BASICS OF

RIC MODEL

CHOOSING AIRFOILS • WING LOADING • CG LOCATIONBASIC PROPORTIONS • AEROBATIC DESIGN

-and much more!

BY ANDY LENNON

~

....... ----_.... ::::.:;"

From the Dublishers of

'0

Aboullhe Aulhor

L ongtime modeler Andy Lennonhas been involved in aviationsince the age of 15, when he

went for a short ride in a Curtis Robin.He soon joined the Montreal FlyingClub and began flying D. H. GypsyMoths and early two-place Aeroncacabin monoplanes.

He was educated in Canadaat Edward VII School, StrathconaAcademy, Montreal Technical School, McGill University and theUniversity of Western Ontario, London, Ontario.

Andy entered the Canadian aircraft manufacturing industry andlater moved to general manufacturing as an industrial engineer.Throughout his career, he continued to stud y all things aeronauti­cal, particularly aircraft design, aviat ion texts, NACA and NASAreports and aviation periodicals. He has tested many aeronauticstheories by designing, building and flying nearl y 25 experimentalRIC models-miniatures of potential light aircraft. His favoritemodel, Seagull III, is a flying boat with wide aerobatic capabilities.

Andy is a valued contributing editor to Model Airplane News , andhe has written for Model Aviation, Model Builder, RC Modeler and RCModels and Electronics. His two other books are " RIC ModelAirplane Design" and "Canard: A Revolution in Flight."

He continues to fly full-size airplanes and is licensed in bothCanada and the U.S. And when he isn 't at his drawing board orin his workshop, he 's likely to be at the flying field testing yetanother model aircraft design . ...

Copyright ~ 1996 by Air Age Media Inc.ISBN: 0-911295-40-2.

Reprinted in 2002; 2005.

All rights reserved, including the right of reproduct ion in wholeor in part in any form. This book, or parts thereof , may not be

reproduced without the publisher 's written permission.

Published by Air Age Media Inc.100 East Ridge

Ridgefield, CT 06877-4066

AirAGEME 0 I A

modela lrplanenews.com

PRINTED IN THE USA

Z THE BASICS OF RIC MODEL AIRCRAFTDESIGN

Introdudion .•........•.....•.4

Chapter 1Airfoil Selection ...•..•....•5

Chapter 2Understanding Airfoils .•9

Chapter 3UnderstandingAerodynamic Formulas .. 13

Chapter 4Wing LoadingDesign •.•.•...............•... 19

Chapter 5Wing Design . ..••. ..••. ..•..21

Chapter 6CG Location and theBalancing Ad ..•.......•...27

Chapter 7Horizontal Tail Design ••32

Chapter 8Horizontal Tail Incidenceand DownwashEstimating ...•.....•..••••.•37

Chapter 9Vertical Tail Design andSpiral Stability .•....•.•..•42

Chapter 10Roll Control Design •.....47

Chapter 11Weight Distribution inDesign .•....•.•.•..••.•.•...••50

Chapter 12Improve Performance byReducing Drag ..•.••••....52

Chapter 13Stressed-Skin Design andWeight Estimating ••.•.•..58

Chapter 14Design for Flaps .•....•...63

Chapter 15NASA "Safe Wing" 69

Chapter 16Landing-Gear Design . ...72

Chapter 17Ducted-Cowl Design 77

Chapter 18Propeller Selection andEstimating Level FlightSpeeds ....•.......•....•......83

Chapter 19Design for Aerobatics .. 90

Chapter 20High-Lift Devices andDrag Redudion .••...•.....93

Chapter 21Centrifugal Force andManeuverability •••••••••• 98

Chapter 22Canards, Tandem-Wingand Three-SurfaceDesign .••.•.•...•...•....•••.. 102

Contents

Chapter 23Tailless AirplaneDesign 111

Chapter 24Hull and FloatDesign ......•.......•......•..119

Chapter 25Basic Proportions forRIC Aircraft Design ....125

Chapter 26ConstrudionDesigns 129

Appendix••••••••••••••••••••134

THE BASICS OF RIC MODEL AIRCRAFT DESIGN 3

Introduction

~,\

- •••~.~••.

so with electric models, which arerapidly becoming popular. They areclean, noi seless and thoroughlyenjoyable alternatives to gas/glow.However, the design process chal­lenges our ability to build strongbut ligh t models with low zero-liftand induced drag and an optimizedthrust system, be it prop or jet.Short of information on the designof electric powerplant systems, thisbook gives you everything you ot h­erwise need , even the impact of car­rying heavy batteries. Perhaps Andywill tackle elect ric powerplants at afut ure date. ...

- Bob KressRetired Vice President, Grumman

The design process begins withweight estimation and structuraloptimization in the name ofreduced weight. The book coversth ese topics for models better thanany sources I have encounteredpreviously. Next in design comesdrag analysis and reduction, whichare covered professionally yet in anunderstandable way for the ama­teur designer. Wh at a treat to seethe consequences of flat-plate dragfrom seemingly small items likelanding-gear-wire legs properlyilluminated. I recently had thistop ic driven home dramaticallywhen I wen t all out to clean up thedrag of my electric fan A-6 Intruderprototype. The improved perfor­mance after the clean-up surprisedme quite pleasantly. What I didcould have been drawn directlyfrom thi s book.

Stability and control, after per­formance, is what we see as animmediate result of our efforts.Result s vary from joy to th e black­ness of the re-kitting process.Andy's book will keep you awayfrom the latt er end of th e bandthrough proper selection, arrange­ment and sizing of th e aircraft com­pon ents contributing to both longi­tudinal and lateral /d irectional sta­bility and control.

The book is ori ented mainlytoward gas/g low-powe red modelaircraft design. With gas models,available power rarely is a problem.Coping with marg inal thrust sim­ply results in using a bigger engineand a tendency to ignore drag! Not

A ndy Lenno n has writtenan outstanding book tha tcovers all required aspects

of the preliminary design processfor model aircraft . Fur the r, muchof the conten t is equally applicableto military RPV and hom ebu ilt air­craft design . Reviewin g the bookwas som ething of a nostalgia tr ipfor me afte r 46 years of designingfull -scal e and mod el aircra ft.Would that I had been able tocarry thi s book with me to anunsuspectin g aircraft industr ywhen I graduated college in 19S1!

My areas of disagreement hereand there as I read were mostly onexotic top ics and did not amountto mu ch . When review ing mynotes jotted down while readingthe draft, I found that many of mycomments simply amplified what issaid in th e text and reflected eventsfrom my own career related to thebook topic at hand. The chapterson pitch and lateral/d irection al sta­bility and control reminded me ofsome Grum man his tory. Weseemed to blow an aerodynamicfuse on every fifth aircraft proto­type-to wit, th e XFSF Skyrocket,mo st of whic h landed in Lon gIsland Sound, and the XF10F,which, about all axes, was said to belias stabl e as an upside-down pen­dulum." The only thin g thatworked flawlessly was th e variablesweep, which we feared th e most!Maybe Andy's book could havehelped. Sadly, Grumman never gotthe chance to go beyond th e F-14and try an F-1SE

4 THE BASICSOF RIC MODEL AIRCRAFT DESIGN

Chapter 1

-.6 .1 ~.,--.""'-:..Z.,,.

Figure 1.Airfoil data for EpplerE197: tift curves(right-hand illustra­tion) andpolarcurves(left).

Figure 2.Taper-wing conecuo« faclorfor non-elliptic liftdistribution.

.25 .50 .75 1.0

CHORD TI P/CHORD ROOT

. 1

. 2 0r-------- - - ,

a::

'"I-~ . 1...I­Z....:;:~ . 0::::J.....C<

. 0

ever, it isn 't necessary to performlabor ious calculations for eachpotent ial airfoil. Direct comparisonof th e curves and coefficients of thecandidate airfoils is more easilydone, without deterioration of theresult s. Th is com parison calls for anunderstanding of the data . Start byexamining th e right -hand illustra­tion of Figure I- Eppler EI97- indeta il.

Eppler E197 is 13.42 percent of itschord in depth. This plot is th eresult of wind-tunnel test s per­formed at th e University of Stuttgartin Germ any under the direction ofDr. Dieter Althaus.

The horizon tal lin e is th e AoA (n,or alph a) line in degrees (measuredfrom th e airfoil 's chord line)­positive to th e right and negative tothe left .

Airfoil

Selectio n

In develop ing thes eairfoil plots, aerody­namics scientists havescreene d out six ofthese factors, leavingonl y the cha racteris­tics of lift, profile dragand pitching momentunique to each indi­vidua l airfo il. Theseventh, Rn, is refer­enced separately onthe airfo il plot.Formulas th at incor­porate all six variablesand these coefficientspermit accurate calcu­lation of th e lift, totaldrag and pitchingmoments for yourwing and choice ofairfoils.

In the airfoil selec­tion process, how-

10 14 18

• Speed . Lift, drag and pitchingmoment are proportional to thesquare of the speed.

• Wing area. All three are propor­tional to wing area.

• Reynolds number (Rn). Th isreflects bot h speed and chord and isa measure of "scale effect."

• Angle of attack (AoA). In the use­ful range of lift, from zero lift to justbefore the stall, lift, profile drag andpitching moment increase as theAoA increases.

• Wing chord(s). Pitch ing momentand Reynolds number are propor­tional to chord.

• Planform, i.e., straight, tapered orelliptical. All impac t lift, drag andpitching moment.

• Aspect rati o (AR). All three areaffected by aspect ratio.

An (ReynoldsNumber)

--100 ,000

=~zoo , ooo

--Z50 ,000

Prollle dragcoe ffic ien t (Coo)

-.2

-. 4

-.6

1 .6

1.4

1 .2

1 .0

it .8

i .6U=.4II

8 .2E...

O ne of th e most importantchoices in model or full­scale airplane design is the

selection of an airfoil. The wing sec­tion chosen should have charac teris­tics suited to th e flight pattern of thetype of model being designed.

There exis t litera lly hundredsof airfoil sections from which tochoose. They are described in "air­foil plots" similar to EI97 (see Figure1). Selection of an airfoil demands areasonable understanding of thisdata so that one can read, under­stand and use it to advantage.

Providing th is understanding isth e subject of thi s chap ter. Referringto EI97 , note that the data is givenin terms of coefficien ts, except forth e angle of attack. These coeffi­cien ts are CL for lift, CDo for profiledrag and eM for the pitchingmoment around the 1/4-chord point.

The actual lift, total drag andpitching moment of a wing dependon seven factors no t directly relatedto its airfoil section . These are:

THE BASICS OF RIC MODEL AIRC RAFTDESIGN 5

CHAPTER 1 A THE BASICS OF RIC MODEL AIRCRAFT DESIGN

1 1 0WING DRAG COEFFICIENT

I-z 0.8wc:;ii:

0.6....wQ

'-'t: 0.4:::;c:I 0.2z3 o

;? 1.0

5 10 15 20 25

WINGANGLE OF ATTACK-DEGREES

AR = 18 AR = 5 AR = 2.51.4

-::.1.2~

I-Z 1.0wc:;ii: 0.8....wQ

'-' 0.6t::::; 0.4c:Iz3 0.2

0

Figure 3.How aspect ratio affects the stallangle ofattack.

Figure 6.How aspect ratio affects drag atagiven lift.

The vertical line, on the left, pro­vides the CL, positive above andnegative below the horizontal line.

On the right of the vertical arethe pitching moment coefficients,negative (or nose down) above, andpositive (or nose up) below thehorizontal line.

In the center are the three Rnscovered by this plot, coded to iden­tify their respective curves.

.25 .---- - - - - - - - - ---,e.20u1f .15!z~ .10....~ .05i3..

1 2 3 4 5 6 7 8 9 10ASPECT RATIO

FIgure 4.Straight-wing correctionfactor fornon-ellipticlift distribution.

In the left-hand illustration,E197's chord line is straight andjoins leading and trailing edges. Thedotted, curved line is the "mean" or"camber" line, equidistant fromboth upper and lower surfaces.

The vertical line is graduated iden­tically with the CL line on the right .CL is positive above and negativebelow the horizontal line, which isitself graduated to provide the profiledrag coefficient Coo'

Now, back to the curves in theright-hand illustration . The liftlines provide the CL data on theE197 airfoil. Note that this sectionstarts to lift at the negative AoA ofminus 2 degrees and continues tolift to 16 degrees, for a total lift spec-

trum of 18 degrees. CL max is 1.17.These lift curves are section val­

ues for "infinite aspect ratios" andtwo-dimensiona l airflow. Forwings of finite AR and three­dimensional airflow, the slope ofthe lift curve decreases as shownin Figure 3. At these finite ARs, theAoA must be increased to obtainthe same lift coefficient. Theseincreases are called induced AoAs.

For example, Figure 3 shows thatif, with a wing of AR 5, you canachieve a CL of 1.2 with an AoA of20 degrees, then with an AR of 9you can achieve the same CL withan AoA of 17 degrees. A higher ARwing will stall at a lower AoA.

In addition, the AoA must beincreased to compensate for thefact that straight and taperedwings are not as efficient as theideal elliptical wing planform.Figures 2 and 4 provide adjust­ment factors (T, or tau) .

The pitching moment curvesquantify the airfo il's nose-downtendency, increasing with increas­ing AoA, but not linearly like the liftcurves.

The curves in the left-hand illus­tration of Figure 1, called "polarcurves," compare CL to Coo' Notethat E197 shows very little increasein profile drag despi te increasinglift, except at the lowest Rn.

Again, these are section values.The profile drag values do notinclude induced drag, defined as"the drag resulting from the pro­duction of lift" and which varieswith AR as shown in Figure 6.

Wing planform also affectsinduced drag . Asshown in Figures 2and 4, the curves identified by 0, or

STALL1.4

1.2 - - . • •~197e ._. ,' - .~-

Rn 200,000 : ,. it.. "... ./ '. E168. ,

" /. ,: ,. .. ,I

:2 " +6 +10 +12 +18:-.t: ,

: .' ANGLE OF ATTACKE214 ... ,.. .4 (ALPHA). .

: ./ -.6,". ,.', . Ii I .

<1'.. ,.,'

Figure 5.Lift curves of three airfoil types. Note thatE168 lifts equally well inverted.

delta, provide the adjustment fac­tor to adjust induced dragto compensate for the wing's plan­form. The total wing Co is thesum of profile and induced dragcoefficients.

___Camber

(~/7;;' :: '~" ~5~

II Heavily cambered

C?7??7iii Moderately cambered-semisymmetrical

~II Symmetrical-no camber

Figure 7.Broad types of airfoil sections.

6 THE BASICS OF RIC MODEL AIRC RAFTDESIGN

l- I-Z Z.... z.o Rn=

.... .ozu uu:: u::...~9.0 mill i on .......

1.5....

0 .......6.0 " 0 .01u uIi: ~3.0 " Cl::;

~0.5 "<

z g; .010

0.1 " z;::: 0u t; .00....en ....

en

-8 -4 0 4 8 lZ 16 ZO Z4

A. SECTION ANGLEOFATTACK- DEGREES-.5 0 .5 1.0 1.5

8. SECTION LIFT COEFFICIENT

Airfoil Selection ... CHAPTER 1

and higher profile drag.The highest Rn in these plots is

Rn 250,000. For a wing chord of 10inches flying at sea level, this isequivalent to a speed of 32mph­ideal for sailplanes, but low forpowered mod els, except at landingspeeds. A lO-inch chord flying90mph is at Rn 700,000 at sea level.

Figure 8 indicates that both liftand drag improve at higher Rns,improving E197'sgood performance.

Figure 8.Effects of Reynofds number onsectioncharacteristics.

In clarification, AoA is the angle atwhich the wing strikes the air (inflight) measured from the chord line.

Ang le of incidence is a drawingreference and is the angle of thewing's (or horizontal tail's) chordline relati ve to th e aircraft's cen­terline or reference line.

AIRFOIL PLOT COMPARISONSThere are three broad types of airfoil(as in Figure 7): heavily cambered(such as E214), modera tely cambered(such as E197) and no camber, orsymmetrical (such as EI68). Eachtype has its own characteristics (seeFigure 5). Greater camber increasesCL max, i.e., moves the lift curve tothe left so that the angle of zero liftbecomes increasingly negative, andthe pos itive AoA of the stall isreduced. Note that symmetrical air­foils lift equally well upright orinverted.

STALL PATTERNSThere are three major types of airfoilstall pattern, as in Figure 9: sharp, asfor E168; sudden lift reduction; andthe soft, gentle stall as for E197.

E168 has ano ther airfoil quirk(see Append ix). At the stall, liftdrops off but doesn 't return to fullvalue until the AoA is redu ced by afew degrees. This phenomenon is

r~"ySharp Sudden Lin Gentle(E1 681 Loss (E197)

FigureS.Types of airfoil stall.

more pronounced at low Rn. This"hysteresis" is caused by separa ­tion of the airflow on the wing'supper surface at the stall that doesnot re-a ttach until the AoA isreduced. Some airfoils have a mo reemphatic version of th is phenome­non.

PITCHING MOMENTCompare pitching moments of air­foils E197, E168, E214 and E184 inthe appendix. The more heavil ycambered the section is, the greaterth e negative pitch ing moment.

The symmetrica l sectio n in E168has virtually no pitching mom entexcept at the stall, where itbecomes violently negati ve. This isa stable reaction. The airfoil strivesto lower its AoA. E168 would be anexcellent pattern-ship airfoil selec­tion ; CL max is good, an d it's th ickeno ugh for sturdy wing structures.

Airfoil E184 has a reflexed meanline toward its traili ng edge. Thisacts like "up-elevator," reducing thepitching-moment coefficient, butalso reducing CL max. In airfoils,you don't get anything for no thi ng.E184 is designed for tailless mod­els-and note the zero lift AoAshiftto th e right at low Rn.

DRAG ANDREYNOLDS NUMBERThe polar curves of airfoils E197,E168, E214 and E184 show theadverse reactio n, in both CL andCo, to lower Rn and to increasingAoA. Each airfoil has a differentreaction-and this should be a seri­ous consideration for narrow wing­tips and small tail-surface chords,particu larly where, at low Rns,th ere's a reduction in the stall AoA

MISSION PROFILEThe final selection of an airfoil foryour design depends on the designand on how you want the airfoil toperform, i.e., its "m ission profile."

For a sailplane, high lift , lowdrag and pitching moment at lowRns is the ch oice . For an aerobaticmodel, a symmetrical section withlow CM and the capacity to ope r­ate both upright or inverted isdesirab le, along with a sha rp stallfor spins and snap rolls and ashigh a CL max as can be fou nd.For a sport model, an airfo il likeE197 is ideal. It has high CL max,low drag and a moderate pitchingmoment. The stall is gentle. Notethat the so-ca lled "fla t bo ttom"airfo ils like the Clark Y (popularfor sport models) are, in fact, mod­erately cambered airfoils.

FORMULASNow for those "dreaded" formulas.Don 't be alarmed; they're simplearith metic with just a touch of alge­bra. Their solutions are easily com­puted on a hand calcu lator that has"square" and "square root" but ­tons.

These formulas have been modi­fied for simpli city, and to reflectmo del airplane values of speed inmp h, areas in squa re inches, chordsin inc hes, pitching mo ments ininc h/ounces and weight, lift anddrag in ounces .

Form ula 1: Reynolds num ber (Rn)

Rn =speed (mph) x chord (in.) x K

(Kat sea level is 780; at 5,000 feet is690; and at 10,000 feet is 610)

Form ula 2: Aspect ratio (AR)

AR = span (in.)2wing area (sq. in .)

THE BASICS OF RIC MODEL AIRCRAFT DESIGN 7

CHAPTER 1 .. THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Figure 10.Method for locating themean aerodynamic chord (MAC).

B: Angle of attack (or incidence)for level flight. CL required dividedby CL per degree of angle ofattack.Knowing wing area, weight andcru ising speed, calculate th e CLneeded as in Formu la 7. Divide th isCL by CL per degree as above toobt ain lift spectrum. Deduct anynegative AoA to zero lift.

where in formulas 6, 7,8,9 and 10:CL = lift coefficient (formula 7);CD = total drag coefficien t (for-

mula 5);V2 = speed in mph squared;S =wing area in square inches;C = mean aerodynamic chord in

inche s (see Figure 10);CM =pitching moment about th e

% MAC at the calculated CL ininch/ounces;

o (sigma) = density of air (sealevel, 1.00; 5,000 feet, 0.8616;10,000 feet, 0.7384).

SPECIAL PROCEDURESA: Lift coefficient per degreeof angle of attack adjusted foraspect ratio and planform.Refer to Figure 1, Part 1 E197. At CL1.00 and AoA of 7 degrees, plus th e2 degrees negative, ao is 9 degrees.Apply Formula 4 to obtain a. DivideCL 1.00 by a to obtain CL perdegree.

REFERENCES

Airfoil Design and Data. by Dr. Richard Eppler,and Profilaren fur den Modellf lug, by Dr. DieterAlthaus, available from Springer-Verlag, NewYorle Inc., P.O. Box 19386, Newarle, NJ07195-9386.

Airfoils at Low Speeds (Soartech #8), byMichael Selig, John Donovan and DavidFrasier. available from HA Stoleely, 1504North Horseshoe Cir., Virginia Beach,VA 23451.

Model Aircraft Dynamics, by Martin Simon,Zenith Booles, P.O. Box I/MN121, Osceola,WI 54020.

C: Stall angle of attack adjusted foraspect ratio and plan.Ad just th e stall AoA for ARand plan­form as in Formula 4. Deduct anynegative AoA to zero lift to obtainpositive value of stall AoA...

SweepAngle

1;4 MAC

Total drag =CD x a x V2 x S3519

Formula 10: Pitching moment

Pitching moment = CM xax V2 x S x C3519

Formula 6: Lift (or weight)

Lift (or weight)« CL x a x V2 x S3519

Formula 8: Model speed

Formula 9: Total profile andinduced wing drag

v = Ii x 3519a x CL x S

Formula 7: Lift coefficient required

CL = lift x 3519axV2xS

"squared";AR = aspect rat io;I) (delta) = planform adjustment

factor (Figures 2 and 4);

If you wan t to know the model'sspeed at a given CL and weigh t:

If you want to determine th e liftcoefficien t needed for a given airspeed and weight:

COEFFICIENT CONVERSIONSUp to th is point, coefficients havehad only abstract values. To convertth ese to meaningful figures, we' lluse the six variables ment ion ed pre­viously in th ese formulas.

1- C-1- C-I-I C/2-

STRAIGHT

CIL --j-";'''''::<':'---i

I-IC

where a = total of section AoA andinduced AoA;

a o =section AoA from airfoil plot;CL= lift coefficient at section

AoA from airfoil plot ;AR = aspec t ratio;T (tau) = planform ad justment

factors (Figures 2 and 4).

Form ula 5: Total of profile (sec­tion) and induced drag coefficien ts

a (alpha)« a o + (18.24 x CJ x (1 + 1)AR

Formula 3: Taper ratio (A.-Iambda)

Formula 4: Total of section andinduced angle of attack (AoA)

Taper ratio = tip chord(in.)rootchord (in.)

where CD = tot al of profile andinduced drag coefficients;

CDo = sectio n profile drag coeffi­cient at CL chosen from airfoil plot ;

CL2 = lift coefficient chosen

CD = CDo + (0.318 x CL2) x (1 + 0)AR

(A straight wing has a taper ratioof 1.)

Mean aerodynamic chord (MAC)Figure 10 provides a graphicmethod for locating the MAC andits lA-chord point. The MAC isdefined as "tha t cho rd representa­tive of th e wing as a who le andabout which the lift , drag andpitching moment forces can beconsidered to act."

8 THE BASICS OF MODEL AIRCRAFT DESIGN

Chapter ·2

Figure 1.Characteristics of NACA 2412 at various Reynolds numbers.

Figure2.Characteristicsof NACA 0012at various Reynolds numbers.

IJ ':'- -

o3ll~r-f-fleLFR1-JC'~S)hJdJ~'9 'fl~nalds NumEer.=t.J' y _c.. ..._

_<>---J. /80.ooo 0.6 J 00 - -- - e.380.000 - V- .8 3 -- .00 1.-- - - -- -- - - - I.J 40.000 -- - /.0 ' ' of ' 0 -- , - .---- 660.000 ! -/.I-· J 001-

7 . - .. - --- 3 3ll000o--·-~·l1aooo

j - - -- . - - - - - -

- - .. ". - - -- -- - -5 - - - - - - - - - -

- - . - -- -'I - - -- -- - ..

-3 - - .. -- - -- - - - -

.. - .. - - - - -"Z . .- -

- -- -I

" . - - - - -

- - - - - -/ - - . - -- - -

-- - - -- - - - - ,x- -- - - - - - - -- -

- - -- - -

A;rfo'~; N.A.C.A OO/e-Dale : J-jJ Tes! :V-OT.IZ j 7-8{k!~ul'~ co r r ec ted 10 in f in/l e Q~Ptlcl r o lio

o

Airfoils

Understanding

j -.

.0

. 1

.0

<Jr:~ .

8.06u

g'.o-Ii".0~° .0

.0

E~~_C '!; ---­

e! 0t'Q -IO

O- - _. -

20 40 60- 80 I~erce,.,1 ofch.:va Z.O

- - - - - - - - - - - .e- - ~ - - - - -- - - -- -

H--H-l}l - - - - - - - - - ---

Airlod, N.A.C.A. OOIZ ~ 0 ~ - .z~tJf\::tS;ze : $ ".(30 " VeL (ft./~t!cJ--68 - Z 3

_ Pre$.(~/hd. alm~: lio ZO . ~ - .

'-'-~ -'-L""",r;hs:,.~v,~;~::7/:"Af':I' :J 'JJ -. '1 ~ - .'1

-8 ·'1 0 4 8 Ie . 16 .1'0 24 ee J< 'Ii -. '1 .Z 0 .Z .'1 .6 .8 10 IZ 1.4 L6 L8An9/e of attock fo r inf/I1IJe aSptlc l r a ti o. a . (deqre" J L i tt coeffic ien t.C..

:i~~I~ H~~ . /J 8fr=11f,Hitt~ . - --1 '1." -n J ~ ~ -,o~U .09.1=L-J,I.1' O. ooo -·t-- -~~ in =i1f 0 0 co ~ 60 80 100 ~.08 ~ _, e~~ :-:-_-= 21.· ~~~. oooooo =_ ___:: _" 661 - 4 10 P~rc.'" o (c /' o ..d IZ.O _ ,.. ~ ,JJ,

~g ~~ -::Jl -· - -- ---- - -· '· ' 8 .§ 1 ~ · ~=:·=:=~5J6J..ggg " '1_=- := - -\:::)0 r 38 - 4 fl - - _.- - .. \! 0 '"

19:n : Jf~ - - - ~ . - ~==:-:-~~~:~gg - - J - - \~ 1~ :~_:~ - - -- - - .. - - - / 6 8,06 - v-_··- - ··41.5 00 I - .= I

lJO J 15 - 1.$11 - - - - I J.90 108 - .&Z - - 1 4 1-

1g-.os - - - - - - - - -- - - - - 1

~ ~!~ (~ 1$ - - ' r~ ~ - - ~;~ .-=- .= _- - _- '.f :: ., _L.C. Rod : I ,SlJ _ ~ i -- . ~ - _ 12 r..:: ~ .04 _ _ _

~;:::-:~ _. - ~ 1. 0 . ~_ ..P.OJ 1"l. _ .. - - ~~ -c"oc", l/lD _ ~~ _ ~~. _. - _ . 8~ . 02 ~ :~ .. -- 1-l~,~,+~.l}__-+-:t;'lJf~UV.. ;,, ~ -

__ _=== ~ ~ :pos~~o~ ~.: :: __ . 6 ~ . 0 / ~~W~~~~~~~~~~r~~_ _ _ _ <>--05" .J.. ~ O.04J _ _ .4 '':;

- ~ -- ~~:-::u - -1- : :~~-- - II -- - - •. - . 9 ' -2 " - - .045 - - .2

- .. - III .- - vi-T " I S · O· - - 05 of- - 0

_ ~ __ A,rfo il.'NA .C.A . c4' c_ Sir e : 5 "u O· V,l(fI./$ ~cJ' 6S .2

l J P,-e $ .($ /nd. o lm ) : 1/-4 lo CO" {! ~ r. sl , y.O T Il 64 001. , 8 "34 _ 4_ _ .!!!..'!..-~'~_~:1 . MA. L._~ ·

- 8 - . 0 • 8 ', 2 16 £0 N <B 32Ang le of olloclf. fo r i,,'i"nil~ 03~C' r o No, ct, (d~9r~ I!.s)

In 1937, NACA issued Report No.586, whic h shows the shockingadverse impact of scale on airfoilcharacteristics (based on tests in avariable-density wind tunnel over awide range of Rns, as shown in

Rn = Chord (in inches) x speed (inmph) x 780 (at sea level) .

A full-scale airplane flying at200mph with a wing chord of 5 feet(60 inches) is operating at Rn9,360,000. A scale model flying at60mph with a wing chord of 10inche s flies at Rn 468 ,000. Whenlanding at 25mph, the model 's Rnis reduced to 195,000.

REYNOLDS NUMBERSA most important consideration inairfoil selection is "scale effect."The measure of scale effect is theRn. Its formula is:

T he selection of an airfoilsection for most poweredmodels is considered not to

be criti cal by many modelers andkit designers. Models fly reasonablywell with an y old airfoil , and theirhigh drag is beneficial in steepen­ing the glide for easier landings.Some years ago , there was a rumorthat a well-known and respectedEastern model designer developedhis airfoils with the aid of the salesof hi s size 12 Florsheim shoes.

In contrast, the RIC soaring fra­ternity is very conscious of theneed for efficien t airfoils . Theirmodels have on ly one powersource: gravity. The better the air­foil , the flatter th e glide and thelonger the glider may stay aloft.

This chapter is in tended to pro­vide readers with a practical, easyunderstanding of airfoil characteris­tics so that their selection will suitthe type of performance they ho peto achieve from their designs. Itdoes not go into detail on such sub­jects as laminar or turbulent flows,turbulators, separation an d separa­tion bubbles, etc . (These are fullydescr ibed in Martin Simon's "ModelAircraft Aerodynamics" and Selig­Donovan and Frasier's "Airfoils atLow Speeds"-see the source list atthe end of this chapter.)

THE BASICS OF RIC MODEL AIRCRAFTDESIGN 9

CHAPTER 2 .... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Figure 4.Aerodynamic characteristics of the NACA 641-412 airfoil section, 24-lnch chord.

reduced from 16 degrees to 11degrees. Both lift and stall anglesare higher than for NACA0012 .

Profile drag increases almostthreefold at th e lowest Rn. Owingto th is airfoil's cambered meanlin e, the pitching moment isminus 0.06 .

For NACA 6412 in Figure 3, th e CLmax goes from 1.7 to 1.35 (79 per­cent). The stall angle is reduced from16 degrees to 12 degrees. Profile dragdoubles at th e lowest Rn.

It should be noted, however,th at camber increase obviouslyimproves CL max and stall angle forthis relatively thin (12 percent) sec­tion at low Rns.

The pitching moment, due to itshigher camber, is 0.135 negative. Ahorizontal tail would need to pro­duc e a hea vy download to offsetthis pitching moment, resulting inan inc reased "trim drag ."

In 1945, NACAissued Repor t No.824, "Summary of Airfoil Data";it includes data on their "six­number" laminar-flow airfoils.NACA 64}"412 is typical (see Figure4). The lowest Rn is 3,000,000.

Th ese airfoils were developedsimilarly to those in NACA ReportNo. 460: a sym me trical sectionwrapp ed around a cambered meanline. However, careful study of pres­sure distribution allowed this typeof airfoil to obtain a very low

},:. - i.J]-, '\' "I' (,' It 1- ~ ~1~ f:: ~(l~ 1, -0"'1'111'" I.. ,", I I I '

['.'. ' ,._ .............. . j'."" _ J_' ,_....,....~t>tI . • 1 I

.'OJrlr: - ~-i I II If I 'Ir I .I I I .

H1LfJTUTITfl III iillU

,

,.

" t.-

, ,~,~

Ii[.

!I.. rt

"

lilt.,

,.: I

, r.t,,."

., , . "

moment, except beyond the stallwhere it's negative (nose down) andstabilizing.

NACA 2412 in Figure 1 is a pop ­ular sport-model airfo il. Comparedwith NACA 0012 , th e ma ximumlift coefficient is slightly h igher at1.6 at the highest Rn. At the lowestRn, with the tu rbulence factoraccounted for (41,500 x 2.64 ,whi ch equals 109,560), th e CL maxdrops to 0.95, or 59 percent of thatof th e highest Rn. The sta ll angle is

- .. - -t -.3 - _ - Ai r fod: N A CA. 641a• _ . Date: 8 -34 Tes/ : V.O.T. ff65~ -. J _~$ COrrec '~d ' 0 " 'inile o SPKf ratio

-. 4 -.2 0 .2 .4 .6 .8 /0 l 2 1.4 /.6 18L i f! coe" jc ienI, C.

. • 0H-H-+t+-J-f-H-H-+t+-J-f-H-t-H~ -.I"~e" .2 - - .- - .. - _.

· 8 - 4 0 4 8 /2 16 eo 24 28 .32An9/~ o f o lJoch (o r infin"'~ o~pee.l r ono. ct, (d~qrt!t!$)

Figure 3.Characteristicsof NACA 6412at variousReynolds numbers.

Figures I , 2 and 3). Note that theRns shown are "test" results andrequire correction for a "turbulencefactor " that wasn 't recogni zed dur­ing th e tests. This factor is 2.64.Each Rn in Figures 1, 2 and 3should be incr eased by th is factor.

The airfoils involved in these fig­ures are "rela ted sections." NACA0012 is symmetrical ; NACA 2412was develop ed by "wrapping" th esymmetrical section around a cam­bered mean line so th at th e upperand lower surfaces were equidistantfrom th e camber line. For NACA2412 , this mean line ha s a camberheight of 2 percent of th e cho rdlength, with its highest point at 40percent.

NACA 0012 in Figure 2 shows ashocking reduction in maximumlift coefficient from 1.55 for th ehighest Rn to 0.83 for the lowest­a difference of 54 percent of thehigher value.

Similarl y, th e stall AoA is sharplyreduced from 17 degrees for thehighest Rn to 10 degrees for th elowest. One very interesting ph e­nomenon is this airfoil's beha viorbeyond the stall at th e lower Rns. Itcontinues to lift up to 28 degrees atalmost full value .

Profile drag at low Rns is almostdouble th at at high Rns and increasesvery significantly at th e stall andbeyond-not surpri sing, consider­ing the post-stall lift beha vior.

NACA 0012 has a zero pitching

10 THE BASICS OF RIC MODELAIRCRAFT DESIGN

Understand ing Airfo ils .. CHAPTER 2

Chord line

c===-=---=-E197

C =====-E168

C ===-E226

c:= ==---=-E374

c= ====--E214

C -------E230

C ==-=-E211

Figure 6.Eppler airfoils.

THICKNESSThicker wings permit strong butlight construction. They may alsoexact a small penalty in dragincrease. Tapered wings with thickroot airfoils that taper to thinner,but related , tip airfoils, are strong,light and efficient. Laying out theintervening airfoils between rootand tip calls for much calculation­or computer assistance.

For high speed , an airfoil suchas E226 shown in Figure 6 is sug­gested. Drag and pitching momentsare low, as is the CL max, and theairfoil performs almost as wellinverted as it does upright. E374would also be a good high-speedairfoil section.

The author has had success with theE197 for sport models. It has low pro­file drag, good lift and a gentle stall,but a fairlyhigh pitching moment.

The E168is suitable for strong hor­izontal or vertical tail surfaces, or forwings of aerobatic models. It performsas well upright as it does inverted.

trailing edge. This produces a posi­tive (nose-up) pitching moment.This airfoil would be suitable for atailless or delta-wing model.Inevitably, CL max is adverselyaffected.

MEAN LINE CAMBERA symmetrical airfoil has the lowestCL max and stall angle. An airfoilwith increased camber produces ahigher maximum CL, but it starts tolift at higher negative angles ofattack with a broader range of liftbefore stalling. Increased camber,however, produces increased pitch­ing moments.

Out of curiosity, the cambermean line for the E197 airfoil wasstraightened out and the envelopewas redrawn as in Figure 5. Theresult was a symmetrical airfoilresembling the E168.

Some cambered airfoils have alower surface trailing-edge "CUSp"created by a localized and increasedcurva ture in the camber mean line,as in the E214, Figure 6. The cuspincreases both CL max and pitchingmoment; it 's called "aft loading. II

E197 in Figure 6 has a slight cusp;airfoils E207 and E209 are similar toE197, but they lack the trailing­edge cusp (reference 12). AirfoilE230 in Figure 6 has an upwardlyreflexed camber mean line near its

100mph is operating at Rn780,000.

The selection of an airfoil for adesign should start with a review ofairfoil plots of the type in this chap­ter. In this author's experience, theplots of the University of Stuttgartpublished by Dieter Althaus are theclearest and most comprehensive.The airfoils developed by Dr.Richard Eppler are favored.

Straight mean line(

E __~

Figure 5.The cambered mean lineofE197 (top) was straightened outandthe envelope redrawn, resultingin a symmetrical airfoil (boNom).

Cambered meanline

s~

profile drag (over a limited range oflower lift coefficients). The P-51Mustang WW II fighter employedairfoils of this type. The "low dragbucket" at CL 0.4 shown in Figure 4shows this drag reduction.

In 1949, NACA issued TechnicalNote 1945. This compared 15NACA airfoil sections at Rns from9,000,000 (9 x 106) to 700,000(0.7 X 106)

The CL max of NACA 641-412 atRn 9 x 106 is 1.67, but it drops to1.18 (70 percent of the highest Rn)at Rn 0.7 x 106. Profile dragincreases from 0.0045 to 0.0072 forthe same Rn range, and the stallang le is 16 degrees, but it drops to12 degrees at the low Rn. Pitching­moment coefficient is 0.063.

This report concluded that at lowRns, the laminar-flow section didnot offer substantial advantagesover those in Report No. 460 andReport No. 610. NASA (NACA's suc­cessor) continued to do researchinto laminar-flow airfoils withmuch success; but at the high Rnsof full-scale airfoils and aided bycomputer analysis.

The worldwide RIC soaring fra­ternity, however, has done muchwind-tunnel testing and computerdesign of airfoils for model gliders(references 10 to 15 inclusive).Though the Rn range of these testsseldom exceeds Rn 300,000, anyairfoil that offers good perfor­mance at this low Rn can onlyimprove at the higher Rns of pow­ered flight . A lO-inch-chord at

THE BASICS OF RIC MOD ELAIRCRAFT DESIGN 11

CHAPTER 2 .... THE BASICS OF RIC MODEL AIRC RAFT DESIGN

PITCHING MOMENTThe airfoil's pitching moment isimportan t both struc tur ally andaerodynamically. In flight-partic­ularly in maneuvers-the pitch ingmoment tries to twist th e wing in aleadi ng-edge-down di rection. Thisadds to th e torsional stress placedon the wing struc ture by theailerons and extended flaps. High­pitching-momen t airfoils requirewings that are stiff in torsion, andth at favors thicker sections and fullwing skin s, particularly for high-ARwings.

Aerodynamically, th e nose-downpitching moment requires a hori­zontal tail do wnl oad for equilib­rium. This adds to th e lift th e wingmu st produce and increases totald rag- called "t rim d rag." Thepit ching moment is litt le affectedby var iations in th e Rn.

STALL BEHAVIOROne reason for preferring wind­tu nnel test data ove r computer-

AIRFOilCONSTRUCnONMost powered model aircraft operate inan Rn rangefrom 200,000 to well over1,000,000. This is above thecriticalrange of Rnsatwhich turbulatorsareconsidered to be effective.

For themorerecently developed air­foils , there is a considerable degree oflaminar flowthat significantly reducestheir profile drag. This flowis easilyupset byprotuberances on the wing'ssurfaces.

For smooth surfaces, full wing sheet­ingis suggested, with afilm overlay­either over a foam-core or built-upcon­struction- that will promote themostlaminar flow and also result in awing­stiff in torsion (seeChapter 13,"StressedSkin Design").

There are large models whose wingshave multiplespars onboth top and bot­tomsurfaces and are covered only inplastic film.

Because it shrinks onapplication, thefilm tendsto flatten between each riband each spar. As a result. multipleridges run both chordwise and span­wise. rendering laminar flow impossible.

Contrast this with thevery smoothsurfaces of high-performance RICsoaring gliders.

12 THE BASICS OFRIC MODEL AIRCRAFT DESIGN

developed performa nce curves isth at th e former provides an accu­rate "picture" of th e airfo il's behav­ior at th e stall and beyond.

In general, th ere are th ree bro adtypes of sta ll (as shown in Figure 9of Cha pter 1, "Airfoil Select ion "):sharp ; sudde n lift drop; and gentle.

For sport models, a gentle stall isdesirable. Sha rp sta lls and th osewith a sudden lift drop are appro­priate for man euvers in whic h theabili ty to stall a wing easily isrequired, such as spins.

ZERO LIFT ANGLEThe ang le of zero lift for a sym­metrical-section airfo il is zerodegrees AoA. Cambe red airfoil sec­t io ns such as E21 4 shown inFigure 6 sta rt to lift at almost 6degrees negati ve AoA, but for thisairfo il, that ang le is unaffect ed byvariations in th e Rn.

Con trast th is wit h airfoil E211.This airfoil 's angle of zero liftmoves closer to zero degrees at thelower Rns.

The forward wing of a canardmu st stall before th e aft wing; but,for longitudinal stability, the aftwing mu st reach its airfoil's zero-liftang le before th e front wing 's airfoil.If th e forepl an e's airfoil reach eszero lift first, a violent dive resultsan d, because th e aft wing is stilllifting, a crash is almost inevitable.

The low-Rn behavior of the E211mea ns th at , at low speeds-or nar­row cho rds- th is airfoil may reachzero lift more readily. Its use as aforwar d-wing airfoil on a cana rd isto be avoided. Airfoil E214 is moresuitable.

MAXIMUM LIFT COEFFICIENTFrom zero lift , h igh er camberresults in a higher CL max andhigher sta lli ng angles. Thisimpacts the mo del's takeoff, sta lland landing spee ds. A high CLma x permits slower flight in allthree points; a lower CL maxreve rses these conditions .

SUMMARYIn aerodynamics, nothing is free. Ingene ral, high lift mean s increaseddrag and pit ching moments; forhigh speeds, CL max is redu ced andso on. The type of performancesough t for a design dictates whic hairfoil characte ristics are signifi-

cant. Having selected these, an yadverse characteristics mu st beaccep ted and compensa ted for.....

NACA AND NASA DATA

1. Report 460* : The characteristics of 78Related Airfoil Sections from Tests in theVariable Density Wind Tunnel; 1933; Jacobs,Ward and Pinkerton.

2. Report 586* : Airfo il Section Characteristicsas Affected by Variations of the ReynoldsNumber; 1937; Jacobs and Sherman.

3. Report 610* : Tests of Related ForwardCamber Airfoils in the Variable-Density WindTunnel; 1937; Jacobs, Pinkerton andGreenberg.

4. Report 628* : Aerodynamic Characteristicsof a Large Number of Airfo ils Tested in theVariable-Density Wind Tunnel; 1938;Pinkerton and Greenberg.

5. Report 824* : Summary of Airfoil Data;1945; Abbott, von Doenhoff and Stivers.

6. Technical Note 1945* : AerodynamicCharacteristicsof 15 NACA Airfoil Sections atSeven Reynolds Numbers from 0.7 x 106 to9.0 x 106; 1949; Loftin and Smith .

7. Technical Note NASA TN 7428*: Low­Speed Aerodynamic Characteristics of a 17percent Thick Airfoil Designed for GeneralAviation Applications; 1973; McGhee, et. al.

8. NASA Technical Memorandum TMX72697* : Low SpeedAerodynamic characteris­t ics of a 13-percent Thick Airfoil Section;1977; McGhee, et. al.

9. NASA Technical Paper 1865* : Design andExperimental Results for a Flapped NaturalLaminar-Flow Airfoil for General AviationApplications; 1981; Somers.

10. Profilpolaren fOr den 1900 Ellflug, Book 1;1980; Dieter Althus, Neckar-Verlag,Klosterring # I, 7730 Villingen-Schwenningen,Germany.

11. Profilpolaren fur den 1900 Ellflug, Book 2;1986; Dieter Alth us, Neckar-Verlag,Klosterring # I , 7730 Villingen-Schwenningen,Germany.

12. Eppler Profile MTB 12; 1986; MartinHepperle. Verlag fUr Technik und HandwerkGMBH. Postfach 1128, 7570 Baden-Baden,Germany.

13. Model Aircraft Aerodynamics, SecondEdition; 1987; Martin Simons.

14. Airfoils at Low Speeds-Soartech 8; 1989;Selig-Donovan and Fraser, Zenith AviationBooks. P.O. Box I , Osceola, WI 54020.

15. Airfoil Design and Data; 1980; Dr. RichardEppler, Springer Verlag, New York, NY.

'Available from U.s. Departmen t ofCommerce, National Technical Informa­t ion Service, 5285 Port Royal Rd.,Springfield. VA 22161.

Chapter ·3

T his book reflects a deep andlifelong interest in aviation;a close study of the vast

amo unt of timeless aerodynamicresearch data, both full-scale an dmodel, tha t is readily available.

This, cou pled with th e practica lapplication of this data to thedesign, construction and flying of awide variety of model airp lanes ,reflects those many years of studyand experience.

(These models perform well , andphotos and 3-view drawings ofthem are incorporated in to thisbook and are compiled in Chapter26, "Construction Designs .")

Layma n's language is used, butinevitably some aerodynamic jar­gon and symbols have to be intro­duced. The many charts, curves andformulas may be intimidating toth ose readers who are not familiarwith the use of the mass of infor­mation they contain. Once actualnumbers replace symbols in the for­mulas, only pla in , old , public-

school arithmeti c is needed . Apocket calculator with "square" and"square-root" buttons simplifiesthe work.

The problem seems to be "ho wand from where to obtain the num­bers." This chapter is designed toan swer this. The various figures aremarked to illustrate the sources ofthose numbers, and th e specifica­tions of an imagin ary model air­plane are used as samples.

The most important formul as dealwith lift, drag and pitching moment.

LIFTThe airfoil plot of Eppler E197 (seeFigure 1) shows thi s airfoil's behavi orfor "infinite AR," i.e., no wingtips.

Airplane wings, even very high­AR glider wings , have "finite" ARsand do have wingtips. Lift is lost atthose tips; th e wider th e tip cho rd,th e greater th e loss.

The wing's AoA must beincreased (induced AoA) to obtainthe CL needed as AR decreases.

Understanding

Aerodynamic

Formulas

Induced drag incre ases at low ARs.Airfoil plots mu st be adjusted to :

• reflect th e AR of your wings; and

• reflect the wing 's planform­straigh t (con stan t chord) ortapered .

An elliptica l wing planform needsonly th e ad justmen t for AR.

The form ula for both AR andplanform ad justments is:

a = ao + 18.24 x CL x (1. + T)AR

a = 9° + 18.24 x 1.00 x 1.17= 12.5°6

wh ere a = tot al AoA (AoA) needed;aD ="sectio n" or airfoil plot AoA;CL =CL at that AoA;AR=aspect ratio;T =Plan form ad justment factor.

Had th e win g been tap ered wit h atap er ratio of 0.6 (tip chord 7.5inches divided by root chord 12.5in ch es, or 0.6), the planform

Refer to Figure 2. E197 produces liftof CL 1.00 at 9 degrees AoA, fromzero lift , for infin ite AR.

A con stant-chord wing of AR 6has an adjustment factor T of 0.17(see Figure 4 of Chapter 1).

Replace th e symbols with thesenumbers:

Cl cL CM

1.6RE 1.6 -.4

C ==--==- 100000 Lift+ 200000 1.4 -.35 CL YS. AoA

1.4 X 250000 /CL max1.17

1.2

.8Profile drag YS. 11ft .8

COYS. CL

.6

.4~

.2 Pitching moment

0CM YS AoA

-.2 .12 .14 6 10 14 18

-.4 Co Angle 01attack-.4 .1 (AoA)

-.6

E197(13.42"/0) -.6 .15

Figure 1.Eppler E197 airfoil plot.

THE BASICS OF RIC MODEL AIRCRAFT DESIGN 13

CHAPTER 3 .... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Figure 2.Eppler airfoil E197 produces lilt of CL 1.00at 9 degrees AoA, from zero lilt , for infiniteAR.

of both wing and tail airfoils set atzero degrees rela tive to thei r fuse­lage centerlines . A symmet ricalairfoil at zero degrees AoA will pro ­duce no lift .

Wh at happen s is that, to takeoff, the pilot commands up -eleva­tor, thus adj usting the wing to aposit ive AoA, and it lifts. The liftproduces down wash th at strikesthe hori zontal tail at a negative (ordownward) angle causing a down­load on th e tail that main ta ins thewing at a posi tive , lifting AoA. Inboth upri gh t an d inverted flight,the fuselage is inclined nose up ata sma ll ang le, and with someadded dra g.

SOLUTION N o. :zThi s method is mo re accurate andin volves one of the "drea de d"form ulas, as follows:

CL of1.00 af 9· from zero""CL 01 0.111 per degree

1.6 -.4HE

===-------- 100000 CL 1.4 -.3 CM+ 200000X 250000 1.2 .25

c

.6

.8

.4

.2

.04 .06 .08 .1 .12 .14 -14 14-.2 Co AoA

' .4Foraspect ratio6-constant chord 1/ .1CL 011.00 at 12.5 Irom zero Ii"

-.6 -.6 .15E197 (13.42%) c.. of .08 degree

1.6

1.4CL

1.2

ad just ment factor would be 0.067 5,reflecting the lower tip lift lossesfrom th e narrower tip chord.

