DOUGLASXT82D-1

SKYPIRATEA lesson to be learned early when

searching for good scale subjects isthat the airplane which looks ideal ina picture or three view more oftenthan not has defects hampering con-version into a practical model. Whenplans are drawn up for building youfind cowlings expanded to sizes thatblanket the flying prop, fuselages thatturn out big enough to double as adog house, or are so divided by wingposition and cabin location it wouldtake a shoe horn to get the equipmentsqueezed in and still leave room forcockpit detail. Spinners and wheels endup as odd sizes obtainable nowhere.

Then there's that old scale bogietail area. Maybe with our present highlydeveloped proportional it really is pos-sible to fly anything in model form thatflew full size, using scale tail area, nomatter how small. Still, and here myold free-flight phobias crop out, itseems that a fair sized fin and stabare definite aids in having a flyablereproduction, particularly in criticallow speed situations. Add landing gearpositioning problems, too short or toolong nose moments, etc. and you getthe picture. There is no such thing asa perfect scale prototype and very feweven close.

So you can understand my enthu-siasm when I came across a 3-view ofthe Douglas XTB2D-1 alias the Devas-tator II and Skypirate, the answer toan R/C scale modeler's prayer. Despiteyears of digging in old magazines andbooks I had never heard of this ail-plane until information was printed inthe January 1961 (International edi-tion) of the English magazine FlyingReview. Subsequent research seems toconfirm that this was the first time datahad ever appeared in public print. Forexample it is not even mentioned inJane's "All the World's Aircraft" of theperiod, an almost unheard of omission.My conclusion is that this "lost" air-plane must be the rarest completedexperimental by a major manufacturerin the World War II period.

It was an exceptionally large 70 footspan, 3 seat ship board torpedo bomber,powered by a 3,000 h.p. Pratt andWhitney XR-4360-8 engine. Intendedfor the big Midway class carriers, onlytwo prototypes were built and flown

during 1943 -45 and the end of thewar brought cancellation of the project.

Prior to discovering the XTB2D-1 Ihad sifted through the plans, photosand specs of hundreds of airplaneslooking for possible scale subjects andsubmit that this is one of the best ofthe lot. It fits into AMA scale competi-tion in more respects than any otherI have ever investigated.

Check some of these advantages:l. High aspect ratio wing with a thick

section. An ideal setup which maybe flown and maneuvered at highwing loadir:gs without having asnap-roller on your hands.

2. Ample tail area. By not requiringany enlargement it will be possibleto gain maximum points in scalefidelity judging.

3. Wide track tricycle gear. Althoughrule changes have been made mak-ing things easier for airplanes withnarrow, ground-locpy 2 wheel gear,you still want to get the airplaneup and down safely and with a littledignity.

4. Scale operations. Few airplanes offerso many scoring possibilities forpoint gains in the gadget depart-ment: Retracting gear, flaps, carrierarrester hook plus [our stores pylonsto carry bombs, rockets, smokescreen, land mines, drop tanks, etc.

5. Low frontal area. Try any otherradial engine monoplane or evenmost any in-line engine type andyou'll see the contrast.

6. Clean, relatively simple lines with-out fillets. Low wing layout for con-test condition flying.

7. Last, but not least, that intangiblesomething that makes military air-planes a little more attractive thancivil craft.

About the only important debit thatcan be chalked up against the designis that the low frontal area brings withit the problem of too small a cowl tocompletely hide the engine. However,by inverting. the damage to appearanceis negligible. With the head outside inthe slipstream a valuable sub- advan-tage is gained in eliminating the over-heating ever present with enclosedengi nes.

I considered for a 'long time tryin~a retracting gear on the Skypirnt<:-

374the lure of 20 points is strong. Com-mercially available units work line onthe relatively short and light weightL. G. of the standard size multi butwould be more than a little strained ona scale of this size and wing loading. Aspecially built layout might tak e thepunishment, but would be heavy anddifficult to build. It comes down to agamble in either case and I have seenas many points lost as gained playingthis particular slot machine - a bellylanding doesn't get many perfectionpoints!

