assembly over a flat pattern. The half axles pivot in the center of the spreader assembly and move like the real thing. The legs are made from brass tubing squeezed in a vise into an oval, to approximate the shape and dimensions on the various drawings. The legs fit onto tabs on the fuselage so that the undercarriage could be test fit, then removed to finish the fuselage underside. Piano wire loops and brass brackets were installed on the spreader bars exactly like the real plane as attachment points for the undercarriage rigging. The bungee cords were simulated with elastic thread.

I was anxious to begin final assembly even before the wings were completely finished. The first step was to attach the elevator. The assembly fixture built way back on 9-26-00 was used again to align everything. Some excess putty and paint on the fuselage had to be ground off so that the elevator mounting piece would fit like it had four years earlier. The control cables were cut to length and attached to the horns and operation from the joystick was verified. The rudder was installed next and then all the rigging and brace rods for the elevator and rudder. I again used .015 piano wire and my in-line turnbuckles for rigging. Whenever smaller turnbuckles were needed I cut the telephone wire insulation shorter and sanded it to a smaller diameter.

Next the undercarriage, top center section and bottom wings were all epoxied in place. The rigging was fabricated and installed from the brackets on the undercarriage to the brackets on the underside of the fuselage. The landing wires were fabricated and installed from the brackets on top of the cabane struts to the brackets on the bottom wings.

I had a Japanese scale drawing of the Salmson 2A2 interplane strut, which I scanned and adjusted to exactly 1:16 scale to make eight paper templates. The struts were then constructed the same way as the cabane struts but I used mahogany instead of styrene, to give a more authentic wood appearance.

Additions were made to the assembly fixture to support and align the top wings while they were dry fit to the center section and the struts. Before final installation of the starboard top wing I had to feed the aileron cable through for later connection. When everything was ready and double checked the wings and struts were epoxied in place.

Twin Lewis machine guns and a Scarff Ring were fabricated to go on the observer's cockpit. I made the side plates from used aluminum printing plates. I chose piano wire for the gun barrel and the gas tube because piano wire stays straight and looks like real metal (because it is). A tiny dot of flat black paint on the end of the barrel made it look hollow. The collars on the barrel and gas tube are aluminum tubing.

The ammunition magazines for the Lewis guns were made from a slice of wooden dowel rod. Paper templates were glued on, then styrene strips super glued right onto the paper template. Vertical strips were applied the same way around the sides, then it was trimmed, sanded and painted. The handles are made from used aluminum printing plate material.

On the Scarff Ring, the things that look like sections of gears are for locking the guns at a desired elevation. Mine were made by cementing paper templates to pieces of used aluminum printing plate material. The gear teeth started as #80 holes, then an X-acto knife was used to carve away the outside of the hole. leaving something resembling a gear tooth. Sanding with folded-over #400 sandpaper helped disguise the errors. Installing the slave trigger was the last step in completing the Lewis guns. As I understand it the operator would use his trigger finger to fire the right gun and the other fingers of the same hand to fire the left gun, either gun or both together. His other hand was used to raise and lower the hoop holding the guns. And this was also the last detail to finish the model after four-and-a-half years of work.

For those wanting to see more photos and information, I documented the build as it was underway. It can also be seen at the WWI Models site