A CL of 1.00 for 12.5 degrees is1.00 divided by 12.5, or 0.08 perdegree. This is the "slope" of the liftcurve at AR 6 and constant chord.

Our example mod el design hasth e following specifications :

• Estimated gross weigh t of 90ounces;

• Wing area of 600 square inches(4.17 square feet);

• Wing chord of 10 inches;

• Span of 60 inches;

• Estimated cruising speed of 50mph; and

• Wing loading of 90 divided by4.17, or 21.6 ounces per square foot .

The three-surface "Wild Goose" wasdesigned to theaerodynamic andstructuralprinciples in this book; specifically thosedescribedin Chapter 22, "Canard, TandemWing andThree-Surface Design." It's anexcellent flier.

There are two solutions to the deter­min ation of the wing 's AoA to sup­port th e plane in level flight at theestimated cruising speed .

SOLUTION No.1Refer to Figure 3. At a wing loadingof 21.6 ounces per square foot andat a speed of 50mph, the wingneeds a CL of close to 0.20.

Our wing develops a CL of 0.08per degree AoA. To produce CL 0.20would require an AoA of 0.20divided by 0.08, or 2.5 degrees fromzero lift, which for E197 is minus 2degrees.

The wing would thus be set at(2.5 minus 2) or 0.5 degree AoA­and at 0.5 degree ang le of incidenceto th e fuselage centerline on yourdrawings.

Note that a symmetrical airfoil'sang le of zero lift is zero degrees AoA.If our wing used a symmetrical sec­tion, its AoAwould be 25 degrees, aswould its angle of incid ence.

This is the "rigging" for a sportmodel, using a cambered airfoilsuch as E197, i.e., 0.5 degree AoA.Most pattern sh ips use symmetricalwing and horizont al tail airfoils;such airfoils have no pitchingmom ent and perform as well invert­ed as th ey do upright, but withlower maximum lift coefficients (CLmax) compared with cambered air­foil sections. (See Cha pter 2,"Understanding Airfoils.")

These agile models have chords

Lift =CL x a x V2 x 53519

Because we want to obtain the CLneed ed, th is form ula is modif ied to:

CL = Lift x 3519a xV2 xS

where CL =CL needed;

1001JY

95 V-'>90 I--

f-- Wing liffcoefficients

~ 11 ~ /85

/ V80

75 II I~/70 / V65 V

I ~ n ./:I: 60 / V... / ./:IE 55

/ / V 4JlV/c 50 1/ / / V 5.~V... ./... 45...

I / V V V I~V'"40

1

/ V / V V V- t!JV35

II / / ""V ~v30 V-

I/V,~ .> V ,../......

~I-(2225 l.--

mph} 20 V/ ,/ .....-: 1.17 ,../V ~~I-

(1615'/' /'" 1:80--l::=I-......

mph) 10 ~ :::::: :::::::-;::;--

5

4 8 12 16 2o 24 28 32 36 40 44 4(21.6)

WING LOADING

Figure 3.Nomograph for quickdetermination of wingloading, lilt andspeed at sea level.

14 THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Understanding Aerodynamic Formulas • CHAPTER J

CL 1.6 RE 1.6 -.4 C Stall CL max

C ~ 100000 CL-.35 M~1.4 + 200000 1.4

Stall climax X 250000 CL 1.171.2 1.2 -.3

1.17

.8 .8Pitching moment at

.6 0.5 AoA ot 0.060

.4Profile drag at CL0.20010.01 3

.2

.12 .14 10 14 18-.2 Co

-.4 Profile dragat CL .1 AoAmax 011 .17 010.015

-.6 -.6 .15

E1 97 (13.42%)

Figure 5.Profile dragandpitching moments.

A modeler living in Denver, CO, at5,000 feet above sea level would usea 0 of 0.8616.

For our model, at sea level, th iswould be CL= (90 x 3519) dividedby (1.00 x 502 x 600), or 0.211.

Our sample wing has a CLof 0.08per degree. The wing 's AoA wouldbe 0.211 divided by 0.08, or 2.64degrees, less the E197's 2-degreenegative to zero lift, or 0.64 degree,rounded out to the nearest 1;4degree, or 0.75 degree.

Lift = model's gross weight inounces;

V2 = estimated cruise speed inmph "squared";

S = wing area in square inches;o = density ratio of air (at sea

level, it 's 1.00; at 5,000 feet, it's0.8616; and at 10,000 feet , it 's0.7384).

our sample model had slotted flapsth at, when extended, increased thewing's CL max to 1.80, the stallspeed would decrease to 16mphfrom the unflapped 22mph , orbecome 27 percent slower.

STALLING ANGLESIn Figure 4, at infinite AR, the E197stalls very gently at about plus 11.5degrees, or 13.5 degrees from zerolift. For our wing of AR 6 and con­stant chord, th is would be:a = 13.5 + (18.24 x 1.17 x 1.17divided by 6), or 17.5 degrees fromzero lift , or 15.5 degrees AoA ataltitude.

For landing, however, this stallangle is greatly modified by:

• At constant CL, changes in wingloading are reflected in the speedneeded for level flight , and viceversa.

• Ground effect. As shown inFigure 6, at 0.15 of the Wingspan(60 x 0.15 , or 9 inches) aboveground, th e stall ang le is reduced to0.91 of its value at altitude, or to 14degrees.

• The level fligh t wing AoA.Because th e wing is at 0.5 degree, itwill stall at 13.5 degrees higher AoA.

18

Stalls

1.6 -.4CL CM

1.4 -.35

good con trol , would be 26.4m ph.Th is no mograph is mo st useful

in th e early stages of a mo del'sdesign . For example:

• At constant speed, it shows theeffect of cha nge s in wing loading,i.e., win g area and /or weight, onthe CL nee ded for level fligh t . Aswing loading increases, so mu stthe CL.

• At constant wing loading, it dis­plays th e effect of the CLon speed(or vice versa). For illustration , if

RE100000

+ 200000X 250000

.12. .14

Stall at inlinite AR -----'"'1-~--...,

Stall at AR 6-constant chord-~Lk~~:i::::~

E197 (13.42%)

1.6

1.4 C1.2

.8

.6CL

.4

.2

0

-.2

-.4

-.6

FIGURE JThis nomograph is one of the mostuseful charts in this author's "bagof tricks." It compares three impor­tant factors: speed (mph), wingloading (oz./sq. ft ) and wing CL. Itreflects the impact of changes inthese factors.

For example, our paper designhas a wing loading of 21.6 ouncesper square foot of wing area; th ewing has airfoil E197, which has aCL max of 1.17. Using Figure 3, itsstall speed is 22mph. Adding 20percent, its landing speed, under

Figure 4.The stalling angles of Eppler airfoil E197.

• High-lift devices. As Figure 7shows, slotted flaps extended 40degrees would cause a further

THE BASICSOFRIC MODEL AIRCRAFT DESIGN 15

CHAPTER:I ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

.c;..~<:­...<:=2...

•

.25Cand .30Csloned flapsI- with leading-edge slot

!--~Plain split

flaps/'

Sioned flap~---<, -- -r- _

'"..<:...u

'":;;'0..e;,<:<

1.1 0 -l.--::::==: E::::=: r::::::::I 1.0-;;:::::V-- V

i-1--8~-::::::""0.9 - ~ 1-- 6 V-~ ........ i- 4~V....-

,/v <,

V'"Wing aspect rallo0.8

/

0.7

0.05 0.1 0.2 0.3 0.4 0.5

12

10zen_...

8......encc~fficcQu l 4z"";;;uQ~z<!:!...

01-0u ...::::> ....Qc:l"'zcc<

-4

o 10 20 30 40 50 60

Height ot wing from groundWingspan

Figure 6.Impact ofground effect onangfe ofattack.

FLAP DEFLECTION ANGLE- DEGREES (@ Rn 250,000)

Figure 7.The effect onflaps andLEslotsontheangte ofattack at maximum lift.

1+-------.23C-------.t

reduction of 4 degrees to 9.5degrees stall ang le. Had the slottedslaps been combined with fixedleading-edge (LE) slots, there wouldbe a gain of 9 degrees, to 22.5degrees stall ang le.

The model's landing stall angle hasa major impact on landing-geardesign. (Chapter 16, "Landing GearDesign," goes in to this in detail.)

Figure 8 shows the geometry of afixed LE slot . Note how the slottapers from the lower entry to theupper exit .

Figure 9 displays the benefits of anLE slot in added CL and additionaleffective angles of attack before thestall. Drag is little affected.

Figure 10 shows the additionalCL to be obtained from varioustypes of flap alone, or in combina-

tion with LE slots.Slotted flaps and fixed LE slots

combine to more th an double theCL of mos t airfoil sections, produc­ing STOL performa nce .

For example, our E197 CL max is1.17. Equipped with deployed 30­percent-chord slotted flaps withextended lip and LE slots, bo thfull-span, th e Wing's CL max wouldbe 1.17 plus 1.25, or 2.42 .

Our sample model so eq uippedwould stall (Figure 7) at 14mph.

Figure 11 shows the added profileCo to be added to the section's pro­file CD' when calculating the totalof both profile and ind uced drags ,discussed un der "drag," as follows.

DRAGThe drag coefficients shown inFigures 5 and 11 are profile drag

on ly. The CL max profile drag of theun flapped E197 is 0.0 15 (Figure 5)and for full-span slo tted flap swould be an additional 0.121(Figure 11), for a to tal of 0.136 inprofile drag. Induced drag is notincluded. Note the very smallincrease in E197's profile drag forCL 0.20 to CL max 1.17.

The formula for calculation oftotal wing drag is:

CD = CDo + 0.318 X CL2 x (1 + 0)AR

where CD = total of both profileand induced drags;

Coo = section profile dragcoefficient at the chosen wing CL;

CL2 = wing lift coefficient

"squared";AR= aspect ratio;o = planform drag adjustment

factors .

Slat--+--.......

.0185C

R.23C

Figure 8.Geometry of thefixed leading-edge slot.

16 THE BASICS OF RIC MODEL AIRCRAFTDESIGN

Our model's wing has a Coo of0.013 at CL 0.20 (Figure 5) and adrag planform adjustment of 0.05(see Figure 4 of Chapter 1).

Replacing symbols with numbersfor the plain wing:

CD =0.013 + 0.318 x 0.22 x 1.056

or 0.01523 .

If our sample wing had full-spanslotted flaps that extended 40degrees and that were 30 percent ofthe wing chord, the total CD' at aCL max totaling (1.17 + 1.05), or

Understanding Aerodynamic Formulas A CHAPTER 3

1.8 .36I I "..

s lOlledlWlng

l---l

/ \/I \

1.6/ I \ .32

1I1l lncrease /I ,

.: V ,1.4

/.28

I

/ <,v:1.2

// / I

.24

"/ 1I // L

1.0'7'"1 I .20

Plainwing

~ t VI f!"u

.8 Lj

.16 ut: I t:l::; I I <a:.... rl QQ

/I ....... .6 I .12 Q

z/ V/ .......zs z....

iI:II -/

c::;....Rn @ 609,000 iI:....

0 .4 .08 ....u / V'"

..../ / CD 0

u

~/ ".. :;;V ADA Increase.2

L :::..---- .04p

0 0.4 0 4 8 12 16 20 2

ANGLE OF ArrACK-DEGREES(Rn 600,000)

.30Cslolled flapwithextended

lip and leading­edge slot

1.80

~ 1.60~...

15 1.40~ 1mI 1+------T~11l 1.20~ m:J ....-_"--,.,p;..- -=l~ 1.00

~:E .80=>:Ei .60~

o~ .40

!!1~ .20

The Wild Goose shown withslotted flaps onboth front and main wings extended forslow, stable landings.

6

FLAP DEFlECTlON-llEGREES (@R250.000)

Note that a tapered wing's roo tchord always flies at a hig he r Rnth an its tip chord at any speed,owing to the narrower tips (whic hcan be pron e to tip-stalls as a result).

Full-scale airfoil research da tamay be used for model airplanewing design-with careful regardfor the major effect of scale on par­ticularly lift, drag and stall angles.

Figure 10.Increments of maximum lift due to flapsandleading-edge slots.

o r---"""T"-..,......~-+-~--I

PITCHING MOMENTSThese have nothing to do withbaseball! All cam bered airfoils haveno se-down, or nega tive, pitch ingmom ent s. Symmetrical airfoils haveno pitching moments, excep t at th estall. Reflexed airfoils may have lownose-down or low nose-up pitch ingmoments.

Nose-down pitc hi ng momen tsmust be offset by a horizon tal taildownl oad tha t is ac hieved byhaving that tail's AoA set at a neg­ative ang le to the downwash fro mthe wing. (Chapter 8, "Hor izon talTail Inciden ce and Dow nwashEstima ting," goes into det ail.)

Rn =speed (mph) x chord (in.) x K

K at sea level is 780; at 5,000 feet,it's 690; and at 10,000 feet , it 's 610.

Our samp le mo del's win g cho rdis 10 inches, and at a landing speedof 26.4m ph and at sea level, its Rnwould be 26.4 x 10 x 780, or 205,920.

In Denver, the Rn would be26.4 x 10 x 690 , or 182,160.

A quicker solution at sea level isgiven in Figure 12. Layin g astraightedge from "speed" left to"chord" right, Rn is read from th ecenter column .

(Note: in Figure 2 of Chapter I , th elower drag correction factor 0 forth e tapered wing , of taper ratio 0.6,is 0.02 compared to that for aconstan t-chord win g of 0.05. )

SCALE EFFECTScale effect is measured by Rn. InE197, lift and pitching momentsare little affected by th e reductionin Rn from 250,000 to 100,000, butprofile drag increases substantially.

The formula for Rn is simple:

Figure 9.The benefits of thefixedleading-edge slot.

Repl acin g the symbols with num­bers for th e plain wing at 50mph :

Drag (oz.) = 0.410 x 1 x 142 X 6003519

Drag (oz.) = 0.0 1523 x 1 x 502 x 6003519

(Figures 5 and 11)

Drag (oz.) = CD x a x V2 x S3519

2.22 (Figure 10), would be:

CD= (0.15 + .121) +0.318 x 2.222 x 1.05

or 13.7 ounces .

or 6.5 ounces.

And for the full-span, slotte d-flapversio n at a sta lling speed of14mph, 30-percent-cho rd flaps at40 degrees:

or 0.410.

The formula for total wing drag is :

THE BASICS OF RIC MODEL AIRCRAFTDESIGN 17

CHAPTER 3 ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

.28QQ

~ .24wuu::: .20....wQ yfQ<.>

.16C> V~~c

~ .12u:::~~~iit" nd 1---:;z-:0

a:"- .08 BO~z

00t-SIOfl flapse ~w .04en....

0en.....zw

0 10 20 30 40 50 60::Ewa:<.>iii!: FLAPDEFLECTIDN-DEGREES (@R250,000)

Figure 11.Increments ofprofile drag coefficient at CLmax or increasing flapdeflections.

As Figure 5 shows , the E197 air­foil has a negative CM of 0.060 atan AoA of 0.5 degree. Note that CM,like c., varies with th e AoA.

Also, the CMapplies to the wing'sV4MAC; on our straight wing of 10inches chord, at a point 2.5 inchesfrom its leading edge.

The pitching moment formula is:

SPEED · MPH REYNOLDS CHORD·NUM8ER INCHES

180 3400000 243,DOd,ooo 23

1602,5000,000 22

140 21120 2,000,000 20

100 1,5000,000 1918

90 1780 1,000,000 1670 800,000 1560

600,000 1450 500,000 13

40 400,000 12

300,000 11

3010

III 150,000 9

20100,000 8

1580,000

760,00050,000

1040,000

6

32,000 5

Figure 12.Nomograph forquick determinationofReynolds numbers.

18 THE BASICS OFRIC MODEL AIRCRAFT DESIGN

PM ill ill-OZ. = CM x a x V2 x 5 xC35 19

where CM = airfoil pitching­moment coefficient at the AoA oflevel flight;

V2 = speed in level flight"squared";

S =wing area in square inches;C =chord in inches;a =density ratio of air.

Our sample Wing's nose-down PM is:

PM = 0.060 x 1 x 502 x 600 x 103519

or 255 .75 in-oz .

A moment is a force times a dis­tance. In our sample, if a tail­moment arm dis tance were 30inches, the tail download to offsetthe nose-down moment would be255.75 divided by 30, or 8.52ounces. (Chapter 8 goes into thi s indetail.)

RPM, SPEED AND PITCHNOMOGRAMFigure 13 was developed to helpmodel designers choose proppitches and diameters suitable forboth plane and engine to ob tainoptimum performance.

This is explained in Chapter 8.Figure 13 should be used withFigure 3, "Wing Loading Lift SpeedNomograph. " Don 't use Figure 13alone to estimate the speed of anyprop/plane/engine combina tio n; ifthe prop pitch and diameter aren'tsuitable for a model's character is­tics, the nomogram will not beaccurate.

It would obviously be poor judg­ment to use a high -pitch, low­diameter propeller on a large, slowflying, draggy model with low wingloading. Similarly, a low-pitch ,large-diameter prop on a low-drag ,fast airplane with a high wing load­ing would be a poor choice.

I hope that this chapter will over ­come any problems some reade rsmay have with formulas in th isbook. To succeed, one mu st try! Noeffort , no success! ...

STATIC RPM LEVELFLIGHT NOMINALX1,000 SPEED (MPH) PITCH

4 18.3 420

255 5

3035

6 40 6

7 7

8 8

9 9

10 10

11 15011

12 12

13 20013

14 14

15250 15

16 300 16

17 350 171818 400 1919

20 450 2021 50022232425

Figure 13.Nomogram forchoosing suitable proppitches anddiameters.

Figure 1.The author proposes theuse ofplain flaps, depicted above, onpattern ships(see text).

60·

W ing loading is simplyyour model's weight inounces (including fuel)

divided by its wing area in squarefeet. It's expressed as "ounces persquare foot of wing area."

In th e initial stages of design of anew model aircraft, man y majordecisions have to be made that willdetermine its ultimate size andconfiguration:

• the size and make of engine (ifany) ;

• th e type of performance goalssought; (basically, is it a sportmodel of moderate speed andmaneuverability or one that's fastand aerobatic? As a glider, is it athermal seeker or a fast, sleek, aero­batic sailplane?);

• the Wing planform (straight,tapered or elliptical);

• the airfoil; and

• the estimated weight.

Your mode l's wing loading is one ofthese major decisions-and should

Gap seal

Chapter 4

be "performance-objective oriented."Wing loadings vary widely; glid­

ers and sailplanes have wing load­ings that range from less than 10ounces per square foot to IS ouncesper square foot. Sport models areusually in the I S to 20 ounces persquare foot range. Pattern modelshave wing loadings from 23 to 26ounces per square foot. Scale mod­els are min iatures of existing air­craft. None of my scale modelingfriends knows or cares what hismodel's wing loading is. They relategross weight, in pounds, to enginedisp lacement to ensure adequatepower.

Scale models don't often involvethe same design latitude as othertypes of model, but some arefantastic examples of excellentworkmanship.

HIGHER WING LOADINGSI personally favor higher wingloadings because they result insmaller, stronger, faster and-ifyou 're careful in the design andconstruction phases-less "draggy"aircraft.

Higher wing loadings, however,result in higher stall and landing

speeds. Level flightrequires a higherangle of attack orgreater speed. Themost serious impactof a higher wing load­ing is on centrifugalloads when engagingin maneuvers thatinvolve heavy eleva­tor action. Such man­euvers include tightturns, sharp pull-upsor dive-recoveries.

An advantage of ahigher wing loadingis that, at any givenspeed , the wing mustoperate at a higherlift coefficient that 's

Wing Loading

Design

further up the slope of th e lift curveand closer to the stall. Entry in tomaneuvers that in volve wingstalling, such as spins, snap rollsand avalanches, is more readilyachieved.

Once you 've est imated yourdesign 's gross weight (with fuel)and decided your wing loading,the wing area (in square inches) issimply:

model gross weigllt (oz.) x 144willg loading (oz./sq. ft.)

LANDING SPEEDSWing loadings and landing speedsare closely related. Refer to Figure2, and read up from th e 16 ouncesper square foot point at the bot­tom of the chart to the CL of 1.00(most airfoils ' CL max is close to1.00). On the left side of the chart,you 'll see that the stall speed is20mph. Do th e same thing on the36 ounces per square foot line,and you' ll see that the stall is30mph. Adding a "safety margin"of 20 percent to each stall-speedestimate results in landing speedsof 24 and 36mph. The latter is toofast for comfort.

CENTRIFUGA L FO RCECentrifugal force is expressed inmultiples of "G", where 1G is nor­mal gravity. Its formula, includingthe model's 1G weight, is:

THE BASICS OF RIC MODEL AIRC RAFT DESIGN 19

CHAPTER 4 A THE BASICS OF RIC MODEL AIRCRAFT DESIGN

• ou tboard ailero ns of 25-percent

4 8 12 16 20 24 28 32 36 4044 48Wing loadlng- Ol./Sq . ft.

chord tha t are 35 to 40 percent ofthe semi span in length;

• plain flaps inboard of the aileronsto th e fuselage; and

• E168 with a CL max of 0.98 , th enthe full y dep loyed flap at 60degrees would provide a wing CLmax of 1.30 and, at 20 degrees ofdeflection, a wing CL max of 1.13.

The pilot could extend th ese flapsup or down at any angle to suit th eman euver in progress. Land ings,with a 60-degree flap deployment,with a high wing loading of 28ounces per square foot , would be at28mph- a comfortable speed.

In addi tion, for sharp-turn ingman euvers, lowering these flapspartially to 20 degrees would pre­ven t high -speed stalls.

At 100mph in level flight , a CL of0.068 is required. For a turn radiusof SO feet at 100mph, th e load fac­tor would be 14.34G's. This calls fora CL of 0.97, whi ch is dangerouslyclose to th e E168's CL max of 0.98 .The 20-degree flap deflec tionwoul d provide a CL of 1.13, whichwould be safer.

With flaps up , the higher load ingwo uld move th e level-flight CLhigh er up th e lift slope, closer to CLmax. In tu rn, th is provides easierentry into any maneuver requiringthat th e win g be stalled.

A .60-powered pattern modelthat weigh s 8 pou nd s (128 ounces),and has a wing loading of 28oun ces per square foot would havea wing area of 4.57 square feet, or658 square inches.

Patt ern ships have evolve d overtim e into beauti ful configuratio nsof startling similarity to one another.It's tim e to consider some freshapproaches to th eir design . Perh apsflaps and higher wing loadings aresuch approaches. A

.. .1"

lilt /' /CO " I lents ~ .15 ./

// I V

V V ~

/ /V

~k/./ V ,40V

1/ / / ./ V ·5!V/ 1/ ,/' V "'!!IV

/ 1/ / /' V V """ ,88',....-

/ 1/ j / v :./ v r 1J1i-

f/ / /.. /' :./ ,....- ...... 1J11 i--

'/ ~ ;..--: -::-V ,....- --..... !-"

-0 r:/. :.- ~ --- ;..;.---~~~ :;:::::. V.~

100

9590858075706560

~ 55! 50:: 45CLen 40

3530

2520

1510

5

Figure 2.From wingloading at thebonom, readvertically to the appliicable liff coefficientandthen move leff (horizontally) to findthespeed in milesperhour. The stallspeed isbased onanairfoil'smaximim liff coefficient.

PLAIN FLAPSPlain flaps (Figure 1), however, inwings with sym me t rica l ai rfoilsections , such as E168 (standa rdon pattern mod els) would func­tion equally well angled down (foruprigh t fligh t) or up (for invertedflight). They achieve their CL maxat 60 degrees of deflect ion andwould add an additional CL of0.62 at that ang le, plu s additionaldrag to slow the mo de l. At 20degrees of deflection, the addi­tional CL would be 0.25.

If we assum e:

1.17). Tighter tu rns are possiblewitho ut danger of a h igh- speedstall. The Swift 's sturdy flaps arestrong enoug h to accept thistreatment.

The Swift wasn' t design ed to bea stu n t mod el; it 's a "sport-for­fun " model with a wide speedrange and low landin g and takeoffspeeds, i.e., with flaps deployed .Its slott ed flap s aren' t suitable forthe wide range of aerobatics thatpattern shi ps per form, bothupright and in verted .

where N =load factor in G's:mph =speed in mph;R = man euver radius in feet;G = acceleration of gravity (32.2

feet/second per second).

N = 1 + (1.466 x rnph)2R x G

Aerodynamically clean model air­craft tha t have powerful engi nesand are correctly "propped" canachieve very high speeds. Theno rm for patte rn shi ps is 100mph.My "Swift" has a top speed of125mph; its gros s weight is 92ounces , and its wing loading is 22ounces per square foot. At 90mph,it flies at a CL of 0.072.

In a stee p tu rn of a SO-foo tradius, th e load factor would be

N = 1 + (1.466 X 90)2 = 11.8 G'S50 x 32.2

SLOTTED FLAPSThe Swift-slotted flaps up-Willland at 30mph. With flaps down40 degrees, at a CL max of 1.9, itslan ding speed is 22mph. Flaps thuseliminate th e adverse effect th athigher Wi ng load ings ha ve onlanding speeds.

In high-speed, short-radius turn­ing maneuvers, 20 degrees of flapdeflecti on would in crease theSwift's CL max to 1.6 (from flaps-up

In th is maneuver, th e Swift's winghas to lift 11.8 x 92, or 1,086,ounces-a shocking 68 pounds.

Just think what thi s means bothaerodynamica lly and structurally.This is why I favor stiff, stro ng,fully shee ted and stress-skinne dstructures.

The lift coefficient in this turnwou ld increase 11.8 times to CL0.85, well within its E197 airfoil'scapacity of CL max 1.17. There's ahea lthy marg in before th e stall.

If the Swift's airfoil were E168with a CL max of 0.98, however,th en th is margin wou ld be greatlydiminish ed . (See appendix forEppler airfoil data.)

It's impossi ble to gauge accu­rate ly th e mo del's turning radiifrom several hundred feet away,hence th is safety facto r is needed toavo id "high-speed sta lls" (whic hwould probably result in un com­manded sna p rolls).

20 THEBASICS OFRIC MODEL AIRCRAFT DESIGN

T he Swift's design is the cen­tral theme in this chapter. Itweighs 92 ounces fueled , has

600 square inches of wing area (4.17square feet), an AR of 6.3 and ispowered by an O.S. Max 0.46 SFengin e rotating a lOx9 or lOxlOAPC prop . Its top speed is 125mph ,and flaps fully extended, it will stallat 18mph. Its wing loading is 22ounces per square foot, and itspower loading is 200 ounces percubic inch of engine displacement.

A detailed ana lysis of th e Swift'sweight of 92 ounces reveals that46.5 ounces (or 50.4 percent) of thatweight can be classified as "fixed."This weigh t, over whic h thedesign er has no control, consis ts of:

• Power unit-spinner, prop,engi ne, muffler, cowl, tank an dfuel;

• Co n t rol unit- receiver (6­channe l), batt ery (700mAh), fiveservos, an on/o ff switch, and foamshock insulation ;

Chapter ·5

• Landing gear-tricycle with 2­inch-diameter wh eels.

The remaining weigh t of 45.5ounces (or 49.6 percent of thegross) is composed of win g, fuselageand tail surfaces. This portion isunder the control of th e design er. Thewing loading he selects will dictateth e wing's area, and generally, th esize of fuselage and tail sur faces. Itwill also influen ce th e structure;lower wing load in gs and lowerspeeds redu ce flight loads, particu­larly tho se du e to centrifugal force,pe rmitting ligh te r, less ruggedstructural design .

It 's poss ib le to design a model of800 square inc hes of wing area(5.56 square feet) with th e samegross weig ht as th e Swift by use ofa more open structure. This mod elwould have a lower wing loadingof 16.5 ounces per square foot andwould sta ll at 18mph.

Thus, flaps for landing wouldn 'tbe needed . The weigh t of the fifth(flap) servo; the additiona l weig h tof the 700 mAh battery (versus500mAh); an d th e add it io na lweigh t of the flaps, their h in gin gand thei r actua tio n would all be"saved." The performance of thismod el would not be as goo d as theSwift 's, however, largely owing tothe increased to ta l drag resu ltingfrom its larger size.

The point of all thi s is th at th etype of performa nce desired by th edesigner dictates th e wing loadingand, to a large extent, th e structure .For the Swift, hi gh spee d andma ne uverabili ty were the obje c­tives, calling for a rugged, stress­skinned and low-drag design. Thus,within reason able limits, win g load­ing governs performance and struc­tura l design .

WEIGHT ESTIMATINGHavin g selected th e power and con­trol units and type of landing gear,it isn' t difficult to closely estimate

Wing Design

th eir fixed weights.Sim ila rly, hav in g decided on

the wi ng loading, the variab leweigh t of wings, tail surfaces andfuselage may be estimated withreasonabl e accuracy. My own esti ­m at es h ave only ra re ly been"righ t on"; the tenden cy was tounderestimat e. In compensation,the Swift 's gross was overestimatedat 100 ounces, whe reas the actua lis 92 ounces- 8 ounces differ­ence. While n ot per fect , thisrationa l but practi cal approachshouldn' t result in a differencebetween the est imate and act ualof mor e than 10 percent .

With weigh t estimates of bothfixed and var iable componen tsachieved and the wing load in gselected, the wing area is easilycalculated:

Wing area in square inches =Weight in oz. x 144

Wi ng loading in oz. per sq. {to

It's useful at the ini tial stages of anew design to have a prelimina ryestimate of th e new model's totalweight and wing area. In Chapter13, "Stressed Skin Design," theweight versus wing area of 14 modelsis analyzed, disclosing a surprisingconsistency in th e weight versusarea relationship of 0.1565 ounceper square inch-or 22.5 ounces persqua re foot . For those ado ptingstressed -skin cons truct ion, th esefigur es provide an easy weight­estima te basis.

For others who prefer lighter,mor e open structures, a study ofconstru ction arti cles and productreviews will help .

A word on tank size. It makes no

THEBASICS OF RIC MODEL AIRCRAFT DESIGN 21

CHAPTER 5 ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

REMOTE FREE STREAM

sense to provide a 16-ounce fueltank on a model powered by a 040to .SOci engine. Most sport flightsseldom last more than 25 minutesso, on landing, the 16-ounce tank isstill half-full. Your model is penal­ized to about 1;2 pound carryingthi s useless weight. A guide to tanksize relative to engine displacementis 20 ounces per cubic inch ofengine displacement. Thus, for aAOci engine, an 8-ou nce tank isright on .

Now, let 's cons ide r the manyother design decisions to be made.It's fun !

WING PLANFORMS• Elliptical wings. Th is is the"ideal" wing planform. lt has thelowest induced AoA and induceddrag and stalls evenly across itsspan. These factors in crease fortapered or rectangular wings. Forexample, a rectangular wing of AR6 would requ ire an induced AoA (T)17 percent highe r and with induceddrag (&) 5 percent higher than anelliptical planform . (See Figures 2and 4 of Cha pter 1.)

Structurally, th e elliptical wing isdifficult to produ ce. Each rib is dif­ferent an d wing skins all have adouble curva ture, chordwise andspanwise. The Spitfire 's ellipticalwing is a classic example.

• Rectangular wings. This is theeasiest typ e to design and build.All ribs are th e sam e, and wingskins ha ve a sing le chordwise cur­vature . Whil e it suffers in compar­ison with th e elliptical, for smallmodels, it ma in tains a constan t Rnacross its span, whe reas a taperedwing of the same area could havetip Rns in th e high drag/lower liftand stalling-ang le ran ge of lowRns, leading to premature tip- stallsat low speeds.

Structurally, the wing roots needreinforcing , owing both to narrowerroot chords and higher bendingmoments. The cente r of lift ofeach win g hal f is farthe r from thecente rline than an elliptical ortap ered wing .

• Tapered wings. A tapered wingwith a tip cho rd of 40 percent ofth e root chord comes closest to theideal elliptical planform in bothinduced AoA and induced drag (see

:Z:Z THE BASICS OFRIC MODELAIRCRAFT DESIGN

Figure 2 of Chapter 1). For wings ofsmaller models, this taper ratioresults in narrow tip chords andundesirably low Rns at low speeds.Increasing the taper ratio produceslarger tip chords. The resulting lossin efficiency isn 't great and is the"lesser of the two evils."

Structurally, the tapered winghas lower root bending moments,and the wider, deeper root chordprovides the greatest strengthwhere it 's needed most-at theroot. A tapered wing can be lighteryet stronger than a rectangularwing of the same area.

• Sweptback wings. This causessimilar behavior to decreased taperratio (smaller tip chord) and leadsto early tip-stalls with a nose-uppitch, since the tips, being behindthe CG, lose lift. lt has a dihedraleffect; 21;2 degrees of sweepback(measured at 25 percent of thechord) is roughly equivalent to 1degree of dihedral. lt also promotesdirectional stability; if yawed , theadvancing wing's center of dragmoves away from the CG, and theopposite, retreating wing's centermoves inward. The resulting dragimbalance works to oppose theyaw. Large sweptback anglesincrease induced drag and lowerthe wing's maximum lift.

Wings of moderate taper ratios(0.5 to 0.6) with straight-across trail-

EFFECTIVE LIFT -

a

Figure 1.The origin of Induced drag.

ing edges and sweptback leadingedges are popular for pattern ships.These wings tip-stall readily for easyentry into wing-stalling maneuverssuch as snap rolls, spins, etc.

Structurally, a sweptback wing'slift tends to reduce the Wingtip'SAoA, particularly at high speedsand high centrifugal force loads. Astiff wing structure will preventpotentially damaging wing flutter.

• Swept-forward wings. Thesetend to stall at the wing root first.The unstalled tips promote goodaileron control at high angles ofattack. The root stall reduces lift aftof the CG, causing a nose-up pitch.

Forward sweep is destabilizing inyaw. The centers of drag and lift ofthe advancing wing panel moveinboard; on the opposite, retreatingpanel , these centers move outboard.The unequal drag moments increasethe yaw, while the unequal liftmoments cause a roll, but in a direc­tion opposed to the yaw. Control ofthis instability calls for increased ver­tical tail surface area and effective­ness, along with generous dihedral.

Structurally, a wing very stiff intorsion is required to overcome thewingtips' tendency to increase theirAoA. Any flexibility could be disas­trous at high speeds.

In full-scale airplanes, modestsweep forward moves the wings'main spar aft , out of the way, and

I 01· INDUCEDDRAG

-

Power loadingoz./cid 2-strl!ke

improves the pilot's forward anddownward vision.

• Delta wings. The triangular shapeof a delta wing is so called because ofits resemblance to the capital letterdelta (d) in the Greek alphabet.These have very low ARs. Low-ARwings stall at high angles of attack­but with high induced drag. Vortexflow is high , since a delta wing is vir­tuallyall "wingtip."

Deltas don't need flaps for land­ing owing to their high AoA capabil­ity, but should be landed with somepower-on to overcome their highinduced drag. Power-off, they havethe glide charac teristics of a brick!

A tailless delta-wing model, withthe whole trailing edge composedof elevons , is highly maneuverableand will not spin , but requires sym­metrical or reflexed airfoil sectionsfor longitudinal stability.

Structurally, deltas are verystrong. The deep, wide centerchord promotes strength, and thelow AR reduces the bendin gmoments at the wing's center.

• Co m bined rectan gular andtapered wings. This planform isrectangular for roughly 50 percentof the semispan (inboard) andtapered for the remaining 50 per­cent to the wingtip. Piper Warriorsand Cessna 172s typify this plan­form. It comes close to the ellipti­cal in shape and efficiency, yet ismore easily produced than atapered or elliptical wing. Th ecommen ts earlier regarding thehazards of low Rns of narrowwingtips apply. The rectangularinner portion is wider in chord,which provides a stro ng win groot, and bending moments arelower than for a rectangular wing .

ASPECT RATIOTh is importan t ratio is th atof wingspan to mean chord. Itsformula is:

Span2 = Aspect ratioA"T'ea

The Swift 's wingspan is 61.625inches and its area is 600 squareinches. Its AR is:

61.6252= 6.3600

IAIUliModel type

The AR of a wing has a majorimpact on its "induced drag"­defined as th at drag caused by th edevelopment of lift-and is sepa­rate from th e drag caused by th ewing airfoil's form an d frictio n,called "profile drag."

As Figure 1 indicates, increasi ngthe AoA causes th e lift to tilt rear­ward, resulting in a horizontal vec­tor th at produces ind uced drag .

The classica l formula for theinduced drag coefficien t is:

Lift coefficient2

st x Aspect ratio

or

0.318 X CL2 = CDiAR

Obviously, th e higher the AR, th elower will be the induced dragcoefficient-and the lower theinduced drag. This is why soaringgliders have such lon g, narrowhigh- AR wings.

An airplane's total drag is com­posed of two types: parasite drag(includi ng profile drag), whichdoesn 't contr ibute to lift; andinduced drag, which results fromthe Wing's produ ction of lift. Figure2 illustrates thi s relation sh ip.

Induced drag has a very signifi­cant difference from both lift andparasite drag. The latter two arepro portional to the square of thespeeds; induced drag, however, isinve rsely proporti on al to the squareof th e speed . It's lowest at h ighspeeds and highest at low speeds .Lift and parasite drag are low at lowspeed and high at high speed .

At 100m ph , th e total of profileand induced drags for th e Swift is22.4 ounces, of which th e induceddrag is 0.215 ounce-or less th an 1

Wing Design ... CHAPTER 5

Wing loading Aspect ratiooz./sq. ft .

percent. At 30mph, total wing dragis 4.3 ounces, of which 2.3 ounces,or 54 percent, is induced drag­useful in slowing th is model forlanding.

It's th is relat ion ship that explainsth e power-off, brick-like glide of adelt a wing. The low AR and highlift coefficients result in very highinduced drag for low-speed deltaflight.

Figure 2 depicts typical airplanedrag curves . Where the induceddrag equals the parasite drag is th espeed of the maximum lift-to-dragratio and of the maximum range.

Range, for model airp lanes, is nota factor of any consequence, exceptin rare instances, since most pow­ered RIC flights seldom exceed halfan ho ur in dura tion .

ASPECT·RATIO PROSFor a given wing area, increasing theWing's AR will reduce th e induceddrag. The narrower chord tips resultin smaller wingtip vortices; the liftper degree of AoA increases so thatth e model flies at a lower AoA.These all favor high ARs.

ASPECT·RATIO CONSLower chords on smaller modelsresult in lower Rns-particularly atlow speeds . Scale effect causes anincrease in wing profile drag, aredu ction in maximum lift an dlower stalling angles .

The centers of lift of each winghalf are farther from the fuselage forhigh-AR wings, resulting in substan­tial increases in root bending loads.In addi tion, long, na rrow wingsmust be stiff in torsion to preven ttwisting un der loads from twosources-pitching-mo ment changesas th e model maneuvers and theopposed action of ailerons. Wingsweak in torsion have been known to

THE BASICS OF RIC MODEL AIRC RAFT DESIGN 23

CHAPTER 5 ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

r

r

STALL-.......

TOTAL DRAG

INDUCED DRAG

IFigure 5.AnIllustration of thewingtip vortex flow.

VELOCITY

Figure 2.Typical airplane drag curves. Parasite drag varies directly asthe speed squared; induced dragvaries inversely asthespeed squared.

STALL PATTERNSFigure 3 illustrates how the variouswing planforms stall at high anglesof attack. Note th at th e rectangularwing stalls root first, perm itt ingeffective aileron control well intothe stall.

There are a variety of ways inwhich tip-stalling may be delayedto higher angles of attack. The bestand simplest form is the NASA­developed and tested partial-spanwing-leading-edge droop. This fea­ture has been used very successfullyon six of my model designs .

Figure 6.The downwash andwake fora convention­al, rear-tailed, air­craft. Note thesug­gested droop fuselagethatwould decreasedrag. Time framesabove thewing arespaced farther apart toIllustrate higher­velocity air.

carefu l drag reduction is neededalong with sound propeller selec­tion. Higher flight speeds result withlower lift and profile drag coeffi­cients and lower induced drag untilthe total drag equals the thrust. Toprovide the optimum strength-to­weight ratio to overcome highcentrifugal force loads, stressed-skinstructu ral design is suggested. Toreduce landing and takeoff speeds,slotted flaps are recommended.

REARWARD ANDDOWNWARD ACCELERATION

TIMEFRAMES

Figure 4.Asair tlows pasta wing from leading edgeto trailingedge, positive pressure iscreated below theWing, while negativepressure exists above. At thewingtip, thepositive-pressure bottom wingair flowsaround thetip andis drawn Into thenega­tivepressure region above thewing. Thisaction gives rise to the wingtip vortex, aswellasto lesser vortices along thetrailingedge.

RECTANGUlAR, ).... 1.0

HIGHTAPER, ).,,, 0.25MODERATE TAPER, ). " 0.5

POINTED TIP, ). " 0

ELLIPTiCAl

Figure 3.Stallprogression patterns for variousplanform wings.

experience "aileron reversal." Thisoccurs when heavy down-goingaileron action twists the wing lead­ing edge down . The up-going twiststh e leading edge up. The modelbanks in a direction opposite to thatintended by its bewildered pilot.

High ARs result in weightincreases, particularly for modelsdesigned for high speeds wherehigh centrifugal loads are encoun­tered. Increased weight results inhigher wing loadings and higherparasite drag. Obviously, there mustbe some compromises.

With his neck "stuck way out,"thi s author suggests th e followingclassifications for radio-controlledmod el aircraft (see Table 1):

From this designer's poin t of view,to obtain the maximum efficiency,

24 THE BASICSOF RIC MODEL AIRCRAFT DESIGN

Wing Design .... CHAPTER 5

Root

Figure 9.Spanwlse loaddistribution of modified wingatCL =DAD.T!!e wing features a 45· swept tip.

Load

Figure 7.The SchuemannwlRg planform.

FlgureB.Modified wing

planformgeometry; 45"

swept tip.

l.E.

Figure 10.Rutan model 81 Catbird. Note three surfaces.

lE.

WINGTIP DESIGNThe major difference in efficiencybetween the elliptical planform,considered the best, and otherplanforms is largely due to wingtiplosses. The elliptical has no pro ­nounced tip-one could say it is"all tip "- whereas the rectangularplanform has the widest tip .Tapered wingtip widths vary withtaper ratio.

Figures 4 and 5 portray the air­flow over and under a wing andparticularly the tip vortex flow.Figure 6 shows the wake and down­wash resulting from the wing 's pro­duction of lift.

Obviously, the narrower the tip,the lower the tip losses with dueregard to stall patterns and scaleeffect, particularly at low speeds. Atip-stall close to the ground may be

damaging to both model and itsdesigner's ego!

Over the years, aerodynamicistshave explored many wingtip con­figurations in their search forimproved wing performance. Twoforms, somewhat resembling eachother, have emerged.

First is the Schuemann planform(Figure 7).

The second is the "sheared"wingtip, largely developed by c.P.Van Dam of the University ofCalifornia. Figures 8 and 9 providean outline of a sheared tip alongwith its spanwise load distribution.Note how close "modified" is to"elliptical" in Figure 9. This form oftip has been, or is being, applied tofull-scale aircraft designed by suchno ted aerodynamicists as BurtRutan and Peter Garrison. Figures

10 and 11 illustrate these designs.This author uses a modified

sheared wingtip that is both simpleand rugged. Figure 13, a top viewof the Snowy Owl's wing , illus­trates this tip form .

FLAP CHORDSEarlier model designs, such as th eSnowy Owl, had slotted flaps whosechord was 26 percent of th e wing'schord and were close to 60 percentof the wing 's semi-span in length(see Figure 13).

After being throttled back andhaving their flaps fully extended,these models porpoised upward sud­denly. Elevator down-trim appliedsimultaneously with flap extensionwould prevent this behavior, whichwas annoying.

Analysis disclosed that the

THE BASICS OF RIC MODEL AIRCRAFT DESIGN :zs

CHAPTER 5 .... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

~0.2566e slolled lIap~

SNOWY OWL csc.

l Oll( . oe--'I'!J ;0.015e

0.30e Fowler lIap; ~ tgap ; O.015e 'O'q,c/

SEA HAWK

I~nooe--.\ g~p ; 0.02e

0.30e slolled lIap ~with extended lip; 'O'a Vgap ; 0.02e t'

SWIFT

X 1.8

'"Eu'1 .6<l

"E 1.4'"'(3

iE'" 1.20u

~ .10E::l

E .8'x'"Ec: .6su

~ .4

- - 0.2566e s lolled lIap•••• 0.30e Fowler tlap;

gap ; 0.015e- 0.30e slolled tlap

with extended lip;gap ; 0.02e

10 20 30 40 50 60

Flap deflection, degrees.

Figure 13.Snowy Owl's flaps were 60% of thewingsemi-span.

Figure 12.Comparison of increments of section maximum tift coefficientfor three flaps ona NACA 23012airfoil.

increase in an gle of downwashfrom the extended flaps was forcingthe tailplane down and creating agreate r force than th e increase innose-d own pitch . The wing 's AoAand lift increased, and the modelzoomed upward until the excessspeed bled off. The model thennosed ove r in to th e flap-d own ,slow glide.

Experi ence with three of mymodels (Sea Gull Ill, Sea Hawk andSwift) has proven that wideningth e flap chord to 30 percent of th ewing chord produces a ba lancebetween these "nose-up" and"n ose-do wn" forces, flap s full yextended. All three models exhibitno change in pitch on loweringflaps-but fly mu ch more slowly.

On landing approach , groundeffect reduces th e downwash angleand increases the nose-down pitch.The glide close to th e ground steep­en s, but appro priate up -elevatoraction raises the nose so th at agentle, slow landing result s. ....

26 THE BASICS OF RIC MODELAIRCRAFTDESIGN

r------- - --~-~~

Eppler 197(9.75" Chord)

~

100%Eppler 197M(10.1 " Chord):----t

-.----

T he location of the center ofgravity has a major impacton longitudinal stability,

the selection of the horizontal tail'sangle of incidence and on themodel aircraft's maneuverability.For sport models, it's customary tolocate the CG at the wing's aerody­namic center (25 percent of MAC).

There is, however, a range of CGsboth ah ead of and behind thewing's aerodynamic center. Thesepositions result in varying degreesof long itudinal stability. The steelball in a saucer is a very graphicmanner of describing pitch stabilityat various CGs (see Figure 1). Notethat at position 4, the neutral point,the ball is on a flat surface and maybe moved in any direction withoutreturning to its original location incontrast to positions 1, 2 and 3,where the ball does return. At point5, the ball will roll off the inver tedsaucer, indicating serious instability.

The following will outline the var­ious CG advantages and limitations.

FORWARD CGThe most forward CG possibledepends on the downward liftingcapability of the horizontal tail.When I designed the Swift, the taildownload needed to offset its wing

Chapter ·6

airfoil's pitching moment was cal­culated at 15.4 ounces at 60mphlevel flight . A CG at 5 percent of theMAC, almos t 2 inches ahead of theaerodynamic center, wou ld furtherincrease the required tail download .This results in three things:

• It increases the weight the air­plane's wing must support.

• It reduces the ho rizontal tail'spitch maneuverability. This isbecause a major part of the tail's liftcapaci ty is taken up with overcom­ing the nose-down combination ofpitching moment and CG.

• This limited capacity makesachieving a full stall attitude diffi­cult, if not impossible, in groundeffect (th is pressure of the groundreduces downwash). Moreover,with slotted flaps fully extended,the wing's nose-down pitchingmoment is further increased evenwith full up-elevator.

However, at this forward CG, themodel's longitudinal stabilitywould be high, and it would recov­er by itself from any pitch distur­bance, returning to level flight . Itwould be easy to fly, but not highly

CG Location

man euverabl e. Moving the CGrearward improves maneuverabili tybut reduces pitch stability.

REAR CG ANDTHE NEUTRAL POINTModern aerod ynamic ana lysis forassessing the stability of an airplaneis based on the fact tha t a wing andtailplane represent a pair of airfoilsin tandem. Each has its own aero­dyn am ic center, but th e combina­tion will also have a correspondingMAC equivalent to th e point wherethe total lift (and drag) forces of th etwo airfoils effectively act. Thi sMAC is called th e "neutral point"(NP). It follows th at th e NP will liebetwe en th e aerodyna mic centersof the two airfoils and closest to thelarger or more effective lift produc­er, i.e., th e wing of conventionalcombinations, or th e aft wing of acanard. Any disturban ce in pitchthat momentarily upsets the nor­mal flight path of th e aircraft willcause a change in AoA of both air-

1. Very Stable

CG Poslt lons---.. 5% MAC

2. Stable

25% MAC

3. Less Stable

30% MAC

4. Neutra l

35% MAC

5. Unstable

BEHIND THE NP

sa ucers ?

Steel Ball

Wing 's MAC ---.

Aerodynamic Center----+25% MAC

Figure 1.In Ihis illustration, a ball bearing in a saucer simulates therelative pitchstability of various CG locations.

THEBASICS OF RIC MODEL AIRCRAFT DESIGN 27

CHAPTER 6 .. THE BASICS OF RIC MODEL AIRCRAFT DESIGN

FIXED WEIGHTS DUNCES PERCENT

WEIGHT ANAlYSIS FOR THE sWln

VARIABLE WEIGHTS OUNCES PERCENT

WING: 600 square inches at0.039 oz/sq. in., Y1 6-inch-thickbalsa skins, two spars, ailerons, slotted flaps(control cables included). 23.4 25.4%

HORIZONTAL TAIL: 120 square inches at0.028 oz.lsq. in.,Y16-inch-thick balsa skins and elevators; 40%(mass balanced.) 3.4 3.7%

VERTICAL TAIL: 40square inches at0.030 oz./sq. in.,Y16-inch-thick balsa skin, one spar and rudder.(Mass balanced.) 1.2 1.3%

FUSELAGE: Length from the engine bulkhead to the ruddertail post is 34.5 inches, 6 inches deep and 4.5 inches wide.This comes to 931 .5cubic inches at 0.017oz.lci assuming31.l2-inch-thick balsa skinsand3116-inch-thick balsa corners(control cables included). 15.8 17.0%

• Experience with several models indicates an averagefuselage weight of0.017ounces percubicinch,given theconstruction noted.