Another good argument can be de-veloped in favor of a fixed gear. Notethat in the Ar-.fA rules the gear isworth 25 points for scale fidelity and25 points for workmanship. For retrac-tion you are pretty much restricted tostark simplicity or run the chance ofoverloading the mechanisms. \Vith afixed super-detailed gear you can notonly gain more fidelity and workman-ship points under the landing gearcategory but also improve the overallgeneral appearance and raise yourscore in that section also. It seemslikely the retracting gear points can bepretty much retrieved in scale judging

. without any "Look, Ma - no wheels!"incidents.

The model is I rl6" equals one footscale. giving an 84" span. Wing area is864 sq. in. or 6 sq. ft. With a full loadof dope and details aboard. this addsup to a 29 ounce wing loading ataround II pounds. I'm sure it could bebuilt lighter by a builder not so ad-dicted to epoxy and fillercoat.

r-.lost notable contest win of the modelwas first place at the 1965 NationalsRjC scale event. Later it took firstin the ReM Design Contest. Mostrecent placing was 3rd at the Hastings.Nebraska meet last September. I hada lead in scale points and was all set toblitz my favorite adversaries, MaxeyHester and Bud Atkinson. But "engine-iris" helped to goof lip my flights onSaturday and Sunday a Nebraska cy-clone grounded scale flying.

We'll not get into any "glue A toB" directions, for if you haven't builtcnouch models to not need this type ofinstruction you shouldn't be trying acomplex design anyway. Construction15 mostly of Sig Contest Weight balsa

with the exception of wing spars, fuse-lage stringers and longerons, which aremedium grade,

The wing and tail are of usualmulti-type construction. The one-piecelanding gear mounting block (getsomeone in your group with wood-working tools to make it for you) isthe main strength of the center section,so don't try substituting standard sepa-rate multi 1. g. mounts - a snappedwing fluttering down is a great thrillfor spectators but doesn't do much forthe fuse. Plank the bottom of the wingfirst. J nstall Williams Bros. 90 degreebellcranks as control horns for theailerons by epoxying one arm to arib and the bottom planking. Run apair of wires up the spars to powerthe navigation lights, terminating inthe center section servo cavity. At thewing tip blocks leave loose ends insidethe hollowed part, long enough toreach outside.

Provide blind nut mountings to retainthe bomb pylons, since they must beremovable to be able to take off thelanding gear when desired. Incidentally,a simple method of pulling the bombrelease toggle string is to fasten it tothe flaps, directly behind the pylon.Running the flaps into "up" position(above neutral- in the prototype thismovement was used as a dive brake sois not unscalelike ) , jerks the release.Different length strings will permitdroppi ng the stores from pylonseparately.

For accuracy, J like to build theailerons and flaps as an integral partof the wing, rather than separately.After planking and sanding, saw themfree using a blade removed from anX-Acto razor saw. It will be an aid inthis process if during planking youeither mark or slit the planking in theproper spots. Cut the open edges to anangle to allow the desired amount ofmovement. Close the open wing holewith 3/32" sheet and face the flaps andailerons, also with 3/32".

Stab-elevator and fin-rudder are alsomade as units and sawn apart afterplanking. Use temporary bracing at theback of the tail ribs until the surfacesarc partially planked.

The fuselage is ruggedly constructedaround a 1/4" sq. "bridge girder." Thismadness has a method, as you havealready discovered if you have triedto plank a flimsy compound curvedstructure - the strength is absolutelynecessary to keep it aligned duringbuilding. The formers are not notcheduntil after assembly on the framework.Pin the 1/8" x 1/4" stringers in place,one at a time and make the spots fornotches by cutting on each side witha razor blade. Note that the front sec-tion (on the outside of the 1/4" frame)is solid soft block balsa directly behindthe cowling to provide material for

carving the cooling and exhaust exitducting. The tail cone and the areaimmediately in front of the rear cock-pit is also made from block, hollowedfor lightness.

On the prototype aircraft, thishumped section of the upper fuselagewas to feature a power gun turret butavailable photos do not show it actuallyinstalled. It is believed that the ideawas dropped when projected utilizationof the aircraft was changed in mid-stream to something like the missionsoon acquired by the XBT2D. proto-type of the AD Sky raider. The turretopening in the first prototype Sky-pirate was covered over in rather crudeslap-dash fashion by the junior appren-tice of the sheet metal shop, but thesecond prototype (Serial No. 36934,distinguished by a white cowl ring) wasdone in neater, contoured fashion, asshown on the plan.

