BUILDING A WORLD WAR I COMBAT FIGHTER
by Ed Kettler 
ed.kettler@verizon.net
 

Updated: Thursday, February 05, 2004

I was inspired by the articles and the pictures from the World War I demonstration events this summer, and decided to join in the fun. I just started flying RC in November 2001, and entered combat in April 2002.  I really like scale, but there isn't enough of it here in North Texas, so I started with Open B and have a P-47 from JK Aerotech that will see battle for the first time in the North Texas Combat Round Robin in 2003. If you are familiar with the JK Aerotech construction method, this article will have a familiar theme.

I chose the SE5 for a couple of reasons: it was a mainstream fighter, had generous wing and tail surfaces, and was basically a box structure, making it ideal for the carved foam construction method. I spent a couple of evenings scaling up the drawings from Scale Aircraft Drawings, Volume I - World War I, by the Publishers of Model Airplane News, and then fiddling around with details about landing gear design, cabanes, interplane struts and the myriad of other details. This is my first plane design, so there was some backing and filling as I changed my approach.

The airfoil is the Eppler E203, a semi-symetrical airfoil commonly used on the 2610 planes. It is similar to the E205 used on my P-47, and flies nicely, so I decided to go with a known commodity. A flat-bottomed airfoil would make creating the cabane saddles easier, but it is nothing that can't be overcome with a jigsaw. Kelvin McFadden graciously hotwired the wing cores and fuselage section for me, so we will begin the project like we are doing a short kit.

The fuselage was cut from blue Hi-Load 40 foam in both profile and top view, so the taper to the tail was already done, as was the horizontal tail notch. A bread knife was used to quickly trim the upper corners, then a drywall screen was used to contour the curved upper deck section. The next picture shows the fuselage after the bread knife has been used.

The drywall screen was a treat to use. When I built the P-47, I used X-acto knives and 60 grit sandpaper. Lou Melancon recommended drywall screen, and I was amazed at how much faster the shaping stage went! From the raw block to the stage below is 30 minutes:

A trip to the scrollsaw was necessary to cut out the lower wing saddle and the slot for the vertical fin. I cut out an area for the cockpit, and the headrest for the pilot will be made of scrap foam and glued on later. The fins were laid out on paper then taped to 4mm Coroplast. The scrollsaw was used to cut the fins out, and a little touch up with sandpaper will be required to remove the rough edges. I ended up cutting the rudder in line with the horizontal stabilizer so that I could slide the two fins together. The elevator is about half chord, so there wasn't any other way. Bamboo skewers will be used to strengthen the fins. A flute was used to make the hinge lines for the rudder and elevator. So after about an hour's work, here is where I am:

The historical SE5 had a short center section, with outer panels that carried a modest amount of dihedral, and I decided to stick with that approach because it made the wing attachment a bit easier by not having to deal with the wing join at the fuselage intersection. The construction technique is on par with Open B standards, which is probably over kill (and heavy) for the World War I speeds, but we will have to wait for a few mid-airs before we decide either way. The leading edge is 1/4" polypropylene rod and there is an upper and lower 1/8" fiberglass rod spar at the thickest part of each wing. The first step is to rout the channels for the spars and leading edge. Tape the wing sections together first. I use a chalk line to mark the line then use router attachment on my Dremel to cut the channel. When I build later versions, I will rout the channel a bit deeper, say 1/4" to 3/8" to keep the spar down in the wing; it popped up due to the dihedral.

The leading edge was also routed to accept the poly rod stock. This is a bit of a challenge, so what I do is put the wings in the shucks and darken the shuck edges with a felt tip pen, then freehand the router between them. After cleaning up all of the "blue sawdust" you end up with a stack of foam parts ready for assembly.

The wings are epoxied together after a little sanding on the center section and outer panel ends to get a good fit. I used the drywall screen, and only a few passes were needed. To hold the dihedral, I cut some 1x2 blocks to place under the wingtips. Tape the wing sections together, mix up some 15-minute epoxy, and be sure to line the shucks with plastic wrap to prevent gluing the cores to the shucks. Once the glue is applied, check alignment of the leading edges and spar pockets, and place the 1x2 blocks under the shucks at each wingtip. The spars and leading edge are glued in with Probond, and the wing sprayed with 3M77 to help the bi-di tape adhere. T-pins were used to help hold the spars down while the glue cured.