VARIABLE SUBTOTALS 43.8 oz. 47. %

called th e "static margin ." For th esame setup, mo ving th e CG aftwould reduce th is static margin(and, thus, th e inherent lon gitudi­nal stability) until a condition ofneutral stability is reached whe nthe CG and NP coincide . Furthermovement of the CG aftward tobehi nd NP would result in seriouslongitudina l instability.

The NP's position governs themargin of stabil ity available (staticmargin, or distance between CG

POWER: Spinner, prop, engine, muffler, enginemount, fuel tank, fuel cowl (3 oz.), fuel tubing, nuts and bolts.

CONTROL: Receiver (6-channel), 700mAh battery,five 5148servos, switch, two extension cables, foam­rubber protection for receiverand battery.

TRICYCLELANDINGGEAR: 2-inch-diameter wheels,51.l2-inch-diametermusic-wire legs, fairings, nose-wheelbracket and steeringarm, nuts and bolts.

FIXED SUBTOTALS

TOTAL WEIGHTS

WEIGHT (gross per square inchof wing area):

foils. This will be tran slated as anincrease (or decrease) in th e totallift at th e NP. The system is longi­tudinally stab le if th is change in liftproduces a correc ting effect, whichit will if the NP is beh ind th e CG. Anose-up disturbance inc reasing liftwould apply this lift inc rease at th eNP, behind the CG, causing th enose to drop and vice versa.

The degree of inhe rent stability isgoverned by ' th e distan ce betweenth e CG and the NP aft of it. It's

:zs THE BASICS OF RIC MODEL AIRC RAFT DESIGN

26.35 28.7%

15.0 16.3%

7.0 7.6%

48.5 oz. .52%

92.3 oz. 100%

92.3/600 =0.1538oz./Sq. in.

and NP). It's also the farthest aftpos ition possib le for th e CG whilestill avoiding instability.

Calculation of the NP's preciselocation is very complex. There areman y factors involved:

• tailplane efficiency;

• areas of wing and tailplane;

• distance between wing's and tail'saerodynamic centers;

• slopes of th e respective airfoil 'slift curves;

• fuselage area distribution in planview;

• downwash varia tions; and

• the many effects of th e pro­peller's rota tion.

Full-scale practice is to calculate theNP's approxima te posit ion andthen to fina lize its precise locationby wind-tunnel tests and/or byactual flight tests at increasinglyrearwa rd CGs.

For practical model design pur­poses, the "power-on " NP is locatedat 35 percent of MAC from its lead­ing edge. The "power-off" NPmoves a few percentage pointsfart her aft, so tha t a mo del is morestab le in an "engine-idling" glide .

With CG at 25 percent MAC andNP at 35 percent, th ere's a healthystability margin of 10 percent. Theminimum suggested stability mar­gin is 5 percen t, or a CG of 30percent MAC.

Locating the CG farther aft, sayat 33 percen t MAC, would be dan­gerous. As fuel is consumed, th e CGmo ves back an d could easily reacha point behind th e NP, leading topitch instab ility under power.

Pattern- ship designers recognizeth is risk an d position their fueltanks on the model's CG. As fuel isconsu med, the CG does not sh ift.Engine-dr iven pumps force the fuelto the carburetor.

These designers use symmetricalwing airfoils (with lower CL max )because of their little or no pitch­ing mo ments and aft CGs close tothe NP. A sma ll tailplane uploadbalances th e aft CG. The result is ah ighl y maneuverable model-but

CG Location ... CHAPTER 6

Using the techniques describedin this chapter, the Swift 'sCG was rightonthemoney. Noballast was needed.

fashioned from pushrods and bell­cran ks. With this setup, radio/i nter­ference hasn 't been an issue for atleast 10 models.

As the ph oto of th e Swift's wingclearly illustrates, the wing centersection is open ahead of the mainspar and behind the aft spar. Thishe lps in providing access.

This aut hor makes the followingsuggestions for the installation ofthe cont rol components:

• Positio n the receiver aft so that itand the antenna are away from thewiring to th e servos-and keep th eantenna as far away from the con­tro l cables as possible.

• Position engine, rudder and ele­vator servos close behind the tank.

Side viewof theSwiftplanwith power, control andlanding­gearcomponents. The balance-line fulcrumis inposition at thelower center. (I used a triangular draftsman's scale asa ful­crum, buta spare piece of3A-inch balsa triangle stock wouldalsowork well.)

one that must be constantly"flown," demand ing in tense con­centration from its pilot.

Since th e stability is close to neu­tral, any distur bance will divert th emodel from its flight path, but th eaircraft will not seek to return to itsorigina l course volunta rily, as a pos­itively stable model would.

IN THE WORKSHOPYou have designed and bu ilt yourvery own model airplane . Wisely,before you go out to th e flyingfield, you decide to check th e ph ys­icallocat ion of your model's CG. Toyour disma y, you find it's well awayfrom its design location . You arenot alone; it has happened to oth­ers, including thi s author.

To correct th is situatio n, you' llfind th at you do n't have as muchflexibility in rearrang ing things asyou might think. Your eng ine , fueltank and servos are in fixed loca­tions. The onl y items that are read­ily moveable are the receiver andbatte ry.

SERVO INSTALLATIONAND CCiQuestions of CG inevi tably lead toa consideration of the arrange mentof in ternal components and link­ages. Bitter experience indicatesthat wiring from servos to receivershould be kept well away from bothreceiver an d an tenna to avoid radioin terference. This author dislikesdowel pushrods from servos to rud­der and elevator, and wire pushrods

plus bellcranks for ailerons andflaps . Such installations requir ethat rudde r and elevator servos belocated near the wing trailing edgeand tha t th e fuselage be "open"interna lly back to th e tail surfaces.In addition, th ey vibrate heavilywhen th e engine is running, doingboth servos and control surfaces nogood. Bellcran ks lead to "slop" atth e contro l surfaces.

Stranded steel cables running inplastic tubing permit the fuselageservos to be moved forward for easyaccess; the cables are run down th einside walls of the fuse lage, orth rough th e wing ribs, out of th eway, and permit direct "no-slop"linka ge between servos and controlsurfaces. No bellcranks are needed;cables do not vibrate as do linkages

• Position servos for ailerons andflaps in the open wing center sec­tion, between the main and aft spar.

• The receiver's battery sho uld belocated so that "major surge ry"isn 't requ ired for its removal andreplacement.

• Finally, all in-fuselage and in­wing equipment should be readilyaccessible.

These objectives have been realizedin the Swift . The front topof the fuselage is rem oved byunscrewing one bolt. Similarly, th elower engine cowl is even easier toremove . All components are readilyaccessible for adjustment, replace­ment or any other reason. The tan k

is fueled with thefuse lage top "off."Straigh tening the nosegear after a hard land­ing is easy (you simplyunscrew the steeringarm setsc rew an dremove th e gear).

Getting back to yournew design; if you areunable to relocate youractua l CG to where youwant it, your onlyrecourse is to add bal­last, either up front fortail -heaviness-or aftfor nose-heaviness.Lead shot, lightly coat­ed with epoxy ordissolved cellulose

THE BASICS OF RIC MODEL AIRCRAFT DESIGN :zg

CHAPTER 6 It. THE BASICS OF RIC MODEL AIRCRAFT DESIGN

• Gat he r all the fixed-weightcomponents that you possess. Forthose you don't have, make "dum­mies" of th e same weight. Yourscale is used here. Expired AA, Cand D batt eries, lead shot, fishingsin kers, etc., are useful for"dummy" purposes.

• Similarly, make dummies for eachof the variable weight items andwin g, fuselage and tail surfaces,both horizontal and vertical.

It may be used on any confi gura­tion, con ven tional, canard, flyingboat, etc. Used for the Seagull 1Ilflying boat during the design stage,th e balancing act resulted in mov­ing the engine nacelle forward 2inches; its weight of 31 ouncescompensated for a substantial tailheaviness. On completion, thismodel required no ballast. Timespent on th e balancing act avoidedmajor and difficult modificationsto the finished model-or additionof a substantial weight of ballast upfront .Here are the steps needed:

off, weighs only 5 ounces.Tank sizes are nominal, influid ounces, wh ich is ameasure of volume, notweight. Use your scale toweigh the tank, bothempty and full. The dif­ference is fuel weight!

A scale is essential forgood design. The authoruses an old beam scale,but th e type used forweighing ingredients incooking is available at lowcost. It is recommendedthat you use one witha lO-pound capacity­graduated in pou nds ,

ounces and ounce fractions.

THE BALANCINCi ACTConcern with correctly locating th eactual , physical CG durin g thedesign process lead to developmen tof th e technique that I refer to asth e "balancing act ." This pro cedurehas been used successfully on man ymodels-and th e resulting CG'sphysical and design location s coin­cided or were very close.

The Swift's wingis bolted inposition. Note thatall components remain accessible.

cements (like Sigment or Ambroid),may be stuffed into convenient cor­ners and is self-adhering. Having toadd much ballast isn't good designpractice, how ever. Added weightdoesn 't improve the model'sperformance.

• landing-gear components.

WEICiHT ANALYSISIn Chapter 5, "Wing Design," ananalysis revealed that over 50 per­cent of the Swift's gross weight wascomposed of three groups of items offixed weight:

Once selected, these are items overwhich the designer has no weightcontrol; the engine is an exam ple.If you don't alread y have th esecomponents on hand, their indi ­vidual weight s are easily obtained.

Don 't be fooled by the tank size.The fuel in an 8-ounce tank, topped

The balance beam is onthe fulcrumand theweight-at theshortend- is positioned so that beam andweight (a drafts­man's "duck") balance onthe fulcrum.

• power components and fuel;

• control components; and

All the actual anddummy, fixed and variable weights in position-andagain thebalance beamis level. The actual and design CGs nowcoincide.

c~

l A N K ySE RVO{ " .

ING co' ;'~U'S f:'" cc \/Efn . 'A.\\.. cc,

i ~ Ec.

<,' B A,T l .

• Position th e CGs of the variabl e­weight items as follows:- wing with flaps: 50 percent MAC- wing without flaps: 40 percentMAC-horizontal ta il: 40 percent MAC- vertical ta il: "eyeball" the CG-fuselage: normally 40 percent ofth e distance from engine bul kheadto rudder post. (Because of th econcave aft contours of the Swift'sfuselage, this was adva nced to 35percent.)

30 THE BASICS OF RIC MODEL AIRCRAFT DESIGN

The Swiff's fuselage is designed for easy access.

CG Location .... CHAPTER 6

Seagull III. The originaldesign hadtheengine nacelle farther back. The "balanc­ing act" indicated that it was tail heavy.The nacelle was moved forward 2 inches;noballast was needed when the modelwas completed.

• Draw a side view, full-scale, ofyour design showing the positi on sof your fixed-weight items. Showyour de sign 's CG clea rly-butdon't detail any internal structure .

• Locate and identify the CGs ofyour variab le-weight items-wing,fuselage and horizontal and verticaltails. Draw vertical lines from th eirCGs to th e board th at will be usedas a balance beam .

• Place a fulcrum, e.g., a spare pieceof 314-inch balsa angle stock, onyour worktable. The fulc rumshould be vertically in line with themodel's CG.

• Place th e "balancing beam" onth e fulcru m and weight the shortend so th at th e beam is balancedon th e fulcrum .

• Carefully posit ion th e fixed andvariable weigh ts, actual compo­nen ts and/o r dummies in theirrespecti ve position s, ver ticallybelow their design positions.

If balance is achieved-good. Ifth e beam tilt s down at the tailend, your design is tail heavy.Slight forward movement ofpower components, nosewheelunit and possibly fuselage servosshould achieve balance. Measurethe distance of this forward move,and elongate the design 's fuselageaccordingly.

lf the beam tilts down at thefront, your design is nose heavy.The best solu tion is to move thedesign's wing forward .

Carefully move the beam and itsweigh ts backward-then movewing, wing servo and landing gear(or dummies) forward to the origi­nal positions relative to your sideview. Some trial -and-error move­ment will achieve balance. The dis­tance the beam is moved backwardwill indicate the distance the wingmust be moved forward to get theactual and design CGs to coincide.

Now that the posit ions of all thecomponents have been establishedfor the correct CG, mark your draw­ing acco rdingly. The fuselage

internal structure then may bedetailed. (See Chapter 13,"Stressed Skin Design .")

The balancing act is not too time­consuming, is certainly dependenton reasonably accurate weight esti­mates for the variable weight itemsand has proven itself to be a valu­able design tool. Having to addgobs of weight, fore or aft, to yourmodel to pin down that elusive CGto its design location is no t goodengineering. The balancing act willsurely reduce the amount of weightneeded, if it doesn't eliminate itentirely. ...

The Swan canard, flaps extended onIts cradle. Twelve ounces of ballastwere needed-andprovidedfor-as a resultof using the "bafancing act."

THE BASICS OF RIC MODEL AIRCRAFT DESIGN 31

Chapter J

1.4

1.2

Figure 1.Polar curves fora flat-plate airfoil at lowReynolds numbers.

R=420,000

.-----m:ooo. .

.2

.4

.6

.8

1.0

ward direction called "downwash."(See Chapter 8, "Horizontal TailIncidence and Downwash Estimat­ing," for furthe r discussion.)

Obviously, no self-respec tinghorizontal tail should find itselflocated in this very disturbed wake.

The angle of the downwashdepends on th e lift coefficient atwhich the wing is flying . An air­plane has many level flight speeds,from just above th e sta ll at lowengine rpm to its maximum speedat full throttle.

At low speed, the wing's angle ofattack must increase, as does its liftcoefficient, and the downwash angleis high. At top speed, the reverse istrue, and the downwash angle is low.

At low speed, the hori zontal tail'sdownward lift must be increased toforce th e wing's airfoil to a higherAoA. Part of this download is sup­plied by th e increase in the down­wash angle. At high speed, the tail 'sdownload must be reduced to lowerth e wing's AoA- but again, sinceth e downwash angle is reduced, th etail download is reduced.

The point of all thi s is that as themodel's level flight speed varieswith the throttle setting from low

reflexed airfoils have no pitchingmom ent. Symmetrical sections arepopular for aeroba tics; they flyequally well upr igh t or inve rted.Reflexed sectio ns are used on tail­less models.

A mid - or sho ulder-wing locationperm its th e centers of lift, drag,thrust and gravity to be closer toeach other. This, in turn, min imizesth e imbalance of forces th at fre­quently oppose one another.

The horizontal tail supplies th ebalancing force to offset the netresult of all th ese forces, and itschord line mus t be at an angle toth e downwash that provides eitherth e upward load or (most often) thedownload requ ired.

• Center of drag. A high-wingmodel has its center of drag abovethe CG. A nose-up reaction occurs.A low-wing model reverses thisreaction.

• Upwash and down wash.Upwash origina ting ahead of th ewing strikes both propeller disk andfuselage at an angle , ahead of th ewing , and th is causes a nose -upreac tio n . Down wash from thewing's trail ing edge strikes both theaft fuselage and th e horizontal taildownward, and thi s also causes anose-up reaction .

• Thrust line. A thrust line abovethe CG causes a nose-down reac­tion. If it is below th e CG, a nose­up reaction result s.

WAKE AND DOWNWASHThe tail surfaces of a conventional,rear-tailed airplane operate in a verydisturbed atmosphere. The airsweeps downward off the wing 'strailing edge as the result of the liftgenerated. This airstream is calledthe "wake." This wake is turbulent,and it influences the air- bothabove and below itself-in a down-

FORCES AT WORKAn airplane in steady level flight isa remarkable "balancing act ." Liftmust equ al th e model's weight;forces causing the model to nosedown must exac tly equal forcescausing a nose-up reaction ; thrustmust equa l drag.

What are these forces?

• Pitching moment. The pitchingmoment of semisy m metrical orflat-bottomed airfoils causes the air­craft to nose down . Symmetrica l or

• CG placement. A CG ahead ofthe wing 's center of lift causes anose-down reaction. Behind thewing's center of lift, a nose-upaction takes place. A CG verticallyin line with the Wing's aerodynamiccenter, i.e., at approximately 25percen t of th e MAC, exerts no nose­up or nose-down force.

T he design of an airplane'shorizontal tail surface raisesmany questions. What area

should it have? How far behind thewing should it be located? Whereshould the tail be located vertically,relative to th e wing? What ang le ofincide nce shou ld it have? What air­foil? What proportion of its areashould the elevators have? Andwhat type of construction shouldbe used? This chapter will answerth ese question s.

Horizontal Tail

Design

32 THE BASICS OF RIC MOD ELAIRCRAFT DESIGN

Horizontal Tail Design ... CHAPTER 7

Lift

• airfoil section; and

• area and tail moment arm;

TAIL AIRFOIL SECTIONSSince th e hori zon tal tail surfacehas to pro vide lift-both up anddown-sym metrical airfoils suchas Eppler E168 are recommended .Many mod el s in co rporat e flatbalsa sheet or flat built-up tailsurfaces. These are less effect ive,aerodynamically, than symmetri­cal airfoil s.

Figure 1 shows polar curves (CLversus Co) for a flat plate airfoil atlow Rns. Lift is greater, and drag isless for E168.

As explained in Chapter 13,"Stressed Skin Design ," symmetricaltail surfaces may be made lighteran d stronger than shee t balsa andmuch stronger th an built-up sur­faces (and only slightly heavier).

TAIL ASPECT RAT IOSThe upper portion of Figure 2 illus­trates the effect of AR on lift andAoA. For AR 5, the stall occurs at a20-degree AoA, and at AR 2.5, thestall is at 27 degrees-both at a liftcoefficient of 1.2. Thus, at AR 5, thetail surface responds more quickly tochanges in AoA th an at AR2.5 sincethe lift per degree of AoA is greater.

For sma ller models, however, th etail 's ch ord sho uld not be less than5 inches to avo id unfavorable lowRn effects. An AR of 4 to 5 withconstant cho rd is recommen ded .

• Th e down was h angle alsoincreases in proportion to th e liftincrease from th e lowered flaps.Thi s increases th e horizon tal taildownload.

• The wing's nose-down pitc hi ngmoment increases sha rply.

Experience with th e Seagull III, th eSeahawk and th e Swift indicates th atthe flap chord (in percen t of th ewing's chor d) influ ences th e model'sflaps-down beh avior.

Flaps with wider chords-up to 30percent of th e wing's cho rd- gene r­ate very little pitch cha nge whenextended. The increase in tail down-

• Both lift and drag increase sub­stantially, and th e model's speeddec reases.

SLOTTED FLAP EFFECTWhen slotted flap s are full yextended, several things occur :

25

.1 .15Wing drag coellicienl Co

4 -=tT~R=1 8 , AR.9 ARcS AR.2 .S-2 Basicsection f,'i / . F ' .o AR=lnfinite-.e.- "/ .>

.. I

8 l5. / ...- I

//

I6 (NOSWEEPBACI I~4

/ /

V,t;" -' I I I I

2/1v '

I I I I0 27

where HTA =horizontal-tail area insquare inches;

TMA=tail-moment arm in inches;WA =wing area in square inches;MAC =Wing's mean aerodynamic

chord in inches.

Based on experience, this authoruses a simp le meth od for establish­ing the horizontal-tail area (HTA). Ifyou have a wing AR of 6 and a tail­moment arm that is 2.5 times thewing's MAC, th en a tail area of 20percent of th e wing area is adequate.

Here is the formula:

ho rizontal tail surface depends onth ree factors:

HTA = 2.5 x MAC x 20% x WATMA

=::; .

1.

) .~ 1 .c..;g -a; .3

Figure2.Effect ofaspect ratio onwingcharacteristics.

• aspect ratio.

AREA ANDTAIL·MOMENT ARMThe tail -mo ment arm (TMA) is thedistance between the mean aerody­namic chords of the wing an d tail.It is, in effect, th e lever on whichthe tail's area wor ks.

For short TMAs, this formulawill increase th e tail area ; for longTMAs, area is reduced , but wha taerodynamicists call "tail vo lume, "i.e., area times TMA, will remainconstant.

• It is out of th e fuselage's bound­ary layer.

For high -wing models, a low-sethorizontal tail brings it well belo wth e wake .

In addition to its verticallocation, the effectiveness of th e

to high-or vice versa- the hori­zontal tail 's lift mu st vary accord­ingl y. On model airpl an es, th is isacco mplishe d by changing thean gle of th e elevators. This angl e iscontrolled by the elevator tr imlever on th e transmitter-literallyat one's fin gertips (a little up­elevator at low speed and somedown for high speed ).

The an gle of incidence of th efixed portion of th e horizontal tail ,i.e., th e stabilizer, is important butno t too critical. For semi symmetri­cal or flat-b ottomed wing airfoils ,an angle of incidence of minus 1deg ree (as measured against theda tum lin e) is appropriate . Forsymmetrical wing airfoils, an angl eof incidence of zero degre es is sug­gested. There are some exceptionsto these rules, as you will see.

VERTICAL LOCATIONOF THE HORIZONTAL TAILIn addition to th e downward deflec­tion of th e air by the wing , result ingfrom its production of lift, both pro­file and induced drags "pull" the airalong with th e wing, so that by th etime it reaches th e tail, it has lostsome of its velocity. (This is easier tovisualize if one con siders the air­plan e fixed with the air passing atlevel flight speed, as in a wind tun­nel.) This reduction adversely affectsth e tail's effectiveness.

The greater th e vertical distancebetween th e Wing's wake and th ehor izontal tail , th e smaller (flatter)the downwash angle is and the lessthe reduction in velocity of the airis. A T-tail location, atop th e verti­cal tail surface, raises it well abov eth e win g's wake and puts it in lessdisturbed air.

Other T-tail advantages are:

• It does not blanket th e rudder, forbetter spin recovery.

• The elevato r may be situatedabove th e prop slipstream .

THE BASICS OF RIC MODEL AIRCRAFT DESIGN 33

CHAPTER 7 ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

In a tight turn at h igh speed, cen­trifu gal force increases the winglift and the weight at the CGahea d of the wing's aerodynamiccen ter. A force couple results thatresists the turn. Th is imposes ahe avy addit ional load on the hor­izontal tail th at , even with full up ­elevato r, it ma y be unable tosuppo rt- and it stalls-limitingthe model's maneuverability.

For a CG vertically in line withthe Wing 's center of lift , these

• In ground effect, and particular­ly for a flapped model, more pow­erful tail downlift is needed toraise th e model' s nose for a flaps­down landing. This is more pro­nounced for wings using cam­bered , i.e. , semisymmetrical orflat-bottomed, airfoils owing tothe Wing's nose-down pitchingmoment. For sym met rical-wingairfoils, the tail download needo n ly balance the nose-downmoment of the forw ard CG andthe nose-down pitch from theextended flaps .

FORWARD CGSee Figure 6. A CG ahead of thewing 's aerodynamic center ha s onlyone advantage: it improves longitu­din al stability, since it increases the"stability margin." (See Chapter 6,"CG Location .") A forward CG hasth ese consequences:

• The model's maneuverability isreduced , particularly when cen­trifugal force comes into play.(More on th is subject further on. )

• With respect to an y maneuverinvolving centrifugal force (andthere are few that don 't ), thatforce acts at the CG and also sub­stan tially increases the load thewing must support. (See Chapter4, "Wing Loading Design .").

• The forward CG should be nofarthe r forward than a point 16 per­cent of the MAC, i.e., measured aftof th e lead ing edge.

• The tail download to balance theforward CG adds to the load thewing mu st support, in addition tothe model 's weight. Profile andinduced drags (called "trim drag")of both wing and tail increase.

level, "ground effect" occurs. Whena plane is in grou nd effect:

Lowering flaps causes an increase inthe downwash angle and in thenose-down pitch; but th e severedownwash angle reduction , du e togro und effect, reduces th e tail 'sdow nlo ad, causing the model tonose-down in a sha llow dive. Thisis part icularly noticeable for modelswith wide-chord (up to 30 percentof th e Wing's cho rd) slotte d flaps .

This beh avior requi res consider­able up-elevator force to stop thedive and to raise th e aircraft's noseto th e near- stall touchdownposture.

CG LOCATIONSThe optimum CG is vertically inlin e with th e wing 's aero dy namiccenter at 2S percent of its MAC.

There are, however, advantagesand di sadvantages in h eren t inpositioning the CG ahead of orbehind the Wing's aerodynamiccen te r.

ELEVATOR EFFECTIVENESSThe larger th e elevator area , inproportion to the horizontal ta il'stot al area, th e more effective theelevato r, as shown in Figure S.

For slotted flapped mode ls, anelev ator area of 40 percent of thehori zontal tail's area is suggested.This prop ortion provides adequa teelevator author ity to achi eve near­fu ll-st all land ings, with fla psextended and in gro und effect.

Wit hout flaps , a proportion of30 to 3S percent is adequate.

Full eleva to r deflection of 2Sdegrees, both up and down , isappropriate. Th is may, at first,prove sensitive but , with practice,has proven to be no problem . Athigh speeds, elevator low dual rateis suggested.

• The most importan t change is asevere reduction in the downwashangle to about half its value athigher altitude .

• The induced drag of the wingdecreases (see Figure 4).

• The wing behaves as though ithad a higher AR; lift increases andth e sta ll AoA decreases (see Figures2 and 3).

0.1 0.2 0.3 0.4 0.5SelSl

...__v »>

/'->

-:I

/

I VI -:

-:V

I ~ ::::::: :;::I ~t _ 10 z>~, :--- 8

'.--' ~ ~ ...--

Wing Aspecl Ratio I./

.2

Q

~

>~ 0.05

~

:i 0.6

Flgure40"Impact ofgrou,,~ effectpn induced drag -

:c 1.0~

i ::c; 0.7

.6

.4

.8

0.70.05 0.1 0.2 0.3 0.4 0.5

Height 0' Wing From G,ound/Wingsp,n

" ~ 0.5c

~_! :::0.05 0.1 0.2 0.3 0.4 0.5

Height0' Wing FromG,ound/Wlngsp,n

1.0

GROUND EFFECTWh en an airplane is on fina lapproach and descends to half itswingspan above ground (or water)

load th at tends to cause a nose­up react ion is equalized by theWing's higher nose-down pitchingmom en t . It is very satisfying tolower full flap, after th rott ling backand have the model continue on itsmerry way, with out nosi ng up ordown , but flying noticeably slower.

For narrower chord (2S percent)flaps, the flap-induced tail downloadis greater th an the nose-down wingpitching moment. When th e flapsare extended, this causes the modelto nose up sha rply and ratheralarmingly.

34 THE BASICS OF RIC MODEL AIRC RAFT DESIGN

Horizontal Tail Design A CHAPTER 7

This reduced efficiency affects theNP locati on . It acts like a reduct ionin tail area : it moves the NP for­ward and reduces the static ma r­gin. The larger the vertica l separa­tio n between wing and tail, thebetter. For models whose wing ison or in the middle of the fuse lage,a 'l-tall is best . For high wingsabove the fuselage , a low tail isindicated.

There is another aspect to allthis. For the same NP, a high, moreefficient tail may be reduced inarea, yet would have the sameeffectiveness as the lower, largertail. If made larger in area , th emore efficient hi gh er tail willmove the NP aft, thereby en larg ing

• The relat ive vertical positioningof the wing and horizontal tail has asignificant bearing on the tail 'seffectiveness, or efficiency. A taillocated close to the wing's wake, inheavy downwash, loses effective ­ness . At this location, the tail is inreduced dynamic air pressurecaused by the drag of both wing andfuselage. This redu ces that ta il'seffectiveness to un der 50 percent. Incontrast, a T-tail is 90 percen teffective.

• Similarly, a longer tail momentarm will move the NP aft .

• The relative size of the areas ofthe hori zon tal tail and wing.En larging the tail will move the NPrearward for a larger static margin .

Slatc:L(. =Slot ..

Horizontal talt-plane section , NACA 23,009, inverted and slotted .

Basic airfoil , NACA 2412, maximum Iltt coellicient 1.00at stall speed of24mph , angle of attack14degrees, Rn 183,000 and wing loadingof 24ounces persquare foot.

/<i/

R~'~~:'·~"IFlap

IIISlotted and flapped airfoil, maximum CL 2.20at stall speed of 17mph,angle of attack 24degrees, Rn 135,000 and wing loading 24 ounces persquare foot , speed reduction 7mph (29%).

Figure 8. :Crane wing andtail sections.

tio ns for stabilityand flight control.

Tail upload

Taildownload

Taildownload

NEUTRAL·POINTMANIPULATIONThe re are ways tohave both a mod­estly aft CG anda healthy stabilitymargin between th eCG and the NP. Thema jor factors in flu­enci ng th e neutralpoint's location are:

• Attempting toredu ce trim dragby moving th e CGtoo far aft cancause problem s.Th is requires anin crease in thetail's positive AoAfor equilibrium. Ina sha llow dive, th ewing's AoA andCl both decrease.Since th e down­wa rd angle ofthe down wash isprop ortional toth e wing's Cl , th edive reduces thedownwash ang le,which becomesmore nearly paral-lel with th e fuse­lage cen te rline .The tail 's AoA andlift increase, result­ing in a some-

times violent "tuck under." Soaringgliders with CGs so located have lostwings in the resulting steep dive.Moving th e CG forward and reduc­ing th e tail's AoA is th e remedy.

• This author is nervous aboutth e use of an aft CG coupled withslotted or Fowler flaps. The largeincrease in down wash angle createdby th e extended flaps could cha ngeth e tail's AoA substantially, conver t­ing a positive up-load (or mild nega-tive dow nload) to aheavy download .The combination of Ban aft CG and aheavy tail down­load might wellresult in a disas-trous stall.

Downwash

CG

CG SymmetricalPitch moment Wing lilt _

r:~-re-d---=----""'·~ :1 download

ws'nglilt NP ___:=-=---=- Down~

Pitch moment CG

/.. Wing lift -I C--,"_--,..:=--NP Downwash ....

Cambered

For symmetrical airfoils , the hori­zontal tail's airfoil is set at a positi veAoA, relative to the downwash, toproduce an upload to offset the aft­CG's nos e-up moment. The wing'stotal load and trim drag are bothreduced.

The disadvantages of an aftCG are:

• The stability margin is redu ced,which could have serious implica-

forces are directly opposed and donot add to the tail 's load.

Figure 6.Forward CG force diagrams.

• Owing to th e nos e-down pit chingmoment of a cambered airfoil, th ehorizontal tail normally has adownload requirement. The aftCG's moment about the wing'saerody namic center redu ces thistail download.

Figure 7.AffCG force diagrams.

• Maneuverability is increased­centrifugal force acting on th e aftCG actually reduces the tail loadsneeded for these maneuvers.

AFTCGSee Figure 7. A CG behind th ewing's aerodynamic center offersadvantages, but has serious poten­tial disadvantages:

THE BASICS OF RIC MODEL AIRCRAFT DESIGN 35

CHAPTER 7 ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

the static margin.Structurally, a tail in the fuselage

presents few problems. AT-tail doesimpose heavy loads on the verticalfin. If thicker symmetrical airfoil s,such as the Eppler 168 or NACA0012, are employed for the verticaltail along with stressed-skin con­struction (see Chapter 13), the finwill have adequate strength.

A simple formula for estimatingthe mo st aft CG locati on, but stillleaving an adequate static marginfor safe, controllable flight is:

.17 + (.30 x TMA x SH x HTE) x 100=MAC SW

CG location, in percentof the MAC,measured from the MAC's leading edge,

where:TMA = tail-momentarm in inchesMAC = meanaerodynamic chord in inchesSH = horizontal tail area in square inchesSW = wing area in square incnesHTE = /lOrizontal tail efficiency, estimatedat between40and 90 percent and basedon the tail'svertical location relative to thewing's wake

This formula reflects the fuselage'scontribution to the NP location.Depending on its size and shape,the neutral point can advanc e up to15 percent of the win g's MACunder th e fuselage's in flue n ce.Calculation of the fuselage's contri­bution is complex and beyond thescope of this article.

Using the Swift 's actual andimaginary modified va lues willillustrate all thi s.

ActualTMA-25.5 in .; MAC-9.75 in .;SH-120 sq. in .; SW- 600 sq . in .;HTE-90 percent.

ModifiedTMA-29.25 in .; MAC-9.75 in .;SH-150 sq. in .; SW-600 sq. in .;HTE-90 percent.

The actual mo st rearward CG isat 31 percent of the MAC. Since thedesign CG is at 25 percent MAC,there is a healthy static margin . Inthe modified version th e most rear­ward CG would be at 37 percent ofthe MAC.

36 THE BASIC S OF RIC MODEL AIRCRAFTDESIGN

Thus, th e modified versionwould also ha ve a healthy stabilitymargin with a CG at 31 percent ofthe model's MAC, well behind thewing's aerodynamic center of lift at25 percent MAC.

CAMBERED AIRFOILSECTIONSSemisymmetrical or flat-bottomedairfoil sections may be used in thehorizontal tail. They have a widerrange of AoAs before the stall and ahigher CL at the stall than symmet­rical airfoils. Where a powerful up ordownload is requ ired, such sectionsare useful. For uplift, the tail airfoil isright side up; for downlift, the airfoilis inverted. It should be noted that acambered airfoil starts to lift at anegative AoA, not zero degrees as forsymmetrical sections. The Eppler205 section and the Eppler 222section are suggested as tail airfoils(see Appendix). Note the shift tolower negative angles of zero lift atlow Rns.

An example of the need for apowerful download, in gro undeffect, is the "Crane," a short take­off and landing (STOL) model.This model had full-span , fixed ,leading-edge slots and, flapsdown, it stalled at 20 degrees AoA.After some trials, this model wasable to achieve full-stall landings.An all-moving tail with an invert­ed , cambered and leading-edge­slott ed airfoil, called a "stabila­tor, " as in Figure 8, was required ...

REFERENCES

Report 648* : Design charts tor predidingdownwash angles and wake characteristicsbehind plain and flapped airfoils. bySilverstein and Katzoff• .1939.

Report 651*: Downwash and Wake BehindPlain and Flapped Airfo ils. by Silverstein.Katzoff and Bullivant, 1939.

*Both reports are available fro m theNational Technical Information Service. 5285Port Royal Rd., Springf ield. VA 2216 1.

Chapter 8

Horizontal Tail

Incidence and

Downwash

Estimating

• Airfoil pi tching momen t.Symmetrical sections have nopitching moment; semisymmetri­cal and flat-bottom airfoils havesuch moments, which are alwaysnegative, or nose-down. Their valueis calculated using Formula 10(pitching moment) in Chapter 1,"Airfoil Selection."

0.9

0.4

0.65

TaileNiciency0.9

Minus M

Plus M1 0.65i'l wing MAC

'1- - ----

• The wing's drag moment. Thewing's total of both profile andinduced drags, in ounces, at thewing 's AoA for the design cruisingspeed, is calculated using Formulas5 ("Total of profile [section] andinduced drag coefficients ") and 9("Total profile and induced wingdrag"), of Chapter 1.

The drag moment is the drag inounces multiplied by th e vertical

Tail Ve MAC

14-------Dislance X

Wake displacemenl

Wake ce lerllne

Wing l/eMAC

• Setting the tail 's incidence,relative to the downwash at the cal­culated AoA to provide th e balanc­ing moment.

MOMENT EVALUATIONThe following de ta ils the fourmajor mo ment sources. There areothe rs, which are beyond thescope of this article, but small ele­vator trim adjustments wouldcompensate for their minorvalues.

• CG location. A CG that's aheadof the wing's 1/4 MAC causes anose-down, or negative, moment.Its value is the horizontal distancebetween the CG and 1;4 MAC, ininches, multiplied by the model 'sgross weight in ounces. Havingthe CG behind the Wing's 1;4 MACcauses a positive or nose-upmoment. Its value is calculated inthe same way as for a forward CG,but it has positive value. In levelflight, a CG that's vertically in linewith the wing's lift (at 1/4 MAC)contributes neither up momentnor down moment.

A n airp lane in level flight atits selected cruising speedis a classic balancing act .

To achieve this balance, both nose­down and nose-up moments mustbe evaluated. The horizonta l tailmust balance the net result of thesemoments. (A moment is simplya weight or force multiplied bya distance-also called "arrn'") Thehorizontal tail 's AoA, relative tothe wing's downwash , should besufficient to provide the upward,or most often, the downwardlift required to provide th isequilibrium.

The penalty for having an incor­rect tail incidence is heavy elevatordeflection at cruise speed . This addsdrag and could result in a lack ofadequate elevator authority tobring the airp lane to a near-stalllan ding posture wh ile in groundeffect, with full flap deflection andwith a CG located ahead of thewing's aerodynamic lift cen ter.

Establishing the appropriate tailincidence calls for:

• An evaluation of the moments, ininch-ounces, both nose-up andnose -down to obtain the ne t result.Nose-up moments are offse t bynose-down moments;

• A determination of which type oftail lift-upward or downward-inounces is required to provide thebalancing moment at the model'sselected cruising speed .

• A calculation of the tail angleof attack required to provide thistail lift.

• An estimate of th e downwashangle at the horizontal tail's location.

Figure 1.Wake and downwashdetermination.

THE BASICS OF RIC MODEL AIRCRAFT DESIGN 'D

CHAPT ER 8 • THE BASICS OF RIC MODEL AIRCRAFT DESIGN

distance, in inches, between the CGand the wing's 1/ 4 MAC on th e air­foil's chord line. If th e wing isbelow th e CG, th e moment is nose­down, or negative. If it' s above th eCG, the mom ent is nose-up , or pos­itive . If th e wing is on the CG,it contributes 110 dr ag pitch ingmoment.

• Thrust moment. A thrust lineabove th e CG promo tes a nose­down (negative) moment. Belowthe CG, th e mom ent is nose-up(positive). The thrust, in ounces, isdifficult to pin down witho ut awind-tunnel test. An educatedguess is a thrust of 40 percent of themodel's weight for level fligh t atthe de sign cruise speed . Themoment, in inch-ounces, is theestimated thrust multiplied by thevertical distan ce in inches fromthrust line to CG. If th e thrust linepasses th rough th e CG, there is nothrust pitching moment.

• Net result. The net sum of th esefour mom ent sources will providethe balancing moment that the

38 THE BASICSOF RIC MOD EL AIRCRAFT DESIGN

horizontal tail plane must provide.Usually, the ne t result is a nose­down , or negative figure.

TAIL LIFT NEEDEDDividing the net mo ment figuregiven in th e previous sectio n by th etail 's lever (or tail moment arm­the distance from CG to th e tail's 1/4MAC in inches) will tell how muchlift, in ounces, the tail mus t developto provide the balance moment. Ifthe n-et moment is negative, ornose -down, the tail must lift down­ward. If positive, the tail lift mustbe upward.

TAIL ANGLE OF INCIDENCEThe tail lift required, in ounces,sho uld be ad justed to compensatefor the tail 's efficie ncy (or lackthereof) . See Figure 1. That ad just­ment would be: lift requ ired divid edby tail efficiency. For a T-tail wherethe lift required is 100 ounces, thiswould increase to 100 divided by0.9, or II I ounces .

To calculate the tail AoA neededto provide that lift, use Formula 7("Lift coefficient required": specia l

procedure A: "Lift coefficien t perdegree angle of attack adjusted foraspect ratio and planform" and spe­cial procedure B: "Angle of attack(or incidence) for level flight" inCha pter 1. Identify whether th eangle is positive (upward lift) ornegative (downward lift).

DOWNWASH ANGLEESTIMATINGThe first step is to determine thelocation of the wake centerline atth e tail (Figure 1) so as to obtain th ewake displacement H. With H andtwo other dimensions from yourdrawi ngs, plus (or minus) M anddistance "X," you can easily locatethe wake cen terline relative to thetail.

• Wake centerline. Factors con­trolling the wake disp lacement are

- wing aspect ratio;- wing planform; and- wing's CL at th e design cruising

speed.If a th orough design job has been

done, th e CL will hav e been dete r­mined in calculating th e wing'sangle of incidence for level flightFor more detail, see Chapter 18,"Propeller Selec tion and SpeedEstimati ng."

Refer to Figure 2, A to F. This wasextrac ted from NACA Report No.648 and is not difficult to use. First,note th at all th e dimensions aregiven as a percentage of th e wing'ssemi-span .

• The column on th e left coversth e wing planforms, both straightand tapered, for aspect ratios of 6, 9and 12. Dihedral and sweepbackmay be ignored. Select th e plan­form closest to your design.

• The center column provide s thewake displaceme nt for each of theplanforms for a CL of 1.00. Note thedecrease with increasing aspectratio. If your wingspan is 60 inches,the semi-span is 30 inches. Ifdistance X in Figure 1 equals 24inches, th en wake displacement is24 divided by 30, or 80 percent ofthe semi -span. In th e center col­um n, Figure 2A, th e wake displace­ment at 80 percent of semi-span is8 percent of the semi-span, for a CLof 1.00. If your win g's CL is, say 0.3 ,thi s displacement would be reduced

Horizontal Tail Incidence and Dow nwash Estimat ing .& CHAPTER 8

to 0.3 multiplied by 8, or 2.4 per­cent of the semi-span (distance H inFigure 1).

Now convert distance M into apercentage of the wing's semi-span.If, for your design, M equals 4inches, wake displacement is 4divided by 30, or 13.3 percent ofsemi-span. Note that M is negativefor tails below the wake centerline.

Adding distances Hand M givesthe vertical location of the hori zon­tal tail relativ e to the wake CL• Inour example, H plus M, 2.4 percentplus 13.3 percent is a total of 15.7percent, and distance X is 80percent of the wing semi-span.

d iffer ence between angu larsetti ngs for uprigh t camber edsectio ns and in vert ed camberedsections .

PATTERN ·SHIP DESIGNPattern ships have evolved intoconfigurations in which th e fourmajor moment sources have beenreduced to a min imum:

• The CG is on or close to th ewing 's lift cen ter (Vol MAC).

• The symmetrical airfoils have nopitchi ng moment.

• With the wing on the CG, thewing's drag moment is no nex istent.

• The thru st line passes throughth e CG. Tail surfaces are generousin area. "More is bett er" is the pre­vailing belief. Thes e large area smove th e NP aft, improving th e sta­tic margin and permit ting th e CGto be beh ind the wing's 1;4 MAC. Inman euvers, centrifugal force, actingat th e aft CG assists; the model ismore agile.

The win g and tail airfoils are bothset at zero incidence-a "no-lift"

,II

I/

Where toposition failplane

I I

'\ WAKEI ~

-I-----..I

increase in the downwash angle. Athighspeeds, the tail's download must be reducedto lower the wing's angle 01 attack; butagain, because the downwash angle isreduced , the tail download is reduced.

The point 01 all this is that asthe mcdel 'sleveillight speed varies with thethrottle set­tingtrem IDW to high-or vice versa-thehoriznntal tall's lill must vary accDrdingly.On mndet airplanes, this is acenmpllshed bychanging the angle 01 theelevators. Thisangle is controlled bythe elevator trim leveron the transmitter-literally at one's finger­tips(a little up-elevator at lowspeeds andsome down lor high speeds) .

The angle DI incidence 01 the fixed part ofthe herizuntal tail, i.e., the stabilizer, isimpertant, but nottoo critical. FDr semisym­metrical Dr flat-bDttDm-wing airfDils, anangle of incidence ofminus 1 degree (asmeasured against the datum line) is appro­priate. For symmetrical-wing airfoils, anangle 01 incidence of0 degrees is suggested.(There are same exceptions to these rules.)

\

DOW~WASHI

"

UPWASH ~ -:.

.....----- -"..,- - -:--- - - - - -.,., "---- -

Droop fuselage Rearwardandfor better streamlining downward acceleration

""~------

The downwash andwake 01 a conventional, rear-tailed aircraft.

Wake and Downwash

The tail surfaces ofa conventional, rear­tailed airplane operate In a very dis­

turbed atmosphere. As the figure Illustrates,the air sweeps downward oil the wing 's trail­ingedge asthe result 01 the lilt generated.This airstream is called the "wake." Thiswake is turbulent, and it influences the air­both above and below itself-in a downwarddirection called "downwash.II

Obviously, no sell-respectinghorizontal tailshould linditself in thisvery disturbed wake.

The downwash angle depends on the 1mcoenicient atwhich the wing is flying. An air­plane has many leveillightspeeds, from justabove the stallat lowengine rpm to its maxi­mum speed at full throttle. Atlowspeeds,the wing's angle ofattack increases, asdoesits lilt coellicient, and the downwash angle ishigh. Attop speed, the reverse is true, andthe downwash angle is low.

Atlowspeeds, the horizontal tail's down­ward lill must be increased to force thewing 's airfoil to a higher angle 01attack.Part 01 thisdownload is supplied bythe

TAIL INCIDENCEIn the example abo ve, the down­wash angle is 1.5 degrees. If thetail AoA needed for balance wereminus 2 degrees, that 2 degreeswould be relati ve to the down­wash angle. Figure 3 diagramsthis relationship and shows thatthe tail 's angle of incidence (rela­tive to the model 's centerline, forthis example) should be minus0.5 degree. CAUTION: for cam­bered airfoils, the angle of zerolift is not the chord lin e as it isfor symmetrical sections, but itcan be sever al degrees ne gat ive asshown in the airfoil plots for thesection concerned. Th is mu st beconsidered when establishing theAoA relative to the downwash .Note also that there 's a major

DOWNWASH ANGLERefer to the th ird vertical columnin Figure 2A. At 80 percent"Distance behind" and 15.7 percent"Vertically above ," the down washangle, for a CL of 1.00, is between5.4 degrees and 4.8 degrees, or 5degrees . For our CL of 0.3, thi swould be 0.3 x 5, or 1.5, degreesand is the downwash angle at thehorizontal tail's location.

In Figure 1, there is a dotted out­line of a tail below the wake center­line-the tail location for manyhigh-wing aircraft. The downwash­angle-estimating procedure appli es,but the difference is that distan ceM would be a minus figure and H apositive figure, which would redu cethe vertical displacement. Notehow the downwash an gles arereduced as the vertical displace­ment is increased.

THE BASICS OF RIC MODEL AIRCRAFT DESIGN 39

CHAPT ER 8 ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

140 16080 100 12060..

I I-n I ,I 1-1-

t---1. I-t-I- 17N r-- • I-i-~q 5< I-

oki\l,. I" t--

20

..

i60

A

,..i20

Distancebehind' , -chordpoint, pen;entsemi-span Distancebehind ' ,-chord point,percentsemi-span

B 60

2O f"-.t--t-+-"l'--4:::++-t--t-t-t-j

i

)i I I I

20 .. 60 80 100

41' 36"

60 80 160

Distance behind ',-chord point, pen;entsemi-span

Distancebehind ' s-croa point, pen;entsemi-span

c

I )~20~~k,~

Distance from C, ' oecen semi-span60 80 100 120 140 160

60 '"..

I '

80 100 120 140 16060o..

o

Do 20 40 60 80 100

Distance trom C" pen;entsemi-span Distance behind I ,-chord point, percent semi-span Distance behind I '-chordpoint, percentsemi-span

Distance behind ' ,-chord point. pen;entsemi-span Distancebehind ' , -chord POInt, percentsemi-span

1.60 16080 100 12060

1

I I -l-

e-- I -I- 1r-, -i I-I- L~ Ik·~ .. Io..

20

60

80

..

160120 1401008060

f4 t.

E

D .o 20 .cJ 60 80 tCQ

Distance from C" pen;entsemi-span

F

--- _:?1601..1201008060..

1 II,

C- I- 3

I-~ •

0

20

60

..

I I80 100

I60

I..Distancebehind ' '- chord point, pen;ent semi-span Distancebehind ' , -chord point, percentsemi-span

40 THE BASICS OF RIC MO DEL AIRCRAFT DESIGN

Horizontal Ta il Incidence and Downwash Estimating ... CHAPTER 8

condition. However, as soon as up­elevator is applied, the wing's AoAbecomes positi ve; both lift anddownwash are produced. Thatdownwash strikes the horizontaltail at a negative angle, producingtail downlift that maintains thewing at a positive lifting angle.In verted, th e same conditionsapply. In both positions, th e fuse­lage is inclined at a sligh t nos e-upan gle to provide th e win g's lift.

TAIL DEEP STALLSome authoriti es state that, at highangl es of attack, the wake from thewing may blanket the horizontalT-tail, and the airp lane will havedifficulty recoverin g from a stall.This condition is called "deep stall. "

Cases of full-scale deep stall haveresulted in fatal crashes. All haveinvolved test-flights of twin or tri-jetaircraft with aft, fuselage-mountedengines and rearward CGs for th etests. In a stalled condition, thewing and engine-pod wakes mayblanket the horizontal tail.

There are many prop- and jet­driven aircraft with T-tails th at haveno deep-stall probl ems .

RECENT DESICN ANALYSISThe following models are furtherdiscussed in Chapter 26, "Co nstruc­tion Designs."

• The Swift. The Swift's thrustline, wing drag and CG are in line,and the CG is vertically in lin ewith the 1/:! MAC of the wing. Theonly significan t moment is theresult of th e win g's airfoil pitchingmoment. At 60mph cruis e speed, atail setti ng of minus 1 degreeproved to be correct.

• Seagull III flying boat. Thi smodel had two major nose-downmoments: th e high thrust line andthe wing's airfoil pitching mo ment.Centers of lift and drag coincidedwith the CG. The pusher enginearrangement was chosen so thatth e horizontal tail would be partlysubme rged in th e powerful propslipstream in th e hope th at pit chchanges caus ed by power (rpm)variations would be minimized.Luckily, th is was successful; themodel exhibits no change in pitchas rpm are varied.