Most instructions for planking gen-erally advise use of many small strips.I don't care for this method, not onlybecause of its considerable tediousness,but equally because of the multiplicityof seams and extra weight of all theglue required. The S. P. fuselage wasplanked in 6 major pieces of 1/8"sheet - from top centerline to toplongeron, then to the bottom longeronand last hunk to the bottom centerline,on each side. First requirement is theproper grain of wood that will easilybend into a curve, well soaked in waterto be very pliable. By soaking I don'tmean immersing the entire sheet. I wetthe outside repeatedly with a cottonswab, it then assumes a fair amountof curl on its own.

Starting at the front and working tothe rear, fasten down the planking tothe stringers, using plenty of 'pins (itmay take several hundred per piece),binding down with tape or bandagesand in some places on the edgewhere possible, with clamps. As youwork to the rear you will come to anarea where it can no longer be forceddown. At the former nearest to thisbulge. remove a "relief" section - ahigh aspect ratio thi n triangular silverwhose bottom width seldom need bemore than 3/32" and at the top dimi-nishing to a razor blade cut. This mayhave to be repeated further along.Later filling and sanding will hide theseseams. Note that the relief cut do notnecessarily have to be made very farinto the sheet to accomplish the pur-pose in most places.

The rear cockpit glazing is flat sheetcellulose acetate but the front canopyrequires a bit of molding. Don't letthis scare you - it is a skill that everyscaler should acquire. Since shapedlenses must be made for the wing tipnavigation lights, let's start there as agood practice project before tacklingthe more exacting fuselage task. On the

-- ---~~~~-

completed and sanded balsa wing tip.mark the outlines of the lens and sawthis piece out. Mount it on a sturdypeg, wing outline edge up. and fastenit to a solid base. J clamp it in a smallvise. Finish the mold with several coatsof fiberglass resin, sanded and polishedto a gloss. Be advised that every flawwill show on the final product - not soimportant on this small item, but some-thing to keep in mind especially forthe canopy. Don't use sanding sealeror fillercoat as a shortcut to a smoothmold finish - the heat will ruin it.

Cut some strips of celluloid, allow-ing some excess length for handlingand set the mold up right next to theburner of the kitchen stove, preferablyelectric. Wearing gloves, pick up theends of the strip with two pair of longnosed pliers and hold several inchesover the red hot element. When thecelluloid begins to smoke slightly. sagand assume a shimmering rainbow or"goldfish" surface sheen. quickly trans-fer to the mold and pull the strip downover it, stretching it in the process andtucking the pliers under the bottomedge (the mounting peg must be smallerthan the base of the lens mold) so thatthe exact shape is impressed into theplastic. Hold until cool.

If this is your first attempt you willundoubtedly waste some material be-fore you get a perfectly shaped andclear molding. You will soon recognizethe effects of too hot, too cool. toomuch stretch. etc. Some strips that failto work out because of being too coldcan be re-heated and tried again. Thesesmall items take so little plastic thatyou can learn economically the tech-nique necessary on the cockpit. Sincethe navigation light covers must fitflush with the wing tip surface it willbe necessary to trim them down slightlywith a razor blade. Reach through theopening in the wing tip block withtweezers and pull out the wiring.Solder on a G.E. 222 3 volt bulb orequivalent and mount in the opening.Don't cement the plastic cover in placeuntil after final paint job and cover theseam with a small strip of ScotchPlastic Tape. Dye the left light red andthe right green with dial light coloringor put some aniline dye in clear dope.

The cockpit job could perhaps bemade smaller by making the slidingportion from flat sheet and mouldingjust the front windshield but as longas you are about it it's nearly as easyto make it all in one piece. But firstthe pattern: I nset balsa blocks intothe planked fuselage by tack cementingthem in the cockpit opening. Cut andsand them to shape. flush with thefuselage contours. Remove and add a3/16" extension all around the bottom- the finished canopy must be biggerthan the hole it is covering, unlike thelens. Fiberglass and polish.