As this was my first scratch built plane, there were several "lessons learnt" that I believe should be passed along. The first one was when to do the trimming of the trailing edges and wing tips. This should have been done before I glued the wings together, as the parts would have been easier to handle. I created a paper template for the wing tip, and traced with a Sharpie onto the wing. The SE5 has constant chord, non-tapered wings, and the upper and lower wing planforms are the same, so one template is used on all panels. After cutting to shape, the tips were tapered with the drywall screen to match the real plane's tips. The ailerons are on the upper wing, and the wing is trimmed back 1" to accommodate them. The lower wing is trimmed to a straight line, creating a 1/16" trailing edge. Any thinner will make it to fragile. The upper wing had a bigger cutout for pilot access, and this is shown marked for cutting.

At this stage of the project, you can now assemble the components and see if it looks like an airplane, so here is the work in progress at this stage.

The first system to get installed is the engine mount. This is a 1/4" ply bulkhead drilled for the engine mount, with blind nuts epoxied into place. A hole is drilled for the fuel and vent lines to pass through the firewall, and a pocket is cut into the fuselage from the left side. To speed things along, I used a 1" drill to quickly rough cut the pocket, then finished it up with an X-acto. You need to leave enough foam to protect the tank from the protruding engine mount bolts in the event of a crash. Before installing the tank, be sure to pressure test it, as a fuel leak will dissolve the foam. I used red fuel tubing for the feed line, and blue for the vent so there wouldn't be any mistakes during assembly.

Note the lines on the fuselage. They should be extended back to the tail, as these are the waterlines for determining proper incidence angles.

The firewall is fuel proofed by covering it with the Zagi tape, then the engine mount is attached. It is best to get all of the drilling done before attaching the mount to the plane. A liberal dose of 3M77 is used to attach the firewall to the foam, then strapping tape is used to crisscross the firewall and forward fuselage to firmly attach the firewall to the plane.

A groove needs to be cut for the throttle flex cable installation, passing through a hole in the firewall. You will need to temporarily attach the engine for this step. I used a Dremel router for this operation. Install this cable and check for routing back to the throttle servo location.

The strength of the forward fuselage is increased by wrapping it in 2mm coroplast, then stapling it to the firewall. The first step is to spray the fuselage with 3M77, then align the coroplast to the fuselage, pressing in place. Once the coro is in place, you can trim the lower wing pocket. Note the markings on the top of the fuselage: you will be cutting foam pockets for the radio gear, and you don't want to cut a hole in your fuel tank! Wrap the fuselage with strapping or bi-di tape to hold the coro in place tightly. This taping step is very important, as this keeps the fuselage from crushing and exploding in a vertical landing.

Grooves should be routed from the cockpit back to the tail surfaces for the rudder and elevator cables. I ran one down each side, and installed the HS-85MG in the cockpit area. An Hs-85 for the throttle and the charge jack/switch assembly were installed forward of the cockpit. The receiver was cut in between the switch plate and the fuel tank (remember I said to mark it?), with a 6" servo extender for the ailerons. The battery was initially installed between the forward landing gear attach pad and the firewall., with the battery cable routed up the back edge of the forward cabane.

These are pieces of 3/16" light ply cut and epoxied together. The wing saddle was a bit of a puzzle, as I needed to cut to accommodate the airfoil curve. After a bit of head scratching, I picked up the shuck for the center section and traced it on the plywood ... Southern engineering at its finest! The end of the wing saddle pieces were notched to allow the rubber bands additional grip in the case of a midair.

Attaching the cabanes was for me the most cumbersome part of the build. The fuselage was leveled and the tailplane, waterline and engine mount were checked. The lower wing was attached, and checked for incidence. It was about a degree off, so back to the jigsaw for some "tuning". Realign everything, recheck, retune, realign, etc. OK, now at 0-0-0, so the upper wing is ready. Big question: how do I hold the thing in place long enough to check the incidences and glue? I've never built a biplane, and several internet surfs and a couple of RC building guides provided some inspiration but not a solution. I was sitting at work during a conference call, fiddling with a paper clip, and there it was: alignment pins! I drilled a 1/16" hole at the base of each cabane, and pushed the paper clip into the coro and foam, pinning the cabane in place. Once the back cabanes were aligned with each other, the front cabanes were aligned to set a 1 degree positive incidence, then the pins were pushed though to mark the location. The cabanes were removed, Goop applied, and then reattached with the paper clip pins to ensure alignment.

In retrospect, I really needed to measure the location of the top wing leading edge, and moved the cabanes forward another 3/4". It looks OK, but the CG issues may be helped by pulling the wing forward.