Downwash angle

Downwash

L..- Tall angle otlncldence

Ta ll angle of anack relative to downwash

Figure 3.Tailplane angle of Incidence.

• Swan canard. The nose-dow npitc h of the h igh thrust line is off­set by th e aft Wing 's drag moment.Pitching moments of both fore andaft wings add to the foreplane'sload. The foreplane downwas hreduces th e wing's AoA and lift inthe area shadowed by the fore­plane. The wing's AoA in th is areawas increased to compensate.

• Seahawk float and tricycle gear.Here , the major nose-downmoments are caused by th e wing'sdrag, below the CG, and the wing' sairfo il pit ch ing moment . A thrustline above the CG adds to the nose­down moment. The 1;4 MAC is ver­tically in line with the CG and pro­duces no moments in level fligh t.

• Osprey tail-dragger and twinfloat plane. The major moment scaused by wing drag and the wing'sairfoil pitching moment opposeeach other. Thrust line an d CGcoincide, and th e latt er is vertica llyin line with the 1;4 MAC in levelflight ....

Horizontal

THE BASICS OF RIC MODEL AIRCRAFT DESIGN 41

Chapter 9

model airplane is concentrated intwo elements, one representing themass ahead of the CG and th eother, the mass behind the CG.There are thus two principal axissystems to consider:

• the inertia axis through the CG,joining the two element masses (seeFigure 2).

• the aerodynamic, or wind, axis,through the CG, in the relativewind direction; and

If, in level flight, the aerodynamicand inertial axes are aligned, noinertial coupling will result fromrolling motion.

If, however, the inert ial axis isinclined to the aerod ynamic axis, asin Grant's th eory, rotation aboutth e aerod ynamic axis will createcentrifugal forces th at , through the

Figure 2.Side viewof "Cloud-Niner" showing estimated aerodynamic andinertialaxis.

Airplane Design. " "Lateral area "refers to aircraft surface areas thatface sideways. This th eory, in a nut­shell, states that if:

• th e model's CLA was at about 2Spercent of the tail moment arm aftof th e CG;

then the model would be spirallystable. Figure 1 illust rates the lay­out required by this theory.

Put in to pract ice by manymodelers, this theory was proventime and time again and wasapplied in th e early days of rudd er­on ly RIC by suchwell-know n andrespec ted modelersas Hal deBolt and BillWinte r. The latter'sbeautiful "Cloud­Niner" (outlined inFigure 1) still reflectsCharlie Grant's ideas.

Today, with th every precise, power­ful and reliabl econtrol provided by Figure 1.modern RIC equ ip- Side viewof "Cloud·Nlner" with estimated CG andCLA locations.ment, which permit sunlimited aerobatics,this theory is lessimportant , but none­theless valid.

• the win g joining the front CLAand the rear CLA sloped upward tothe front ,

INERTIAL ROLLCOUPLINGThis autho r surmisesthat inertial couplingin rolling plays as biga part as side areasin underst and ingGrant's CLA ideas.

The ma ss of a

• a line through CLA and CG washorizontal; and

Design and

CENTER OF LATERALAREA CONCEPTIn 1941, Charles Hamp ton Grant,then editor of Model Airplane News,publi shed his cen ter of lateral area(CLA) th eories, in his book "Model

V ertical tail desig n is mo recomplex than one mightimagine. It involves con­

sideration of wing dihedral, fuse­lage and landing-gear side areas,CG locati on and the importan tvertical tail area .

A brief summary of model air­plane history is timely. In the 1930s,models were light, tissue-coveredand rubber-band powered. To flyproperly, they depended solely ontheir inherent stability.

The small, single-cylinder gasolineengine, such as the Brown jr., with itsfuel tank, ignition coil, condenserand battery, revolution ized modelaviation. Gas models were bigger,heavier and flew faster and longer.They still depended on inherent sta­bility to avoid damaging crashes.Radio control was still ahead.

Early RIC "rudder-only" modelsstill relied on the model's inherentstability. Rudder control really only"steered" the mo del. It becam eappa rent that there was a seriousspira l instability problem. Modelswere spiral-diving into the ground .

Vertical Tail

Spiral Stability

42 THE BASICS OFRIC MODEL AIRCRAFT DESIGN

Vertical Tail Design and Spiral Stability ... CHAPTER 9

profile to reflect the additional pairof surfaces. Add the necessary lay­ers of cardboa rd as shown cross­hatched in Figure 3. Note that forthis configura tion, at the wing ,three layers would be needed, twofor the wings' side areas (becauseof dih edral, each wing has a leftand right "lateral surface" com­prised of the vertical rise in thewing , as seen from the side) andon e for the canopy outline .

Size your vertical tail sur face toan area th at looks righ t. You' llsoo n find out how accura te yourestimate was.

To locate the CLAof th is profile,simply establish its CG. It is easilydone by inse rting a pin throughth e profil e at pinhole no. I , inFigure 3; push the pin into somevert ica l surface , door jamb, oredge , and allow the profile tohan g free under gravity. Make aloo p at one end of a 3-foo t lengthof string, and slip it over the pin­head; to the other end of thestring, tie a small weight, e.g., anut, key, or paperclip. Allow it,too, to han g free under grav ity.

The profile's CG will be some­where along the thread line ; markthis line on th e profile. Repeat thisprocedure from another point ,somew ha t distant from pinho le 1.In Figure 3, th is is shown as pin­hole 2.

Where the two thread linesinte rsect is the cardboard profile'sCG and your model's first CLA.The CG (and CLA) will not, in allprobability, be at 2S percent ofTMA; redu ce or add to your verti­cal surfa ce area until it does. Youmay have to repea t this processseveral times to get th e righ t tailarea/Cl.A relation ship-unless youare smarte r than th is au thor(which cou ld well be!).

Figure 4.Blanketing of thevertical tall In a spin, asaffected bythepositionof thehorizontal tall. Notice the absence of blanketing In a T-tallconfiguration (0).

DIHEDRALWith today'smodern radiocontrol an dailerons, thehigh dihedralangles th atwere built intofree-flight andrud de r-o nlymodels are nolonger needed.For poweredR/C mod elswith ailerons,th e followingdihedral angles

(25to 28percent01 distance Irom CGto vertical MAC)

are suggested:

LOCATING THE CLAThe following procedure has beenused by this author for many yearsand on ma ny models-all success­ful fliers-to determine vertical tailarea . It is applica ble to all configu­ratio ns, flying boat s, canards, float­planes, etc.

In his "full-scale" book, "TheDesign of th e Aeroplane ," Britishaerodyna micist and author DarrolStin ton recommend s a very similarprocedure.

Cut out a cardboard profile ofyour desig n, full size, tha t repre­sen ts the late ral surfaces of the air­craft. For two lateral surfaces, e.g. ,for the right and left sides of thefuselage , a sing le cardboard profilecutout will su f-fice. If there aremor e than twostacked lateralsur faces (viewingthe plane fromthe side), e.g., thewing's d ih edral,landing -gea r orvertical-tail sur -faces, an addi tio n­al piece of card­board will have tobe layered on th e

Sweepback also has a dihedraleffect. Two to 3 degrees of sweep­back is equivalent to 1 degree ofdihedral.

• low wing-4 degrees

• high wing-2 degrees

• mid wing-3 degrees

Pinhole 2Vertical surtace map

Distance D

Rear CLA

Figure 3.The center of lateralarea (CLA) relative to theCG.

The Wasp was another .15-poweredmodel-a tandem-wing biplane with4degrees of dihedral oneach wing. The CLAwas originally at 25 percent of the VTMA,butowing to doubts about theforward fuse­lage's Impact ondirectional stability, thevertical tall area was Increased to bring theCLA to30percent. The Wasp was spirallyunstable andunpleasant to fly. Cutting offthefin tops to therudder toplevels (flnoto­my!)andaddingsmallstreamlined capsImproved thespiralstabilityandthemodel'sbehavior. The CLA was then back to25percent of VTMA as originally planned.

actio n of inertial forces, cause apitching mom ent. This is "inertialroll coupling" (see Figure 2).

Sin ce th e inertial axis slopesupward to th e fron t, a nose-up pitchwill occur when the model rolls.This prevents the fatal spiral dive.This type of spiral stability is greatfor sport models, but th e inerti alcoupling must inhibit any maneu­ver where rolling is involved.

Look at the side view of theauthor's Swift (see Cha pter 26,"Construction Designs") . The CLAis at 2S percen t of th e tail-momentarm , as per Grant, but th e position sof th e two element masses makethe aerodynamic and inertial axesalmost coincide. In rolling, noiner­tia coupling (tha t could in terferewith aerobatics) will occur. Patt ernmodels have similar configurations.

THE BASICS OF RIC MOD EL AIRCRAFTDESIGN 43

CHAPTER 9 ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

turning up to 270 degrees of its cir­cular path, it is spirally stable. Therap idity with which it rights itselfis a measure of its degree of spiralstability.

• Neutrally sp irally stable . If itcontinues to turn without the angleof bank increasing, it is neutrallyspirally stable.

• Spira lly un stable. If the angle ofits bank slowly increases as it turnsand its speed gradually increases ina descending spiral, it is spirallyunstable. The rap idity with whichit increases its bank angle is anindex of its degree of instability.

impression that its designer knowswhat it is all about!

A dorsal fin area of 10 percent ofvertica l tail area is suggested .

ALL·MOVINGHORIZONTAL T·TAILSFigure 5 sketches an all-movingT-tail or "stabilator" that's suitableboth for powered models and forsailplanes; for the latter, massbalancing may not be required ifth e glider is in tended to fly at a rel­ative ly low speed. A "T" stabilator'sarea may be reduced 10 percentfrom th at of a conven tional stab­elevator hori zontal tail plane.

»:% chord .Jf

(HINGE LINE)

, Alternate rudderMass balance

Center portion fixed"'.' " . to top 01 vertical tail

.. ". 1 Outer partsr ". 'pivot aboutStabilator hinge lineMass balance

RudderMass balance

Figure 5.Perspective drawing of anall-movinghorizontal T-tail or stabilator.

VERTICAL·TAILASPECT RATIOThe AR of horizontal and verticaltails (and wings) bears on th eir effec­tiveness . Vertical-tail ARs of 2.5 to 3are suggested. To determine your ver­tical tail's AR, use thi s formula:

ARv = 1.55 x Bv2

Sv

whe re ARv = vertica l tail aspectratio;

Bv2 =heigh t of vertica l tail fromfuse lage bottom, in inc hes,"squared"; and

Sv = vertica l tail area in squareinches, incl ud ing fuse lage belowthe fin . •

A T-tail cap ping th e vertical tailsurface, as in the "Swift," effec­tively increases th e vertica l tail 's AReffect .

Figure 4 shows how th e horizon­ta l tail could dan gerou sly blanketthe vertica l surface in a spin. TheT-ta il in Figure 4D is not blan ketedin th is way.

DORSAL FINSThe Swift has a sma ll do rsal fin . Ithas three useful func tio ns :

• inc reases fuselage stab ility at highside slip angles;

• reduces vertical tail stalling; and

• just plain looks good!; it gives the

RUDDER POWERFor powered sport models, a rudderarea of 30 percent of the vertical tailarea, with ang ular travel of 30degrees eithe r side of neutral, is sug­gested. For sailplanes with high-ARwings and for pattern sh ips, a rud­der area up to 50 percent of thevertical tail area is recommended.

RUDDER AILERON EFFECTA rudder that has its "area-center"above a horizontal tail line throughth e CG will act like an aileron whenused. It induces a roll th at isopposed to th e rudder-forced yaw.

To avoid thi s, the rudder's areacenter sho uld come close to or fallon th e horizontal line through theCG. The portion below th e CGopposes and neutralizes the rollingaction of the portion above the CG(Figure 1), and the rudder actioncauses yaw only.

Upwardly dihedral V-tails havepro nounce d anti-yaw roll actionwhe n th e ruddervators act as rud­ders. Downwardly dihedralled(anhedralled) V-tai ls have rollingactio n in the same direction asth e yaw.

SPIRAL STABILITYTo assess an existing model air­plane's spiral stability-or lack ofit-is easy. In level flight, at themo del's normal cru ising speed andat a reasonable altitude, put it in a15- to 20-degree bank, then neu­tralize th e contro ls and watch itsbehavior closely.

• Spirally stable. If it returns tonormal level flight, upright, in

LEVELS OFSPIRAL STABILITYHigh spiral stability is needed forfree-flight models (for obvious rea­son s) and for trainers. When anovice pilot gets into trouble, if hismodel has good spiral stability, heneed only neutralize his controlsand the model will, on its own,recover, provided it has enoughaltitude.

For sport models, a moderatedegree of spiral stability is desirable.This applies also to flying boats,floatplanes, canards and partic­ularl y to rudder- and elevator-onlymodels, both powered and gliders .

For pattern and aerobatic models,neutral stability or mild spiral insta­bility is needed for good maneuver­ability. The spiral dive is slow todevelop , so the expert pilot has noproblem controlling th e model.

A high degree of spira l instability

Snowy Owlwas a AD-powered model with5degrees of dihedral, slotted flaps, aT-tailanda CLA at 25 percent of its VTMA. It flewwell, butin slow, nose-high, flaps-down,levelflightat lowrpm, it developed a mildOutch roll. Theorizing thatturbulence, fromboth the nose-uppostureandtheloweredflaps, was blanketing thevertIcal tail, Idoubled the dorsal-fin area; thiscorrectedthe problem. The 5-degree dihedral wasfound to betoo highforgood inverted flight.

44 THE BASICS OF RIC M ODEL AIRCRAFT DESIGN

Vertical Tai l Design and Spiral Stability ... CHAPTER 9

is not desirable, nor is too muchspiral stability, which in h ibitsman euverability.

Testing the spi ral stability ofan existing mo del as noted aboveis hindsigh t. The old saw that,"Foresight, as good as hindsight, isa damn sight better" applies. Weneed a way to incorpo rate th edesired degree of spiral stability in adesign while it is still on the draw­ing board.

LATERAL ANDDIRECTIONAL COUPLINGSpira l stability requires a bala ncebetween lateral (roll axis) and direc­tional (yaw axis) forces. Theextremes are:

• Large dihedral angles on the wingalong with a small vertica l tail arealeads to "Dutc h roll" (characterizedby tail wagging cou pled with aslight side-to -side roll) or even astall-spin crash. The lateral forcesare too high .

• A large vertica l tail area alongwith little or no dihedral leads tosides lip; the large tail resists theslip, and a killer spiral ensues. Thedirectiona l forces are too great.

Somew he re between these extremeslies th e correct balance of lateraland directiona l forces that will pro­duce the degree of spiral stabilitythat suits the designer's pe rfor­ma nce objectives.

BALANCE OF FORCESSince spi ral stability req uires abalance between lateral and direc­tional forces, i.e., a balan ce betweenth e effects of dihedral ang le andvertical tail surface area, th e designprocedure is to establish the lateralparameters (di he dral) first, andthen to bal ance the directionalparameters (vertical tail area) tomatch , at the chosen CLAposi t ion.

• Lateral stab ili ty- Dihedral. The Wing's dihedralang le is a major con tr ibutor tolat eral stability. See th e chart"Sugges ted Dihe dra l Angles."

The relative positi ons of wingaerodynamic center (AC)-2Spercent of the MAC-and CG bearon the di he dra l angle. A high wing

enjoys some pen­du lum stabilitythat's absent frommi d- an d sho ul­der-win g posi­tions. With CGabove the wingAC (as in a low­wing setting)the re is pendu­lum in stab ility,he nce, the di f­feren t dihedraldegree figures.- Sweepback actslike dihedral. Inlevel flight, 2 to 3degrees of sweep back are equivalentto 1 degree of dihedral. The dihe­dral effect increases both with angleof sweepback and CL and so, unlikenormal dihedral, it increases withhigher AoAs.

Many pattern ships use taperedwings wit h straight-across trailingedges and sweptback leading edge s.The angle of sweepback on thequarter-chord line is about 7degrees on a wing of AR6 and taperratio (roo t to tip) of 1:0.6 andnee ds no dihedral. Without dihe­dra l, there are no side area s pro ­jected by th e wing ahead of theCLA, and that reduces the verticaltail area needed.

High sweepback angles on full­scale aircraft increase latera l stabilityto such an extent that negativedihedral (anhedral) is introduced toreduce lateral stability for betterlateral control. The LockheedGalaxy is an example.- Forward sweep. Heavy forwardsweep (20 degrees or more) is verydestabilizing both laterally (in theroll axis) and directionally (in theyaw axis). When yawed, one wingadva nces and the other retreats; thecen te rs of lift an d drag of theadva nci ng wing panel have reducedmoment arms to th e CG . Themo me nt arms on the retreatingpane l are increased. The differentialin drag momen ts increases the yaw;but the lift-momen t differentialcauses a roll in a direction that'sop posed by the yaw. The model will"co rkscrew" and probably crashun less there is sufficient vertical tailarea and/or vertical-tail momentarm to prevent th e yaw.

This requires: 1) an area that'ssufficien t to bring the CLA to the

30 to 3S percent of vertical-tailmoment arm (VTMA) position ; 2)h igher dihedral (as discu ssedabo ve); and 3) a limit in the for­ward sweep to not more th an 30degrees measured on th e quart er­chord line.

In addition , the model will bespirally un stable. The major advan­tage of forward sweep is th at th ewing stalls at the roo t first. Rolldamping and effective aileron con­trol continue to high AoAs beforethe wingtips stall. Th is permitsslow, high-AoA flight.

• Directional stability. The majorfactors are th e amo unt of verticaltail area and its moment arm to th eCG (i.e., verti cal tail volume). Thevertical-tail AR, like that of a wing,is a contributing factor. Higher-ARvertical tail s have stee per lift-cur veslopes; th ey are th erefore moresensitive , but stall at lower AoAs. Athigh side-slip ang les, a high-ARvertica l tail can sta ll, resultingin reduced contro l. A dorsal fin isrecommended to overcome a lackof vertical-tail effectiveness at highAoAs, such as wh en flap s areextended and at high sid eslipangles.

Sweepback aids directional stability.When yawed, th e advancing wing'scen ters of lift and drag have greatermom ent arms th an th ose of th eretreating wing. The drag-momentdifferential reduces the yaw, and thelift differential promotes a roll inthe direction of th e yaw.- Ailerons. Good aileron design, withdifferential, reduces or elimina tesaileron-induced adverse yaw. (SeeCha pter 10, "Roll Contro l Design .")- CG location. If th e CG location of

THE BASICS OF RIC MODEL AIRCRAFT DESIGN 45

CHAPTER 9 A THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Verlical tail lf4MACmotion is opposed by th e effect ofth e dihedral, th at dihedral sho uldbe no larger th an is necessary tomeet othe r criteria.

• Adverse yaw. The effects ofadverse yawing moments on rollingvelocity may be decreased by increas­ing directional stability, or bydecreasing dihedral.

In Frank Zaic's 1935/36 yearboo k,under th e heading, "Determinationof Rudder Area," a similar profilemethod is described. In it, the CLA iscalled th e "directional center." It wasintended for use on rubber-powered,free-flight models. Grant's procedurewas a refin em ent of this earlymethod. Thanks to Martin Simon sfor bringing th is to my attention.

Those who are interested shouldread NACA Technical Note No. 1094of 1946: IIExperimental Determin­ation of th e Effects of Dihedral ,Vertical Tail Area and Lift Coeffi­cien t on Lateral Stability andControl Characteristics ." A

• Directional (weath ercock)stability. Modification s that increasedirectional stabili ty, such as anincrease in vertical tail area, perm itgreater roll rates to be obtained andmake th e perform ance of a givenbanking maneuver possible withdecreased aileron deflection .

The effect on later al man euver­ability of changing th e tail lengthwhile maintaining th e same direc­tion al stability, i.e., th e same tailvolume, and th ereby incr easing th edamping in yaw, is negligible.

The increase invertica l tail arearequired to movethe CLA aft is sur­prisi ngly large. Forone model, theSkylark, an increasein vertical tail areaof 60 percent wouldhave been neededto move the CLA aftfrom 22 percent to30 percent of its ver­tical-tail momentarm-a distance of1.65 inches.

POSTSCRIPTWhile reading an old (1947) NACAReport No . 868 -"Summary ofLateral Contro l Research"-I foundsome very significant data (the datain NACA reports are timeless).Though expressed in general terms,without specifics, they reinforce th eideas expressed in this article andGrant's CLA theories.

CONCLUSIONThe profile methodfor balancing lateraland directional fac­tors, at the selectedcenter of lateral area,is certainly not high­tech, but it's simple,effective and applica-

ble to the great majority of conven­tional planform configurations.

The CG/CLA relationship and th eSSM bear a remarkable resemblanceto the CG-neutral point and static­margin concept in the longitudinalstability considerations outlined inChapter 6, "CG location" and thematerial discussed here will wellreward the model airplane designer.These techniques have worked wellon a variety of designs built andflown by the author, and they're agood stepping-off point for furtherexploration of stability considera­tions in model design.

Verlical tail lf4MACCenter 01lateralarea (CLA)---l--r?.4~~

Areas .>-""'----.j~

doubled

Center 01gravity (CG ) "::::'~-=:::::::~-::-::--4~:::::::"'-i.:

Center 01gravity (CG)

Figure 6.Conventional profilemodel.

Figure 7.Canardprofilemodel.

SPIRAL STABILITY MARGINRefer to Figures 6 and 7. These staticstability margins are suggested:

an existing model is moved for­ward from a position that's verti­cally in line with the wing's AC, itlengthens both the VTMA and thedistance from CG to CLA (spiralstability margin or SSM). Forexample, the Swift has a VTMA of24 inches, and with the CG underthe wing's AC, the SSM is 25 per ­cent of the VTMA, or 6 inches.Moving the CG forward 1 inchincreases the VTMA to 25 inches,and the SSM becomes 7 inches, or28 perc ent of the VTMA. Thi sis enough to change the spiralstability from mildly positive toneutral.

If the CG is mo ved aft of thewing AC by 1 inch, both VTMAand SSM are reduced. For the Swift,the VTMA would be 23 inches andthe SSM5 inches, or a CLAlocation21.7 percent aft of the CG. This is avery spirally stable location.

SSM as % of VTMA

Super spiral stability

Good spiral stability

Neutral spiral stability

Mild spiral instability

Very spirally unstable

22252830

33 and up

• Lateral stability. High , positive ,effective dihedral combined withweak directional stability, i.e., smallvertical tail area, result s in a largeopposing action to an y rollingmotion (experienced with theSkylark) and can lead to a pre­dominance of lateral oscillation,i.e., Dutch roll. Since the banking

Powered by a .45 converted to diesel opera­tion, Osprey wasdesigned as a trainer with3 degrees of dihedral, slotted flaps and agenerousdorsal fin. CLA wasat 25 percentof the VTMA, and tail-dragger landing gearwas used. It was a stable, yet maneuverablemodel to fly. Banked15 to20 degrees andcontrolsneutralized, it would returntoupright level flight in about 90degrees ofacircle. Flaps down, it was stable, andonfloats, it was pure fun.

46 THE BASICS OF RIC MOD ELAIRCRAFTDESIGN

Chapter 10

Figure 1.Outboard ailerons.

B. PLAINAILERON

C. TOP HINGEDAILERON

farther from the model's CG, theyhave more leverage. One serious dis­advantage is that, with equal up anddown movement, th ey producegreater adverse yaw than do stripailerons. The downgoing aileron hasmore drag than the upgoing, andth is unequal drag tends to yaw th emodel in a direction opposi te to thetu rn commanded.

A remedy for thi s condition isaileron dif feren tial, where theupgoing aileron's angular travel istwo to three times th at of thedo wn going. This author uses amodified Frise, top-hinged aileronwith a differential of 2.5:1. Theextended lower, forward lip projectsin to the airstream below th e wingwhen th e aileron is raised, produ c­ing drag that favors th e turn (seeFigure IA). Turns are made withoutuse of rudd er. Figures 1Band 1Cshow two ot her forms of barn-doorailerons.

The outboard locat ion permitsuse of flaps spanning 60 to 65percent of the wing's semi-span.This wide, short type of ailero nsho uld be mass balanced for flutte relimination.

Two othe r forms of aileronsdeveloped to overcome adverseyaw are slotted and Frise ailerons.Use of differential ailero n is moreeffective in producing desirable yawmoments than is the use of eith erof th ese two aileron types. Bothslotted and Frise ailerons require

Design

Roll Control

CONVENTIONALAILERONSIn general, th is typefalls into two cate­gories: outboard, or"barn door," and stripailerons. Outboardailerons (see Figure I),usually are 25 percentof the wing chord inwidth and 35 to 40percent of the semi­span in length. Being

25° up

\of

25' up

• all-moving horizontal tails(stabilators).

NONE (OR MINIMAL)This form of rudder-only lateralcontrol is popular for sailplanesand some powered sport mod els.Wings for this type need additionaldihedral. For powered models, th iswould be 5 degrees for high wings,6 degrees for mid wings and 7degrees for low wings.

Thermal glid ers have po lyhe­dral-typically 5 degrees from rootto 3/5 of the semi-span, with anincrease of 3 degr ees from thepolyhedral joint to th e wingtip.On this type, whe n rudder isapplied, the model yaws. Airstrikes the wing at a sligh t diago­nal. For the win g on th e outside of

a turn, th e wind thatstrikes th e wing atan y given point onthe LEexit s from theTE at a point slightlycloser to the fuse­lage. Because of thedihedral, there is aneffective increase inAoA. This situationis reversed on theopposite wing. Bothcause the model toroll. It is importantthat such modelshave good spiral sta­bility.

i-*-1 0' down

~ 25' up

~i

..... -'+1--,~· I J 1 0· down17~ I"- 25% chord--I

Massbalance

A. MODIFIED FRISEAILERON

• spoilers and slot lip ailerons;

• flaperons;

• external airfoil ailerons;

• conventional ailerons;

• none or minimal (via roll cou­pling)-on rudder and elevator­only models;

• all-moving wings (pitcherons):and

D esirable roll or lateralcontrol characteristics areimportant for good and easy

maneuverability.There are several types of roll con­

tro l in use on today's model aircraft:

THE BASICS OF RIC MODEL AIRC RAFT DESIGN 47

CHAPTER 10 A THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Hinge

~ -~ \'f - -- 20' up-.,~--'J - 25~ """"""

20' dawn- ...!

---.. 10"/0 Chard ~

Figure 2.Strip aileron.

more deflection than plain aileronsfor the same roll rate.

Strip ailerons (see Figure 2) arelong, narrow and almost full span.They simplify wing construction,and they produce less adverse yawthan outboard ailerons, since theircenter of area is closer to the CG.

Most are actuated by servosmo ving horns on their inboardends so that differential is easilyintroduced . Made of solid balsa TEstock, they are prone to flutter andshould be mass balanced at th eoutboard end to avoid thisproblem.

EXTERNALAIRFOILAILERONSExterna l aileronswere a Junker'sdevelopm ent andmay be seen onsome full- scal eultraligh t aircraftflyin g today. AsFigure 3 shows,these co nsist ofsma ll, separa te

win gs that aretucked under the

main wing's TE, which provides aslot effect over the small wing. Theseare full span; the outboard portionsform ailerons, and the inboard forma type of slotted flap. Hinged exter­nally, the y should be mass balancedfor flutter elimination.

FLAPERONSFlaperons are a for m of plainaileron that can be ope rated asailerons and drooped simultane­ously as flaps. They extend for mostof th e Wing's semi-span, like stripailerons. When in th e fully loweredposition as flaps, and th en used as

ailerons, there is a high degree ofadverse yaw that cannot be over­come by aileron differential action.Rudder control, either manual orelectronic, must be introduced tocounter the adverse yaw of thistype of roll control. Mass balancingis recommended.

SPOILERS ANDSLOT·LlP AILERONSFigure 4 shows a typica l spo iler.Provided its leading edge is beyond70 percent of th e wing cho rd, thereis no lag in th e contro l's aerody­namic actio n . Only one spo ileroperates at one tim e-the on e onth e inside of th e turn. The oppositespoiler stays retracted. They pro­vide posit ive into-the-turn yaw,work inverted, and require no massbalan cing. A version of the spoiler,some times calle d the "slot-lipaileron" is shown in Figure 5.

This form of roll contro l provedvery effective on both my Crane Ian d n. The roll rate was fast andworked inverted. With flaps low­ered, roll contro l was very crispat low speeds, since raising thespoiler destroyed the slot effectove r th e flap, reducing its addi­tional lift . Yaw was favorable. Thismodel's performance, at low speedsparticularly, was spectacular.

Hinge~

~ ~'" --- ~_\~~---c:::~~2~~S~5:' ~~ ~ Neutral.. .... ... ........

t10' dawn

Figure 3.External airfoil aileron.

PITCHERONSThese are a recent development forRIC sailplanes. Each wing panelrot ates aro und spanwise pivot slocated at the wings 1/4 MAC. Bothare contro lled by one servo, butconsiderable differenti al is neededto offset adverse yaw.

Very few degrees of rotation areneeded since each wing panel

Hinge-.'- 10"/o -JI Chard__ 20"/oChard

Figure 4.Spoiler.

48 THE BASICS OF RIC MODEL AIRCRAFTDESIGN

cBasic airfoil

_

-----:-::-_::::=--~A~ Nate gap

SI~"'.;1'(" ........~Spailer·up '" .- Slat

Retracted_ ~~~' 4r

Pivot paintsFlap

Figure 5.Slotted andflapped airfoil .

Roll Control Design A CHAPTER 10

ot her types of hingin g, some formof gap seal is advised.

Figure 7 provides sugges ted pro­por tions for ailerons, strip aileronsand spo ilers that were deve lopedby NACA. They are goo d sta rtingpoi nts when yo u are creating yourown designs. A

Downward aileron

II T I

Upward aileron

I ~Servo armtravel ~==---t--ji>'.......

SPAN B

Figure 7.Typical control-surface geometries.

Suggested aileron and spoiler geometrieslor model alrcratt, Irom NACA Report No. 605. Resume and Analysis01 NACA Lateral Control research . Weick& Jones. 1937.• 25"10 of 60"10 01the lull chord.

.25x60C·

tI

y- .16 C

TI

I

.r,.25 C

I

~ .15C

+

i

! .10 CI-I

I

:1. .10 C..I

AI

1J .15x6C

"II .20x6C

~ III( B/2 ~

Figure 1 aileron

I I1+ .40 B/2 ..;

Figure2 aileron

I+-- .60 B/2 ~

Figure 3 aileron

~ .80 B/2 -.i

Figure 4 aileron i tI Cx

- ' I:-l 1 .25 x Cx --M -~ ~.40 B/2

Figure5 spo iler

1.I .2C

t~ .60 B/2 -+l

Figure6 spoiler

~

.2' ~ .50 B/2 -J

r Figure 7 spoiler i .~+Cx

~I -

.20xCx --..~Il' I k- .50 B/2 ~ I

;r-.60 C

1.

...60 C

Y

SERVO

Figure 6.Aileron differential (schematic for oneaileron linkage).

AILERON DIFFERENTIALFigure 6 shows how to use aservo's rotation to produce ailerondifferen tial.

GAP SEALINGWind-tunnel tests have proventhat a 1132-inch gap on a lO-inch­chord wing will cause a loss ofrolling moment of approximately30 percent. A gap seal for all con­trol surfaces is suggested. The side­bar "Flap and Aileron ActuationHinges" of Chapter 14, "Designfor Flaps," provide s a hingingmethod that has proven durableand inherently gap sealing. For

rotates in its entirety. The wing­fuselage joint would need specialattention to avoid local separationand increased drag.

STABILATORSSome recent jet fighters use suchtails . They move in opposite direc­tions for roll control, and up ordown for elevator action---or anycombination of the two. They seemvery effective and, for a model,higher ARs would provide longermoment arm s. Adverse yaw wouldbe small.

Pivoting on the spanwise pivotsat 1;4 MAC wou ld result in lowoperating loads, as for all movingwings. This form of roll controlmight have app lication on patternships, leaving the wing free for full­span flaps .

THE BASICS OF RIC MODEL AIRCRAFT DESIGN 49

Chapter 11

Weight

Distribution

in Design

A n anal ysis of the weight ofth e average .40 to .50 glow­powered, rad io-con trolled

model aircraft with ailerons disclosesthat th e power and con tro l uni ts,combined, weigh very close to SOpercen t of the aircraft's gross weight.

The power unit (PU) is composedof spinner, pro p, engine, muffler,engine mou nt, fuel tank, fuel, cowl,fuel tubing and nuts and bolt s. Thecontro l unit (CU) is made up ofreceiver, battery, servos, switch ,extension cables, foam protectionfor receiver and batt ery and servoscrews. In th e design of a model,

Figure 1.Three-view drawing o(Granville canard.

50 THE BASICS OF RIC MODEL AIRCRAFTDESIGN

the distr ibutions of th ese heavyunit s alon g th e length of the fuse­lage has a major effect on th atmodel's maneuverability.

Massing both units as closetogeth er an d as close to the CG aspossible while keeping th at CG inits design location will result in ahighly man euverable model.

Moving th e power unit forwardby elongating the fuselage ahead ofthe wing requi res that the controlunit move aft to keep th e CG at itsdesign location. Man euverabilitywill be redu ced as a result . A fewsim ple definition s will h elp inunderstanding thi s reduction:

• Moment. A force times a distance.

• Inertia. The resistance of an objectto any change in its motion or tobeing moved from a state of rest.

• Moment of inertia. The inertiaresistance tim es its distance fromsome related point. In our case, that"related point" is th e model's CG.

• Momentum. An object in motionhas momentum equal to its masstimes its veloci ty. In maneuvers,both th e PU and CU acquiremomentum in a direction differentfrom th e original line of flight.

The PU's weight multiplied by itsdistance from th e mod el's CG is its"moment of inertia. " The sameapp lies to th e CU.

Obviously, th e greater th e dis­tance of both th e PU and CU fromth e model's design CG, the greaterthose mom ents of inertia will beand th e greater th e resistance to theman euver.

Also, lon ger moment arm s (inth is case, distance of the PU and CUfrom the CG) requ ire bo th PU andCU to move th rough greater dis­tances, for a given angular displace­ment, as the aircraft maneuvers.

Lon gitudinally, the moment toovercome the moments of in ertiaof both units for maneuvers is themodel 's TMA multipli ed by theforce gene rated by deflecting theelevato rs. Th e model's TMA ismea sured from CG to 1/4 MAC ofthe hor izontal ta il. For a givenTMA and elevato r force , thegreater the moments of inertia ofth e PU and CU, th e slower themodel's reaction . Loops will ha vegreate r diameter, and th e modelwill be less agile.

With th e man euver underway,both the PU and CU acquiremomentum. To stop the maneuver,thi s momentum mu st be overcome.Larger mom ents of inertia producelarger momentum and slow therecovery from that maneuver.

Directionally, the same applies.The rudder will have less effect inyawing the model. Also, asexplained in Chapter 9, "VerticalTail Design and Spiral Stability,"elongating the fuselage ahead ofth e CG increases its directionallydestabilizing side area, requiringincrea sed vertical tail area for stabil ­ity and control, further aggravatingthe situation. Greater moments ofinertia have one advantage: theyoffer more resistance to any distur­bance. In level flight , the modelwill "groove."

SPINNINGIn a ta ilspin, one wing panel is fullystalled, but th e opposite panel con­tinues to lift. The model rotatesrapidl y, nose-down, around a verti­cal axis through its CG. Up-elevatorand rudder into the spin maintainthe rotation.

Centrifugal force acting on themodel's components comes intoplay. The long er moment arms ofboth the PU and CU result in theseuni ts rotating at higher speeds, gen ­erating greater centrifugal forces,which act horizontally, away from

Weight Distribution in Design ... CHAPTER 11

Figure2.Three-view drawing of Long-fl.

dangerou s ailero n flutter greatlyoutweighs the small reduction inmaneu verability that 's occasioned b yth e ma ss -balance weights. Thesame comments apply to mass bal­ancing of elevators and rudd er.

REAR·ENGINE CANARDSFor con vent ion al designs, it is notdifficult to position both poweran d contro l un its so as to minimizeth eir mom ents of inertia .

Rear-engine canards, without aftwing sweep, are a different matt er.Such aircra ft ha ve the ir CGsbetween fore and aft wings, closerto the latt er. The PU at or behindthe aft wing is balanced by locatingthe CU as far forward as possible. Inmost cases, additional ballast isrequ ired up front to locate the CGcorrectly. The moments of iner tiaof both uni ts (and ballast) couldnot be greater.

My Swan canard was not intendedto be aerobati c, but in level flight , itgrooved beautifully. The re arecanard configur ation s that havelower moments of inertia.

the spin axis. This action flatt ensthe spin .

The lon ger mom ent armsincrease th e momentum, reduceth e rudd ers' effectiveness in sto p­ping the spin and delay th e spinrecovery, which could lead to adamaging crash.

LATERAL CONTROLInertia roll coupling is a con sidera­tion in lateral control. For thosedesigns in which the aerodynamicand inertia axes coincide, axialrolls are little affected by largermoments of inertia . In snap rollsand barrel rolls, centrifugal forcecomes into play, as it does for spins ,resulting in slower initiation of andrecovery from these maneuvers.

The model's wing is a factor, as itweighs close to 2S percent of themodel's gross weight. For good lat­eral maneuverability, keeping th ewing panel's CG as close to thefuselage center line helps. Thisresults from :

• Tapered wing of moderate AR.

• Ailerons, mass balanced to avoidflutter, permit aileron and flap

servos to be positioned in the wingcenter section.

While aileron mass-balance weightswork against lat eral maneuver­ability, keeping th e ailerons lightreduces the mass-balance weightcorrespondingly. Freedom from

Figure 3. .Three-viewdrawing of the Miles M.39SLibel/uta.

• Granville canard (Figu re 1).Both PU and CU (the pilot) arelocated close to the CG for goodmaneuverabi lity. A modernizedversion of th is clever design wouldbe interesting.

• Rutan 's Long-EZ (Figure 2).The sweptback aft wing perm its th ePU to move forward , shortens th efuselage and permits th e CU (pilot)to move aft, close to th e CG. Thebig wing-root strakes house th e fuelon th e CG. The wingtip vertica lsurfaces have reason able mom entarm s for good direction al control,but th eir loca tio n increases th ewing 's mom ent of ine rtia, reducinglateral maneuverab ility.

• Miles Libellula (Figure 3). Thiswas a British wartime design. Thetwin engines ahead of th e moder­ately swept aft wing br ing th epower units closer to th e CG longi­tudinally. Both fore and aft wingshave flaps. Note the high-AR fore­planes on bot h the Long-EZ an dth e Libellula. ...

THE BASICS OF RIC MODEL AIRC RAFT DESIGN 51

Chapter 12

looks rep resentative of ma ny oftoda y's fuselage shapes. From its CDof 1.261, deducting th e prop CD of.577 and adding th e extra drag of.260 for tricycle gear/tires and of.336 for the fully exposed engine,results in a worst-case CDof 1.28. At40mph, this would generate a 19­ounce drag; at 50mph, a 30-ouncedrag. Surprised? This doesn 'tinclude wing and tail-surface drag.A good drag-reducing design couldlower this to a CD of .38 (5.7ounces) at 40mph but, again , th iswouldn't include wing and tail-sur­face drag. Figures 3 and 4 from

.269

.340

.225

.198

.458

.261

.242

.237

.775

1.034

1.261

Elastic band

Fuse lage

c==--­C ~

~======c=========-C -========-

Nose

4

2

3

7

6

5

1 o~C)

~

o(Q)

o

:~ £========­1°f!\F11~~ ~

area and speed, it will accurately pro­vide the actual drag in ounces. Forour purposes, the CD provides therelative drag value of each shape .Analysis of the Cos in Figure 1 willprovide some surprising results.

Deducting the .198 CD of fuse­lage 1 from that of fuselage 8 (0458)gives a CD of .260 for the landinggear on ly-or more than the drag offuselage 1. This gear was lI8-inch­diameter music wire , and thewheels were the thin, symmetrical,cross-sectioned type that was popu­lar at the time. Current tricyclelanding gear with their large, fattires would, con -servatively, doublethe CD to .520­or more than2112 times that offuselage 1.

Deducting the.198 CD of fuse­lage 1 from that offuselage 9 (.775)provides a CD of.577 for the sta­tionary propeller.From fuselage II'sCD of 1.261,deducting theprop CD of .577,the landing gearCDof .260 and the.340 CD of fuse­lage 2, resul tsin the exposedengine-cylinderdrag of CD .084.A fully exposedengine, mufflerand firewallwou ld, conserva­tively, have a CDfour times asgreat : .336.

Fuselage 11,which is 48 incheslong and 33square inches Figure 1.in cross-section, Drag coefficients of various fuselages.

• sha pe of the fuselage.

The CD for each reflects the dragvalue of that shape. When used in aformula that includes cross-section

• airspeed;

• cross-section area; and

Reducing Drag

I t will come as a surprise to mostmodelers (and some modeldesign ers, too) to find how

much air resistance, or drag, theirminiature aircraft generate in flight .The sources of much of it are suchthings as expos ed or partiallycowled eng ines; wire landing-gearlegs; fat tires; dowels and rubberbands that are used to hold downth e wings; large, exposed controlhorns and linkages; and thick TEson wings and tail surfaces.

This doesn 't impl y that the mod­els don 't fly well; they do! In fact,th e high drag is beneficial: it causesfairly steep glides-engine throt­tled-that ma ke the landings ofth ese relat ively low-wing -loadingmodels easy to judge. Their perfor­mance suffers in all other flightaspects, however.

Many years ago, Model AirplaneNews published a very significantarticle by Hewitt Phillips and BillTyler, titl ed "Cutt ing Down theDrag." It was based on wind-tunneltests conducted at the Massach u­setts Institute of Technology Aero­nautical Laborat ory at modelairplane speeds of from 15 to40mph. The test models were48 inc hes long and of typicalmodel airplane con struction.

Figure 1 summarizes the results,which are given in terms of theirCos. The actua l drag in ounces of amodel fuselage depends on threefactors:

52 THE BASICS OFRIC MODELAIRCRAFT DESIGN

Reducing Drag ... CHAPTER 12

Station Fuselage Fuselageno. S nO.1

0% 0.0000% 0.0000%

5% 0.0475% 0.0750%

10% 0.0660% 0.0980%

20% 0.0920% 0.1130%

30% 0.1080% 0.1 030%

40% 0.1130% 0.0750%

50% 0.1030% 0.0520%

60% 0.0900% 0.0390%

70% 0.0710% 0.0325%

80% 0.0490% 0.0250%

90% 0.0250% 0.0180%

100% 0.0000% 0.0000%

Figure2.Fuselage diameter asa percentage of fuse­lage length for feast-drag circular fuselages.

This chart permits accuratescale construction of the

fuselages depicted in Figure 1.

fuselage. The air has to expa ndfrom the hig h point of the wing tothe TE and also fill the re-entrantcorner formed at the TE and thelower fuselage. The resultant turbu­lent flow cause s high drag andreduces tail -surface effectiveness.The cure is wing-root fairings , e.g.,those on the Spitfire, but they'redifficult to make.

CJ<:)

~--------- ---­---~----- --------

--~.-----~.--=-------~-------------- -

Figure 3.High-drag airflowaround wirelanding-gear leg.

TYPES OF DRAGHere's a list of the various types ofdrag an d their causes:

Phillips and Tyler's article illustra tethe high drag caused by unfairedlanding-gear legs. Figure 2 providesdata for reproducing fuselages 1 andS in Figure 1.

Figure 4.These twoobjects give thesame drag.

• Upwash/downwash. In levelflight, ai r doesn 't flow hor izontal­ly on to th e wing's LE, or from itsTE. Ahead of the wing , th e airflows upward to the LE (calledupwash) and downward off the TE(downwash) .

To achieve this negative lift, the hor­izontal tail surface must be at a neg­ative angle to the wing's downwash;this would result in increasedinduced drag. Since th at extra 4ounces must be supported by th ewing, its induced drag also increases.

The re are othe r forces that causenose-up or nose-down act ions and,to achi eve level flight, th e horizon ­tal tail mu st ove rcome the netresultant force:

• Wing-pitch ing moment. This isa nose-down moment, except forsymmetrical or reflexed trailing­edge sections, which have little orno pitching mom ent.

• Trim drag. Consider a l Ou-ouncemodel, which has its CG 1 inchahead of its wing's cen ter of lift. Anose-down moment of 100 oz.-in.results . To maintain level flight, th ehorizontal tail must lift downward.Using a TMA of 2S inches, thatdownload would be 100 -;- 2S = 4ounces.

• Powerplant drag. This is causedby exposed engines, cylinder heads,mufflers and tuned pipes.

• Induced drag results from theproduction of lift, and it depends onseveral factors: the wing area, thewing AR, the wing planform, theflight speed and the CL at which thewing (and the tail surfaces) operate.It's normally less than th e wing-pro­file drag.

• Wing and tail-surface profiledrag. These are similar to skin­friction drag and depend on theshapes of the airfoils and on the Rnsat which they fly.

t7"

t

!AI" dla. music wirej -

J""'\ro c_==-~--=---=J

• Skin friction is proportional to th eamo un t of exposed surface area andits roughness as well the Rn at whichthe model flies. The smooth,reflexed, pressure-recovery shape offuselage 1 in Figure 1 has the leastsurface area, and this contributes toits low drag.

• Separation drag. An exampl e ofthis is a th ick, low wing on a round

• In terference drag is caused by thebreakdown of smooth airflow owingto such things as landing-gear legs,bracing struts, dowels, open cockpits,etc., that disturb the air flow over theaircraft aft of the cause (Figure S givesexamples ).

THE BASICS OF RIC MO DEL AIRCRAFT DESIGN 53

CHAPTER 12 ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

of both its ap pearan ce andperformance.

WINGS AND TAIL SURFACESThere are three ma jor considera ­tions in wing design: wing cross­section or airfo il; aspect ratio ; andplanform.

• Aspect ra tio. This has an impacton induced drag; th e higher th e AR,

• Airfoils . Of th e three, airfo ilselection is th e most crit ical. Selectfrom th ose ai rfoil sectio ns forwhich th ere are wind-tun ne l testcurves at model airplane Rns.

In th e Eppler E197 section (seeappendix), th e lift curves show amaximum Cl of 1.17 with a gentlestall. The pitching mom ent is fairlycons tant for all AoAs. The pola rcurves show th e profile Co versusth e C l . Note th at th e profile CD islow despite th e increasing C l ,

except at th e low Rn of 100,000. Awing of 6 inc hes in cho rd flyin g at20mph would be operating near Rn100,000 . Table 1 provides the datafor rep rodu cing E197 for any ch ordlen gth . Th is airfoil is 13.42 percen tof its chord in dept h, permitt ingstrong, but light, wing structures .

For tail surfaces, see th e curvesfor the symme trical Eppler E168section. Not e th e h igh er profiledr ag at Rn 60,000. A 4-in ch cho rdflying at 20mph would be operat­ing at Rn 60,000. Avoid chords ofless than 5 in ch es on tail surfaces.Table 2 provid es data for duplicat ­ing this secti on.

Th e foll owing deals wit h dr agreduction for win gs and tail sur­faces and th e eng ine and muffl er.

• Flying a low-d rag, slot ted­flap- equipped mod el provides anew and th rilling experience.

Righi

• The quicklyand easily removedengine cowl andupper fuselage makeservicing of theeng ine , fuel tank,servos , etc. , verycon ven ient.

• The use of slotted flaps as outlinedin Chapter 14, "Design For Flaps,"will provide very quick takeoffswhen half extended, and slow, steeplanding approaches and gentletouchdowns when fully extended.By selecting the angle at which th eflaps are deployed (from 0 to 40degrees) and adjust-ing engine rpm ,it's possible to fly at any chosenspeed from just above the stall at20mp h to th e maximum speed; forthe Swift, that's at 138mph .

• The fully balsa -cowled engineand muffler are distinctly quieter.

• At slower speeds and lower rpm,fuel consumption is reduced.

• The model willlook sleek and fasteven standing still ;one can be proud The Seagu/lllf is an example of a low-drag airplane design.

Wrong

--~

Figure 5.Causes of interference drag.

Upwash causes a nose-up forceon the fuselage ahead of the wingand on th e propeller, because airflows into th e pro peller disk at aslight, up ward ang le. Downwashim pacts on the aft fuse lage and onth e horizontal tail surface , and itcauses a nose -up action.

• Thrust-line location. If it 'sabove th e CG, it produces a nose­down couple; below th e CG, anose-up coup le.

• Center-of-drag location. If it'sabo ve th e CG, it causes a nose-upforce; below th e CG, a nose-downforce .

Som e readers may question the valueof th e drag-redu ction techniquesoutlined in this chapter, particularlysince they involve extra time, effortand cost to achieve. Reduced draghas th e following benefits:

• Improved accelerati on and, withprop er propeller pitch and diameterselectio n, h igher flight speeds andbetter vertical performance. A dragof 30 ounces at SOmph increasesth e model's weight by th at amo untwh en climbing.