Cut your celluloid (.020 or .030) ina strip that will allow about 1/2" lee-way all around and make handles ofstrip aluminum to bolt to the longedges and allow an extra inch on eachof these sides. Same idea as the lenses,but the strip is too wide for pliers.With the mold set up on a stand asthough it were in the fuselage at a 45degree nose up angle, repeat the pro-cedure used on the navigation lightcovers. Occasionally you will get whit-ish areas like dope blushes, due toincorrect temperatures and if not toobad, these can be removed with aPlexiglas scratch remover polish. Cock-pit framing is done with strips ofScotch Plastic Tape (not Electricians,it's not oil proof.) Get black, whichholds paint better than colored tapeand dope it the same color as the restof the airplane, after it is applied tothe canopy. Don't stretch it anymorethan necessary when applying or it willshrink back in the sun.

For the dual nose wheels I usedGeoff Franklin's excellent ready madesteerable unit, bulkhead mount type.(Address: 101, Jarrom St., Leicester,England.) The aluminum block intoWhICh the two landing gear legs areplugged and retained with set screws,was drilled out for 5/32" instead of1/8" wire U have been known to landless then lightly), with coil springs atthe top of each leg, out of sight, justinside the cowling. If you haven't gotanyone in your club with a wire bend-

, ing device (thanks, Maxey! ) to make':i these for you, commercially made coils"I can be adapted or the 1/8" legs that

come with the unit used. Since thepaired wires were filled out to scalediameter with a brass tube cover,plugged in place with extra wire andfilled with solder, they are a bit heavy- but you'll bend something else be-fore you will the nose gear. Neverworry about a little extra weight in thefront of a scale model. Better some-'thing useful than a hunk of lead. Frontwheels are 23/4" Graupner Germanmade solid sponge.

The main gear is a 3/16" wire tor-sion bar and to keep weight down thisfar back, polystyrene plastic tubingwas used to flesh it out to scale diam-eter. To paint this tubing, sand offthe gloss and use Hobbypoxy paint.Wheels are English Keilkraft 4" pneu-matic, light but durable. (Wheels areavailable from Polk's or Geoff Frank-lin.) The wheel bushing will have tobe drilled out to 3/16" and this mustbe done carefully on a lathe for there isnot much leeway - but enough,

The cowling was commercially cus-tom-spun from aluminum, a beautifuljob, accomplished with the help of aceOmaha R/Cer, Dr. Bill Clark. If youwant to make your own, I'd adviseturning a wooden form on a lathe,

finishing it with fiberglass, polishedand waxed and making a plaster cast.One good trick is to leave a "peg" ofseveral inches in diameter out of thefront of the wooden pattern --'--therewill be one there anyway, just don'ttake it off. Leave this sticking throughthe bottom of the plaster. When dry, afew whacks on this with a hammerwill drive the form out of the mold.Coat with release agent and make aone piece fiberglass cowl in the usualmanner with 7.5 ounce boat cloth.

A handy product for the scale fanis Len Purdy's Air-O-Sheet, the ma-terial used in the Lanier ready-builts.It is a soft, easily worked and heatmoldable plastic. Many detail items onthe model were made from 30 gaugeAir-O-Sheet; the landing gear coverdoors, wheel hub covers, instrumentpanel and cowl cooling flaps.

For this last item a strip of A.O.S.was wrapped in place around a can thesame diameter as the back of the cowland put in the oven at 350 degrees for10 minutes. Individual cowl flaps werecut from this and epoxied in place. Alittle extra work, but so much morerealistic looking than painted on ver-sions. Air-O-Sheet can be painted withAero Gloss or butyrate dope, but it isbetter to spray rather than brush anduse thin coats. It can be sanded wherenecessary, as is required to match thecowl flaps exactly to the cowl, but hasa smooth surface that needs no fillingor preparation for painting ordinarily.A special liquid cement is availablewhich welds pieces of the materialtogether quickly and in this mannersuch exacting little dinkies as the"scissors" on the sliding part of theLG. oleo strut were quickly made.