The rules call for a "rise off ground" capability, so landing gear are required. A 1/4" ply block was installed to attach the 1/8" landing gear wires, and the diagonals were made up of 1/16" wire that loops around the aft rubber band dowel. Plastic straps hold the main legs, and the diagonals are soldered to the main legs. The gear can be folded to make transportation easier.

The tail wheel was made up of 1/16" wire, a piece of brass fuel tube and some wheel collars. The brass tube was epoxied to the vertical stab, and the coro trimmed to allow movement. The tiller arm is poked into the rudder. Simple, works great.

White Zagi tape was used throughout so I wouldn't have to paint the wing and fuselage bottoms. At this time, I installed the upper wing servo and ailerons. The hinges are Zagi tape, and very simple. Use three pieces of strapping tape to hold the aileron 1/8" from the trailing edge. Fold the aileron on top of the wing and tape in place temporarily. Run a strip of Zagi tape so that it covers the aileron, the edge of the aileron and trailing edge, and adheres to the bottom of the wing. You may want to spray the wing with 3M77 to improve adhesion. Untape the aileron, crease the joint, flip the wing over, repeat the process. Use a hex driver to push the upper and lower tapes together, and you have a hinge!

I scuffed upper surfaces with steel wool, then sprayed Olive Drab Mono-Kote paint. A bit of masking to keep overspray off the white, and it couldn't be simpler.

The cowl is a 2-liter Seven-Up bottle shrunk with a heat gun onto a wood mold. Next time, a 3-liter bottle will be used, as the 2-liter was just barely able to be used.

Oh, here's another place where a little planning would have helped. I forgot to deal with the interplane struts until I had finished covering and painting. I used aluminum strips that I bent into an "L" that I slotted through the wings, then connected with pushrods using clevises and EZ-links. This required a recover and repaint, so it should be done before covering next time.

The local hobby shop did not have RFC roundels, so I settled for US markings, as some US squadron flew SE5s. Tail colors were done in Monokote trim strips.

Despite best efforts to get the CG right, it was still tail heavy. A Higley heavy hub and nut helped, but another ounce of lead attached to the firewall was added before the initial testing. The high speed taxi trials indicated it was still too tail heavy, and the thrust wasn't sufficient, so it was back to the shop. The battery was removed from aft of the firewall and moved into the engine compartment, and a heavy Dubro hub replaced the Higley hub. The 9x4 MAS was replaced by a 9x5 APC recommended for the OS .25LA, but I didn't like the perceived lack of thrust, and tried a 10x4 APC instead. This pulled better, so I decided to use that for first flight. I have a handful of props left to try to see what works best.

My friend Alan Shannon did the honors of the first flight, as I didn't want a case of numb thumbs or a rookie mistake to destroy the plane. Acceleration was quick, and the plane was airborne within 25-30'. Weight is around 3 pounds 13 ounces, and the plane can fly quite well at that weight with the OS 25LA. The elevator was too sensitive, and needed to be adjusted, while the ailerons and rudder needed more throw. The angles were fine, according to Alan, and mild aerobatics were performed. The sensitive elevator did cause it to snap, so I let Alan finish the flight while I savored the experience. The plane looked like an SE5 buzzing around the sky, and I can hardly wait for other planes to join up as we go to fight the Flying Circus!

The tail needs to be strengthened, as the foam in the tail section was damaged when the plane flipped over on landing and banged the top of the vertical fin into the ground. I took a look at Lee Liddle's SSC Mustang, which has a 1/8" light ply keel/crutch that serves as an engine mount and runs to the tail post. Another approach would be thickening the tail section, or adding 1/32" light ply reinforcements.

The CG needs to be watched, and I think moving the upper wing forward about 3/4" will help greatly. This will allow the plane to "go on a diet" and lose some of the counterweight, improving its flight performance. The polypropylene leading edges are likely to go as well, and I will just use bi-di reinforcement.

The cabanes will probably be made narrower in width but out of 1/4" ply. They are a bit too wide as is.

Overall, I'm pretty happy with the prototype. It isn't going to win any Scale NATS awards, but it looks like an SE5 and flies like an SE5. I learned a lot about building and design during the process, so it was an enjoyable experience. Total construction time was around 11 hours, which included a lot of think time as I tried to engineer on the fly. Subsequent planes will probably shave an hour or two off the build time, now that I know what to do.

Length: 36"

Wingspan: 43"

Wing Area: 294*2 = 588 square inches = 4.08 square feet

Weight: 3 pounds 13 oz.

Wing Loading: 0.9344 pounds per square foot or 14.95 ounces per square foot