54 THE BASICS OF RIC MODEL AIRCRAFT DESIGN

The Canada Goose is a canard thatusesthe low-drag techniquesdescribed in this chapter.

the lower tha t drag. This is whysoaring gliders have lon g, slender,high-AR wings . For models, highAR results in narrow chords tha t

Table 1: Epp er 191AerodVDamic zero-2.1 Degrees

Chord Upper Chord LowerStation Surface Station SurfaceXU YU XL YL

.000 .000 .000 -.200

.318 .789 .279 -.6401.104 1.683 1.1 64 -1.2782.335 2.633 2.555 -1.8933.996 3.600 4.438 -2.4546.075 4.556 6.797 -2.9458.551 5.478 9.610 -3.36511.402 6.345 12.852 -3.70614.599 7.139 16.493 3.95518.112 7.844 20.495 -4.12521 .902 8.442 24.818 -4.1 9525.933 8.918 29.414 -4.18530.1 59 9.250 34.231 -4.08534.551 9.413 39.236 -3.85539.085 9.394 44.415 -3.53543.735 9.191 49.723 -3.16548.474 8.806 55.091 -2.76553.282 8.246 60.447 -2.36558.146 7.542 65.718 -1.96563.028 6.752 70.834 -1.59567.860 5.920 75.725 -1.26672.575 5.079 80.323 -.96577.105 4.254 84.564 -.71581.384 3.466 88.388 -.50585.349 2.733 91 .738 -.32588.939 2.068 94.572 -.18592.096 1.478 96.864 -.07594.778 .960 98.572 -.00996.960 .530 99.637 -.00598.604 .219 100.000 .00099.642 .050100.000 .000

have higher profiledrag at low Rns. Thisdefeats the lowerinduced drag bene­fits of the high ARs.

Long, slender wingsimpose greater stress­es at the wing rootsand require strongerstructures. In aero ­batics , they slow anymaneuvers involv­ing rolls.

For RIC spor tmodels, ARs of 5 to 7are suggested-a nim­

ble airplane results and, on smallermodels, prevents narrow chordsand low Rns.

• Planform. This is the wing 'sshape as viewed from above . It maybe straight, tapered , a combinationof straight and tapered, or elliptical.It may also be swept back or sweptforward.

The elliptical is the most efficientplanform, but it 's difficult to make .In addition, the tips fly at low Rnand are prone to tip-st alling.

Tapere d wings with taper ratios(ratio of tip chord to root chord) of.5 to .6 are close to elliptical wingsin efficiency. Each rib is different,and laying them out is time­consuming. The wing is strongestat the root , but, on small wings, thelower tip chord results in lower Rns,higher drag, and risk of tip-stallingat low speed.

This also applies to combinedstraight and tapered wings, in whichth e wing is straight for 50 to 60 per­cent of the semi-span and the out­board 40 to 50 percent is tapered.

A modest sweepback of 5 to 10degrees is popular in pattern modelsbecause it improves aerobatic per­form anc e. Sweptback wings tend totip-stall more readily. Forwardsweep reduces tip-stalling, but itimposes heavy tors ion loads on th ewing structure.

Straight, untapered wings of ARof 6; use of th e NASA "safe-wing"LE droop ahead of the ailerons (seeChapter 15) and holl owed balsablock wingtips are recommended.

Horizontal tail surfaces sho uldhave lower ARs (4 to 4.5) to keepchords above 5 inches and to avoidlow Rn profil e drag. Streamlined

Reducing Drag .& CHAPTER 12

Table 2: Eppler 168Chord Upper LowerStation Surface Surface

NR xrr YOIT YUIT

1 1.00000 0.00000 0.000002 0.99893 0.00006 -0.000063 0.99572 0.00027 -0.000274 0.99039 0.00071 -0.000715 0.98296 0.001 42 -0.001426 0.97347 0.00238 -0.002387 0.96194 0.00352 -0.003528 0.94844 0.00477 -0.004779 0.93301 0.00609 -0.0060910 0.91573 0.00754 -0.0075411 0.89660 0.00914 -0.0091412 0.87592 0.01094 -0.01 09413 0.85355 0.01293 -0.0129314 0.82767 0.01513 -0.0151315 0.80430 0.01754 -0.0175416 0.77779 0.02014 -0.0201417 0.75000 0.02293 -0.0227318 0.72114 0.07588 -0.0258819 0.69134 0.02898 -0.0289820 0.66072 0.03219 -0.0321921 0.62941 0.03547 -0.0354722 0.59755 0.03879 -0.0307923 0.56526 0.04210 -0.0421024 0.53270 0.04535 -0.0453525 0.50000 0.04848 -0.0481826 0.46730 0.05143 -0.0514327 0.43474 0.05415 -0.0541528 0.40245 0.05650 -0.0565829 0.37059 0.05865 -0.0586530 0.33928 0.06027 -0.0602931 0.30866 0.06146 -0.0614632 0.27006 006211 -0 0621133 0.25000 0.06220 -0.0622034 0.22221 0.06169 -0.0616935 0.19562 0.06057 -0 0605736 0.17033 0.05881 -0.0588137 0.14645 0.05640 -0.0564038 0.12408 0.05335 -0.0533539 0.10332 0.04971 -0 0497140 0.08427 0.04555 -00455541 0.06699 0.04094 -0.0409442 0.05156 003595 -0.0308344 002653 0.02535 -0 0253545 0.01704 0.01980 -0.0198046 0.00961 0.01444 -0.0144447 0.00428 000910 -00091048 0.00107 0.00460 -0.0046049 -0.00000 0.00000 000000

DickelT... = 0.124 RuecklagelT = 0.250WoelbunglT = 0.000RuecklagelT =0.001Profiletiefe... =T

THE BASICS OF RIC MODEL AIRCRAFT DESIGN 55

CHAPTER 12 .... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

LANDING GEARThis necessary, but drag-producing,appendage provides a significantopportunity for reducing drag.Aluminum landing-gear legsshould have rounded LEs and TEstapered to an almost knife-edge asin Figure 6.

OVERALL DESIGNGood overall design will do much toreduce trim drag. Ashoulder or mid­fuselage wing location, along with ahigh thrust line (inverted engine),will bring the centers of lift, thrust,

flight at that speed. At other speeds,the increase in fuselage drag mustbe accepted. Figure 2 of Chapter4-the lift, wing loading and speedchart-is very useful in this con ­nection. Using that chart, proceedas follows:

From the wing loading of yourmod el at the bottom of the chart,read upward to the cruise speedyou've selected. Where th e verticaland horizontal lines intersect, you 'llfind the CL needed. For example, awing loading of 24 ounces persquare foot at 60mph needs a CLroughly halfway between CL 0.15and CL 0.2o-say CL 0.17.

Refer to the lift-drag curves forthe wing airfoil of your choice, anddet ermine the AoA for CL 0.17.Using Eppler E197 as an example,an angl e of minus 0.5 degree willproduce CL 0.17. To adju st for thewing 's AR of 6, another 0.5 degreeshould be added to this and the rec­tangular planform, bringing theAoA to zero degrees.

In your design, the angle of inci­dence of th e wing to th e fuselagecenterline would be zero degrees toobtain the lowest fuselage drag atthe 60mph cruise speed.

The Seahawk at rest, flaps extended.

The basic low-drag features may,however, be incorporated. Such amodel is shown in the photo of theSeahawk. Another photo displaysthis airplane on its single float. Themodel 's large Youngman flaps, fullyextended, are very effective.

At a gross weight, on wheels , of110 ounces, powered by a .46engine turning an llx8 prop, thismodel's per form an ce is thrillingand justifies the drag-reducingtechniques in this chapter.

In plan view, the fuselage sidesshould be straight and parallel atthe wing-fu selage intersection toavoid separation drag. Reflexingstarts just afte r th e wing TE.

The angle of incidence at whichthe wing is set relative to the fuse­lage centerline is important. It'ssafe to assum e that the fuselage'slowest drag occurs when it 's flyin g,

in level flight , with itscenterline hori zontal.

The wing's beingfixed to the fuselage willcause variations in thefuselage 's centerlineattitude. At low speed,the wing must operate ata higher AoA to provideadequate lift for levelflight. At high speeds,lower AoAs furnish theneeded lift. Hence, thefuselage's centerlinedeparts from the hori­zontal, nose up at lowspeeds, and nose downat higher speeds, bothwith increased drag.

The solution is toselect a level-flightcruising speed and toad just the wing's angleof incidence to providethe lift needed for level

1t.lz" ply care

\ Aluminum landing-gear leg

Music-wire landing-gear leg

E

~f-.

ft 5T ~ ISand to streamline shape

Figure 6.Streamlining landing-gear legs.

FUSELAGEThe fuselage with th e lowest CD'fuselage no. 1 in Figure I , isn'tentirely practical for an RIC modelthat seeks to simulate the appear­ance of its full-scale big brothers.

ENGINE A ND MUFFLERExposed engine cylinders and muf­flers are major sources of drag. Fullyexposed engines, firewalls and muf­flers are even worse.

Some mufflers permit cowling ofboth engine and muffler completely.This type of cowl has been used onseveral mo de ls powered by .40to .45 and .46ci engi nes withabsolutely no coo ling problems.

The two cooling air outlets areat points of reduced air pressureon the sides of the fuse lage.Remember, only th e air that actu­ally hits th e eng in e cylinder do esth e coo ling . This thick balsa cowlalso acts as a sound damper. Enginenoi se is noticeably reduced. (SeeChapter 17, Ducted Cowl Design.)

forms such as E168 have lower dragthan Y4-inch-th ick sheet-balsa sur­faces. By use of stress-skinned tech­niques, th ey can be lighter andstronger. .

For both wings and tail surfaces,avoid thick TEs; sand them to YJ.6inch thickness with rounded edges.Thick TEshave the same drag as wirelanding-gear legs and are longer.

56 THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Reducing Drag .. CHAPTER 12

Figure 7.Swift airfoil selections.

It's important that your engine beadju sted to its lowest, con tin ualidle- around 2,500rpm. At an y­thing high er, say 3,00 0 to3,500rpm, it may be necessary tostop the engine in flight once th efinal approach has been established.

The model's struc ture is ofstre ssed-skin construction. You'den joy flying a model such as this! ..

ENGINEIIDLEFOR LANDINGAn aerodynami­cally clean modelsuch as the Swiftis capable of land­ing, flaps down ,at air speeds inthe 20 to 25mphrange. It doesn'tneed much propthrust to fly atvery shallowangles, makinglandings difficult.

sen to remove itfrom the fuselageboundary layerand the propelle rslipstream intoundisturbed air.Since this locationresult s in only twocorners, instead ofthe four of an in­fuselage location,drag is reduced.

The receiver andtransmitter shouldhave one extra

channel of "proportional" natu re sothat flap extension may be tailoredto the flying speed desired.

Figure 7 provides wing and tail­surface airfoil profiles and control­surface th rows.

Ailero ns, elevators and rudderare mass-balan ced for flutter pre­ventio n. In a dive, this model'sspeed would be high.

A feature of this mod el is theremovable fuselage top, from fire­wall to just aft of the wing. It's heldby dowels at the front and onenylon bolt at the rear. Its easyremova l provides access to all ser­vos, receiver, fuel tank and nose­wheel linkage, etc. This is a real con­venience.

Note that th e flap width is 30 per­cent of th e wing's chord, ratherth an 25 percent. Thi s pro vides

greater dragwhen it's extend­ed for a landing.The Swift is veryclean aerodynam­ically, and th eaddit ional dra gof th e wider flapwill prove bene­ficial.

Hinge _;';

--_"" • - : ;; -20~ Up

C =at: ,'..- , 20~ DnStab and elevator ' ",.f,

Fin and rudder ' '~,-{

~ ~?Pler 1 97 and s l o~ "

Below theSeahawkonits single float. Note the subfin below thehorizontal tai/plane.

gravity and drag very close to oneanother, thu s minimizing the hori­zontal tail's load, reducing its, andthe wing's, ind uced trim drag. Themodel will also be more nimble.

THE SWIFTThis model aircraft's design wasbased on the concepts in th ischa pter and Chapter 14. See the 3­view of the Swift in Cha pter 26 .

This is a sma ll, fast, highlymane uve rab le mo del, but withflaps down 40 degrees, th e planewill sta ll at 17m ph . A "safe" land­ing speed would be 25 perce ntgreater, or about 21mph. Top speedis 138mph. Total drag at 50mph isestimated at 12.5 ounces, in clud­in g wing and tail surfaces . At90m ph , th is would increase to 42ounces . The T-tail location was cho-

THE BASICS OF RIC MO DEL AIRCRAFTDESIGN 57

Chapter 13

The Canada Goosecanard features stressed-skinconstruction. Power is a .35ci engine.

Stressed-Skin

Design

I t's a sound engineering principlethat, to maximize strength andto minimize weight, the structur­

al material shou ld be located as farfrom the "neutral axis" as possible.

This chapter will explain, in sim­ple terms, what this neutral axisbusiness is all about and how toarrange the structure of your modelfor maximum strength withoutadverse weight penalty.

To start with, a nod ding acquain­tance with basic forces is needed.There are only four :

• Tens ion. Pulling on an elasticband puts it under tension.

• Com pression . Opposite of ten­sion . A column supporting a roof isunder compression.

• Shear. Forces opposed to oneanother. Cutt ing paper with scissorsis "shea ring." Each blade opposesthe ot her.

• Leverage. A 90-pou nd person sit­ting 2 feet away from the balancepoint of a seesaw will be exactly bal­anced by a 6O-pound person sitting 3feet from the same point , but on theopposite side. The greater leverageon the lighter person 's side offsetsthe other's greater weight. Both sideshave 180 foot/pounds of leverage.

BENDINGThese forces exert th emselves in avariety of ways. Figure 1 shows al-inch- square balsa strip beingbent; all four forces come in to playhere. The fibers on the outside ofthe bend are bein g stretched­under ten sion . On th e inside ofthe bend, th ey'r e bein g pushedtogether under compression. Theseopposing forces develop shea r. Inour balsa strip, tha t shea r acts on aline mid way th rough called th e"neutra l axis."

Now look at Figure 2, illustrationA. This shows th e end view of thel-inch-square balsa stick. The neu­tral axis and the leverage from thecenters of th e balsa areas above and

below th e neutral axis are shown.Consider Figure 2, illustration B.

The beam is composed of balsalx7/i6-inch upper and lower flangesjoined by a V16-inch-thick balsa webwith its grain vertical. Both A and Bhave the same cross-section areas.

Obviously, th e "leverage" fromth e neut ral axis to the flange centersis greater in B than in A. B will besubstantially stro nge r than A inbending because the material is far­ther from tile neutral axis.

The balsa web in B is under shearin th e bending of th e beam . Balsa ismuch stro nger in shear across thewood grain than along the grain;and stronger along th e grain in bothtension and compression .

Consider Figure 3. It displays thesame beam as B in Figure 2, butwithout th e balsa shea r web-andas part of a wing structure underflight loads. The upper flange isunder compression , and the loweris under tension .

Failure will occur by the upperflange buckling as shown in Figure3, illustration B; and in the absenceof the web, th e opposing forces willdistort th e structure.

With th e vertical-grain shear webin place, the buckling is resisted , asare the shear loads. These webs addmu ch strength for littl e additionalweight.

Obviously, th e farthe r apart theflanges are, the stronger th e beam;or, by reducing flange size andweigh t, obta in the same strength .

A thicker wing can be madestrong but light; its spar flanges arefarther apart and smaller.

TORSIONTorsion is composed of shear andtension. In Figure 4, a tube is beingtwisted in opposite directions at itsends. The arrows in th e center showopposi ng shear forces; th e twistingtends to elongate th e fibers intension .

58 THE BASICSOF RIC MO DELAIRCRAFT DESIGN

Stressed-Skin Design • CHAPTER 13

The Sea Loon-a .15ci-powered twin-boomflying boat. Flaps arefully extended.

WINGS, A ILERONSAND SLOTTED FLAPSFigure 7 details the wing structureof th e "Swift"-a model with slot ­ted flaps that's designed for lowdrag . Its aerod ynamic design was

The Snowy Owl has anexternal glow-plugpower plugin thejack. Plug removal issalelyaway from the dangerous rotallngprop. It's AUcipowered,

wood grain; and th e skin aids thespar s in tension and compressionloads parallel to the grain .

Horizontal and vertical tail sur­faces have to contend with , princi­pally, bending loads as elevatorsand rudd er operate, The same struc­tur al principles apply.

Fuselages enco unter a wide vari­ety of loads in flight and particular lyon landing. A tubular structure isbest able to resist the heavy bend­ing, twisting and tension loads, Inbalsa, a tubular or oval well-stream­lin ed fuselage is difficult toprod uce. In fiberglass, it can bedon e, but th e molds requ ired areexpe nsive for "one-off" models.The compromise, in balsa, is flatsheet sides, top and bottom withgenerou s corner radius, This comesclosest to th e local round or ovalcross-section.

It always surprises me to findhow stro ng stressed-skin structuresbecome after assembl y of pieces offlimsy balsa . Built straight, they donot warp . Models built 10 years agoare in flyable condition today.

impo se loads th at, on larger models,require a seco nd spar in fron tof th ese surfaces, with some tor­sion-resi sting structure . Full balsashee ting in YI6-inch balsa skins ,top an d bottom of the wing, main­tain s the airfoil sectio n and adds

little weight, but con­siderable st reng th.Ribs ma y be "cap­stripped" bet weenspars with th e cover­ing sagging betweenthe ribs, reducingthe airfoil's integrity.Both fully an d par­tially shee ted wingsare covered wit hyour choice of mate­rials . The grain of th e1/16-inch skin runsparallel to the spanto resist torsion anddrag loads across the

result fro m theairfo ils' pitch in gmoment and from thetwist in g act ion ofailerons in opposited irecti ons and thenose-down loads offlaps when extended.These loads are allsubstantially increasedin high-speed maneu-vers such as steepturns , sharp pull-ups,etc. where centrifugal

forces come into effect.The D-spar structu re of Figure 6 is

designed to resist all th ese loads. Itcombines a cylinder and a beam.Note that the materi al is as far fromthe neutral axis as possible and thatth e beam is close to the wing 'sthi ckest point.

Ailerons and flaps, as mentioned,

A I~ 1-- 1' ~ r-'/Y]'

I ' /y]' , . . - 1&r-1'--j I.iM!r •1 .: ' ' , ' 1 I

Neutral Axis- 1'

..

1 Ii. . uLI' ,B --L

Flanges /

B.

A.

Figure 1.Bending.

Figure 3.Flanges buckling under load.

In Figure 5, A is a solid cylindri­cal rod; B is a hollow cylinder withthe same cross-sectional area ofmaterial as A. Again, obv iously, B ismuch stronger in torsion and bend­ing than A because of the mat erial 'sgreat er leverage from th e neutralaxis.

Th ere is a limit to this leveragelength, i.e., the point at wh ichyou can still retainthe same cross-sec­tional area of materi­al; beyond this limit,the ou ter skin wouldbecom e so thin that itwould fail by localbuckling under load.Full-scale airplaneshave thin-skinned fuse­lages rein forced bylateral frames and lon-

Figure2.gitudinal stringers to Beam construct/on.resist buckling.

A beam such as thatin Figure 2, illustration B, is weakin torsion. Figure 6 illustrates thisbeam in a win g. An airp lane win g,in addi tion to bending loads fromlift, must resist drag and torsionloads . Drag load s are due to airresistance or drag . Torsion loads

THE BASICSOFRIC MODEL AIRCRAFT DESIGN 59

CHAPTER 13 ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

9 .75" chord

Y1." plywood slot lip

.i>:

Aap p ivot point

Sldnjolnts

A. Swift Inboard wing and flap sectio n

y1." Balsa skins -E197 lop and bottom

Figure 6.O-spar wing structure.

A . Spanowhawk wing and flap section

10 ° down

Hinge

V32" balsaskins

V32" balsaskins

Double MonoKole 1h Inge 20 up

'I." balsa

V. " A alsa

V." A balsa

Vo" bal sa

Vo" balsa

B . Swift outboard wing and aileron section

Webs - GR vert.

E193

E193M

10.1 " chord

...- Yl. ·· X t~4 ·· ca

I ("P!ionalJPa

- .

I

V,o" balsa skins

V,o" balsa skins

E197 with NASA''droop''

Figure 7.The wingstructure of theSwift.

described in Cha pter 12, "ImprovePerform an ce by Reducing Drag."The Swift's structure is based onthe principles outlined previouslyin this chapter. "A" is a section cutthrough the flapped portion, and"8" is cut through the aileron andNASA "drooped" LE.

Figure 7A shows th e Swift's two­spar wing with vertica l-grainedwebs running from top to bottomflanges and between th e wing ribs.The 3/i6-inch-square LE spar addslittle strength but prov ides gluingsurfaces for joining top and bottomVI6-inch balsa LEskins. The aft sparabsorb s th e flap drag and lift loadswhen flaps are extended.

Figure 78 sho ws th e structure atthe ail erons design ed to resistaileron twistin g loads. The diago­nal VI6-inc h balsa sheet runningfrom the lower flang e of th e aftspar to th e upper skin stiffens theaileron attach me n t point. Th eailerons and flaps are simple boxstructures.

B . Spanowhawk wing construction

Figure 8.The wingstructure of the Sparrowhawk.

Figure 4.Tube under torsion.

Figure 5.Round structures.

The Swift's ailerons are of modi­fied Frise design. With equal up anddown travel of "bam-door" ailerons,th e downward extension producesmore drag than th e upward one .This uneven drag pulls the wrongway-out of the turn-and require scoordinated rudd er to correct theresulting adverse yaw.

The Swift's aileron s have differen­tial travel-the upgoing moves twicethe angle of the downgoin g. Also,the lower forward lip of th e upgoingaileron projects into the airstreambelow the wing, producing favorabledrag as in Figure 7B. .

These two factors combine

to pro duce "in to-the-turn " yaw.Rudder action isn 't needed; th emodel turns on aileron action .

Both aileron s and flaps of thi scon struction are stro ng, stiff units.Note the lead-wire, aileron massbalance.

The win g cente r sectio n is open,with the center section main andaft spars runnin g across the fuse­lage. This leaves the cen ter sectionfree for in stallati on of aileron andflap servos where th ey're accessib leby removal of th e canopy as inFigure 10. It also provides access toth e elevator, rudder and en gin eservos in th e fuselage.

60 THE BASICS OF RIC MODEL AIRCRAFTDESIGN

Stressed-Skin Design .... CHAPTER 13

Removable canopy!.- 5.1 " chord j

IY," b,balsa ~ ~~~~~~

",AA~f~~-JY1. ·· ba lsa sk ins V,"= re ...... .::::: ~

A. Swift stab and elevator construction

A

A2 3 4 5

B

6 7 - B ulkheads

Swift fuselage construction

0/,,"00 1sa si des,top an d botto m

I

Fuselage section BoB

th e fuselage top edges (Figure 10)reinforce these edges, along with tri­angular gussets at th e upper-fuselageto bulkhead corners, as shown.

LANDING GEARBoth main and nose-gear struts ares/32-inch-diameter music wire. Fair­ings have to be added and shaped tostreamline cross-sections.

The nose strut has a shock-absorb­ing coil that's entirely inside th efuselage for low drag. The mainstruts have a square "U" in that

All bulkhead partsYa" balsa

(No te grain direction)

Fuselage section A-A

0/," b,

~;::=~===:::::::::::::L~~~ t........ gusset,'/2" thick 0/,,"001sa

sides,

---~tt';,~

Figure 10.Swlfffuselage construction.

Figure 11.Typical fuselage sections for models with .40 to .BOci engines.

.60ci engines.The sides, topand bottom areall 3/32-inch firmbalsa sheet withth e grain run­ning length wise

of th e fuselage. The generouslyradiused corners are of 31I6-inch balsasheet and are as far from th e neutralaxis as possible.

The typical bu lkhead is com­posed of four pieces of lAl-inc h balsath at are cemented together at theove rlap p ingco rners. Noteth e wood-grainorientation .

The firewall Canopy

/

pa rti ngis 31I6-inch ply- 0/,," line

wood and does dou~tlX/7Jfh:::::====':::::::=JHmrt1 "trip le duty. Infront are motormount andcowl, and land­ing-gear nose­wheel bracketsare on the rear.The wing andlan d in g-gea ra ttac h me n tbu lkheads arebalsa with ply­wood reinforce­ment. The eas­ily removablecano py andtop in Figure10 weaken th efuselage struc­ture. Ben eat hth e wing, thefuselage is rein ­forced by th efour-bolt, wing­t o - f u s el a g eassembly.Doublers along

Y." x V. " balsa-, J'" ...

...... ::,g ra in Va' bal sa .... ....... ::::: ,.

B. Swift typical fin and rudder construction

Figure 9.Typical cross-sections of IheSwlff's tail.

This open center sectio n leaves itrelatively weak in tors ion . How­ever, th e wing is firmly bolted tothe fuse lage struc ture at fourpoin ts. The torsion load s areabsorbed by th e fuselage structure,as are th e main landing-gear loads .

FUSELAGEFigure 10 provides an outline of th eSwift's fuselage construct ion andFigure 11 shows typical fuse­lage sections for models with .40 to

HORIZONTAL ANDVERTICAL TAIL SURFACESFigure 9 details typical cross-sec­tions of th e Swift' s tai l. "A" displaysth e stab and elevator sectio ns. Thestab has one spar with tri-stockrein forcing th e up per skin at th eelevato r's doubl e Mon oKote hinge.

Eleva to rs are composed ofJ"s-inch balsa L.E. spar and l!J6-inchbalsa skins, top and bottom . Ribsare 3/32-inch balsa sheet.

Because th e horizontal tail ismounted on top of th e fin, th e finstruc ture incorpora tes a spar andshea r web, as in "B," to absor b theloads imposed by this T-tail loca­tion . The rud der const ructio n issimilar to th at of the elevator's.

Figure 9, illustrat ion A'scon struc­tio n has been used successfully onsma ll model wings of up to 7-inchchord, as shown in Figure 8A and B.Flaps, ailerons, stabs, elevators , finsand rudders of th e sma ll models areall skinned in 1/32-inch balsa shee twith llI6-inch balsa ribs.

THE BASICS OF RIC MODEL AIRCRAFT DESIGN 61

CHAPTER 13 ... THE BASICS OF RIC MODEL AIRC RAFT DESIGN

Notes: theOsprey was notfully sheet covered and, hence, was lighter. The Swan had12ounces of leadballast in thenose toposition theCG in thedesign location. The Wasp hadonlyfourservos, not tive.

Over the years , this author hasdesigned, built and flown 14model aircraft, all RIC, and all ofthe type of stressed-skin structuredescribed in this series. These aredetailed in the Table, "14 StressedSkin Designs" and plotted on theaccompanying graph.

The total weight of the 14 was1,151.75 ounces, and their com­bined wing areas totalled 7,359square inches; the weight persquare inch of wing area was0.1565 ounce. Power loadings(ounces per cubic inch of enginedisplacement) varied from 200 tojust over 300 ounces per cubic inchdisplacement. A model that has 625square inches of wing area wouldweigh an estimated (625 x 0.1565)or 97.8 ounces.

Obviously, the lower the powerloading, the greater the power-to­weight ratio , and the better theclimb performance and top speed .

Anyone interested in designing amodel to these structural princi­ples, in the 0.15 to .46ci range , willfind this tabulation a useful guide.Stressed-skin design results in theoptimum weight-to-strength ratio.

There are logical justifications forall the Swift's design features­except one-the styling of thelower rudder TE. The author justlikes it that way! ...

WEIGHT ESTIMATINGEstimating the weight of a modelairplane while it's still in the con­ceptual stage is an importan t anddifficult decision.

14 Stressed-Skin DesignsGross Wing Wing Power

Eng. Model weight area loading loadingModel disp. type (oz.) sq. in./ oz./sq. ft. oz./ci

sq. 't.1. Seahawk 0.46 Sporttrike 110.0 655/4.54 24.22 239.0

2.Seagull III 0.46 Ryingboat 112.0 694/4.81 2328 243.0

3. Swift 0.46 Sporttrike 92.0 60014.16 22.11 200.0

4. Osprey 0.45 Tail-dragger 113.0 768/5.33 21 .2 251.0

5. Swan 0.45 Canard 115.0 669/4.64 24.78 256.0

6. Crane 0.45 STOl trike 101 .5 643/4.46 22.75 226.0

7. Gull 0.40 Sport trike 93.0 643/4.46 20.85 232.5

8. SnowyOwl 0.40 Sport trike 104.0 643/4.46 23 31 260 0

9. Canada Goose 0.35 Canard 75.0 44413.08 24.35 214.0

10. Flamingo 0.35 Flying boat 74.0 500/3.47 21 .32 2110

11. Sparrowhawk 0.15 Sport trike 38.0 25011 .73 21 .96 253.0

12. Wasp 0.15 Tandem wing 36.3 30012.08 17.42 242.0

13. Sea Loon 0.15 Flying boat 42.0 250/1.73 24.27 280.0

14. Skylark 0.1 5 Spor trike 46.0 300/2.08 22.11 307.0

800

700

600

500

400

300

200

100

0 10 20 30 40 5 60 70 80 90 100 110 120oGross weight, In ounces

Average weight: .1565 ounces per square Inch of wtng area

portion in th e fuselage; the horizon­tal legs are shock-absorbing torsionbars that distribute land ing loadsover the same two bulkheads thatabsorb wing loads.

62 THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Design for

Flaps

Chapter 14

result s from th e high flap dragwhen the flaps are fully extended.

Stalls-flaps down-are at 17mph.On a low-wind day, full-stall slowland ings are pure fun-like a birdlanding on a branch-and groundroll seldom exceeds 4 feet.

With a lOx7.S prop , the SnowyOwl's top speed is estimated at7Smph. It's fully aerobatic, but itrefuses to do more than one or two

turns of a spin, which is then con­verted into a fast spiral dive (cour­tesy of the NASA droop) and fromwhich recovery is prom pt upon neu­tralizing the controls.

On a windy day, it will hover,alm ost moti on less, flaps fullyextended, engine throttled back andwith full up-elevator. Aileron con­trol is still effective in th is nose-highaltitude, and no tip-stalls occur.

A n RIC mod el designedspecifically for flaps ope nsup a new and exci ting

dimension in sport flying. This air­plane will be fast , structurallyrugged and well-streamlined, and itwill have a higher-than-usual wingloading; but with flaps lowered, itwill land at tra in er speeds ofaround 20m ph. It will also have avery wide speed range!

This chapter will first deal withthe design of a model th at will useflaps; then it will detail th e designand actuation of th e flaps th em­selves and give tips on flying withthem.

To illustrate the featur es of amodel designed for flaps, considerthe Snowy Owl (see sidebar). Thisplane was built IS years ago and isstill flying. Powered by an old 040engine, it weighs 104 ounces, has awing area of just under 41;2 squarefeet, a wing load ing of sligh tly lessthan 24 ounces per square foot, anda power loading of 260 ounces percubic inch of engine displ acement.It features th e NASA "safe wing"droop modification.

This model's performance hasproven to be bett er than any otherAO-powered model encoun tered sofar. Takeoffs- flaps half extended­from grass requ ire no more than10 feet with a fast steep climb.Landing approac hes- flaps fullyextended and engine idling- maybe very steep (almost vertical) with­out significant acceleration . This

THE BASICS OF RIC MODEL AIRCRAFT DESIGN 6 3

CHAPTER 14 • THE BASICS OF RIC MODEL AIRCRAFT DESIGN

The Seagullll/-anamphibious flying boat.Powered by an0.5. Max 46SF engine, itweighs 113 ounces andis anexcellentperformer. Its large slotted flaps arefullyextended for landing.

It's great fun to make a lowpass-flaps fully exte nded; enginethrottled; nose-high attitude­at about 2Smph, followed byanother pass with flaps up andengine wide open .

The Snowy Owl's speed rangeis remarkable. Maneuvers-flapsdown-are very tight indeed. On aday with littl e or no wind, do n 'tattempt to land the Snowy Owlflaps-up , because the glide is fastand very flat, and you could easilyovershoot the flying field.

On the other hand, landings ona very windy day should be madeflaps-up . The high wing loadingprovides good penetration, and thehigh airspeed gives good control.Thanks to the NASA droop, thereare no wing-tip-stalls when maki ngnose-high landings.

SLOTTED·FLAP DESIGNLet's make a bold stab at design inga wing for a slotted -flap-equippedmodel call ed the "Swift ." Toa greater exten t than the SnowyOwl, it will take advantage of th elift-increasin g capacity of theextended flaps.

For th is project, the chosenwing loading is 25 ounces persqua re foot of wing area. This ishigher than Snowy Owl's andshould result in a smaller, ligh te rmodel with even lower drag. Bycomparison, a gross weig ht (withfuel ) of 100 ounces see msreasonable. The wing area wouldthus be 100 divided by 2S to eq ua l4 square feet, or S76 squareinches. The Swift is powered bya .46 engine, and its power load­ing is 217 .3 ounces/cubic inchdisplacement.

For this project, test-fly withlOx9 and lOxlO props to select th e

64 THE BASICS OF RIC MO DELAIRCRAFT DESIGN

one that performs best for thismodel. At 1l,000rpm, a 10x9 propwould produce an estimated topspeed of 90mph.

Figure 2 shows th e actual dimen­sions of the Swift's wing and theproportions of its features. With a

wing loading of 2S ounces persquare foot and a CL max of 1.933,this model will stall at just under18mph at sea level. If you have theCL for a particular airfoil and wingloading, stall speed can be estimat­ed quick ly by using th e curves

Design for Flaps ... CHAPTER 14

Sparrowhawk is a 15-powered airplane with a wingarea of250square inches and a wing load­ing of 22 ounces per square foot. It's nimble and fun to fly.

ward surface of th e flap and theunderside of th e slot lip shouldconverge or na rrow stea di ly fromthe slot en try in th e wing under­side to th e exit over th e flap topsurface. Th is accelerat es the air­flow ove r th e flap, del aying itssta ll and im proving its lift. It 's thereason the slotted flap is superiorto either th e split or the planeva riety.

• Air flowin g from th e slot shouldmerge smooth ly into th e air flow­ing around th e wing and the flap .

• Having an appreciable length ofslot lip on the upper wing surface isadv antageous.

shown in Figure 3 of Chapter 3.Add 20 percent to this sta ll speedfor a safety margin, and this mod elwould be capable of tou ch in gdo wn , no se-hi gh at 22mph under"no-wind" conditions. This is acomfortable landing speed .

Well -developed flap s o n amodel designed specifically forflap s will produce an aircraft thatha s high top speeds and is verystro ng and rugged . It will alsoha ve a very wid e speed ran ge, andthis will permit slow landings(flaps-down) and flight at anyspeed desired within that speedrange. The plane will be more ver­satile than the ave rage sport .40and much more fun to fly.

CUIDELINES• With flap extended, th e slotformed between the upper for-

The Osprey is a tail-dragger. Powered byanO.S. Max45 FSR, it weighs 113ounces andhas awing loadingof26.5 ounces persquare foot. Under "no-wind" conditions, it takes off fromwater in less than40 feet onfloats.

- - - - - - - 29.25"--- - - - - - -.1

IC .80 C

_1 I __ L _.25 C

...----.65 A- -'-- - i. }.'V\!...I+- ---''---

Proportions

.35 A---t1+-- - - - - - - - A- - - - - - - - I

.15C-j

1+-------------- - - -'-- - - ---- 58.5 01..-- - - - - - - - - -1

Figure 2.Outline of the Swift's wing (576 square inches: aspect ratioof5.94).

THE BASICS OF RIC MODEL AIRCRAFT DESIGN 65

CHAPTER 14 ... THE BASICSOF RIC MODEL AIRCRAFT DESIGN

t----------C-----------,1-- - - - - - - -.80C- - - - - - - - -

HORIZONTAL TAIL SURFACESWhen slotted flaps are extended inflight, a num ber of things happ en:

• Wing lift increases substantially.

the bottom. Each flap has twopivot ribs and one horn rib-allmade of plywood; the rest of theribs are made of 313z-inch-th icksheet balsa .

The form of slot entry shown inFigure 3 was used on Snowy Owl.Although this smoothes the airflowinto the slot , it leaves a drag-pro­ducing gap when the flap is retract­ed. Later designs simply have thelower wing skin extended to theflap's LE (see Figure 5) without anyapparent adverse affects.

.25C-

Slot lip~_.£Arc of a cir Ie

/ #1/16 Rad ius

Slot

Figure 3.Slotted flapproportions.

Figure 4.Flap In the40-degree-down posit/on. • Wing drag also increases, and this

slows the model.

• The nose-down pitching momentincreases .

The outcome of these force changesis some degree of nose-up pitch.This is overcome by applying nose-

1/16" balsa

.....Pivot

3132" ply

1,.;6.112" plyslot lip

Cross-seellon/cutaway line

• The angle of the downwash fromthe wing and the lower flapincreases sharply; this impacts onthe horizontal tail at a negativeangle and leads to a tail downloadthat induces a nose-up pitch.

3132" music-wire pivot

Rib31.l2" balsa

Enlarged cross-section of flapsupport-pivot rib.

Figure 5.Wingandslotted-flapconstruction andhinging.

1/16" ply

FLAP CONSTRUCTIONAND OPERATIONFigure 5 details the structure ofboth the wing and the flap. The1116-inch-thick plywood flap sup­ports and the 313z-inch-th ick ply­wood pivot and horn ribs areshown in Figure 6. The enlargedsection of the "Flap support-pivotrib" shows the sanding required tostreamline this assembly.

The flap has 1/16-inch-thickbalsa-sheet skins on the top and

Figure 3 provides the proportionsof a slotted flap for the Eppler 197airfoil that conform to these guide­lines. This is based on proportionsdeveloped in the wind-tunnel testsoutlined in NACA Report 664, FlapType lb.

This flap extends by rotatingaround a fixed pivot, to 40degrees. Note that only the topfront and LE curves are added toform the flap's profile; the rest areprovided free by th e wing profileitse lf.

Figure 4 shows th e flap in the40-degree-down position and pro ­vides the proportions of th e slotgap and the slot lip overhang.These proportio ns are importantfor good flap performance.

Positioning the pivot point sothat the flap-up and 40-degree­down positions coincide with thoseshown on the drawi ng is done by asimple trial-and-error method.

Trace the flap profile and chordline on translucent material suchas onion-skin paper, tracing paperor drafting film. Lay this tracingover the flap drawing in the lipposition. Using a pin as a pivot,rotate the tracing so that the flapextends. Trial and error will guideyou to a pivot point where thetracing coincides exactly with thedrawing of the flap, in bo th the upand the 40-degree-down posi­tions. Mark this position carefullyon your drawing.

66 THE BASIC S OFRIC MODEL AIRCRAFT DESIGN

Design for Flaps ... CHAPTER 14

TAIL SURFACE AIRFOILAND STRUCTUREFigure 9 shows details of the tail­surface airfoil and the structuraldesign used on several successfulmodels. The depth of this sectionprovides a very strong, light, simplestructure with low drag. The sameprinciples of airfoil and structureapply to the fin and the rudder.

wing and flap downwash angledecreases to roughly half of theangle at higher altitude. Th isreduces the tail download propor­tionately. This occurs at a badpoint; the tail download should beincreasing to raise the nose to ahigh angle for a slow landing.Powerful elevators are needed toproduce the tail download required.

An elevator area of 40 percent ofthe total horizontal tail area with atravel of 30 degrees up and down isrecommended for a model that 'sequipped with slotted flaps .

In normal flight-flaps up­these large elevators may be sensi­tive at first, but with experience,you'll adjust to them.

FLUTTER PREVENTIONWell-streamlined model aircraftwith fairly high wing loadings andpowerful engines can achieve veryhigh speeds, particularly when div­ing . This invites the very real dan­ger of control-surface flutter, whichcould destroy that surface veryquickly and would probably resultin a disastrous crash .

This is particularly true of thewide-chord control surfaces inher­ent in "designing for flaps." Theonly certain way to prevent flutteris to offset the weight of the controlsurface behind its hinge withweight in front of the hinge, withboth weights balancing at thehinge line.

The modified Frise aileron shownin Figure 7 lends itself to mass­balancing very easily. Shielded hornbalsa tips on rudder and elevatorpermit this mass-balancing (seeFigures 9 and 10). Flutter preven­tion for flaps has proven to beunnecessary. Thanks to theirstressed-skin construction, wingsand tail surfaces are torsionally verystiff and free of flutter.

See also Chapter 20, High-LiftDevices and Drag Reduction," for

Hornrib

~'116 D (2 required)

A T-tail operates in air that'sonly lightly disturbed by thedownwash . It's thus more effectivethan a lower tail , which is in airthat's disturbed by the fuselage, inheavier downwash and in theprop's slipstream. The T-tail ismore affected by the increasein downwash angle on loweringth e flaps.

GROUND EFFECT ANDELEVATOR DESIGNIn ground effect , at an altitude ofless than half the wingspan, the

(4 required)

Pivot ribFlap support

Figure 6.Flap plywood componenls.

down trim by means of the elevatortrim lever while simultaneouslylowering the flaps. With a littlepractice, this becomes almostautomatic.

The nose-up pitch varies withthe speed at which the model isflying when you lower the flapsand the extent to which they'relowered.

Experience has proven that T-tailmodels, e.g., the Snowy Owl, pitch­up to a greater degree than those inwhich the horizontal tail is in thefuselage, e.g., the Osprey.

THE BASICS OF RIC MODEL AIRC RAFT DESIGN 67

CHAPTER 14 ... THE BASICS OF RIC MODEL AIRC RAFT DESIGN

./

Sullivanstandard cables (.056" Dla.)

\'<

1f16-inch balsa rib

th e benefi ts of O.30c chord slottedflaps.

Flying RIC m odel ai rcraft ischallengin g, exciting and fun . Ihop e th at "flapp ed flying" will addto your enj oyment of th is sport . Itha s for mel s,SIDE VIEW

sl16-inch triangle (cutaway)stock balsa

Double 1__---- .40 CMonoKote hinge elevator

• "'-...: -----::0 'l1I-inch balsaspars

'l1I-inch-dia. lead wire~

1,!16-inch-thick balsaskin-,..

Tiphorn detail

Figure 9.Typical tail surface construction- E168 airfoil.

Stabilizer

Elevator

Balanceat hingeline

'111" dia.lead wirebalance

---- __-.J TOP VIEW

Figure 10.Typical shielded horn andmass balance for elevator andrudder.

68 THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Chapter 15

The SnowyOwl in slow-speed f1ighl with flaps extended. The increasing leading-edge droopahead of theailerons is clearlyvisible.

Figure 1.Classic stall/spin flight path, frequently fatal. Wrong way to "hit"therunway.

markedly. Lowering an aileron tointroduce a roll input at this angleincreases the wing 's AoA at thataileron, and may cause it to stall­just the opposite of the action com­manded by the pilot.

A TRAGIC SCENESuppose an inexperi enced pilot isflying a high-wing aircraft . He's in aleft-hand pattern for landing at abusy airport, and a light crosswindis blowing from left to right. Afterturning onto th e base leg of hisapproach, he slows th e airplane bythrottling back and increasing itsAoAby applying up-elevator. Whilescanning the area for other traffic,he lowers the flaps, trims the air­craft and announces his intentionto land.

At an altitude of 300 feet, heturns left again onto finalapproach, and our inexperiencedaviator finds that th e crosswind hasmade the plane drift well to theright of the centerlin e. To correct,he cranks in more left aileron tosteepen his bank, and he adds up ­elevator to accelerate his turn; bothincrease the centrifugal load. As theaircraft is realigned with the run ­way, the pilot applies heavy, rightaileron to straighten up. The down­aileron (left) wing stalls, and overhe goes to the left as the planestarts to spin . Unable to recover atthis altitude, he becom es anotherstatistic.

In an attempt to remedy thespin/stall syndrome, a variety ofwing modifications were tested by

NASA

··Safe Wing"

0)1

speed. At the same AoA, doublingthe speed increases lift fourfold.Also, lift varies directly with theAoA, from the airfoil 's zero lift angleto its stalling angle . In high-speedflight, the wing operates at a lowAoA; at low speed, that angle mustbe increased to maintain levelflight . The stalling angle of thewing's airfoil determines the low­est speed limit .

Centrifugal force plays a signifi­cant part in stalls and spins becauseit increases the weight that the wingmu st support. It's encoun-tered whenbanking steep ly, sharply pulling upin to climbs, and when you panicand use full-up-elevator when pulling

out of dives at lowaltitude.

For example,a full-scale Cessna172 at gross weightstalls at S7mph.In a 6O-degreebanked turn, its stallspeed increases by42 percent to81mph, and this isdue entirely to theextra load imposedby centrifugal force.As a normal wingapproaches the stall­ing angle, aileron­control effective­ness deteriorates

//

-Actual lllght path

H ere's a grim statistic: roughly30 percent of all fatal acci­dents involving light, full­

scale airplanes are caused bystalling and spinning at low alti­tud es, and ground impact occursbefore th e spin fully develops.Several members of my club havediscovered that R/C model aircraftare also prone to this insidious fail­ure. What's happening?

As a private pilot, I've been inter­ested in wing modifications thatwill improve th e stall/spin charac­teristics of both full-scale and R/Cmodel airp lanes. Most modelersknow that a model's wing lift is pro­portional to the square of its air-

THE BASICS OF RIC MODELAIRCRAFT DESIGN 69

CHAPTER 15 .fa. THE BASICS OF RIC MODEL AIRCRAFT DESIGN

r--- ----- 23'-3" ----------.,

NASA's LE droop has been suc­cessfully inco rpora ted in to sevenR/C model aircraft: th e .IS-po weredSparrowhawk; the AO-poweredSnowy Owl II; the .IS-powered SeaLoon (a flying boat); the Swift; theSeagull III; the Seahawk; and theOsprey, which is a AS-poweredcraft designed to be used with bothwhee ls and floats.

While th e smaller models can beforced to spin, onl y one or two turnsare achieved before the spinbecom es a spiral dive, and recoveryis inst an taneous when the controlsare neutralized. Aileron control isgreatly improved in the stall, withth e flaps up or down. Despite man yatte mpts, I haven 't been able tospin th e larger models.

As the illustration of the airflowover the NASA wing shows, the out­boa rd, drooped panels become very

high insur an cep rem iu m s ,they're buildingfewer, full-scaleligh t airp lan es.Verilite AircraftCo. Inc. hasdeveloped anew design thatincor pora tesNASA's LE mod i­ficatio ns . TheSunbird (Figure2) is th e first air­craft designed toprovid e spinresistance andth ereby reducestall/s pin acci­den ts. NASA hasrun extens ivewin d - tun n e I Figure 2.tests on th is air- Verilite AircraftCo. Inc. Sunbird.craft, and it hasbuilt and tested a sma ll scalemodel, a lA-scale R/C model and afull-scale version . A 28-degree AoAwas recorded before th e stall wasencountered.

On th e previous page, a photo ofmy Snowy Owl (one of my earlier

models) is in slow­speed flight withits flaps extended.The increa sing LEdroop ahead of theailerons is clearlyvisible, an d itreached its maxi­mum at the wing­tips. This modifi­cation succeededin delaying th estall, but theaile ron s provedineffective in th eatt itude shown.

The Sea Loon in its natural element-water. The leading-edge droopstartsat the inner-wing stripe.

The Osprey, powered by a .45 diesel, about to start its takeolfrun.The leading-edge droop shows clearly.

NASA'S SOLUTIONIn th e late '70s , NASA's AmesResearch Center initiated a programto develop an improved LE thatwould be inexpe nsive to manufac­ture and would require no mainte­nance. After determining the bestwing modification th rough exten­sive wind-tunnel tests, NASA incor­porated these design changes into anR/C scale model. Stall/spin character­istics were significantly improved,and , to confirm these R/C modelresults, four, full-scale light aircraft­a Grumman American Yankee,Beech Sierra, Piper Arrow and CessnaI 72- were mo dified and flownextensively.

Because ma nufacture rs pay such

aerona utica l eng inee rs: fixed orretractable LE slots; wing washoutto reduce tip angles; greater camberat the wing tips and slot-lipailerons. While modi fication s didimprove stall behavior, they alsoagg rava ted spin ch aract erist ics.Many of th ese changes worsenedaircraft performa nce and increasedthe complexity and cost of con­struction and maint enan ce.

70 THE BASICS OF RIC MODEL AIRCRAFTDESIGN

NASA " Safe W ing" ... CHAPTER 15

th e cross-hatched section, and alight LE spar. Cove r th em withbond pap er or th in balsa, and gluethi s unit to th e outboard wing LE. Ihaven 't tried th is droop on sym­metrical airfo iled wings , but itmight delay th e stall in bothupri ght and inverted flight (seeFigure 6C).

Congratulatio ns, NASA, for yourmaj or contribution to avia tionsafety. I hope th is "safe" wing willbe incorp orated in future aircraftdesigns. ...

B. Semisymmetrical airfoil

A. Flat-boUom airfoil

Chord line

Chord line

t------------C -----------~

...--- - --- - - - - - C- - - - - - - - - - - --l

-------- - - - C - - - - --1

Chord line

C. Symmetrical airfoil

Figure 6.NASA droop (cross-hatched areas) onvarious airfoils.

low-AR wings, with a stall that'sconsiderably delayed. The droopitself, which delays the stall toapproximately twice the stall angleof the basic wing, permits effectiveaileron contro l at th e higher AoAs.

If you fly models with flat ­bottom or semisym metrical airfoil s,

you could modi fy th e wings byadding droop . (See the cross­hatched areas in Figure 6 A and B).For evaluation purposes, I've do nethis by using Styrofoam, wh ich isheld in place with transparent tape.

As an alternative, you could addbalsa ribs like th e ones sho wn in

0.388/2=1 Vortex8/2

.........0;..~ f-----------,.--- - ---,

ICenterline Centerline

Figure 7.The wing planform showing theproportionsof theadded leadlng­edge droop. Note that the corners formed by theInboard end ofthe droop mustbesharp where the droop addltlon meets thenormal alrfoll.

Figure 8.The airflowover theNASA wing at highangles ofattack. While theInboard, undrooped section Isstalled, thesharp-cornered notch In theleadingedge produces a chord-wise vortex that effectively separatesthetwo areas.