The color scheme of the' Skypiratebelongs to a short period which manypublished articles and books on Navycolor schemes ignore, usually statingthat the matte finish graduated darkto light blue top with white bottomwas replaced by overall gloss midnightblue. In fact, from May to Sept. 1944at least some Navy aircraft (Hellcats,Corsairs, other experimentals such asthe XFI5C-l and XBT2C) werepainted overall gloss sea blue. Thesecan be detected in black and whitephotos by the fact that the insigniasurround can be clearly seen on anall gloss blue airplane - later theypainted the whole airplane the colorof the insignia border, and hence therewas no longer any border.

The nearest thing to this color suit-able for the model commercially avail-able is Aero Gloss Curtiss Blue, Usingthis as a base and more referencesthan can be detailed here, I mixed myown according to a scientific formula- a dash of gray, white, black, Cor-sair blue, etc. The idea is to give it aslightly darker, grayish cast. For a

check in stirring up this brew, godown to your friendly neighborhoodhobby dealer and borrow a piece outof the Monogram SBO Dauntless kit.(If he complains, tell him Bill Nor-throp sent you.) The color of theplastic is about what you are shootingfor. Mix enough (say 5 pints or even3 quarts) so you will have some leftover for repairs; to duplicate it againwill take a spectroscope.

A former habit I had of piling ondope, coat after coat, came to ascreeching halt the time I found J hadmanaged to add 11/4 Ibs of finish to awing. The finishing procedure used onthe XTB2D was simple but effective,giving a much more scale-like surfacethan a super high gloss. The plankedbalsa was first brushed with one coatof Sig Superfill to close pores andgrain and sanded down considerably,Next a coat of clear dope, lightlysanded. Then wet silk over everything.doped in place and followed by anothercoat of clear. Then two or three coatsof heavily brushed on Aero GlossFillercoat (it can be sprayed, butwastes a lot), sanded extensively butwith care not to get through to thesilk. Last coat fine sanded and polished,This gives you a uniform surface forspraying on the color dope.

You can use commercial star andbar decals but I consider painted insig-nia more permanent. Patterns are onthe plan and come back here while Itell you how easy it is to paint themwithout going to art school. First, spraythe areas where the insignia will appearwith white dope, feathering the edgesof the patches so they will not showthrough. With the white Fillercoatbackground, two coats will be plenty.Lay a light cardboard pattern of theinsignia in place and trace aroundit lightly with a soft pencil. Usingan ordinary draftsman's rulingpen and compass (a flexible rulerwill also be handy), with Corsair bluedope in them instead of ink, drawthe outlines of the insignia. You mustwork quickly because the dope startsto dry in the pen immediately. I keepa cup of thinner close by, dip the penin it and wipe clean on a rag beforeeach refill. Sometimes the dope will notstart flowing from the pen or compasseasily. Mark on your finger instead ofthe hard surface of the airplane andtransfer over to the work at once. Tokeep the compass point from punchinga hole in the model when drawing thecircular parts. scotch tape on a small.square of celluloid and place the pointon this.

With the blue thus sharply outlinedit is a job no more difficult than akid's coloring book to paint in theareas between the lines with a smallcamel's hair brush. Or for really per-

_ fect results, mask out, using the drawn

line as a guide and spray paint, withan airbrush. This will require cuttinga circular mask for the round parts.

Test flying was placed in the capablehands of Maxey Hester. The modelflew, you should pardon the old cliche,right off the board, requiring only themost minor of adjustments. It is not atricky model to fly, considering it'swing loading. Keep turns fairly sha\1owuntil you get the feel of the ship - thelong wing span makes it a little slowerto come out of a bank. Anyone capableof managing a low wing multi wouldhave no trouble. It makes particularlysmooth and scale like lift-offs from therunway and handles easily on landing.In fact, Maxey says I shouldn't getany points for landings because themodel lands itself. Maybe I should getminus points then for the ones I man-age to louse up a bit.

Since scale fidelity material on theXTB2D-l is so hard to come by, 1'\1undertake to supply copies of the itemsI have co\1ected at the cost of repro-duction and mailing. Send a self ad-dressed envelope to Claude McCullough,Rural Route 5, Ottumwa, Iowa 52501for a list of available data and informa-tion on the spun aluminum cowling andready made canopy. I'd also be gladto answer any questions you may haveon the construction of the model.There's one I won't be able to providean answer for and that's how come Ihad the wing wheel well covers onbackward when the pictures were taken!

Lf

II