THE BASICS OF RIC MODEL AIRCRAFT DESIGN 71

Chapter 16

Landing-Gear

DesignT he landing gear of a pro­

peller-driven aircraft hastwo major functions. The

first is to provide adeq uate clear­ance betwee n prop tips and theground. The second, and no lessimp ortant, is to permit th e plane torotate on both takeoff and landingso that the wing's AoA comes closeto the stalling angle of its airfoil. Atthat AoA, the wing is near the air­foil's CL max . This permits the low­est landing and takeoff speeds ofwhich the model is capable.

On the gro und, however, itshould no t be possible to rotate toor beyond the wing's stalling ang le.Such a stall on takeo ff or landingcould be damaging, both to themodel and to its designer's ego!

For windy-day flying, good judg­ment dicta tes flaps-up landings,and at a lower AoA for good con ­trol. The wind's speed reduces th emodel's ground speed accordingly.

This chapter dea ls with th e land­ing-gear function . Int elligent deter­mination of the AoA for landing

1. 6

and takeoff requires considerationof the following:

• The airfoil's characteris tics andthe Rn at landing and takeoffspeeds.

• Adjustment of "section values" tothose for your wing's AR and plan­form .

• The impact of ground effect.

• The effect on the stalli ng angle offlaps when extended.

• The wing's AoA in level fligh t. Ifth at angle is 3 degrees and the land-

The Wasp tandem wing. The prop'sposition,just behind themainlanding gear, hasnoclearance problem.

6 10 14 18

Angle ot a lta ck

(URlmsity of Stuttgart,

.05

·. 6 . 15

. ... •1

1 .6 -..Pitch ingmoment

1.' -.3/ coelllclent(CM)

.1' 1 .2 · .3~"

1 .0 - .25

Rn ,=

--1 00 ,000

= 200 ,000

- -250, 000

Dra g coeff ic ient (CD)I

.06 .08 .1 .12

...·.e

-.2

1. 0

1 .2

1.'

~ .8

e

=::;.2

=.! .6..::..8 .4

Figure 1.Airfoil data for Eppler 197.

72 THE BASICS OF RIC MOD EL AIRCRAFT DESIGN

Landing-Gear Design ... CHAPTER 16

5040

Drooped leadingedge

/

GROUND EFFECTThis phenomenon starts at half themodel's wingspan above thegro und (or water) and becomesmore intense closer to the ground.Both landings and takeoffs, hence,are made in "ground effect." It actslike a subs tantial incr ease in AR.A reduction in the stall AoA and in

HIGH·LIFT DEVICESSlotted flaps reduce takeoff andlanding AoAs (as shown in Figure 7of Chapter 3). A 20-degree flapdeflec tion causes a reduction of 1degree , but for the full 40-degreedeflection, it is 4 degrees . Sincelandings are more cri tica l thantakeoffs, use 4 degrees . As one for­mer jet fighter pilot pu ts it,"Takeoffs are optional; landings areun avoidable."

II

" I" "-i/'1'- --

IIIIII

The Canada Goose Canard'stricycle landinggear. Propellerclearance ontakeoffsandlandings is critical for rear-enginecanards.

Angle 01altack-degrees

TIII

Basic airfoil

Sum ma ry :our AR 6straight- wingwith airfoilE197 wouldrequire a 12­degree AoAto achieveCL 1.1.

:£: S'. A'A~~1 1~ o.03Chord ~~ --. BaSICwing

Leading-edgedroop

A

. 2 -

1 .0

"0 .6CGO

:~=cii .4o

<.1

_ .8-'!:.

Figure 2.The NASA safe-wing droop.

a "fi n ite" AR andwingtips. In addition,the wing's pl anform(straigh t or tapered) hasan impact . The formulapreviously discussed inChapter 3 will help youto adjust the Wing's AoAto provide th e lift coeffi­cient selected and com­pensate for both AR andplanform.

Using the data inFigure 1 and noting thatth e E197 airfo il starts tolift at minus 2 degrees and achievesCL 1.1 at plus 8 degrees, th e sectionAoA would be 10degrees. Using anAR of 6 (this depends on yourdesign, of course), th e tot al AoAequals 13.91 degrees. Let's say 14degrees-less the minus 2 degrees(since it starts lifting at minus2 degrees), or 12 degrees for th ehori zontal.

1/

"''---- - --+- --..

- 0.57 0.38_

1_ -+-__ Semi·span ---_

ing/takeoff angle is 12 degrees,th en the plane has to rota teth rough on ly 9 degrees to reach th e12-degree angle .

LANDING GEARFor conventional models, the wingcharacteristics control the land­ing/takeoff AoA. For canard or tan­dem-wing models, lift is generatedby both wings. Well-behavedcanards or tandem wings havefron t wings th at mu st stall first, sotha t for landing-gear design , onlythe fore-plane 's characteristics areto be considered, not the aft wings.

Now, about those six factors :Figure 1 provides the lift, drag andpitching-moment character istics ofthe Eppler 197. On the left, CL 1.1ha s bee n selected as th etakeoff/landing CL atan 8-degree AoA. Thisis well below this sec­tion's stalling angle of16 degrees, and thestall is gentle with nohysteresis. Figure 1 ofCha pter 14, "Designfor Flaps," gives th eadditional lift coeffi­cient that slotted flapsdevelop.

If yo u know (orcan reasonably esti ­ma te) your mode l'swing loa di ng inounces pe r sq ua refoot , and if you calc ulate yourWing's "c1ose" to CL max., asabove, with slotted flaps deployed20 degrees for ta keoff and 40degrees for lan ding, Figure 3 ofCha pter 3, "Unde r sta n dingAero dynamic Form u-las ," willprovide th e means to estimate bothlanding and takeoff speeds in mph.With the Rn under your belt, selectthe appropriate Rn curves of yourairfoil. Note th at Figure 1 offers dif­ferent curves for different Rn num­bers. For E197, lift is littl e affected,but profile drag increases at low Rn.

• Wings incorporating the NASA"droop" will have an increase inlanding/takeoff ang les.

SECTION VALUEADJUSTMENTSThe values in Figure 1 are called"section values" and are for "infi­n ite AR." A model's wing has

THE BASICS OF RIC MODELAIRCRAFTDESIGN 73

CHAPTER 16 ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Figure 3. The dynamics of tricycle landing gear. With the CG aheadof the main gear, the inertia of the CG tends tokeep the model mov­ing straightforward. Figure 4. The dynamics oftail-dragger landinggear. With the CG behind the main gear, the inertia of the CGtendstoexaggerateanydivergence from a direct path straight forward.

A.

AirplaneCL

A.

CG

« I .=E-Airplane CL

Runway~

ships of 2-stroke or 4-stroke models,but not 2-stroke versus 4-stro ke.

The formula is sim ple:

1 x 'gross weigllt (oz.) = power loadingengine cid

A trainer that weighs 80 ouncesand is powered by a .40ci 2-stro keengine would have a powe r loadingof 1 divided by .40 x 80 = 200ounces/cid. The crane's power load­ing of 225 ounces/cid with a2-stroke engine shows that it hasgreater weight for its power thanthe trainer.

CG AND LANDING GEARThe CG location, in both th e hori­zontal and vertical senses, is th efocus around which th e landing­gear geometry is established. Formodel aircraft, th e only cause of aCG shift during flight is th e reduc­tion in the weight of the fuel as th eflight progresses. For a conventionalmodel, th is causes a rearward shift ofabout 3 percent of the MAC. For arear-engine canard, the fuel tank istypically behind the CG so th at asimilar, but forward, CG shi ftoccurs. The vert ical CG location isusually "eyeball" estimated. lt is bet­ter to get it a bit higher th an lower.

There are two major types oflanding gear:

Ai rp laneCLCG

B.

Tire drag ~

:+:cGMomentum

Figure 3.

Tail angle, -200

POWER LOADINGPower loading in ounces per cubicinch of engine disp lacement is auseful "ru le of thumb" for evaluat­ing the weight-to-power relation-

THE "CRANE" IIThe Crane II, a STOLmodel, had a verynose-high landingposture. It had anll-inch-diameter variable-pitchprop; full-span LE slots and slottedflaps. Spoilers on the wing 's uppersurface provided roll control. Thehorizontal tail had an inverted andLE-slotted lifting airfoil to providethe high tail download that is need­ed to achieve the very high AoA (20degrees ) provided by the Wing'sslots and flaps.

The Crane II had a fueled weightof 101.5 ounces and a wing loadingof 22.75 ounces/square foot ; powerwas a .45 engine; power loadingwas 225 ounces per cubic inch orengine disp lacement (cid).

the stall for takeoffs/landings.

The remedy wouldbe to lower the aftfuselage to reducethe tail angle so as toavoid the stall. Thiswould affect spiralstability as discussedin Chapter 9,"Vertical Tail Designand Spiral Stability."The longer gearwould increase bothweight and drag .

induced drag results. For a modelwith a span of 60 inches, and withits wing 8 inches above the groundon touchdown and AR6, this reduc­tion would be 10 percent of our 12­degree AoA, or 1.2 degrees.

Using the Swift as an example,the wing's AoA for level flight iszero degrees, so no adjustment for apos itive AoA is called for.

NASA SAFE·WING DROOPThis is recommended for sportmodels (see Figure 2). lt delays tip­stalling and provides effectiveaileron control in the stall. Sincethe droop occupies 38 percent ofthe sem i-span , it is estimated that itprovides a full 4 degrees more inthe takeoff/landing AoA.

Summary: the adjusted AoA forCL 1.1 of airfoil E197 is 12 degrees;slotted flaps reduce this by 4degrees; ground effect makes a fur­ther reduc tion of 1.2 degrees; andthe NASA droop adds 4 degrees fora ne t AoA of 10.8 degree s.

For the Swift, this was increasedslightly to 11 degrees to provide a2-inch prop-tip ground clearancewith a lO-inch-diameter prop. TheSwift illustrates the benefit of ahigh thrust line provided by aninverted engine (see 3-view inChapter 26) . If the engine wasupright and still fully cowled, theth rust line would be lowered byroughly 2 inches. A landing gear 2inches longer, to preserve the2-inch ground clearance, would benecessary. This could entail a sub­stantial increase in the "tail angle,"bringing the Wing's AoA to above

Wheelbase

Figure 5.Fuselage upsweep required toobtain a hightall angle and a short landinggear. This drawingshows the Crane, which was designed bythe author.

• Tricycle. The CG is ahead of themain wheels, and the nose wheelis steerable.

• Tail-dragger. The CG is behind

74 THE BASICS OF RIC MODEL AIRC RAFTDESIGN

Landing-Gear Design ... CHAPTER 16

C Caster

%~ :' - - - -B r~~kelSteeringarmIf-diameler

inwheelCoil

ail wheellY<"diameter

DETAIL DESIGNFigure 6 illustrates the procedurefor positioning the main landing­gear wheels for both trikes andtail-draggers. Take th e tail angledescribed previously and, on a side

rudd er application is needed fordirectional control on takeoffs andon landings.

As the tail comes up, propellertorque and gyroscopic precessioncause the model to veer .Compensating rudder is applieduntil the aircraft is just airborne.

If liftoff is forced by hea vy up­elevator action, the model ha sample dihedral and coarse rudder isstill applied, a sudd en snap roll mayoccur. Unless your reflexes are veryquick , a damaging and embarrass­ing crash will occur. It has hap­pened to this author!

Another disadvantage of a tail­dragger is its tendency to nose over,which is hard on props! Moving thewheels farther forward to reducethis tendency aggra vates themodel 's directional instability onthe ground. To avoid nosing over,taxiing, particularly on grass,should be done ho lding full up­elevator.

view of your design, draw a linethat defines the tail-angle to thehorizontal, originating either at theta ilskid or at the tail wheel.

• Tricycle gear. To prevent themodel from sitting back on its tail,

Fuselage outliner- ,.- --- - --

Firewall- ::~6'-"

plywood ::

Plywoo0 :wedge :

Wheel <:collar/ : :

Figure 7.Nose-andtail-gear detail (twoarrangements fora nose wheet andonefora tail wheel).

A.

B.

3 degrees, asshown in Figure 6,is sugges ted. Onlandi ng, after thenose-wheel h asmade contact withth e ground, th isno se-down ang lewill bri ng thewing close to itsang le of zero lift.The mo de l willtend to cling tothe ground . Thepotential for nose­gear dam age isreduced, and expe­rience has provedthat this nose­down atti tude hasno adverse effecton takeoffs.

Figure 9 illus­trates th e trike geometry for a rear­engine canard such as the CanadaGoose. Obviously, a very highthrust line is needed to avoid theneed for an un dul y lon g landinggear for prop-tip protection . TheSwan canard illustrates thi s point.For such craft, add 5 degrees to th etail angle.

Figure 5 shows how fuselageupsweep may beused to reducethe length of th eland ing-gear legsfor models thatrequ ire large tailang les, suc h asth e Crane .

This high tailangle moves thewheel axles far­ther behind theCG and requiresheavy up-elevatord efl ecti on t orotate the modelfor takeoff; but asth e tail goesdown, the wing'slift ahead of theCG aids th emodel's rotationfor quick takeoffs.

• Tail-draggers. As soon as a tail­dragger's speed, on takeoff, permitsthe tailwheel to lift off, it becomesdirectionally uns table (Figure 4). TheCG wants to get ahead of the mainwheels (see "B" of Figure 4). Coarse

r-_~-,-+ _

~ diameler A..I..j:~~~L::::::==----,~_~~_

Figure 6.Thegeometry of tall-dragger tanding-gear design (above) andtricyctetanding-gear design.

the main wheels, and the tailwheel is steerable.

Bicycle landing gear is a varian t oftricycle gear; a single rear wheelreplaces the normal tricycle mainwheels; the front wheel is steer­able, and tricycle geometryapplies.

The single-wheel CG of somesailplanes is a variation on tail­dragger style and geometry. Thehigh tail an gle is not neededbecause there is no prop, and th esegliders land in a nearly horizontalattitude.

LANDING·GEAR DYNAMICS

• Tricycle gea r. On the landing ortakeoff run, tricycle landing gear­with the CG ahead of the mainwheels- is self-correcting direction­ally (see Figure 3). The nosewheelsteers, prevents the plane from"nosing over" and protects thepropeller.

Wh en a "trike"-geared modeltips backward so that the tail skidrests on the ground, the CG rotateswith it . If this rotation brings th eCG behind the wheel axles, themodel will stay tail-down-a mostundignified posture! Shifting thelanding gear rearward from the CGby 5 percent of the MAC, as sho wnin Figure 6, prevents this fromoccurring.

Most trikes sit with their longitu­dinal center line parallel to theground. A nose-d own angle of 2 to

THE BASICS OF RIC MO DELAIRCRAFT DESIGN 75

CHAPTER 16 ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

follow this procedure. Draw a verti­cal line through the point that is 5percent of the MAC behind the CG.Draw a second line through th ispoint that defines the tail angle tothe vertical line just drawn (seeFigure 1). Notice that th is tail angle isthe same one as that defined by theline drawn from the wheel to theskid. Where these two tail-angle linesintersect, draw a horizontal line for­ward to the nose-wheel position, andthen draw a short vertical lineupward from the same intersection.The main wheel axles should be onthe short vertical line , with th ewheels' ou tside diameter resting onthe horizontal line. Decide whethera nose-down angle is to be used, andif it is, draw the nose angle at 2 to 3degrees to the horizontal line. Noseand tail gear will be discussed later.

• Tail-dragger. Draw a line at 15 to20 degrees from the CG, in front ofth e vertical, as in Figure 6A. Whereth e two lines in tersect, draw bothhorizontal and vertical lines . Themain wh eels' outside diam etersshould rest on th e hori zon ta l line,with their axles on the vertical.

TREAD WIDTHBoth trik e and tail-dragger land­ing gear should ha ve a lat eralspacing ("tread width," or the dis­tance between the centerlines ofeach tire) of 25 percent of th ewingspan of an AR 6 win g (seeFigure 8).

If the wing has a higher AR, cal­culate what the span wou ld be for

AR 6 with the same area. The for­mula for AR equ als span squareddivid ed by the area. Knowing thatth e AR is 6, the imaginary spancan be easily ca lcula te d; thewh eel-tread dimension will be 25percent of that span .

STATIC LOAD SQUATModels with mu sic-wire or alu­minum landing-gear legs originat­ing in the fuselage and sitting onth e ground bearing the model'sgross weight (iG) will "squat." For.40 to .SOci-powered models, thissquat is about 1;2 inch and reducesthe tail angle for takeoff. To com­pensate, reduce your landing gearlegs' "included angle " (see Figure 8)to lower the wheels and compen­sate for the squat.

WHEEL DIAMETERSmaller wheels have less air drag.For paved runways, a 2-inch diame­ter is the recommended minimum;for grass, a 21;4- to 3-inch diameteris suggested.

NOSE· ANDTAIL·WHEEL DESIGNSteerable nose- or tail-wheel gearshould incorporate a modestamount of caster. A modest amountof offset, as in the case of a grocery­cart caster wheel, facilitates steer­ing. Similarly, in th e case of landinggear, such gear tracks well and per­mits easy steer ing. Too much offsetinvites "shimmy." An offset of 20percent of the wheel 's diameter issufficient. Figure 7 illustrates two

1:; load

W squat

Wheel lread-25%ofaspect ratio6span

Figure 8.Wheel tread andsquat detalf.

nose-wheel arrangements (A and B)and one for a tail wheel (C) .

The nose-wheel gear is mountedon the rear surface of the ply engine ­mount bulkhead. For a conventionaldesign, this determines the positionof the nose gear. For a canard with arear engin e; the nose wheel shouldbe well forward, as in Figure 9. Notethat, in Figure 7, A and B, the shock­absorbing coil is totally enclosed inthe fuselage to reduce drag.

For tail-draggers, this authorprefers a somewhat forward tail­wheel location, with the tail-wheelleg supported internally by nose­wheel brackets bolted to plywood,as in Figure 7C.

MAIN LANDINCi-CiEAR LEGSMain landing-gear legs should be acontinuous piece of metal fromwheel to wheel so that bendingloads do not have to be absorbed bythe fuselage structure, but arecontained in the landing-gear legsthemselves. ...

~'~:"'--Whee l base •'--

Caster action,0,10 xdiameter 1(min.)

Outboard stabilizing wheels; diameter 3(bicyclelandinggear on ly) Diameter 3 025 ciamete 1

~::e~~~ft:;::~:~~fl~se~moSi cenJte ~ lol grav;: . .!'w~eel r"', ....... a I Propellerl ange - _ diame,?~ plus 3° s-:

10° toW Wheel -:;-_.1 Wheel diameter 1 ~iameler 2

~'+¥......

Figure 9.Layout geometry fortricycle Dr bicycle landing gear fora pusher canard.

76 THE BASICSOFRIC MODEL AIRCRAFTDESIGN

Pusher enginesInletarea (Ax B) x140%Outletarea (Ax B) x 140%

Figure 1.Sizing cooling-air inlets andoutlets.

O ur model airplane engines,by th emselves, are beauti­ful, powerful examples of

precision machining and enginetechnology.

Hung on the front of a model air­plane and left uncowl ed, they arehideous from a drag poin t of view.Even when partially cowled but withthe cylinder sticking out, they makea model look like a full-scale Cessna172 with a garbage can above theengine just behind the prop-ugly!

A well-designed cowl greatlyreduces drag, improves a model'sappearance and actu ally improvesengine cooling. Why are th ere sofew cowled engin es amo ng th e

Chapter 11

many models, both kit-built andoriginal designs , at our flying fields?This author surmises that there arethree major objections:

• Removing a cowl to service theengine is a nuisance to be avoided. Inmost cases, it is necessary to removethe spinner, the prop , the need le­valve needle and up to a half-dozensmall, easy-to-lose screws. Replace­ment reverses this boring sequence .

• Cowls are difficult to make.

• Fear that a cowled engine will notbe adequately cooled .

The design, con struction and fas­tening of the cowls described in thischa pter responds to and overcomesall three objections:

• The remo vable portion of eachcowl described is almost ridiculous lyeasy both to remove and to replace.Taking off the spinner, the prop andth e needle-valve needle is unneces­sary, and there are no screwsto laboriously unscrew (and lose).The engine is easily accessible forservicing.

• Such a cowl is easy to make, asth is chapt er will demonstrate.

Dueled-Cowl

Design

the cylinder and muffler does thecooling. Air passing 1 inch awayfrom the cooling fins does no thi ng.

A good , low-drag cowl designrequires:

• An inlet;

• an expanding chamber, or"diffuser";

• the item to be cooled: radia tor,or cylinder and muffler;

• a contracting part, or "nozzle(s) ";and

• outlet(s) into the passing air streamat point(s) of low air pressure.

Prop-driven air enters the diffuser,slows down, cools th e cylinder andmuffler, expands because of the heat

The Swift's cowl; note the jacklocation.

• Cooling is adequate, asproven by test runs onhot summer days at fullrpm with the model sta­tionary and consumingfull tanks of fuel.

DUCTED·COWLDESIGNFor minimum drag, thecooling-air entry shouldbe as small as possible,yet large enough for ade­quate cooling. Bear inmind that onl y th e airthat actually contacts

%" ... balsa

Figure 2.Cowl top view-internal muffler.

THE BASICS OF RIC MODEL AIRCRAFT DESIGN 77

CHAPTER 17 ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Figure 3.Cowl section A-A (see alsoFigure 6-internal mUffler).

The pusher nacelle ontheSeagull III flyingboat. The NACA inlet andtheoutletbelow thespinner show.

absorbed, speeds up in the nozzlesand exits at considerable velocity.British WW II Hurricane fighters'ducted-eng in e coolant radiatorswere based on these principles; theycontributed thrust, not drag. Thehot, expanded air exiting the duct'snozzle provided some jet-likepropulsion. This is not to suggestthat these cowl designs will con­tribute thrust, but there will certain­ly be substantial drag reduction.

INLET AND OUTLETSIZING-TRACTOR ENGINESFigure 1 shows the side view of amod el engine . An em pirical rule ofthumb, based on experience, is topro vid e an air-entry area that'sequal to th e area of th e finned por­tion of th e cylinder, as shown.Whether th e opening is round,square, or rectangular makes no dif­ference provided the entry has thearea described.

The cooling air exit(s)' rule ofthumb is th at th e total exit area be140 percent of the entry area. Forexample: an entry area of 1.25square inches requires an exit of1.75 square inches for one, or 0.875square inch each for two exits.

ENGINE ANDENCLOSED MUFFLERFigure 3 shows a horizontal cross­sectio n through the Swift's cowlwith a mu ffler. Both the engine andthe muffler are wholly enclosed. Ithas an inlet, a diffuse r, a cylinder,muffler and nozzles; and the exitsare at points of reduced air pressureon th e fuselage sides (they look likegills on a fish!). The fuselage mustbe widened to accommodate th eengine and muffler as in Figure 3.

The "teardrop" fuselage wasdescribed in Chapter 12, "ImprovePerformance by Reducing Drag."

Cowl "box" outlines

__ .---- !~:~Iy, , ,,,

I :I ,,,,, ', I

, ', ', ', 'r - -, I

\~ri- --\l -))f' balsa she••

Figure4.Spinner ring/entry and rearhold-down detail.

This type of fuselage lends itself toa wider forward section witho ut adrag penal ty. Figures 3 and 6 detailth e cowl installation.

Exha ust stacks may extendth rough the cowl, and the neces­sary holes mu st be elongated side­ways 1;8 inch for cowl removal.They may also end just clear of th einside of the cowl with slightly larg­er, round hol es.

ENGINE ANDEXTERNAL MUFFLERFigure 7 shows the cross-section ofa cowl for an engine equipped witha stock muffler. While the muffler(and pressure tubi ng to the tank) isexposed, its drag is largely over­come by th e jet-like exha ust gasessquirting backward. With an exter­nal muffler, th e fuselage may benarrower, as shown.

COWL FASTENINGThe rem ovable portion of thecowl is he ld in position by three"flat hold-downs" (FHDs). One isin th e cooling air-entry former infront, and two are at the rear ofthe cowl (see Figures 4 and 6). Allth ree engage n o. 2 shou lde r

On

.....---- 2.56 bolt & nut ---..,;

________ . 2shoulder screw

all

Figure 5 A andB.Goldberg flat hold-down (FHD) installation.

78 THE BASICS OF RIC MOD ELAIRCRAFT DESIGN

Ducted-Cowl Design ... CHAPTER 17

struction described previo usly hasbeen used on at least seven modeldesigns. Th e sound-dea deningproperties of thick balsa shee t are adefinite advantage. In this chapter,I will give more details on ducted­cowl construction and also touchon design considerations for a cowlmounted in a pusher configuration.

That portion of the cowl behindthe spinner and surrounding theengine crankcase is solidly CA'd tothe engine bulkhead. The other,removable, portion surrounds thecylinder. The level of the partingline between these two par ts isimportant. It mus t be horizontal,and it mu st separate through thecenter of the needle-valve need le-

___- - - - .. Nozzle exit

ASSEMBLY AND SHAPINGPhoto A shows balsa sheet, tri-stockand plywood components partiallyassembled into the cowl 's twoparts . Carefully trim the length ofboth parts of th e cowl's balsa to suitthe length of your installation, asshown in Figure 6.

At this stage, the fuselage shouldbe finis hed (but not covere d).Temporarily install the engine (lessthe needle-valve needle) and muffleron the engine mount so that thecowl can be shaped inside as shownin the photos and drawings. The plyparting-line separator guides thiseffort. A Dremel sanding drum anddrill will do this quickly and easily.

The cowl structure around thecrankcase requires only minorinternal contouring to clear themuffler; the removable portionneeds con siderably more in ternalshaping to clear the cylinder andmuffler .

The three flat hold-downs areboth CN d and bolted (2-56 bolts

either just above or just below it .Obviously, a suitable slot or slots(half above and half below the part­ing line ) is essential to clear theneedle.

If an external muffler is used,then suitable cutout(s) must bemade to clear the portion from theengine exhaus t to the muffler. InFigure 9, no te the I13z-inch plywoodparting-line separator that guidesthe shaping of the cowl both insideand outside. It is firmly cemented tothe removable portion of the cowl.

External muffler

~~"----.:J~:.:.....i-._ Nozzle exit

1M" gap

Inlet

Parting line

»; ' t :Grain vert.. i/ ,, ,

Cowl box sides3/4" balsa

B

~1I11 11...-- Plybulkllead

Figure 7.Cowl section A-A; external muffler.

J'Tec muffler ---Ir----'~~

Aluminum tube =::::= ~:::~7:.z:!!

Figure 6.Cowl sideview-tractor engine; internal muffler.

CONSTRUCTION HINTSOver the years, I have designed andbuilt many types of cowl. Theyranged from laboriously hollowed­out solid balsa to fiberglass-and­epoxy lay-ups on dissolvable foammandrels. The ducted-cowl con-

screws; two are screwed into theplywood eng ine bulkhead, an done in to the plywood spinnerring.

Ini tially, this author used theseFHDs as shown in Figure SA. Aknife blade inserted at the partingline and then twisted, detachedthe cowl. On smaller models, thismethod was satisfac tory. On largermodels- and after losing severa lde tachable portions in flight(none was ever found de spitelen gthy searches)-it was evidentthat this form of cowl attachmentwas unsatisfactory. It was belatedlyrealized that the wrong end of theFHDs was being used, and thearrangement shown in Figure 58was employed very satisfactorily­no more lost cowls !

A useful byproduct of thischange was that removal requiresonly a sha rp knuckle rap on theremovable portion's side oppositethe muffler. Replacement requ iresthe alignment of the "hooks" onthe FHDs with the shoulder screwsand a rap on the cowl 's mufflerside. It is amusing to have a star­tled onlooker exclaim, "How didyou do that!"

THE BASICS OF RIC MOD ELAIRCRAFTDESIGN 79

CHAPTER 17 ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Figure 8.Cowl box detail; lJ2-inch balsa sheet; internal muffler.

3111" •balsa

....- Trimto length 01cowl~Ishown in Figure 6

__ - / 7- - - -/-- -;--:)

W' • balsacementingstrips

th e fuselage contour, as shown inPhoto C.

Next, rem ove th e cowl and takethe engine and muffler off themotor mount. Epoxy the rear FHDply assembly in the rem ovable por­tion of the cowl as shown in th ephoto. This requires some trim­ming of both the ply and the balsa.Note that the open side of all threeFHD's "hooks" sho uld face awayfrom th e muffl er side.

Now clamp th e cowl into positionas you did before, carefully align ingit with th e spinner and fuselage.Through th e air-entry hole, usingthe rear flat hold-downs as guides,mark the positions of th e no. 2should er screws on the engine bulk­head. Remove the cowl, drill VI 6­

inch holes in th e bulkhead, putsome CA in th e holes, and installth e two screws.

and nuts) to their plywood parts.(Note th e bolt-orientation nutsinside.) File th e round bolt headslevel with th e bottom of the screw­dri ver slo t after th ey've beeninstalled in th e plywood .

Install and lightly tack-glue th ecowl "box" to the engine bulkheadas shown in Photo B, with thespinner ring cooling-air entryassembly cem ented to both por­tions of th e cow l.

Using an old spinner backpl ateof the correct size, clamp th e boxinto position by ins talling th e propnut and washer, putting a Y32-inchbal sa-sheet spacer between th espinner backplate and th e ply spin­ner ring.

Shape and sand th e ou tside sur­faces to match th e spinne r; th ecooling-air en try plywood part ingline; th e 1132-inc h ply separator and

Photo B.The cowl "box" hasbeen clamped intoposition for external shaping.

Photo A.Cowl components areshown partlyassembled.

80 THE BASICS OF RIC MODEL AIRCRAFTDESIGN

Figure 9.Top view ofpusher engine cowl.

Ducted-Cowl Design .. CHAPTER 17

ENGINE PRIMINGPriming a fully cowled engine iseasy. Invert th e model on your fieldbox to bring the engine upright.With a squirt bottle, in ject a fewdrops of fuel in to the carburetor. Ifth e carb is closed, the carb entryforms a small cup which, whenfilled, provid es adeq uate priming.The cooling-air entry hole permitsthis method of priming without

GLOW·PLUG ENERGIZINGWith the engine enclosed, theglow plug is energized by mea ns ofa two-conduc tor, closed-circui ttype, Radio Shack phone jack. Toenergize the plug, a mating, 1;8­inch, Radio Shack plug is wired tothe ex te rnal power source andinserted in to the jack. This is amaj or safety feature because thejack ma y be located well awayfrom th at deadly, rota ting prop forplug removal. Figure 6 details thebronze glow-plug clip that 's easilydisengaged from the glow plugwhen plug replacement is necessary.

The jack is mounted th rough a713z-inch-diameter ho le in a smallsquare of 1/16-inch plywood. Bothare epo xied to th e inside fuselagewall so that the jack's knurled nutpro jects through a 5/ 16-inch­diameter hole in th at wall. Figures12, 13 and I S provide a wiring dia­gram and engine-servo detail for an"onboard" glow-p lug energizingsystem that hea ts th e plug in flight,but on ly at low rpm. The systemensures a reliable idle , parti cularlyfor 4-stroke engines.

Short stacks

Permanently install theengine and muffler, con­nect th e carb-to engine­servo linkage, rep laceth e needle-valve needle,install the fuel and mufflerpressure tub ing from theengine to the fuel tank ,and connect the glow-plugclip to the glow plug.

Solidly CA the fixed por­tion to the engine bulk­head, an d clamp thewhole cowl into positionas before, as shown inPho to C. In Photo D, bothparts are ready for paint­ing. The engine's accessi­bility is evident.

Cowl bOItopand bO"O~-=-_~"~+: ~o~3~"_Sha: ~xt~nsio n

-- -----------,IIIIIIII

Figure 10.Side view 01 pusher engine cowl.

Photo C.The shaped andsanded cowl. The upper portion has beenCA 'd to the engine-mount bulkhead.

Section C-C t:-- ---~r-r-~T~----~

- I

i ~I :tI I

- ---;---!- -r-I 31.12"I Plywood

?'::-'~7>1-:-:_=:===:IIII,

I ' :I I'-----~t L - - - -'

II

- '"' I:s I 3J.l2" plywood

I1

Figure 11.Pusher engine cowl sections andhold-down detail(see Figure 10).

Photo O.This cowl detail shows thatservicingtheengine is easy.

THE BASICS OF RIC MOD EL AIRCRAFT DESIGN 81

CHAPTER 17 .... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Figure 14.Details andordinatesof NACA submergedintake.

Lip

~ 0

~Ramp floor~-_-=DePth

A

Outer surface

Ramp length "L"

Section A-A of 7° ramp

A'~~

__-"i.- - -+_~.... '-Wt th- W

•1.0 ~

Entry ratio ofwidth/depth =3 to 5

Table ofordinates

'YL Y;Yf

0 00.1 0.0040.2 0.0840.3 0.2340.4 0.3860.5 0.5340.6 0.6120.7 0.6880.8 0.7640.9 0.8421.0 0.917

Glowplug

Glow-plugclip

l8-gaugestranded

hookup wire(#278-293)

Figure 12.Onboard glow-plug wiring diagram.

Tothrottle

Section A-ACutlromsix-armservo wheel A

Figure 13.Smallengineservo (Futuba 533-5133).

A

Trimonly ~

Engine cutoff -_W'''

Pushrodconnector

Cut outof roundservo wheel

Figure 15.Engineservo lengthwise in fuselage.

Tothrottle

cowl removal. If, after a flight, theengine is stopped by closing thecarb, subsequent engine starts don'trequire priming. To avoid "hydrauliclock"-having fuel trapped betweenthe piston and the cylinder head­apply your electric starter with themodel inverted (engine upright ).

PUSHER ENGINEINSTALLATIONSFigures 9, 10 and 11 show thepusher installation of the SeagullIII-a flying boat. The engine sits ina nacelle above the hull.

For improved streamlining, a 3J4­inch crankshaft extension wasused, as shown in Figure 12. Anenclosed muffler is mandatory,because the external muffler wouldexhaust the wrong way, facing for-

ward , and it could not be reversed,because that would foul the propand prevent the propeller fromrotating.

Cooling air enters the cowlthrough two, NACA-developed,low-drag, submerged air intakesrecessed into the nacelle (or fuse­lage) sides ahead of the enginebulkhead. The combined areas ofthese intakes is the cylinder areadescribed in Figure 1 plus 40percent. The exit slot under thespinner has the same total area asthe entries . The rotating prop"sucks" cooling air out of thiscooling slot.

Construction, shaping and fas­tening the removable portion andglow -plug energizing are identicalto the tractor installation.

NACA CooLING·INLETDESIGNFigure 14 shows how to develop theshape of the NACA submerge dintake . Note the intake width-to­depth ratio and the ramp floor at 7degrees to the outside surface.

Over the years, I've used pusherengines cowled as described on fivemodels . Cooling problems have no toccurred.

Throughout this chapter, illustra­tions and photos show invertedengines (author 's addiction). Forupright installations, simp ly turnthe photos and drawings upside­down! ....

82 THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Chapter 18

T he wide variety of propellermakes, shapes, mat erials,diameters and pitches avail­

able toda y can be somewhat con­fusing . The choice of a prop to suityour model, its engine and yourstyle of flying requires some under­standing of how a propeller func ­tions. It also requires an appraisalof the weight, wing area and aero­dynamic drag of your airplane andof the power loading of th emodel-plus some insight into itsengine's power characteristics.

In addi tion, the propeller's h igh ­speed rota tio n leads to effects thatevery modeler should be aware of.These are:

• Slipstream;

• asymmetrical blade effect;

• propeller pitching moment;

• torque; and

• gyroscopic precession.

This chapter will cover th ese pointsand he lp to narrow propellerchoice for a given model to one ortwo diameters and pitches.

and that tapers to th e tips . Thesesmall airfoils have all th e cha racter­istics of a wing's airfoil. They have:

• A chord line;

• an angle of zero lift;

• a stalling ang le;

• increasing profile and ind uceddrags as th eir AoA increases;

• a pitching moment; and

• upwash ahead, and wake anddownwash behind the blades .

Propeller blades differ from thewing's airfoil in that they operate atmuch higher speeds than the wing.A 12-inch-diameter propeller thatadvances 5 inches per revolutionand turns at lO,OOOrpm has a tipspeed of 360m ph, while th e modelit propels flies at only 47mph.

A wing normally flies at the samespeed across its span. A propeller,however, operates at differentspeeds: high at the tip and progres­sively slower from tip to root . At halfits diameter, its speed is half that atthe tip . Stresses on the propeller are

Propeller

Selection and

Estimating

Level Flight

Speeds

high, particularly at its center. Thesestresses result from a combination ofcentrifugal and thrust forces, plusth e blade's airfoil pitching momenttrying to twist them.

DIAMETER AND P IT CHPropellers are sized in both diame­ter and pitch in inches. Diameter issimpl y the length of th e prop, tipto tip . It identifi es the size of th eimaginary cylinder in which theprop rotates and advances. Increas­ing th e diameter increases the load

PROPELLER ACTIONA propeller generates thrust by forc­ing a column of air backward­called the "slipstream" as in Figure1. In the slipstream, the air's velocityis increased above the aircraft's for­ward speed, and its pressure isreduced. In addition, a substantialpart of this increase occurs ahead ofthe prope ller. This slipstream swirlsaround the fuselage in the samedirection as the propeller rotation.

A PAIR OF WINGSA two-blade "prop" is actua lly apair of small wings; each has an air­foil cross-section that is th ick closeto the hub for strength and rigidity,

Direction olllight

.1

Figure 1.The propeller'saction.

Direction 01 propeller rolallon

Propeller disk

THE BASICS OF RIC MOD ELAIRCRAFTDESIGN 83

CHAPTER 18 ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Tip speed for th is prop turning atlO,OOOrpm would be:

A 12-inch-diameter prop, advanc­ing 5 inches per revolution , wouldhave a hypotenuse of:

-,)(12 x 3.1416)2 + 52

or 38.02 inches.

imagine cutting the cylinder length­wise down one side, from startto finish of that one revolution.Imagine opening and flattening it.

Figure 3 shows this flatt enedcylinder along with the geometricand actual pitches and blade cross­sections at 100 percent, 75 percent,50 percent and 25 percent of theblade's length.

Note how the geometric angle ofthe blade varies from tip to root sothat the re is a constan t AoA.Calling suc h a pro p "cons tan tpitch" is abit of a misnomer; th e blad e isobv iously twisted. "Constan t angleof attack" is more accura te.

To calculate the propeller's speedat any point along its leng th is easy.Take th e prop tip in Figure 3; in onerevolution, it moves from A to B;AB is the hypotenuse of a right­angle triangle. Recalling highschool geome try: "the square of thehypotenuse of a right triangle isequa l to the sum of the square ofthe other two sides." In formulaform and Figure 3:

AB = -,) (ACZ + BCZ)

Nominal pilch

CONSTANT·PITCHPROPELLERSEach point on a propeller blade­rotating and simultaneously advanc­ing-describes a helix inside animaginary cylinder. Consider oneblade advancing one revolution;

resistance. Under these conditions,the propeller must operate at higherAoAs or slip, with increased profileand induced drags. This reduces theengine's rpm. It shou ld be notedthat, while pitch is a major factor inspeed, a plane obviously can' t flyfaster in level flight than a speed thatis close to tha t permitted by its geo­metric pitch mul tiplied by the rpm.

In a dive, with the engine at fullrpm, the actua l advance per revolu­tion may increase to a point whereth e prop's airfoil is operating at avery low or a negative AoA. The pro­file and induced drag reduce sub­stantially, the prop "unloads" andthe engine over-revs-which does itno good! Experienced fliers throttleback in dives for this reason .

Figure2.Propeller pitches.

Blade section

-.Plane 01 rolation

on th e engine and reduces its rpm.For each prop diameter, there are

several different pitches available.For example, a lO-inch-diameterprop is typically offered in pitchesfrom 6 inches to 10 inches. Thehigher th e pitch, th eoretically, thegreater th e adva nce per revolution,and the higher the engine load­again , reducing its rpm .

Thus, both diameter and pitchmu st be consi dered in propellerselec tion. For high-speed flight,reduced diameter and increasedpitch apply; for slowe r flight,increased diameter and lower pitchprevails.

There are several variations for agiven pitch dimension, as follows(see Figure 2).

• The "no minal pitch " is measuredacross th e flat back surface of th eblade-usually measured at 75 per­cent of the diameter. This is wha tyou buy!

• The "geometric pitch" is mea­sured across the airfoil's chord line.

- - - - ---Flatlened cylinder--------

Figure 3."Constant pitch" propeller.

Geometricpilch

I

True advanceperrev

B

c

·1i

Root

25%

f+----- - - - - - Diameter x 3.1416 - ------

tTip

A

100%

PROPELLER AS AIRSCREWA prop eller has much in commonwith a screw. In fact, they are fre­quently called "airscrews." A screwbeing turned in a threaded hole willalways advance its full pitch for eachrevoluti on. A propeller "screws" intoair that is fluid. The advance per rev­oluti on is not fixed. A heavy modelwith high air drag and in a steepclimbing attitude will offer high

• The "t rue pitch " is th e actua l dis­tance the prop advances per revolu ­tion. The difference between geo­metr ic and tru e pitch angles is th eAoA at whic h the prop airfoil istru ly operating and is called theprop eller "slip."

84 THE BASICS OF RIC MODELAIRCRAFT DESIGN

Propeller Selection and Estimating Level Flight Speeds ... CHAPTER 18

A. Blade at 75% diameter

Figure 4.Liff, drag andthrust vectors at 75%and25% diameters.

they are close enough for all practicalpurposes.

Figure 5.Nomograph forquickdetermination of wingloading, lift andspeed.

• The weight-to -power ra tio, orpower loading. A large engi ne pow­ering a small , light mo del will obvi­ously outperform a heavier, largermodel powered by a smaller engine .

With the large variety of bothmodels and engines available, some

• The model 's aerodynamic d rag.A "clean" model such as the Swiftwill offer much less air resistancethan one with an exposed eng ine,large flat Windshield, large roundor rectangular (in cross-section)wheels, unfaired landing -gear legs,dowels and rubber bands for wing­to-fuselage attachme n t, and other"built-in headwinds."

Parasite drag inc reases in propor­tion to the square of the speed.Doubling the speed results in a four­fold drag increase. High drag mea nsincreased "slip" (the prop will oper­ate at higher AoAs) and rpm an d fly­ing speed will suffer adversely.Lower pitches and larger diametersare appropriate. While Figure 15does not reflect th e im pact of highdrag, it will pu t you "in the ball­park" as far as rpm and pitch areconcerned.

100 .,.95 V"" /90 Coefficients H :- ./85 Y:80 '1<1 / ;/ ",

75 V V .,..70 / V65 V/ ,,-

.c:: 60 / ;7~55

H / V .9- ......~ 50 / 1/ 1/ ~ ,,-V .51. .......... 45

/ l/ V V ......~ / .9-40 II / 1/ / V V V

:!I- ....-35 / / / V y / V ...... .lJI....-30 f/ / 0-t/"',,-VV

,..- .!JI--~-25 r./~ »:~- I-"" I-"" '-"~-20~~V~':::::--I::-:.--15V~~ r:::::--10 .5~

4 8 12 16 20 24 28 32 36 40 44 48

Wing loadlng- oz./sq. tt.

B. Bladeat 25% diameter

in ounces per square foo t of wingarea and the faste r it must fly inlevel flight (or at h igher AoA withh igh er dr ag).

Most models, in level flight, fly atCL of 0.2 to 0.3. If you know themodel's weight and calculate its wingarea in square feet, its wing loading iseasy to arrive at. Figure 5 provides aqu ick way to estimate the model'sflight speed. Say the model's wingloading is 20 ounces per square foot;reading upward from 20 to CL 0.2an d 0.3, level flight speeds are, onthe left, 40 to 48mph. These speedsare minimums; something more isrequired for climbing and othermaneuvers. Adding 25 percent givesspeeds of 50 to 60mph and a meanspeed of 55mph.

Now refer to Figure 15 (page 89):the rpm/pitch/speed nomograph .Place a straightedge at 55mph inthe central, level-flight-speed col­umn, and read off the static rpmand corresponding pitches thatwill provid e 55mph. For example: a7-inc h pitch at 7,OOOrpm or an 8.5­inch pitch at 6,OOOrpm both pro ­vide 55mph .

The nomograph in Figure 15 isbased on a 1O-percent increase overthe nominal pitch advance per revand on a gain of 10 percent in engi nerevolutions as the prop "unloads"from a static position at high AoAstothe level flight speed at much lowerAoAs. This graph will enable you toarrive at a reasonably close estimateof your model's top speed, based onthe engine's static max rpm and itsprop's nominal pitch. These resultswill never be 100 percent accurate, asthe model's weight and drag willhave an unavoidable impact, but

Drag

38.02 in. x 1O,OOOrpm x 60 min. /hr.12 in ./ft. x 5,280 ft./mi.

THE AIRPLANEThe design of the model has amajor bearing on the selection ofits propeller diam eter and pit ch .The factors are:

• The weigh t an d wing loading.The heavier the model, for a givenarea, the highe r its win g loading

or 360. 12mph .At 50 percent of th e blade length ,

the speed would be 50 percent of360 .12mph or 180.06mph. Thoseblades are lethal; take care!

Figure 4 sh ows blad e cross­sections at 75 percent (A) and 25pe rcen t (B) of the blad e lengthfrom the hub. Both are operati ngat the same AoA. No te that at 25percent, because of the bla deang le, the lift is more incline d, thedrag vec to r is increased and thethrust vec to r is reduced in co m­par ison with the 75-percent point.This in ne r port ion is less efficien t,and from 25 pe rcent to the propcente r on ly worsens. A sp in ne r ofro ug h ly 25 percent of the prop'sd iamete r wou ld cover th is porti onand wo uld smooth out the airflowmoving backward. For a 1O-inch­diameter prop, a 2l12-inch-diame­ter spi nner does just that .

In Figure 4B, the high er bladeangle, reduced thrust and increaseddrag reflect the effect of h igherpitches for the prop as a whole. Theincreased drag reduces engine rpm;lower diam eters are indicated . Thereverse is also true ; lower pitcheswith larger diameters.

THE BASICS OF RIC MO DEL AIRCRAFTDESIGN 85

CHAPTER 18 ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

simple way of establishi ng the"weight-to-power ratio " is neededto perm it ready comparisons. Oneway is to calculate what the weightin ounces would be if both engineand model were scaled up (ordown) in proportion to 1 cubicinch of engine displacement (cid).For example, th e Swift is poweredby an O.S. Max .46 SF engine, andweighs, fueled, 92 ounces. Itsweight-to-power ratio is 92/0.46, or200 ounces per cid.

Another example is of a modelweighing 300 ounces, powe red bya 1.2ci engine. Its power loadin g is300/ 1.2, or 250 ounces per cid.This comparison has obvious limi­tat ions. It assumes that power out­put of various sizes and makes ofengines is proportional to theirdisplacements-th is assumptionisn 't too far off the mark. It'sinvalid for comparing 2-strokewith 4-stroke eng ines . Each classmus t be separately evaluated, e.g.,2-strokes should be compared with2-strokes and 4 strokes with4-strokes. Experience indicatesthat 2-stroke models with a 200­ounce per cid powe r loading thatare well "propped" will have excel­lent performance. Higher powerloadings, up to 300 ounces per cid,will resul t in diminished, bu t stillacceptable, performance.

• The type of perfo rm an cedesired. In designing a model,selecting a kit to build, or choosinga model to scratch-build from mag­azine plans, the modeler has perfor­mance objectives in mind thatprobably reflect his or her flyingskills. The design goal may rangefrom a slow, stable , easy-to-fly air­plane (for a beginner) to a fast,h igh-powered, aerobatic model (forthe expert). For th e beginner, lowwing loadings and a higher weight­to-power ratio of 275 to 300 ouncesper cid would be in order.

At the other end of the scale,consider the Swift. Designed as asport model with a wing loading of22 ounces per squa re foot of wingarea, a power loading of 200 ouncesper cid and with the least drag thatcould be reasonably expected­short of retracts-it is fast, maneu­verab le and fun ! It has flown withtwo propellers. The first, a lOx9,has a static rpm of 12,000. The sec-

86 THE BASICS OF RIC MODEL AIRCRAFT DESIGN

TO CHOOSEAPROPThis procedure is recommended forsefecting propelfers forYOllr modef.

O For a given coefficient of liltand wing loading, lind theesti­

mated airspeed as indicated in thenomograph (Figure 5). Increase thespeed by 25percent to allow forclimbing and any appropriate aero­batic maneuvers, e.g., convert a50mph estimate to 63mph.

a Look at the rpm/speed/pitchW nomograph (Figure 15), andpick out a pitch and rpm that willgive you the airspeed you want.

A Look at a published evaluation~ of the eng ineyou arelIyingand see the reported rpm for variousprops tested onthe eng ine. Also lookfor the rpm range where torque ismaximized, if this information is pro­vided. Pick a few props that providerpm within the high-torque range andachieve thedesired speed range .

o Test these props at thelIyingU field and stickwith the one thatprovidesthe best performance.

ond, a 10xl 0 (a "square" prop)turns 11,000rpm sta tic.

From Figure 15, level flightspeeds are estimated to be 125 and130mph-very close! This model 'svertical performance is that ofa "homesick angel"; it performsvert ical 8s with ease and grace.

ENGINESToday's model aircraft engines arefine examples of modern enginetechnology and precision machin­ing. Most are "over square"- thebore diameter is larger than th estroke. This author prefers 2-stro keeng in es because th ey're simpler,more rugged, lighter, more power­ful and less costly th an the 4-strokeversion s of the same displacement.

Engine-evaluatio n articles, suchas th ose by David Gierke an d MikeBillint on in Model A irplane News,and Clarence Lee in RIC Modeler,provide performan ce data on cur­rently avail abl e en gines an d

insight in to their design and con­struction . They provide tabula­tions of static rpm of an enginewhile it is powering various diam­eters and pitches of propellers.Table 1 shows Billinton's recordingof rpm for the Fox Eagle 74 (ModelAirplane News, October '91) andTable 2 shows that of Lee for thisengine (RIC Modeler, March '91) . Inaddition, Billinton provides per­formance curves of the 74 inFigure 7. Note that with silencerand standard .330 carb , the brakehorse-power (b .hp) peaks at15,000rpm, and the maxim umtorque is in the 7,000 to11,OOOrpm range.

Data of this type- and theengine manufacturers' recommen­dations-provide very useful guidesin selecting the diameter to matchthe pitch and rpm determined fromFigures 5 and 15.

MATCH THE PROPAspreviously noted, for a 20-ounces­per-squa re-foot wing loading, a55mph speed is indicated, and a 6­inch pitch prop turning 8,OOOrpm isone possible selection . Look at Table1 (Figure 6) for the Fox Eagle 74. AIS-inch diameter by 8-inch pitchprop would tu rn at around8,OOOrpm. Figure 7 indicates thatthese rpm aren 't too far off the peakof the torque curve for this engine.Another choice could be a 12xlOprop also turni ng in the 9,OOOrpmrange. Like low gears on a car, thelower pitch of 6 inches would pro­vide quicker acceleration and betterclimb, but lower top speed.

TOOLSThere are two items of equipmentevery serious modeler should pos­sess. First is a pho tocell tachometer,either digital or analog, to measurethe static rpm of your engine. It isuseful to compare the performanceof props of various diameters andpitches with the published data asdescribed above. These tachometersmay be used safely from behind theprop, and they aren't expensive.The second too l is a propeller bal­anc er, the type with two sets ofoverlapping, free-turning disks .Balance every prop-you'll be sur­prised how many require balanc­ing-to avoid vibration. On rein­forced plastic props, a coat of silver

Propeller Selection and Estimating Level Flight Speeds .... CHAPTER 18

paint (after a gentle surface rough­ing with fine sandpaper for bett erpaint adhe rence) will aid th e ph o­to cell to "see" the prop. Anyimbalance is easily corrected byadding paint to th e lighter blade .

All thi s will narrow th e cho ice totwo or three props. However, thereis just no substit ute for actual flighttest s in your fina l selectio n toobta in th e performan ce soug ht andthe optimum out put of prop andeng ine.

PROPELLER MATERIALSProps are available in wood, nylonand reinforced plastics. This authorfavors the reinforced plastic prop sbecause of th eir ruggedness andefficiency, even th ough they weighroughly twice th e weight of theirwood en equivalen ts. Avoid unrein ­forced nylon props; th ey lackenough rigidity for use during highpower.

P ROP ELLER EFFECTS• Slipst ream. The slipstream (seeFigure 1) moves as a helix rotating

around the airplane in th e samedirection as th e propeller's rotation ,but at higher th an flight speed. Itstrikes body, wing and tail surfacesat angles and increases the drag ofany obstacle in its path. Its mostunfavorable impact is on th e verti­cal tail surface- it causes yawingth at calls for rudd er-trim correc­tion.

The increase in th e velocity ofthe oncoming relative wind (i.e.,ahead of the pro p) reduces theprop's effective pitch, as does oneblade's dow nwash on the next .Such dow nwash further redu cesth e prop's efficiency. The situatio nis made worse with three or moreblades. For model airpla nes, suchmulti-blade props aren 't recom­mended, except for scale models ofaircraft so equipped.

In full-scale aircraft, multi-bladeprops are used to absorb the highpowe r of modern piston and turbo­prop eng ines. They also reduce thepropeller's diameter so as to avoidcompressibility effects from tipspeeds close to the speed of sound.The loss of efficiency in this reduc­tion must be accepted.

• Asym metric blade effect. Whenth e plane of the propeller isinclined to the direc tion of flight asin Figure 8, the advancing bladeopera tes at a higher AoA than theretreating blade . Th rust on theadvancing side is higher than onth e retreating side. This causes apitching or yawing couple.

• Pitching moment. When thethrust line is tilted as in Figure 9, avector is introduced that causes apitching mo ment. It may combinewith the asym metr ic blade effect.

• Torque. The resistance to rota­tion caused by the prop's drag triesto rotate the whole airplane in theopposite direction. This is particu­larly true in a steep climbing atti­tude at low forward speed and max­imum rpm whe re the prop is oper­ating at high AoAs, such as justafter liftoff. A to uch of oppositeaileron input may be needed to offset th e torque.

• Gyroscopic precession. Like agyroscope, a rotating propellerresists any effor t to cha nge the

direction of its axis. The hea vierth e pro peller and the higher therpm, the greater this resistance. If aforce is applied to tilt the plane ofthe prop's rotation, it is "precessed"90 degrees onwa rd, ill the direction ofthe prop's rotation.

This effect shows up markedl yon tail-dragger takeoffs if the tailis lifted too soon and too high .Precessio n causes a yaw to the left(for props ro ta ti ng clo ckwise,viewed from behind) that couldresult in a ground loop unlesscorrected by rudder action.

The author's flying-boat design,Seagull III, was in itially flownwit h a Grau pn er llx8 prop thatwas mo unted in a pu sher configu­ration with the propeller's plan eof rotation di rectly over the CG(the thrust line was 6 inc hes abovethat CG). Coming ou t of a left ­han d turn, the model would en teran uncommanded, gentle right­han d turn, nosin g down slightly.It was easily corrected, but annoy­ing . Replacing the Graupne r (anexcellen t prop) with a Zin gerwooden equivalent of half theGraupner's weight elimin ated thispeculiarity.

NOISEMany clubs are experiencing prob­lems because of noi se that origi­nates from two sources: th e engineitself and the prop eller. Enginemu fflers and tuned pipes now avail­able go a lon g way to reduce eng inenoise to acceptable levels.

tf:rn metre

Figure 7.Performance curves for IheFox Eagle 74.

THE BASICS OF RIC MODEL AIRCRAFT DESIGN 87

CHAPTER 18 ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Figure 8.Asymmetric blade effect.

Figure 9.Propeller pitching moment.

ated at those rpm. The nomographwas based on two assumptions:

• In top-speed flight , there wouldbe a gain of 10 percent in rpm ,since the prop is operating at alower angle of attack, with lessdrag, th an it would if the modelwas stationary.

• A loss of 15 percent in advanceper revolutio n of the prop com­pared with the prop's nominal pitchadvance. This was incorrectly basedon the oft-repeated statement that aprop /en gin e combina tion devel­oped only 85 percent of the engine'soutput in terms of thrust.

Figure 108.Too little anangle of incidence.

­Nose-up

Down-elevator

• Knowing static and flight rpmallows you to evaluate the gain inrevolution s in flight.

DAVID GIERKE'S INITIATIVESDavid Gierke's "Real PerformanceMeasurement" (RPM) repo rts inModel Airplane News on engine andpropeller performance are, in th iswriter's opinion, outstan ding-a realbreakthrough and a major contribu­tion to model airplane design.

For each engine under study, heprovides not on ly horsepower andtorque curves and details of itsconstruction and hand ling, but alsostatic and level-flight rpm and themodel's actual airspeed at those rpm.He uses a variety of prop makes,diameters and pitches that are suit­able for the engine being evaluated .

.. . -_ --

Figure 10A.Too greatanangle of incidence.

effect of too much incidence or toolittl e. In both cases, fuselage andhorizontal tail drag is higher.

The probl em is to estimate th emodel's level-flight cruising speed.Som e chaps like to fly around th e"pea patch" at maxi mum rpm andtop speed; others, such as yourstrul y, are more conserva tive anden joy flying at something less thantop speed-say, 75 percent of th emodel's highest speed . Either way,evalu at ion of the aircraft 's topspeed is requ ired.

Some years ago, a nomograph wasdeveloped for quickly determining amodel's speed based on its engine'smaximum static rpm and the nomi­nal pitch of the propeller being rot-

LEVEL FLIGHT SPEEDSFor both full-scale and model air­planes , good design practicerequi res that the ang le of incidenceat which th e wing is set (on thedrawing board) result in th e lowestfuselage and horizontal tail drag atthe aircraft's selected cruising speed.

At lower speeds , the aircraft mus tnose-up, through elevator trim , toachieve the AoA that provides ade­qu ate lift. At higher speeds, th ereverse takes place; down-elevatortrim reduces the AoA.

To determi ne the wing's angle ofincidence, you need the wing's air­foil and its lift/drag curves; th e air­craft's gross weight in ounces; theWing's area in square inches; andlast , but not least , the selectedlevel-flight speed in mp h.

It is assumed th at th e lowestdrag will occur when th e modelflies with its fuselage cen terlinehorizontal. The wing's angle ofincidence, relative to that center­line , will then be th e same as thecalculated AoA.

Figures lOA an d lOB show th e

Regarding prop noise , th ere's atrend to long-stro ke engines th atdevelop their highest torque atlower rpm so th at, for example,they can swing pro ps wit hincreased pitches. Higher pitchesand lowe r diameters reduce tipspeeds and prop noise . Propellerswith pitches equal to their diameteror greater (over square), such asll x ll s, llxl2s, ll x13s andllx14s, are now widely available.

88 THE BASICS OF RIC MO DEL AIRCRAFT DESIGN

Propeller Selection and Estimating Level Flight Speeds .. CHAPTER 18

5

4

Nom'n.'Pitch

15D

2110

Z51

31D1514114515111

L ev.1 FlightSpeed IMPH)

18.3ZI

Z5

3835

I 40

58

817881IIlDI

ThiS g rap h will enable you to arrive ata reasonably close est imate of your

model's to p speed

4

5

8 I

i7 I

TIi1

11

11

lZ13

141518171111ZIZ1ZZ +ZI "Z4 •Z5

Static RPMx 1 ,000

Figure 15.This nomograph will enable youto arrille ata reasonably close estimate of yourmodel's topspeed. Aligna straightedgefrom rpm(left) to propnominalpitch(right). The speed in mph Is readoff thecenter scale.

somewhere betwee n th e "nominalpitch " and "zero-lift" ang les. Thenominal pitch is measured, with apitch gauge, on the blade's rear sur­face, at a point 75 percen t of theblade's length , measured from theprop's cen ter. The blade's airfoil,the leading-edge radius and its posi­tion relative to the nomina l pitchall have a bearing (see Figures 11,12, 13 and 14). ..

Figure 12 is a prop blade section .For th e actual advance per rev toexceed th e nominal pitch advance,the blade's actual AoA mus t be

• The assumption of a lO-percentgain in rpm from static to levelflight was no t too far off.

• The big surprise was that th eadvance per revolution exceededthe pro p's nomina l pitch by any­where from 7 to 18 percent.

. ..... .. ...... ~ ....-.-

~l l~ Zero lift-4°ff <A J: .Geo~et ric••••••••••• ••••.• - I ' , pitch

I.. , ' 1Nomin~~

I pitch

Propeller airfoil sections

Zero lift -6°

~ j =r:s-u;~'"'• •• • _ I I 0.75

I ._. NominalI pitch

Figure 11.Graupner prop section.

Flight speeds

Figure 12.APepropsection.

Analysis of David's figures broughttwo facts to light:

• Knowing in-fl ight speeds andrpm allows you to calculate th eactua l advance per revolution andcompare it with th e prop's "nomi­nal" pitch advance.

This calculation is:

Advance per rev =Speed x 5,280 (ft./m i.) x 12 (in./ft.)

rpm x 60 (min./hr.)

Figure 13.Master Airscrewsection.

~ ~ Z,ro n,,'I -y.; '!ieometricI ••••••••••••• ••••••• pitch•••• 0.75'

I !NominalI pitch

Figure 14.Wooden "power" prop section.

THE BASICS OF RIC MODEL AIRCRAFT DESIGN 89

Chapter 19

Design for

Aerobatics

I n the design of an aerobaticmodel airp lane, the first consid­eration must be for th e heavy

loads- both aerody na mic andstructural- imposed by centrifugal

Model1-the Swift.

force in high-speed, shar p, turningmaneu vers. These loads are in addi­tion to the model's own weigh t.

A patte rn ship flying at 100mphin a 120-foot-diameter (60-footradius) turn will sustai n loads ofmore than 12 times its grossweight. If th e combina tio n of wingarea and th e airfoil 's CL max is inca­pable of sup po rting this load, ah igh-speed sta ll will result. A pan ­icked pull-up from a steep dive, atlow altitude, that results in such asta ll co uld be ve ry da mag ing.Simil arly, th e model 's st ruc turemu st not fail un der such heavyloads (see Chapter 13, "StressedSkin Design ").

It's true th at at the higher AoAsneeded to suppor t th ese loads, the

90 THEBASICS OF RIC MODEL AIRCRAFTDESIGN

model's drag will increase enor­mously; this slows th e model andreduces th e load. The highest load ,th erefore, occurs at the start of th emaneuver-before drag slows th emodel appreci ably. The problemlies in selecting th e wing area andairfo il sect ion th at will suppo rtth ese heavy loads. To better under­stand th is, five model aircraft withwing areas of from 400 to 800square inches were analyzed.

The basis for this an alysis ismode l 3, which reflects th e specifi­cations of th e author's Swift. Thismodel has a wing area of 600squa re in ch es and gross es 92

ounces with a full tank (a glow­powered airp lane with an emptytank cannot fly!).

All five have the same 0,46ciengine , RIC equipment and land­ing gear. Anal ysis of th e Swift'sweight discloses th at th e power andcontrol uni ts, plus landing gea raccounte d for 48.5 ounces. It wasestimated th at for each 100 squareinches of wing area added to or sub­tracted from th e 600 square inche s,there would be a weigh t cha nge of5 ounces; a 700-square-inch-areamodel would gross 97 ounces, anda 500-square-inc h versio n wouldweigh 87 ounces.

The Swift's power loading of 200ounces per cubic inch of engin e dis­placement permitted sustained ver-

tical climbs and vertical 8's withlittl e discern ible speed change.

All five wings used for this com­parison have AR 6 and taper ratiosof 0.6, i.e., tip chord = 0.6 x rootchord, and were unswept (see"Wing Area Ana lysis" chart).

AIRFOIL SELECTIONSymmetrical sections performequa lly well invert ed and upright,have zero pit ch ing moments andare ideal for aerobatic models. Theair foil used in th is study wasNACA 641-012-an early laminar­flow airfo il. NACA Technical Note1945 pro vides data on th is airfoiland NACA 00 12 at Rns down to700 ,000 (0 .7x lO6) . A lO-inch­chord wing flying at lOOmph atsea level is operating at an Rn of780,000.

The disadvantage of symmetri­cal airfoils is their low maximumlift capability compared with cam­bered airfoils. Thi s ha s two effects:

• At high-G load s, additional wingarea is needed.

• Landing speeds will be higher,unless slotted flaps are used.

At Rn 700,000, NACA's 64r-012 air­foil has a CL max of 0.9 and a min­imum CDof 0.007.

NACA 0012 has CL max of 1.05and minimum CD of 0.0065 at Rn

Model2-the Wasp tandemwing.

Design for Aerobatics A CHAPTER 19

Figure 1.Section 11ft andpllchlng-moment characteristics of theplain NACA641-012 airfoil section, 24-lnch chord. Figure 2.

Section drag characteristics andsection pitching-moment characteristicsabout the aerodynamic center of the plainNACA 641-012 airfoilsection.

.028

R XJC YICo 0.1 .106 .253 .059o 1.0 .251 .031o 1.5 .261.020!'c,2.0 .265.005

\1 3.0 .256 .DZ6I> 6.0 .259 .011

'0 9.0 .262 -.ODZ

0-0 ' ; 0 ::e:c

u Jc ' I.. .024 ,

~ .020 \ ) ,;1.y~ 0 \ I / .~ ::. .016 , . /~

~ R r:! 012 .'. \. j d~ 0 0.1 . 10' g . '~>~" '~'.~..-../d~30 ~ 1)1

0 1.0 li .008 ~. • 1>6.00 1.5 .. '0 9 0

Flaool'llJl11bolld...te "" 2.0 '" .004 AawodIJl*b--.. .....,;....standard roughness 0 I I I I

-1.2 -.8 -.4 0 .4 .8 1.2 1.6

Section lillcoelliclent-.8 -.4 0 .4 .8 1.2

Section Ii" coetticientc....::..ouC •E " . : 1~o

:lO -.1_.8 -.4 0 .4 .18 1:2 11.6

Section Ii" coellicient

.028

~ .024..:: .020..ou .01601

.!; .012c:ii .008

~ .004

>'<..\~ .

Section angle 01 attack

-16 -8 16 24

_2.0c~1 .6

~1 .2u~ .8

~ .4.. 0

'" ' .4

C -.8..:2-1•2

~ .1

C 0..g-.l:lO -.2

-24

Section angle 01 attack

R XJC YICo 0.1.10' .253 .059o 1.0 .257 .031

o 1.5 .261 .0206 2.0 .265 .005

v 3.0 .256 .DZ8I> 6.0 .259 .011

<J 9.0 .262 ' .002

.026

~~~

DO~_~

~.~~~

~........-.-.

~~

o-.8 ' .4 0 .4.8 1.2Section Ii" ecefticlant

.008

.016

.004

.020

c..;gCDou

.028

cCD <'1--1 ; ; l , 1 ; 1C

Eo

:lO -.t.8 -.4 0 .4 .8 1.21 .6

Section Ii"coellicient

I .024

CD301

~ .012"0

co~..'"

16 24

Flagged symbols denote standard roughness

c --=-

c -·4..3! -.8:.;;g-1.2C .1..E 0

~ -,1

· .2 '--------,-- ---,- - -r-- -,-- ---,- - -r-- -,-- ----,

-24 -16 -8

1j2.0

51.6..31.2:::= .8c~ .4u..'" 0

Figure 3.Section 11ft andpitching-moment characteristics of theplain NACA0012 airfoil section, 24-lnch chord.

Figure 4.Sectiondrag characteristics andsection pllchlng-moment characteristicsabout the aerodynamic center of the plainNACA 0012 airfoil section.

700,000 and wou ld have been abett er cho ice considering the Rns ofthese mod els. However, 641-012was used in the calculatio ns (seeFigures I , 2, 3 an d 4).

Model 3-the Canada Goose canard.

DRAGOther imp ortant considera tio ns arewing drag, profil e drag and particu­larly induced drag. A model withh igh wing drag in both level flightand under high G-force will notperform as well as one with lowerdrag under both. The cha rt showssome startli ng comparisons oflevel-fli ght drag to h igh-G-forcedrag .

This study considers only totalwing drag; it does not include th edrag contributions of fuselage, ta ilsurfaces and landing gear. Altho ughthe tail feathers would var y inproportion to each model's wingarea, th e fuselages would all haveth e same cross-sectiona l area and

would change on ly sligh tly inlength; th e difference in th eir con­tr ibutions to each model's to taldrag would be minimal.

COMMENTS• Mode l 1-400-square-inch area.The CL of 0.874 is dangerous ly closeto 641"012's CL max of 0.9. Sincethis model's level-flight drag is thelowest, it could exceed th e 100mphspeed, despite its high-G wing dragof 77 ounces, and it could stall athigh speed. Its small size wouldadversely affect its visibility, and itslanding speed is high.

• Model 2-S00-square-inch area.Much the same as for model l , with

THE BASICS OF RIC MOD ELAIRCRAFTDESIGN 91

CHAPTER 19 A THE BASICS OF RIC MODEL AIRCRAFT DESIGN

1 400 82 994 77 6.6 29.5 178 0.874 35

3 600 92 1,115 67 9.7 22 200 0.654 29

Model 5-the WildGoose three-surfaceairplane.

Profile CD+ 0.318 x lift CL2 x (1 + 8*)

Aspect rati o*8 (delta) is th e wing planform cor­rection factor. For a wing of taperratio 0.6, it is 0.5 .

• Wing-drag coefficientThe profile Co of airfoil 64 1-012 at aCL of 0.654 is 0.0155 (see Figure 2).The total of both profile andinduced drags is:

210 0.590 27

189 0.742 33258

11.2 201,175 67

1,054 69500 87

4 700 97

2

Wing Area Analvsis

0.0155 + (0.318 x 0.6542x 1.05) =0.3936

Plug in the numbers, and the for­mulas may be solved using simplearithmetic. Happy designing! A

• Wing drag (ou nces)Drag (oz.) =

Total wing CDx speed2 X wing area3,519

the exception that the lower CL athigh G's of 0.742 compared with theCL max of 0.9 provides an improvedsafety ma rgin agains t h igh-speedstalls. Landing speed is high .

• Model 3-600-squa re-inch area,wh ich is the optimum in th isauthor 's opinion . At 0.654, itshigh-G lift coefficien t provides agood safety margi n . Its level-fli ghtwing drag of 9.7 ounces is good ,and its high -G wing dr ag is reason­ab le. Landing speed of 29 mph isacceptable. Its power load ing of200 ounces per cubic inch dis­placement proved satisfactory onthe Swift , and it is large enough tobe read ily vis ible.

• Models 4 and 5-700 and 800­square-inch areas, respectively. Bothhave the same hig h-G wing drag; butlevel-flight wing drag increases withthe added wing area. Combi ned withthe models' grea ter weights, th iswou ld adverseiy affect maneuver­abilit y. The greater wing area result sin lower landing speeds and bett ervisibility.

FORMULASIn developing this comparison, for­mu las pub lished in previous articleswere used and are repeated below

with exa mples for any fellowdesign er to follow.

• Cen trifu ga l forceG's = 1 + (1.466 x speed- mph)2

Turn radius (feet) x 32.2

At 100mph and turn rad ius of 60feet,1 + (1.466 X 100)2 = 12.12 G's

60 x 32.2

• Lift coefficient neededCL =

Gross weight (oz.) x 35 19 x G*SpeedZ x Wing area (sq. in.) x K

At sea level , K is 1.00; at 5,000 feet ,0.8616; and at 100,000 feet , 0.7384.* If greate r than 1G,

CL =92 x 3,519 x 12.12 =0.6541002 x 600 x l

Model 4-lhe Swan canard.

At 12 G's,0.0393 x 1002 x 600

351967 oz .

92 THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Chapter 20

The Crow at rest. Note the wing's high-liftdevices (HLOs).

ing of 40 ounces per square foot, stallspeed increases to 33mph. If thewing max CL could be increasedwith the HLDs to 2.40, the stallspeed would still be 20mph at 40ounces per square foot. (See Figure 5of Chapter 18, "Propeller Selectionand Estimating Flight Speeds.")

U.S. Federal Air Regulations(FARs) specify a stall speed of notmore th an 60 knots (or 69mph) foraircraft weighing less th an 12,500pounds of gross takeoff weight .Sixty-nine miles per hour is as fastas some models can fly at topspeed! Most ligh t, sing le-engine ,full-scale aircraft stall, flaps extended40 degrees, power-off and at grossweight at about 50mph. This is stilltoo high for model aircraft. A"scale" speed is needed!

In "scale realism" (lv[odel AirplaneNews, September 1993 issue), KentWalters' suggestion that scale speedsbe calculated using "the square rootof the scale factor" is explained . Thisis a very sensible suggestion . Most

Devices and

High-Lift

Drag Reduction

MODEL B

The Crow in level IIight.

current models are reminiscent ofthe high-drag aircraft of the '30s.

Very few modelers take advantageof HLDs and drag reduction . Flapsare limited largely to scale models ofaircraft so equipped. Hopefully, thisarticle will persuade modelers toinco rporate flaps and drag reductionin new and innovative designs; thebenefit s justify the effort.

STALL ANDLANDING SPEEDLanding speeds have not beenmuch discussed in the model air­plane press, but are a major consid­eration in full-sca le design .Landing speeds are a fun ction of

the model'ssta lling speed,whic h in turn,depends onweigh t, wingarea and theairfoil's maxi­mum lift capac­ity. Weight andwing area arecombi ned inthe form of"wing loading"in ounces persqua re foo t ofwing area.

At a wing load-ing of 16ounces persquare foot andwing max CL of1.00, the stallspeed is 20mp h.At a wing load-

MODEL ASPECIFICATIONS

Wing area (sq. in.) .750 500Fueled weight (oz.) 96 88Wing planform Constant chord Constant chordAspect ratio 6 6

Span (in.) 67 54.75Chord (in.) 11 .2 9.13Wingloading (oz/sq. ft. ) 18.4 25.3Wing airtoil : E197 E197Tail airtoil .Flat E168Airtoil Cl max 1.17 1.8 (flaps at40°)Power (cid) .•................................0.46 0.46Power loadings (ozJcid) 208.7 191.3Propeller 11x6 10x9

........., 11,000 11 .000

(mpht ..·..······ ..··..·75 1oo..1 19.5 18

...............5

H igh-lift devices (HLDs) on amodel specifically designedto take advantage of the sub­

stantial lift and drag increase theyprovide, coupl ed with good dragreduction techniques, will result insmaller lighter, more nimble air­planes, with a greater range ofspeeds, from stall to top speed. Theirappearance will be sleek-very simi­lar to today's full-scale planes-yetthey will be sturdy and capable ofsustaini ng high -G loads of centrifu ­gal force in their maneuvers.

The homebuil t movement, incooperation with the ExperimentalAircraft Association (EAA), has devel­oped man y superb full-scale, single­engine airplanes of composi te con­struction. They have excellent per­formance on relatively low horse­power. These are the "Lancairs.""Glassairs," "Swift Lightning" and"Pulsars," to name a few. Their out­standing performance is due to gooddesign and careful drag reduction.All have flaps to permit acceptablelanding speeds. In contrast, most

THE BASICS OF RIC MODEL AIRC RAFT DESIGN 93

CHAPTER 20 ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

.40- to .50-powered models will beabout 1;6 or 1;; of the size of their bigbrothers. The square roots of thesescale values are 0.408 and 0.378,respectively. Multiply 50mph byth ese numbers: 50 x 0.408 =20mphand 50 x 0.378 =18.9mph. A model'sstall speed of 20mph seems reason­able. FAR no. 23 stipulates thatapproach speeds sho uld be 1.3 timesth e stall speed, or 26mph. Twenty­five to 30mph are sensible speeds­fast enoug h for good controlresponse, but slow enough for goodpilot response.

In the absence of an airspeed indi­cator, it is not possible to judge amodel 's exact speed. If the glide istoo flat and slow, most models willalert th eir pilots by gently stallingand nosing down (a signal to apply abit of nose-down elevator trim).

A model with slotted flaps flyingon a windy day lands into the windflaps up for more airspeed withbetter pen etration and controlresponse. The higher wing loadingsare less affected by gusts, and thetouchdown speed is reduced by thewind's velocity. An unflappedmodel , with a lower wing loading, iseasily disturbed by gusts, makingland ings more difficult.

MAXIMUM LIFTCOEFFICIENTTo determine the CL max for anunflapped win g, a simple and rea-

Flapsdown. theCrow is descending.

son abl y accurate method is to usethe CL max of the wing's airfoil. For£197, this is 1.17. For a wing withpartial-span slotted flaps of 30 per­cent of the wing's chord in width,the flapped portion will produce anadditional CL of 1.05 at 40 degreesdeflection (see Figure 10 of Chapter3, "Understanding AerodynamicFormulas"). Using £197 again , theflapped portion provides 1.17 +1.05, or a CL max of 2.22. Theunflapped area has a CL max of1.17. To obtain the avera ge CL max,proceed as follows :

Unflapped area (sq. in. ) x 1.17 =xFlapped area (sq. in. ) x 2.22 = YTotal area = x + y

To find the average CL max, div ide(x + y) by the total area. That por­tion of the win g in or on the fuse­lage is considered as unflappedwing area .

Obviously, a tap ered wing ofequal area and aspect ratio,compared with a co ns ta nt-chordwing and the same lengthof slotted flap , would have a higherCL max, since a greater portion is"flapped" (see Figure 1). To deter­mine the stall speed, flaps down,refer to Figure 3 of Cha pte r 1,"Airfo il Selection"; kn owin g themodel's loading and CL max, thestall speed is read off the vertica lleft-hand scale for sea-level condi­tions; otherwise, use this formula(WA = wing area; OF = densit yfactor):

Stall speed mph =

weight (oz.) x 3519CL max x WA (sq . in .) x OF

The density factor at sea level is1.00; at 5,000 feet of altitu de, it's0.8616; and at 10,000 feet , it 's0.7384. This is one variation of thelift formula; involved are four fac­tors: weight, wing area, speed andlift coefficient. Knowing three, thefourth is easily calculated as follows:

Lift (oz.) =

CI.X speed2 (mph) x WA (sq. in.) x OF3,519

Wing area (sq. in .) =

Lift (oz.) x 3.519

Cl

Tapered Wing

Figure 1.Desirable flap proportions forstraight-wing andtapered-wing designs.

DESIGN COMPARISONSTo illustrate the advantages ofHLOs and drag reduction, the spec­ifications of two models (A and B)are outlined-both designed forstall speeds close to 20mph. Bothare powered by .46ci engines andhave the same control unit, butmodel B ha s an extra (fifth) servofor flap actuation.

Model A is typical of many mod­els seen at an y flying field: exposedengine; small spinner (or none);bare music-wire landing gear leg; bigfat wheels, flat windshield; squarecross-section fuselage; dowels; andrubber-band wing hold-downs; flat

Speed - (mph) x WA (sq. in .) x OFLift (oz.) x 3.519

Lift coeffic ient =

CL x speed- (mph) x OFStraight Wing

t.... Span ~Drooped LE N ~~ Flapped areasSpan 13r4••2•• _. __• __•

+t ~ --- I /c , LEslot

•.._-_... .. _.. .. . .. .. . .. ..

I.. ... .... ...........

Aileron Flap : Flap Aileron

t 35%-40% 60%-65% f------ Fuselage~"'~.25C ...... 100%o

0.03Cr .....:'-- _

tDrooped LE!f!!!.. 138%- ~- - - - - -- - _ . - - - -

o03C

94 THE BASICS OF RIC MO DELAIRCRAFTDESIGN

High-Lift Devices and Drag Reduction .. CHAPTER 20

Figure2.The Crow's wingairfoil section.

further 8 ounces (at 70mph) for atotal drag reduction of 12 ounces,permitting a higher top speed formodel B. This is confirmed by expe­rien ce with oth er previo us designs.

Figure 4.Geometry of theretractable Lf slat.

Retracted

FLYING FLAPPED MODELSWind y-day lan dings, flaps up , havebeen discussed. On a quiet day,wind-wise, th e mod el may beslowed, flaps fully deployed, andnosed down as steeply as 45 degreesto th e horizontal. The flap drag willlimit the model's term inal velocity.There is no possibility of a sta ll an d,at a reason able height above th eground, the model is flared for asho rt-field landing. Landing flaps­up on such a day will be tricky; th eglide is fast and flat , and overshoot­ing th e landing area is a real possi­bili ty. Maneuvers under power,flap s exte nded , can be almos tin cred ibly tig ht, and th e flap sth emselves are sturdy eno ugh toperm it th is treatment.

• Takeoffs. Assuming rotation atliftoff to 8 degrees AoA, un flappedmodel A would becom e airborne at24mph. Model B, flaps extended to20 degrees and similar ly rota ted to8 degrees, would be airbo rne at20mph with a shorter takeoff andsteeper climb, flaps still exte nded.With its lowe r power-to-weightratio (power loading) of 191. 3oz./cid, mod el B's lower drag wouldpermit sustained vertical climb .

1o-- -.3c

ounces , leaving 51 ounces for th estructure of fuselage, wing and tailsurfaces. Model B's wing area istwo-thi rds that of model A; it is rea­sonable to estimate that model B'sstructural weight would be two­thirds of model A's, or a weightreduction of 17 ounces.

Model B's weight would, however,be increased by the du cted cowl,large spinner, landing-gear leg fair­ings, full balsa stressed skins, flapsplus th eir servos and linkage, massbalancing of control surface s and a700mAh battery replacing th eusual onboard unit of 500mAh.This is estimated to add 9 ounces,leaving 8 ounces, reducing modelB's weight to 88 ounces. The Crowat 500 square inches of wing area,grossed 87.5 ounces, confirmingmodel B's estimated weight.

As for model A, the Osprey had awing area of 768 square inches andweighed 113 ounces. It had slottedflaps, six servos, a ducted cowl andheavy landing gear weighing 14.5ounces The fuselage was heavilyreinforced for use with twin floats.The fuselage, wing and tail surfaceswere no t fully balsa-sheet-covered.By comparison, model A's fueledweight of 96 ounces for 750 squareinches of wing area is conservative.

• Drag comparison. At 70mph,model B's win g would ha ve 4ounces less profil e and induceddrag than mod el A's wing; but that'snot all! The engine cowl, spinner,shorter rounded fuselage, smallertail surfaces, landing-gear leg fair­ings and small streamlined whee ls,overall smoother surfaces andabsence of dowels and rubb er bandsholding the wing are conservativelyestimated to reduce drag by a

Flappivot polnt- +

R.23c

~R.23c

Figure 3.Geometry of thefixedleading-edge slot.

mum exposure of con trol horns. Ithas slotted flaps, 30 percent of thewing chord in width and 60 percentof the semi-span in length.

Because of its sleek, low-dragdesign , similar to the Swift's, it iscapable of high speeds. Mass bal­ancing of ailerons, elevator and rud­der is incorporated to avoid flutterthat could be very damaging.

WEIGHT ANALYSISLook at the chart on page 93. Thepower and control units and land­ing gear of model A weigh 45

balsa tail surfaces; exposed controlhorns; lots of "built-in headwinds"(beneficial for steepening themodel's glide and making landingseasy). It has no flaps. The wing is D­spar construction, plastic-film-cov­ered; the fuselage is lite-ply; and thetail surfaces are V4-inch balsa sheet.

Model B has a ducted cowlenclosing the engine; a large spin ­ner; landing-gear leg fairings; smallstreamlined wheels; concealed winghold-downs; balsa-sheeted, stressed­skin structure with a film overlay;streamlined windshield; and mini-

THE BASICS OFRIC MODEL AIRCRAFT DESIGN 95

CHAPTER 20 ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

.32

.24

.20

.16

.12

.08

.04

04 8 12 16 20 24

Angle 01 attack In degrees

o

and a delay in stall to a 9-degreehigher AoA, with only a small dragincrease.

The retractable versions are self­opening at higher AoAs, but theydemand smoothly operating, non­jamming mechanisms and shouldbe linked so that th e slats of bothwing panels extend simultaneouslyfor obvious reasons. They may alsobe servo ope rated .

To this author, the added com­plexity of the retractable slat is notjustified by its benefits. The Crowhas full-span, fixed LE slots, asshown in Figure 2.

o

0.2

0.4

1.6 H-+-+-+-+-

1.4 H-+-+-+-+ -HH -

E~ 0.8

'iii8 0.6

• Wing t railing-edgeHLDs. Figure 1 of Chapter14, "Design for Flaps" andFigure 12 of Chapter 5,"Wing Design ," describeand show the additionallift provided by five typesof flap: plain, split, slotted, d 1.2

slotted with extended lips g-and Fowler. 0 1.0

The most practical type,giving the optimum addi­tional lift with lowestadded drag, is the 30 per­cent of chord slotted flapwith extended lip. Theseare easily operated by onestandard servo; th ey'rerugged and very effective.Becauseof their low drag at20 degrees extension, theymay be used for takeoff

Figure 5.advantage. Figure 2 iIIus- The benefits of thefixed Lf slot.trates the flap design for theCrew's wing. The only dis­advantage is the longer streamlinedarms from flap to pivot point neededto provide the backward movementfrom 0.7 percent of chord to 0.85 or0.9 percent of chord.

Though the Fowler flap providesgreater lift, its backward and down­ward motion demands complexpivoting arms or other mechanismsand powerful servos.

• Wing LE high -lift devices: LEslots . Figure 3 illustrates fixed LEslots; Figure 4, retractable LE slats.Figure 5 shows the benefit of fixedLE slots: an increase in CL max of 0.4

*centrifugal

For model B, lift exceeds load atall speeds . Note the loads themodel's structure must sustain athigher speeds. In a tight turn at90mph, the load is 880 ounces, or asurprising 55 pounds.

One advantage of the "30 per­cent of win g chord flap s withextended lip" is that there is verylitt le pitch change when loweringthe flaps. The Swift continued onits merry wayan lowering fullflaps, but it flew appreciably moreslowly.

• Centrifugal force. One concernwith higher wing loadings, such asfor mod el B, is that in a tight turnor sharp pull-up, centrifugal forceplu s the model 's weight couldexceed the wing 's maximum liftingcapacity. This could result in a dan­gerous, high-speed stall, particularlywhen pulling out of a steep dive ata low altit ude. Assuming a turningradiu s of 60 feet (l20-foot diame­ter), the following tabulates theG-forces involved compared withmodel B's maximum lift capaci ty,also in G, at various speeds.

Speed WI. + cent.* Wing max.(mph) lift (G) lift (G)

60 5.00 6.8070 6.45 9.2580 8.11 12.0090 10.00 15.30100 12.12 18.90

Drooped leading edge

5020 30 40Angle 01attack- In degrees

10o

.2

E.. .6uE:g .4u

.s 1.0§o .8

,.~. ;,,;~=<::..,Leading-edge droop

1.2

1+--+----8/2-------+l

Figure 6.WingLf modification for Improved stall/spin resistance.

96 THE BASICS OF RIC MO DEL AIRCRAFTDESIGN

High-Lift Devices and Drag Reduction ... CHAPTER 20

Inverted LE slot

Mass balance

Figure 7.The Craw's stabilator section.

• NASA LE droop . As shown inFigure 6, these de lay the stall byabout 8 degrees; they provideextra lift at higher angles ofattack; and they have low drag.Used as shown for 38 percent ofthe semi-span, ahead of theailerons, they greatly improveaileron control effectiveness ath igh AoAs. The "droop" was usedon the Swift to advantage.

• Horizontal -tail LE slots. Toobtain the high AoAs, before thestall, of the wings with LEslots andslotted flaps, a powerful downforceon the horizontal tail is needed toraise the model's nose. The Craneneeded inverted LE slots on its hor­izontal stabila tor to ach ieve thisattitude. Similarly, the Crow STOLmodel's horizontal stabilator isequipped with inverted LE slots asshown in Figure 7.

• Slot-lip aile rons. Illustrated inFigures 2 and 8, these replace nor-

mal ailerons when full-span flapsare used. On both the Crane andthe Crow, these have proven to bevery effective , and they workinverted. At anyone time, only oneworks- that on the inside of theturn; the opposite one lies flat. Theraised aileron reduces lift and hasinto-the-turn yaw. Both are lightlyspring loaded to hold them downwhen they aren't being actuated.With flaps extended, they are evenmore effective. Raised, the sloteffect over the flap is destroyed,redu cing flap lift and adding into­the-turn drag . They provide crisproll control at lower speeds of flap­extended flight-when most need­ed! The dimensions of these slot lipailerons on the Crow were: width­15 percent chord; length---60 per­cent of semi-span.

• Landing-gear design. Landing­gear design for models with HLDs isthoroughly discussed in Chapter16, "Landing Gear Design." The

"tail ang le" (also called the "tip­back ang le") must be large enoughto permit the model to land at veryclose to its stall ang le of attack andits slowest speed.

• Control unit . Flap operationrequires an extra servo, which maybe operated by the retract switch ona 5-channel (or more) radio , butthis provides only full-up or full­down flap positions-no in between!An auxiliary channel is desirable,controlled either by a three­position snap switch that providesfull-up, 20 degrees down and 40degrees down-flap positions; ora proportional slide switch that per­mits a choice of any flap positionfrom full-up to full-down.

A TRIBUTEDick Murray and Ken Starkey-twofriends and fellow club members­have test-flown each of thisauthor's new designs . Both arepilots of consummate skills; andboth offered valuable, constructivecomments on the flight characteris­tics of each model. For lending metheir skills and for their friendship,I am deeply grateful. Do try HLDsand drag reduction. Models of thistype are highly versatile, and flyingthem is pure fun-well worth theextra effort their design and con­struction entails. Above all, they aresleek and beautiful. ...

A. Flaps down

Figure 8.Slot-lip aileron act/on.

Drag

Drag

Drag

B. Flaps up

THE BASICS OF RIC MODEL AIRCRAFT DESIGN 97

Chapter 21

Centrifugal

Force and

This chapter describes the evaluationof CFand analyzes various center-of­liftjCG positions for conventional(tail-last), tandem-wing, canard andthree-surface configurations.

increase to provide the additionallift needed .

• Drag. Both profile drag andinduced drag increase .

Maneuverability

I n aerobatics, cen trifugal force(CF) imposes both aerodyna micand structura l loads on an air­

plane that may be many times themodel's weight . It deserves seriouscon sideration. CFacts at th e plan e'scenter of gravity (CG). The centerof lift may be ahead of, on , orbeh ind th e CG in man euvers.

CENTRIFUGAL FORCEEVALUATIONIt's easy to evaluate the maneuver­ing loads brought about by CF. Twoimportant maneuvers will be con­sidered: turns in a vertical planeand turns in a horizontal plane.Most aerobatics invol ve a combina­tion of these.

• Turns in a vertical plane-a seriesof loops. The CF will be evaluated atthe bottom of the loop where weightand CF act downward.

• Downwash. The increased liftcoefficient causes an increase inthe downward deflection of thedownwash striking either the hor­izontal tail or the aft wings of thetandem, canard, or three-surfaceconfigurations.

• Pitching moment (PM). Forcambered airfoils, the wing's PMmay increase with increase in itsangle of att ack (AoA). The chartsfor the airfoils involved must beconsulted.

Pitchingmom~~ t..... lifl (2G) Downwash

, : Neutral point-.35 MAC '\~\~~~ C J 0/ ----= ~ake Download......'

CG & AC~ L....- Weight and centrifugal force 2G ...25 MAC L~Static margin-.10 MAC

Figure 1.Loads in a vertical turn (loop).

• If th e center of lift is behind th eCG, th e force coup le will cause th emodel to nose down and resist themaneuver.

• If the center of lift and CG arevertically aligned, weight and CFare neut ralized by lift and do notaffect maneuverabili ty.

• Drag moment. If the center oflift is above the CG, the increaseddrag will cause a nose-up effect. Ifcenter of lift and CG coincide, theresult is neutral. If the center of liftis below the CG, a nose-downaction result s.

• Thrust moment. If the thrustline is above the CG, a nose-downmoment results. If the thrust linepasses through the CG, the result isneutral. If it is below the CG, anose-up moment occurs.

• Structure. The model's structuremu st withstand the substantiallyincrea sed load without failing .

HORIZONTAL TURNSSee Figure 2. With a plane flyingat SSmph in a steady, level, coordi­nated, 200-foot-radius turn, CF actshorizontally; to provide lift tooppose it, the model must be

• Maximum lift coefficient. If thecombined weight and CF in small­radius, high-speed turns exceedsthe wing's maximum lift capacity, ahigh-speed stall will occur.

• Lift. TheWing's AoAand CL mu st

VERTICAL MANEUVERSAssume that a plane flying at SSmphis at the bottom of a continuing 200­foot-radius (400-foot diameter) loop(see Figure 1). The combined weightand CF total 2G's, or twice themodel's weight, and this force acts atthe model's CG. The increase in theload the wing must support is mod­est. Had the loop been flown at90mph,with a lOO-foot radius,

the CF wouldhave increasedto SAG's, plusthe model's IGweight, for atotal load of6.4G's.

Referring toFigure I , theresulting forcechanges are:

• Turns in a horizontal plane-asteady, level, coordinated turn inwhich weight acts downward butCF acts horizontally.

Centrifugal force-1 GModel's weight-1G

Total load- 2G's

Speed-55mphTurnradius-200 fl.

• If the center of lift is ahea d of theCG, lift is upward; CF and weightpull downward at th e CG. A forcecouple is created th at causes themodel to nose up, and thi s assists inth e turn or climb.

98 THE BASICS OF RIC MODELAIRCRAFTDESIGN

Centr ifugal Force and Maneuverabil ity ... CHAPTER 21

• CG aft of the ae rodyn am iccenter (Figure 5). In this configura­tion, the CG is slightly behind the

• CG on the aerodynamic cen te r(Figure 4). The wing's lift, at itsaerodynamic cen ter, is vertically inline with th e CG. In turns, CF nei­ther adds to nor reduces the hori­zontal tail's load.

If the wing's airfoil is cambered,the tail must compensate for thenose-down pitch ing moment. If it issymmetrical, there is no pitchingmoment; this increases th e horizon­tal tail's effectiveness. The increasein the downwash angle that resultsfrom the wing's increased lift coeffi­cient aids the maneuver.

Elevators of 30 percent of thehorizontal-tail area are suggested.The Swift typifies this arrangement.

longitudinally. In man euvers, how­ever, a force couple is created; CFandweight acting at the CG pull down­ward; wing lift at the aerodynamiccenter pulls upward ; both cause theairplane to move away from the loopor turn, resisting the maneuver.

A substan t ial in crease in taildownload is required to overcomethis. Elevators whose area is 40percent of the total horizontal tailarea will have adequate authority,but at high CF values, they simplycan 't provide adequate download,and the tail stalls. This limits themodel's high-speed, low-radiustu rning capability and itsmaneuverabilit y.

The increase in the downwarddeflection of the down wash strikingthe horizontal tail does assist, butth is brings th e tail closer to itsstalling angle.

Downwash~ Lin at .25 MAC .."*-

/ NP at .35 MAC Light=-_-,W~a~k!!..e _-.l~ download

~I~~~~rg l n .. to balanceCG at ---.. . 0 MAC PM.25 MAC

Pitching Increased downwash anglemoment~ V .... L1n 2G ..~

~NP 11\111

~~ ..0(~ake ~o~~{:~~Load 2G 0 orce c~ ~

• Forward CG. The CG is at 15 per-cent of thewing 's MAC,ahead of thewing's aerody­namic centerof lift, whichis at 25 per­cent MAC.The generousstatic marginof 20 percentMAC ensuresthat themodel will beeasy to fly andvery stable

horizontal tail controls theWing's AoA and compensatesfor moments caused bythrust, drag, pitch and CGlocation.

Figures 6, 7, 8 and 9 dis­play configurations in whichtwo surfaces actively providelift , share the model's weightand provide additional liftto overcome th e variousmo me nts listed abo ve.

Elevators for planes shownin Figures 3, 4 and 5 are on thehorizontal tail's trailing edge.For th e tandem wings shownin Figure 7, elevators may beon th e trailing edges of eitherthe fore or the aft wing .

Canard elevato rs are usually onthe fore plane's trailing edge(Figure 8).

For the three-surface designsshown in Figure9, the elevatorsare on the hori­zontal tail's trail­ing edge.

In all cases,the CG must beahead of thene utral poin t(NP) for longitu ­dinal stability.

Note the rear­ward shift of theCG from Figures3 to 9 as themodel's configu- Figure 4.rations change. Loading withCG at .25 MAC in a 2Gturn.

The followinganalyzes each configuration and itsresponse to CF and other forces,both in level flight and under a2G load.

.---_---,., B

CGat-.. ..•15 MAC

Pitching Increased downwash angle

moment '~~L1n_2G ..~'-....... ,. ,~~.Igher

"'---- NP 11\~ down·..-::=~ Wake load

Nose·down ..Load-2G---.- forcecouple

CG

.------- - - -.~-~,...!!... ....,..L-+I

;(Centrifugal force-1G

Speed-55m phTurn radlus-200 n.Centrlfugallorce-1G (see Fig. 1)Model's welght-1G

IIIIIIII

Model's Welght-1 lt'l"""

Figure 3.Forward CG loading In 2G turns.

Figure 2.Loads In a horizontal turn.

CG LOCATIONFigures 3 through 9 illustrate sevenpossible stab le CG location s.

Figures 3, 4 and 5 are for conven­tional airplanes where only thewing's lift supports the model; the

banked as shown. But the wing's liftmust also overcome the model'sweight. As in Figure 1, line CF repre­sents IG, and it must be opposed bya centripetal force of IG. This resultsin a force diagram that is solved byvector analysis. In Figure 2, line ACis th e centripetal force of IG andline BC is the model's weight of IG.

ABC is a right-angle triangle inwhich our old friend , "the square ofth e hypotenuse is equal to the sumof the squares of the other twosides" applies . As Figure 2 shows,the result is 1.414G's, and the ang leof bank is at 90 degrees to line AB.

Obviousl y, in terms of turn radiiand speeds, the hor izontal turn isless demanding than the verticalturn. These comments on lift, drag,etc., for vertical turns, however, doapp ly to horizontal turns .

Ip't hi ~ Lin at .25 MAC Downwash .. cr==-

- ~o~e~gV NP-.35 MAC ' __ n,

- Jf ~Download

Wake

THE BASICSOF RIC MODEL AIRCRAFT DESIGN 99

CHAPTER 21 .... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Wake

Lift at .25MAC

Oownwash.~lifting tall -..+

support th e mod el, plu s addition­al forep lane lift to compensate forthe nose-down pitching momentsof both wings' cambered airfoils.The combined center of lift of thetwo wings is thus ahea d of theCG. Application of down-eleva toron the foreplane does two thin gs:it increases the foreplane's lift ,and the downward ang le of th edown wash reduces the aft Wing'slift. Both act to move thecombined center of lift far the rforwa rd .

CF acting at the CG aft of thiscombined center of lift gre atlyaids the ma ne uver. In ret rospect ,the moment arm from CG forwardto the foreplane's 25 percent MACis short. A better option wouldhave been to place sma ller eleva­to rs on the aft Wing's t railingedge, bet ween the vertic al sur­faces, with ailerons on the fore­plane. Flaps, if used , would berequired for bo th wings.

Pitching moment ~ Lift at .25 MACf--.-i N:=_~ .45MAC

....: LowPitchingmoment Increased downwash angle .......:- upliftf -e-unzn ---~-----~ / NP "IUIIl______

c ~~~Load 2G~ Nose-upforce couple

• Increase themoment armbetween thiscombined cen­ter of lift andthe CG, aug­menting thenose-up force.

The combina­tion of in­creased winglift, reduced orreversed ta illift and theincreased force

Figure 6.lifting tail loadin a 2G turn.

percent of thehorizonta l tailis adequate.

The configu­ration is uns uit­able for a mod elequipped withflaps on thewing. Fully ex­tended, the flapswould:

• Sharply in-crease the down-ward angle of thedownwash striki ng the horizontaltail, reducing its lift or reversing it todownlift.

• Substa ntiallyinc rease t h ewing 's lift andlift coefficient.

• Move the combined center of liftof the wingand tail for-

D ~w uplift ward.ownwash..

NP at .40 MAC-r--~~

~ke.10 Static margin ----..

Load-2G --..

CG at _-~ _.30 MAC

Pitching Imom!!!l- ncreased downwash angle;: ~Lift-2G ---h~_~~l<~NP "IUIIl~ ~ download

~Nose-up force couple

wing's aerodynamic center at th e 25percent MAC location by 2 to 5 per­cent MAC. A modest increase in thehorizontal tail's area of 3 to 5 percentof th e wing's area will move the neu­tral point aft and maintain a healthystatic margin of 10 percent MAC.

Under CF loads, the force couple isupward at the aerodynamic centerand downward at the CG behind theaerodynamic center, and that helpsthe elevator action (as does theincrease in downwash deflection).An elevator area of 25 percent of thehorizontal-tail area is adequa te.

LIFTING TAILSSee Figure 6. This type could almostbe classified as a tandem-wingmodel ; both wing and hor izontal tailshare in lifting the model's weightand in compensating for the variousmoments. It's an old free-fligh tsetup, typified by the late Carl

Figure 5.CG aft of .25 MAC loading in a 2Gturn.

Elevato r~""

Load-20-'"

~Increased downwash angle

-----.::~ \CG NP~ke P~ .....,l;-Reduced~,, _ lift

- Lift .25MAC 1""'-- Combinedcenter of liftPitching / Nose-upforce couple opposingmom",!r / wing's pitchingmoment

f~ ! Wake PitchingC, J °lNP .. momentr ~Statl c margin -c !.....,l;-LJ~

1G-... - Oownwash t"~

Pitching ~ ""{- CCofL ~~mom~t Increased<; f _ li ft

Figure 7.Tandem-wing loading in a 2G turn.

couple betweencen ter of liftand CG wouldrender the air­plane dan ger­ous ly unstablein pitch whenth e flaps wereextended.

TANDEMWINGSSee Figure 7.This configura­tion is shownin the Wasp.Both wingssha re th e lift to

Goldberg's classic Comet design andadvocated by H. deBolt.

The lifting tail has a flat-bottomairfoil and is 35 to 40 perce n tMAC of the wing in area . Thismo ves th e NP aft to 45 perce n tMAC, permitti ng a CG at 35 per­cen t MAC, well behind th e wing'saerody nam ic center at 25 percentMAC, but pro vides a healthy sta t­ic margin of 10 percent MAC.

Up-elevato r reduces the tail'supward lift. CF acting at the CG isbehind th e center of lift , and theresu lt ing strong force coupleactively assists up -elevator action,as does the increased angle ofdownwash . An elevator area of 20

100 THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Centr ifugal Force and Maneuverability At. CHAPTER Z1

50 100 150 200

Speed (mph)

5

10

15

....~ 25o...Cl

30

35

20

40.------ - -r----,

elevators are sens it ive; a rat io of 20percent elevat or area to total tailarea is adequate.

• The aft plane mu st achieve zero

INVERTED FLIGHT ANDMANEUVERABILITYOf th e seven configura tio ns dis­cussed so far, on ly Figures 1, 2 and 3will easily fly inve rted. The rest relyon two wings for suppor t. Inverted,th ese types wou ld not satisfy th etwo critical requ irements for longi­tudinal stability:

• The foreplane mu st stall first.

results, and thishelps with theman euver.

The CanadaGoos e and theSwan had slottedflaps on both foreand aft wings .

THREE·SURFACEDESIGNSSee Figure 9 .The Wild Gooseshown in thephotos illustratesth is design. Thehorizon tal tailcontrols pi tch ,and both wings

have slott ed slaps for slower land­ings. The tail's area moves th e neu­tral point aft, and that permits th eCG to move aft as well.

The closer spacing (longitudinal­ly) of the wings results in a shortmoment arm from CG to foreplan eAC. This results in a higher load onthe foreplane to overco me thepitching moments of the tw owin gs. The combined cen ter of liftis thus ahead of th e CG.

Up-elevator reduces the for e­plane's load but doe s not reduce itslift . The combined center of liftmoves forward; CF acting at th eCG produces a nose-up for cecouple.

The combined elevator down­load and the reduced foreplaneload are very effective in pitch . The

P1"h'"'~.-u:::'Combined momecenter Oll~ Downwash

Pitching L ".25 MAC t CG NPmomeiJY " y Stali c margin

f Wake-..:..:.=--).~ 1G Nose-up lorcecouple

PitchingII/creased d mo~nt "-Red uced lin

oWl/washaI/Ole

,.- Increased lin

,~~C7h~g Combined t CG NPcenter-. 0

01 lin ""I( Increased nose-upcouple

1f~ 1( ' - 1utYElevator / ~~

2G load

Figure 8.Canard loading In a 2G turn.

CANARDSSee Figure 8. Like in th e tandem­wing version , th e foreplane mu stlift its share of th e model's weigh t,plus provide additional lift to offsetth e cambered airfoils' pitch in gmoments; this puts the combinedcenter of lift ahead of th e CG. Sinceth e distance from CG to foreplaneAC is greater than for the tandemtype , the canard foreplane's pitch­ing-mo ment load is less than for thetandem foreplane .

Depressing the forep lane's eleva­tors increases its lift and increasesthe downwash deflection; thi sredu ces the rear plane's lift in theportion "shadowed" by the frontwing. Both move th e combinedcen ter of lift forward. Under CF, agrea te r nose-u p force co up le

CG 0 NP

"-1I".25 MAClift first. For conven tional tail-la sttypes , optimum man euverability isobtained by having a symmetricalairfoil and ensuring th at thrust,drag and lift forces run through th eCG. Th is arrangement neutralizesthe disturbing moments and allowsthe tail full effectiveness, particularlyif it is T-mounted.

Except for its airfoil, which issemisymmetrical, th e Swift's designcomplies with these stipulations. At.

Figure 10.Gforces in pullingout of a verticaldive atvarious speeds andturnradii, includingmodel's' 1G weight. Example: at 100mphina100-foot turn radius, Gforces are 7.7limesthemodel's weight.

3C:=::=­No load

2

..- 1I".25 MAC

Downwash

\ ?Y~ r --+---~

---+-~..:-

'- Increased nose­upcoup le

Combinedcenter 01 Ii"

----------...Pitching

mom?:",J~:" MAC yeo .l " "."·up1,,,,,,..le1G load-rt~'-=-+, Stalic margin

1 Wake.. -

Figure 9.Three-surface loading in a 2G turn.

THE BASICS OF RIC MODEL AIRCRAFT DESIGN 101

Chapter 22

Canards,

Tandem Wings

and Three-

Surface

Designs

H istory repeat s itself. The firstsuccessful powered flightswere made by canards; sub­

sequent design s incorpo rated botha can ard forepl an e and a tailplanebehind th e wing, i.e. three surfaces.

Even tually, th e wing and rear tailversions predominated, and th ey'renow th e conve ntional configura­tions. Recently, however, large lyowing to Burt Rutan 's efforts, thecanard, th e tandem-wing and theth ree-surface versions have reap ­peared (Figure 1). Today, Burt's lat ­est designs are more conventional,but still unique, and in this chap­ter, I'll discuss th e design of th esethree configurations.

The Swan canard pusher.

102 THE BASICS OF RIC MODEL AIRC RAFT DESIGN

ADVANTAGES• Increased safety. For well­designed, full-scale canard, tan dem­wing and three-surface aircraft, th emajo r adva n tage of th eir design istha t it frees th em from th e too­often-fata l, sta ll-spin -at-low-alti­tude crash . Though the foreplanemay stall, th e main wing does not.

• Shared load; reduced main-wingarea. In a conventional aircraft, th ewing does all the work; th e horizon­tal tail is lightly loaded (downwardin most cases) and simply contro lsthe wing's AoA. On th ese three typesof front-wing aircraft, th eir forwardsurfaces work hard and share theload with th e main wing, whichmay, as a result , have a reduced area.

• Main wing spar may be out ofthe way at th e rear of th e cabin; th econventional version's spar goesthrough the cabin an d interfereswith passenger seating (particularlytrue of low- an d mid-wing types).

• Smaller, lighter, more compactairplane-achieved by dividing therequ ired win g area between twolifting surfaces.

DISADVANTAGES• Heavily loaded foreplane. Forstabili ty, the foreplane mustbe mu ch more heavily loaded (in

terms of ounces orpounds per squarefoot of wing area).The foreplane'sloading con tro lsth e aircraft's stallspee d, which isco nsi de rab lyhigh er th an themain Wing's stallspeed. Canard andt and em - w in gtypes take off andland faster andneed a longer run-

way than conve ntional aircraft . Thethree-surface design is better in th isrespect because its foreplane loadingmay be reduced, but t hree su r­fac e s mean mor e interference drag.

• Lim ite d aerobatic capabilities.The high foreplane loading, com­bin ed with the inability to stall theaft wing, lim its th e aerobatic capa­bilities of these three classes. (SeeChapter 4, "Wing Loadi ngDesign .")

AIRFOIL SELECTIONFor all three types of forward-wingaircraft, airfoil selection is very criti­cal. There are three broad categoriesof airfoil: heavily cambered (such asE214); moderately cambered (suchas E197); and no-camber, symmetri­cal type (such as EI68). (See Figure 7in Chapter 1, "Airfoil Selection. ")

Figure 2 compares lift with AoAcurves for th ese three airfoils. Noteth at , th ou gh th e heavily camberedE214 sta lls at a lower AoA, it startslifting at a higher negative angleth an the othe r two . The symmetri­cal E168 sta rts to lift on ly at a posi­tive an gle, and its max CL is th elowest of all th ree. (See th e appen­dix for the sectio n characteristics ofthese airfoils .)

Since all three co nfigur ationshave both forwa rd and main wingssharing th e lift, two requireme ntsare of critical importance for suc­cessful, stable flight:

• The front wing must stall befor eth e main win g stalls. If the mainwing stalls first, th e scenario depict­ed in Figure 3 will result; at low alt i­tude, a crash is inevitable.

• The main wing must arrive at itsang le of zero lift before the fore­plane ach ieves zero lift. If th e fore­plane ceases to lift while the ma inwing still lifts, th e behavior shownin Figure 4 results.

Canards, Tandem W ings and Three-Surface Designs ... CHAPTER 22

Figure 4.Steepdive as foreplane hilszero-lift angle first.

lob-:A. Foreplane reaches zerolill angle first

DOWNWASH ANDTIP VORTICESDow nw ash is tho ro ughl y dis ­cussed in Chapter 7, "Hor izon talTail Incidence", and cha rts for esti­ma ting downwash angles areprovi ded . Each of the th ree, for­ward-wing aircraft is affected bydownwash.

Chords of less than 5 inche s are tobe avoided . (For more on thesesubjec ts, refer to Chapter 1.)

Low aspec t ratios increase th esta llin g angle (desirable for th emai n wings) of all three types.Shorter main wingspans im pro veroll respon se.

A mild forwar d sweep on th eforep lane promotes roo t-stallingfirst (see Chapter 5, "WingDesign "). The result is a gen tle ,progressive stall as th e angle ofattack in crea ses. Such forwardsweep should no t exceed 5 degreeson th e 1;4 MAC line. On a three­surface design, forwa rd swee pwould also ben efit th e hor izontaltailplane.

• The negative angle of zero lift isincreased.

• The sta ll angle is redu ced.

REYNOLDS NUMBERS.ASPECT RATIO ANDPLANFORMHigh aspect ratios red uce thestalling an gle (desirable for fore­plan es) but result in lower Rns,parti cu larly at land ing speeds .

• CL max is increased substantially.

I

~~stalls firsl ---

B. Bolhwings stall

Slallangledecrease

Figure 3.Nose-uppilchasaft wingstalls first.

..Pos.

Angle 01 attack

Basic airfoil E214RE 200,000

-.

-'0

II

I

2.00

- f-- - '~ --;:1Add itional lill " /

Irom flapat 20' /

/_1__~'"E214 with .40C I

slottedflap /depressed 20' 1.

/~egative angle /Increase /

/I

/

Figure 5.Impact ofa 40%chord slottedflapdeployedto20degrees onairfoil section 214.

t rouble. In thelanding flare, if th eforeplane were tosta ll sudde n ly,landing would bevery hard andwould probablydamage the nose­whee l landing gear.

For the three­surface airp la newith a hori zontaltail and elevato rs,

a sharp foreplane stall is desirableto prevent up- eleva-tor action fromstalling both thefront and m ainWings . Elevatoraction would pre­vent a sudde n nosedrop . See Epple rE2 11-a forepl aneairfoil with a sharpstall at low Rn-inthe appendix. Notethe reduction in th enegative AoA ofzero lift as Rn isreduced.

Using slotted flapson th e foreplanes ofcanard and tandem­wing models forpitch control has three effects (seeFigure 5):

A-=-~E1971----B-"fiU4

Figure 2.Lift curves of three airfoil types.

Figure 1.Rutan 'saround-the-worldVoyager.

..

RE 200,000

) /? .IIi -.

\ . /il ,\l_/_ NegatlYe-Angle01 Attack-PosltIYe _

"Alpha"

With these considerations in mind,look again at Figure 2. Obviously,airfoil E214 would be an excellentchoice for the front wing. Its earlystall and high negative angle of zerolift satisfy both requirements, andits stall is gentle.

For the main wing, airfoil E197would again be excellent. Itshigher AoA at the gentle stall and itslower negative angle of zero lift com­ply with both manda-tory require­ments. E168 would not be suitablefor either front- or main-wing air­foils, but it would be a good sectionfor the horizontal tail-plane of athree -surface design.

An airfoil's stall pattern at CL maxand at the wing's flight Rn is anotherimportant consideration. Obvi-ously,for a canard or tandem-wing fore­plane to have sudden-lift-lossor sharply stalling airfoils invites

THE BASICS OF RIC MODEL AIRCRAFT DESIGN 103

CHAPTER 22 ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Figure 7.Downwash impact ona tandem wing.

plane's level-fl ight downwas hangle. The part of th e wing th at 'sout of downwash is left at the AoAcalcul ated to produce adequate lift.This calls for a "jog" in th e wingand was used on th e Swan .

A variatio n of th is is to use th eNASA dro op for th at part of th ewing th at 's out of downwash, sothat th e inboard ends of the droopare just behind th e foreplane tips.

A simpler method, where th eforep lane span is roughly half tha tof th e main wing, is to increase th ewhole ma in wing's AoA by half theforep lane level-fligh t downwashangle. The main wing outboardporti on s will have higher lift coeffi­cients, closer to the stall. TheCanada Goose used this method.

A third method is wing washoutwit h increased root AoA andreduced tip AoA. An accura te built­in twist is needed, but it resu lts inan increase in wingtip stall marginand is stabilizing on a sweptbackmain win g.

In all cases, th e net lift sho uldequal th e calculated lift needed.

• Overall we igh t esti mation.Obtaini ng a rough prelimina ryweight estimate while th e model isstill in the conceptual stage is essen­tial but not easy. The data on weightestimating in Chapter 13, "StressedSkin Design and Weight Estimat­ing," will help. When th e model'ssize and proportions have beenestablished, a more accurate weightappr aisal is advisable. Chapter 5,"Wing Design ," also providesinsight into obtaining thi s estimate.

To avoid th e impact of foreplane­tip vortices on th e main wing , avert ical gap between foreplane andmain plane of half the aft wing'sMAC is suggested-eithe r th e fore­plan e low and th e main plane high,or th e reverse may be used. Theforeplane -tip vortices will th en passunder or ove r the main wing.Longitud inal separation or "stag­ger," between 1/4 MAC points ofeach wing, of two to three times theaft wing's MAC, is ap propriate .

For the th ree-surface design, it issuggested that th e horizon tal tail be"T"- mounted on th e fin where itwill be mo re effective, and th e stag­ger be 1 to 2 times the aft wing'sMAC.

LOGICAL DESIGN STEPS• Power and contro l unit selec­tion. The powe r and control unitstogether weigh SO percent or moreof most models' total weight . Thefirst step in design is to choosethese units and obtai n theirweights.

An plane

-- TReduced

angle 01atlack

Forep lane

Tip vortices

Figure 6.Downwash impact ona canard.

• Can ar ds: foreplane downwashimpacts on a portion of th e aftwing (equal in span to that of theforeplane), reducing th e angle ofattack and lift in the downwashedarea (Figure 6).

• Three-surface models: th e mainplane is affected as in th e canard(Figure 6); and th e horizontal tail isaffected by th e downwash from thatportion of th e main wing that's"shadowed" by the foreplane down­wash. The reduced AoA of th e "shad­owed" port ion of th e main wingmay be compensated for as follows:- For tand em wings of equal span:for level flight at th e designed cruis­ing speed, the aft wing's AoA shouldbe inc reased by the down washangle generated by the foreplane .-For canards and three-surface air­planes: shadowe d portion s of th emain win g should have an increasein AoA th at 's equa l to the fore-

• Tandem-wing aircraft: thewhole span of th e aft wing is simi­larly affected (Figure 7).

Area B Area 01reducedeffectivenessin downwashat 80%

DistanceN=area Ax separationtotal 01 areas A+ B

DETERMINE1. Area A2. Area B-1ess20% lor downwash

impactonarea affected3. Longitudinal separation

NP

CG

,...-------l~-- Longitudinal separation

Static margin 25%01 alf-wing MAC

Area A

DETERMINE1. Area A2. AreaB-1ess20% tor

downwash impact onareaaffected

3. Longitudinal separation

Area 01reducedeffectiveness

___ 1-_ in downwashat80%

~~~,*~gj"'I------ Area B

~----....,_---Longitud i na l separation

Area A

25 MAC 01 alf-wingstatic-margin -

CG

Distance N= areaAx separationtotal 01areas A+ B

Figure 8.Locating a canard's NP andCG.

Figure 9.Locating tandem-wing NP andCG.

104 THE BASICSOF RIC MODEL AIRCRAFTDESIGN

Canards, Tandem Wings and Three-Surface Designs ... CHAPTER Z2

area relationship of fore and aftwings . A pu sh er-engin e designwou ld require an aft CG, a smallcanard and a large wing . A front­engine design would reverse thissituation.

If flaps are used, they mus t pro ­vide balanced lift when extended.Too mu ch additiona l lift fromeither fore or aft wings wouldresult in very serious pitch prob­lems- either a dive or a stall.Obviously, both sets of flaps mustbe extended simultaneously forbalance.

With a small canard of 15percent of th e aft wing in area,flaps on the aft wing would bemuch more powerful than thoseon the foreplane. Another disad­vantage of a small canard and rear­ward CG is the reduction inmoment arm to the MAC of thevertica l tail surface(s); it necessi­tat es very large vertical areas. BurtRutan solved this problem by usingaft-Wing sweepback and placingthe vertical surfaces at the wingtips(Figure 11 ). This substantiallyincreases the moment arm . TheCanada Goose design, with amod est 5 degrees of aft -wingsweepback, had the same philoso­phy applied to it.

Sweepback reduces lift. As modelairplane designer John Roncz putit, "You get around 14 percentmore lift per degree of ang le ofattack at zero sweep than at 30degrees of sweep."

The Swan had a straight aft wing,

Figure 12.Three-viewdrawing o( theRutan Quickie.

tion and effective­ness. Figure 8covers NP andCG locations forcanards, Figure 9for tandem-wingdesigns and Figure10 for three-sur­face models. Thenormal static mar­gin for stability is10 percent of themain wing's meanaerodynamicchord (MAC). Useof a 25-percentstatic margin assuggested leaves a15 percent mar ­gin of error. Test­flying the modelwith cautious rear­ward CG move-

ment will confirm your calculations.

• Sizing of fore and aft wings.The total wing area, having beenestablished, must be dividedbetween the two lifting surfaces.

Figure 11.Three-view drawing of theRutan Long-EZ.

CANARDSFrom the discussion of NP and CGlocations, it is apparen t th at th esmaller the foreplane, the fartherback NP and CG will be and viceversa . The area relationshipbetween the two lifting surfacesdetermines NP and CG.

The heaviest component is thepower unit . Its location dictates th e

Area C

Area ot--~-"1Il'II--reduced

ellec''tivenessIn down­wash at80%

_----4-,+-- Longitudinal separation

1-__-ti~~::::::;::::"'71;4 MACs

Distance N= (area A lOP) + (area 0100)+ (area ClOR)total of areas A+ 0 + C

I+--+I pDETERMINE1. Area A2. Area Hess 20% for

downwash shaded

I ----=~~=~!!:!!!!t;tR:- arearoo 3. Area -less 15%(H ail)

4. Separation and tall­moment arm

• Wing loadi ng selection. Thetype of performance desiredgoverns the choice of wing load­ings. Chapter 5 suggests wing load­ings in ounces per square foot ofwing area.

If th e design is to incorporateflaps, then higher wing loadingsare in order. When deployed, theiradditional lift and drag will pro­vide reasonable landing speeds .With weight and wing loadingestablished, the wing 's total surfacearea is easily calculated:

Weight (oz.) x 144Wmgloading (oz.jsq. ft.)

Wing area (sq. in. ) =

• Level-fligh t speed esti mate. Thisis essential in determining theangles of attack of the fore and aftwings.

• Th e neutral po in t and CG loca­tion. The NP concept is discussed inthe Chapter 6, "CG Location." Forthe three types of forward-wingmodels , both CG and NP will fallsomewhere between the two liftingsurfaces. Precisely calculating theirlocations is very complex andbeyond the scope of this article. Infull scale, the calculations are con­firmed by wind-tunnel tests or actu­al flight tests with the CG at variouslocations.

A simplified method is proposed;it considers areas and their separa-

Figure 10.Locating tnree-sutteee design NP andCG.

THE BASICS OFRIC MODEL AIRCRAFT DESIGN 105

CHAPTER 22 .A THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Figure 14.Rutan model 81 Catbird (VSAEROmodel); note three surfaces.

• Longitudinal and vertical sepa­ration. Longitudin al separation

chapter and flies very well. All fourillustrate the added flexibilityoffered by this three-surfaceconfiguration.

• Airfoil selection. As previouslyexplained, thi s is critical for stab leflight. Additional information andformulas can be found in Cha pter1. The horizontal tail airfoil of athree-surface design should be ofsymmetrical section

(stagger) measured from the 25­percent-MAC points ranges from 1to 3.25 times the aft wing's MAC.

Verticalseparation (gap) shou ld be1;2 the aft wing's MAC as discussed.

Tail surfaces of a three-surfacedesign should have a tail-momentarm as ou tlined in Chapter 7. AT-tail design is favored.

• Aspect ra ti o and p lanformselection. In addition to determin­ing the areas of the wings, you mustalso select their aspect ratios andplanforms as previously discussed.

Figure 15.Piaggio P 180 Avanti three-surface twin.

LEVEL FLIGHTIn level flight, at the selected cruis­ing speed, the fore and aft wingsmust support the model's weight.The calculation of the weight distri­bution , leading to loadings for bothwings, is shown in Figure 16. Theforeplane must, however, support

100 x 600130

or 461.5 square inches in area .The designer needs to take the

area relationship into consideration.

TANDEM WINGSThis type has wings with close toequal area. The NP and CG are wellforward . A pusher engine behindthe aft wing would pre sent animpossible CG problem.

Rutan's Quickie (Figure 12) illus­trates a front-engine tandem-wingve rsion, with its vertical tailmounted on an extension of thefuselage.The Wasp is another tandem-wing

version. The pusher engine is justbeh ind the front wing. The aft wingand vertical surfaces were supportedon booms, This model was very sta­ble, but it had no flaps owing to itslow wing loading .

THREE·SURFACE AIRPLANESThe comments on wing sizing for acanard appl y to the fore and mainplanes of the three-surface type .The presence of a horizontal tailcauses both NP and CG to moverearward (compared with acanard). The tail 's elevators providepitch control. Slotted flaps on bothfore and aft plan es permit higherwing loadings with reasonablelanding speeds.

Figure 13 shows John Roncz's"Eagle"- a successful trainer thatpro ved safe and easy to fly. Itsforward wing area is 67 percentof the main wing area , andboth win gs are equipped withslotted flaps.

Rutan's "Catbird" (Figure 14) isanother three­sur face design .Note the sligh tforward sweep ofboth canard andhorizontal tail.The PiaggioP180 "Avanti" isa twin-pusher­engine, three­surface, slotted­flap airplane(Figure 15). Theauthor's "WildGoose" was builtaccording to th edesign approachoutlined in this

and aft

For a total wing area of 600 squareinches, foreplane area would be:

30 x 600130

or 138.5 square inches;wing area would be:

but its vertical surfaces projectedbehind the wing . Twelve ounces ofballast were needed to correctlyposition its CG- as had been antic­ipated after doing th e "BalancingAct" (see Chapter 6) for this model.The minimum canard area is 15percent of that of the aft wing. Fora front-engine aircraft, such as th eill-fated "Pugmobile," a foreplanearea of close to 60 percent was used.

The Canada Goose had 31 per­cent foreplane; the Swan had 37percent. Using a foreplane of 30percent as an example, total wingarea would be 130 percent.

Figure 13.Roncz'sEagle three-surface trainer.

106 THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Canards. Tandem Wings and Three-Surface Designs .... CHAPTER 22

The Wild Goose, a successful three-surface design.

tance for longitudinal stability are:- The foreplane must stall first.-The aft plane must hit zero-liftfirst.

Now that the angles of attack ofboth wings have been calculated, itis time for this test:

Using "Special Procedure" C inChapter 1, determine the stallingangle for each wing and the zero-liftangles from the airfoils' curves at thelanding speed Rns.

Compare the spread from AoA tothe stalling angle, but before estimat­ing the downwash compensation.Raising the foreplane's lift by lower­ing its flaps will bring it to its stallattitude; the increased lift producedby both the foreplane and its flapwill increase the angle of downwash,increasing the aft wing 's stall margin ,but only for that portion of the aftwing in the foreplan e's downwash;that part out of downwash isn 'taffected. If your foreplane's calculat­ed angle of attack is 3 degrees and itstalls at 12 degrees, there's a spread of9 degrees. With an aft wing at 1degrees, stalling at 14 degrees, thespread is 13 degrees so that the fore­plane stalls first.

Similarly compare the spread fromzero-lift angles of attack to yourcalculated angles for both wings .That of the foreplane should be sub­stantially higher than that of the aft

1;4 MAC 1;4 MAC

"L"

CG

"D"------47-"C"-»: -Fortplan. toadinl- Gross weight I C An plan. loading. Gross wllghtI 0

L L

1;4 MAC 1;4 MAC't="L. Thrusllines-T

CG PM2~ High

"D" A-. /- ! \ F1 Level

Fortpl.l •• P"dl'mllh~;;I.'dl.1JS [.~\ \ F2 Low

High t!lnJsl PM1 + PMZ+(ll F1) Low thrust PMl. PM2 • (T I F2) L•.,elthrust~0 0 D

(see Formulas 5and 9 ofChapter 1).

Figure 18 pro­vides simplefor mulas forestab lishingthe effect ofdrag momentson the fore­plane load inounces. Thetotal foreplaneload is com-posed of itsshare of the

model's weight plus the ne t sum ofthe moment source loads, pitchingmoments, thrust moments anddrag moments (in ounces). Boththrust and drag loads may be posi­tive or negative; take care to iden­tify each so that the net value willbe correct.

LIFT COEFFICIENTSHaving determined the wings' areasin square inches and th eir loadingsin ounces, the level-flight designspeed estimated (see Formula 7 inChapter 1) permits calculation ofthe lift coefficients required foreach wing's airfoil. Applying"Special Procedures" A and B willdetermine the angles of attack toprovide those lift coefficients.

Decide whichof the proce­dures will beused to com­pensate for thereduction inAoA caused bythe downwashaffecting the aftwing behind theforeplane.

The foregoing Figure 16.provides condi- Calculation of wIngloadingsdue to weIghtonly.

tions for levelflight at thedesign speed;any variationsfrom that speedwill requirethe same trimadjustments asfor a conven­tional model.

• Stability test.Two points of FIgure 17.critical impor- Additional foreplane loadIng from wing pitching moments andthrust.

an additional load beyond thatresulting from weight alone. Thisresults from :

• The fore and aft wing's pitchingmoments always being nose-downor negative.

• Drag moments of both fore andaft wings .

• Propell er thrust loading.

Explanation and evaluation follows:Pitching moments are explained

in Chapter 1, and Formula 10 ofChapter 1 permits the calculationof these moments in inch-ounces.Symmetrical airfoils have no pitch­ing moment.

If the propeller thrust is above animaginary horizontal line drawnthrough the CG, a nose-down (ornegative) moment results. Belowthat horizontal line, thrust pro­duces a nose-up moment thatreduces the foreplane load. If theCG is on the thrust line, there is nothrust loading. The thrust, inounces, required to propel themodel at the design's level flightspeed is difficult to evaluate; anestimate would be 40 percent ofthe model's gross weight. For aweight of 100 ounces, th rust wouldbe 40 ounces.

Figure 17 provides formulas forcalculating the wing pitch andthrust-related foreplane loads inounces. Fore- and aft-plane dragmoments consist of the total of pro­file and induced drags, in ounces,multiplied by the distance, in inches,the wing's Y4MAC is above or belowthe CG. If it's above the CG, themoment is nose-up, or positive, andbelow it, it is nose-down, or negative

THE BASICS OF RIC MODEL AIRCRAFT DESIGN 107

CHAPTER 22 ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

• Elevator pitch control was verysensitive.

• Landing speed, flaps-up, was morein keeping with th e aft Wing's lowerloading and comparatively slow­an estimated 25mph.

The explanation of this surprisingbehavior was reasoned as follows: aconventional, tail-last, airplanewith its CG well ahead of its wing 'scenter of lift requires a tail-download (up-elevator) for level flight .The CG of the three-surface designis well ahead of the aft Wing'scen ter of lift, and in level flight, the

A. ForeplanB high1;. MAC 1;. MAC

~"L'. Drag . 1

CGP "D" -cp- "CO 1

Plus Minus Q.Drag . 2

IB. ForBplanB low

1;.MAC 1;. MAC"L"I

.1 !1 .Drag . 2

Q"D" ,cp,

"C" I·Drag #1P fCGi IForeplane load= (orao-'1 I PI+ lPraa+' 2 I 0) Minus Plus

0

• Unique be­h avior of thethree-surfaceconfiguration .Flight tests of theWild Goose dis­closed uniqu ebeh avior th atrelat es directlyto th e threeoptions outlinedabove. Option 1had been select­ed for th ismodel. Duringits design, th eairplane's wingloadings werecalculated to be

Figure 18.46 ounces. per Foreplane loading from fore andaft wing-drag moments.square foot forth e foreplaneand 22 ounces. per square foot forth e aft plane in level flight at60mph.

The foreplane's loading consist­ed of 18 ounces. per square footfor its sh are of th e model 's weight,plu s 28 ounces per squa re foot dueto the nose-down load from theairfo ils' pit ching and the air ­plan e's thrust and drag moments.This high forepl an e loading was ofconcern; but slott ed flaps on bothfore and aft wings were calculatedto bring takeoff and landingspeeds to reason able levels.

During test flights, two unusualcharacteristics became very evident:

plane. As the foreplane movestoward zero lift, its downwash angleis reduced, increasing the aft wing'slift in the down washed area andincreasing the spread from zero liftto actual AoA.

Eppler £214 has a zero-lift angle ofminus 4.75 degrees; if set at 3degrees, as above, the spread is plus3 degrees to minus 4.75 degrees, or7.75 degrees. Eppler E197 has a zero­lift angle of minus 2 degrees. Set atplus 1 degrees, the spread is plus 1degree to minus 2 degrees or 3degrees, leaving a healthy margin of4.75 degrees.

THREE·SURFACE AIRPLANEThis type presents more optionsthan either canard or tandem wingconfigurations as regards the liftdistribution between all threesurfaces.1. The canard and main wing pro­vide all the lift needed. The hori­zontal tail pro vides no lift at theselected speed, but its elevatorscontrol pitch and trim.2. Have the canard pro vide most ofits share of the needed lift with th ehorizontal tail providing a com­pensating download.3. Have all three surfaces share thelift . This author's choice would be1/ 11/ abov e---canard and main wingdo ing all the lifting. Calculation ofwing loads would be that forcanards and tandem wingsdescribed previously.

To all flapTopview

Fuselageat sectionA-A

Front view

Bush with11ll-lnch-o.d. brasstube

V4" plyservo -- ~-

mounts n----+--i?cii;i~~18=i:);~~~~

To thef1apevator

~~

1---+ A

Figure 19.Elevator-flap servo Installation.

108 THEBASICS OF RIC MODEL AIRCRAFT DESIGN

Canards, Tandem Wings and Three-Surface Designs A CHAPTER 22

16

'0

'" 12;;,'"< .."'''' 8- ",

"'-....... ",

tiT 4.5 ..._u0 ...si 0..'" -4..,'"a:

10 20 30 40 50 60

Flap detection-degrees (@250,000)

Figure 20.The effect of flaps andleading-edge slots ontheangle ofmaximum lift.

foreplane 's lift provides th e balanc­ing upward lift. Up-elevator down­loads the tail and unloads the fore­plane, reducing its wing loadingsubstantially. The forepl an e'ssurplus lift is then adding to th eup-elevator action , causing th e ele­vator sensitivity.

This results in a very beneficialreduction in landing and takeoffspeeds, both flaps-up and flaps­down. This unique beh avior hasan impact on the three optionslisted abo ve.

Option 1 is cons idered above;option 2 would reduce th e fore­plane's wing loading, its angle ofattack, its lift coefficient and itsdownwash angle. The aft wing'sloading would increase, requiringan increase in its angle of attack.This would bring both wings' air­foils closer to dangerously unstableconditions, but it could reduce ele­vator sensitivity.

Option 3-having the horizontaltail lift upward-would add to theforeplane 's loading and would resultin even greater elevator sensitivity.

In this author's opinion, option1 is best. Elevator sens it ivity maybe overcome by use of the eleva ­tor's low dual rat e, or by redu cin gthe elevator's area to 20 or 2S per­cent of the horizontal tail 's areainstead of the Wild Goose 's40 percent.

• Longitudi nal control methods.The dominant pitch control forcana rds is a slotted flap on thecanard. Another method is a flapon the forep lane and simultaneousup or down action of ailerons on

the aft wing. The major method fortandem wings is a plain flap of fullor part ial span on the foreplane.The horizontal tailplane's elevatorsare the sole pitch control for three­surface designs .

If option 1 is cho sen and foreand main planes provide the neces­sary lift, th e horizontal tailplane'sAoA should be zero degrees to thedownw ash from the main wing .That downwash angle is based onth e level-flight lift coefficien t gen­erated by th e main wing , which is,itself, in the foreplane's downwash!Chapter 7 provides charts for esti­mat ing downwash.

• Directional control. Chapter 9,"Vertical Tail Design and SpiralStability," provides the basis forobtaining good directional control.For tandem-wing and three-surfacemodels, the moment arm from CGto MAC of the vertical tail surfaces islarge eno ugh to permit reasonablysized surfaces.

Canards, particularly those withsmall foreplanes and pu shereng ines , do not ha ve adequatemoment arm s. Recour se is:-Larger vertical surfaces-Booms or fuselage extensionssupporting smaller surfaces.-Aft wing sweepback and wingtipvertical surfaces.

FLAPSFlaps were previously mentioned,and their limitations were brieflyoutlined. Since both fore and mainwings share th e provision of lift,the additional lift provided on flapextension must not upset the lift dis­tr ibution between the wings. Toomu ch lift from either win g wouldresult in dangerous nose-up ornose-down pitch. Both sets of flapsmu st be lowered simultaneously forth e same reason .

Both of this author's canarddesigns-the Swan and the CanadaGoose-had slotted flaps on bothwings. The foreplane flaps also pro­vided pitch control as "flapevators."On both models, one servo actuatedthe foreplane slotted flap for pitchcontrol, but it was mounted on aslide th at permitted it to move back­ward under control of a secondfixed servo (Figure 19), loweringboth th e fore and aft plane flapssimultaneously- foreplane flaps to

20 degrees deflection and aft-planeflaps to such deflection as balancedthe increased foreplane lift.

Slotted flaps provide their maxi ­mum additional lift at 40 degreesdeflection so that the forep laneflap, still under control of the firstservo, may move up to neutral ordown to the full 40-degree deflec­tion from its 20-degree position forpitch control. Deflecting the fore­plane flap results in a substantialincrease in downwash on the aftwing, reducing its lift and that ofthe aft flaps in the area "shadowed"by the foreplane's downwash.

Any attempt to calculate the aftflap deflection angle to balance thefront flap 's 20-degree deflectionwould have been very complex.Instead, cautious flight tests wereperformed, progres sively increasingaft flap deflection on each flight ,until balance was achieved. Bear inmind th at the foreplane flap couldbe raised or lowered to correct anyminor imbalance, and if the imbal­ance was major, retracting both setsof flaps would restore the model tonormal, flap s-up , flight . Thisworked ; the Swan's aft wing slottedflaps, of partial wingspan, wereextended to 3S degrees in balancingthe foreplane's full-span slottedflaps deployed to 20 degrees .

In flight, lowering the flapscaused the model to "Ievitate"­at much slower speed, but with noup or down pitch-and the fore­plane flap continued its function as

1.2......~

'E 1.0..~

'ii 88 .

0. .6~<;;

~ .4'i;..,< .2

o 10 20 30 40 50 60Flapdellection-degrees

Figure21.Additional flapCL example: .40slotted flapdepressed20 degrees provides t:,. CL of 0.80to 11ft of basic airfoil sect/on.

THE BASICS OF RIC MODEL AIRCRAFTDESIGN 109

CHAPTER 22 ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

elevator under control of the firstservo. Almost full foreflap deflec­tion was needed, in ground effect,to raise the nose for a gentle landing.

Flap deflection reduces thestalling angles of both fore and aftwings and greatly increases theforeplane's angle of zero lift(Figure 20) . For three-surfacedesigns, the same commentsregarding balanced flap lift andsimultaneous extension of bothsets of flaps apply. However, theforeplane flap serves only as aflap; pitch control is effected bythe tailplane's elevators so thatthe foreflap may be deflected 40degrees.

Slotted flaps on a tandem-wingdesign would present the sameproblems as canard flaps. Slottedflaps with chords of up to 40 per-

I I1 1

~educed : : Increasedhit L: : lilt !C<~·• •••• • •• • ••• ••••~•••• ••r.• ' I " I W

' I' .. / '/ - ::.--/ 11 II I

Figure 22.The aymmetric canard downwash due tosideslip.

cent of the wing 's chord may beused on foreplanes, as shown inFigures 20 and 21. Use of suchwide-chord flaps on the aft plane isnot recommended. Chapter 14,"Design for Flaps," provides insightinto flap design, construction andactuation.

• Dihedral. Foreplane downwashimpacting asymmetrically on theaft wing in a side slip creates a pow­erful dihedral effect when the planeyaws (Figure 22). John Roncz 'sthree-surface "Eagle" has no dihe­dral; its wings are "flat." Flight testsconfirmed that dihedral was notrequired. The same would apply tocanards and, to a lesser extent, totandem-wing design

• Landing-gear design. Chapter16, "Landing-Gear Design" coversthis subject. The stalling characteris-

110 TH£ BASICS OF R.OC MOL " A,RCRAFT D£SlGN

tics of the foreplane govern landing­gear design, for all three versions.

• Structural design. The discus ­sion of stressed-skin design inChapter 13 applies to all three typesof front-wing-first airplanes. Use ofthis type of structure would simplifyweight estimating and provideoptimum weight-to-strength ratios .

GLIDER EXPERIMENTAt first glance, the "Plover" appearsto be a tailless glider; in fact it's acanard. The forward-swept innerpanels are the aft plane, and theunswept outer panels are thecanard. The inner and outer panelaerodynamic centers are shown inChapter 26, "Construction Designs,"as are the area 's airfoil sections'neutral point and CG locations.

First test glides, with a verticalsur face of normal size, were a disas­ter and the treacherous behavior ofswept-forward wings was forciblyrevealed .

When yawed, the retreatingpanels' centers of drag and liftmove outboard. The advancingpanel's centers move inboard. Thedrag imbalance greatly exaggeratesthe yaw, and the lift imbalancecauses a violent roll in the oppositedirection. After a couple of damag­ing crashes and some pondering,the vertical surface was enlarged by300 percent of its original area. Themodel then flew well.

The forward panels were readilydamaged on landing. After a sum­mer of repeated flying and repair­ing , it was put to one side. Thebasic concept has merit; it avoidsthe impact of foreplane downwashon the aft plane. A powered ver­sion wou ld be an interesting designchallenge....

The Plover glidercanard.

Chapter 23

AIRFOIL CHARACTERISTICSWith their limited tail-momentarms, tailless airplanes-with theexception of forward-swept ver­sions-can't tolerate airfoils thatproduce high nose-down pitchingmoments; such airfoils includethose that have heavil y camberedmean lines.

See the lift, drag and pitchingmoments for cambered airfoils £197and £214 in the appendix. Such air­foils, when used on a tailless air­plane, call for a substan tiallygreater balancing force. Some early,

LiltBa lancing lorce Stat l c~argin

~ I ~~1~~_ ~~~~~~~~~~_ ~ ~~~ _?-ACfllP ~Wash-in CG

10--- Tail-momentarm

Weight

Figure 1A.Plain taillessforce diagram; Eppler 184 airfoil .

Figure 18.Sweptback taillessforce diagram; Eppler 168airfoil.

Lm

Static p-----' Aerodynamic center-neutral pointmargln - /

C!~ /~ -- --

-~ -

\ Center 01gravity _ Balancing

Tail-momentarmlorce

Weight

tance behindthe CG to pro­vide a longmoment arm,so that a rela­tively small tailarea does thejob.

For a taillessaircraft, thewing itself mustprovide this bal­ancing force.On a straightwing (FigurelA), the mo­ment arm isshort, so a largerbalancing forceis required toproduce themoment need­ed. To increasethe length ofthe momentarm , designersha ve resortedto using widechords , forwardand backwardsweep and deltawings (an ex­treme exampleof sweepback).

I Figure tc.• For pain Swept-forward taillessforce diagram.sweptback anddelta wings, thebalancing force acts downward,reducing the wing's lift and requir­ing additional wing area to com­pensate (Figures lA and lB).

Owing to the high balancing forcesneeded, a tailless airplane is espe­cially sensitive to CG location.

• For a forward-swept wing, the bal­ancing force acts upward, increasingthe wing's lift. This allows less wingarea and higher wing loadings(Figure l C).

CENTER OF GRAVITYLOCATIONFor longitudinal stability, the CG ofany type of airplane must be aheadof its neu tral point (NP). On a con­ventional (with tail) airplane , thehorizontal tail's area and its distancefrom the wing (both horizontallyand vertically) determine the NPlocation. It is possible to have theCG ahead of the wing's aerody­namic center (which lies at 2S per­cent of the wing's MAC) or behind itand still maintain an adequate static(stability) margin between the CGand the NP behind it (see Chapter 7,"Horizontal Tail Design").

On a tailless aircraft, the wing'saerodynamic center (AC) and the NPcoincide. For longitudinal stability,the CG must be ahead of the AC/NPlocation. This results in a nose-downimbalance. For equilibrium, thewing must provide a balancing forceas shown in Figures lA, lB and LC,

For a conventional airplane,th is balance is achieved by thehorizontal tail, which is at some dis-

T he flying wing has intrigueddesigners since the early daysof flight. Its structural sim­

plicity, graceful flight and lowweight and drag potential ha vemajor appeal. Despite this, no full­scale, tailless airplane or flying winghas ever been produced in quantitiesthat could rival those of conven­tional aircraft. This chapter exploresthe pros and cons of tailless design.

Tailless

Design

Airplane

THE BASICS OF RIC MO DEL AIRCRAFT DESIGN 111

CHAPTER 23 .& THE BASICS OF RIC MODEL AIRCRAFT DESIGN

full-scale, tailless designs thatemployed cambered airfoils hadsweepback and inve rted, washed­out airfoil sections toward thewingtips. This provided the balanc­ing force, but certainly did notimp rove th e wing's lift.

To reduce or elimin ate theai rfoil 's nose-d own pitchingmoment, symmetrical airfoils orairfoil s with reflexed mean lin eswere used. In the appendix, E1 84and E230 are two reflexed airfoils;E184 has a low nose-down pitch­ing moment, and E230 has a nose­up moment. An E184 airfoilplac ed inboard with an E230 air­foil placed outboard on a swept­back wing could pro vid e suffici entbalan cin g force. E168 is a symmet­rical airfoil that ha s no pitchingmom ent, exce pt at the stall duringwhich the airfoil becom es nos e­do wn and is stabilizing.

Reflexed and symmetrical air­foil s have substa n tially reducedmax lift coe ffients; E214 has a CLmax of 1.25, whereas E230 has aCL max of only 0.78 . Since bothsta ll and lan ding speeds are direct­ly related to the airfoil 's CL max,these redu ced values result in sub­stantially h igh er landing speeds orthey necessitate an in crease inwing area (lowe r wing loadings) toach ieve those lower speeds.

112 THE BASICS OF RIC MODEL AIRCRAFT DESIGN

HIGH·LIFT DEVICESThe lift that a wing generates is equalto the square of its flying speed.Assuming a constant AoA, doublingthe speed increases lift fourfold.

At high speed, it 's obvious thatless wing area is required (seeChapter 5, "Wing Design"). At highspeeds, less wing area meansreduced drag-both profile andinduced-but substantially higherstall and landing speeds . The GeeBee racers of the '30s reflected thisphilosophy, and they landed "hot."

To provide slower landing speedswith reduced wing area, the mod­ern approach is to use high-lift(HL) devices (such as split, slotted,or Fowler flaps) on the wing's trail­ing edge (combined, in some cases,with leading- edge slots and flaps).Use of these devices results in verylarge increases in the wing's CL max.

Under the conditions describedabove, the wing's area is determinedby its HL-device-assisted CL max andthe landing speed desired. Unfortu­nately, when deployed, these high­lift devices produce heavy nose­down pitching moments that arebeyond the capability of tailless air­craft (with the exception of forward­swept types). To overcome this, smallsplit flaps, which produce more dragthan lift, are sometimes used.

On conventional "tailed" air­planes, the increased nose-downpitching moment is compensatedfor by the hea vy downwash angl eincrea se provided by the deployedHL devices striking the tail ,and by stabilizer/elevator action.Obviously, on a tailless airplane,the Wing's downwash provides nosuch compensating force.

For tailless airplanes (exceptswept-forward configurations) allthree factors-CG location, reducedairfoil CL max and limited use ofHL devices-require an increase inwing area compared with con ven­tional aircraft , and this reduces th etailless craft 's efficiency.

This author's Swift has 600 squareinches of wing area and weighs 92ounces (gross) for a wing loading of22 ounces per square foot. Its airfoilis the E197, and it is equipped withslotted flaps who se chord is 30 per­cent of wing chord, and which occu­py 60 percent of the wing's trailingedge. The CL max (flaps extended 40degrees) is 1.80; stall speed is 17mph.

For an aircraft with a wing CLmax of 0.90 to achi eve the Swift'sstall speed would requ ire a wingloading of 11 ounces per squarefoot . Because of the lower load ing,a substantial increase in wing areaand weight would result. It is notimprobable that this increasewould equal the weight savingsthat would result from usin g ashorter fuselage and absence of ahorizontal tail. Using the Swift'sgross weight of 92 ounces, toachieve the 17mph stall, th e wingarea for a tailless model would be1,200 squa re inches-a 100-percentincrease. Top-speed performancewould be adversely affected.

SWEPT·FORWARDTAILLESS AIRCRAFTOf the taille ss configurations , onl ythe swept-forward (SF) ha s anupward lifting balancing force,which adds to the Wing's overalllift, rather than th e downward, lift­reducing balancing force of theother configurations.

Very few SF taill ess aircraft­either full-scal e or model-havebeen design ed and built, owing totwo major factors :

• The SFwing has a strong tendencyto twist under load, increasing itsAoA. Unless the wing is torsionallyvery strong, thi s tendency leads toflutter and disastrous failure. A stiff,heavy structure is need ed. Modern,composite, stressed-skin design haslargely overcome this problem.

• An SFwing is directionally unsta­ble and requires large verticalsurfaces for directional stability.

Since lift is all upward, the nose­down pitching moment of cam­bered airfoils is easily overcomewith an SF wing. Such airfoils, withtheir higher CL max , may be used.

High-lift devices, such as slott edflaps, may be incorporated at theinboard trailing edges. Elevators aredepressed at the wingtips to increaselift forward of the CG and offsetboth the added lift and the nose­down pitch of the extended HLdevices that are behind th e CG. Inthis condition , both elevators andflaps add to the wing's total lift.

An SF wing characteristicallystalls at the wing root first . Because

Tailless Airpane Design A CHAPTER 23

Figure 2.The 1922Arnoux "Simplex" racing mono­plane designed byCarmier.

this area is aft of the CG, such a stallcauses the airplane to nose-up. Topermit the SF wing to stall ahead ofth e CG first (causing nose-down ),an increase in the wing 's angle ofattack toward the tip (wash-in) isdesirable. This adds to the wing'stwisting tendency and reinforcesthe need for torsional strength.

It does not require much imagi­nation to see a parallel between thisSFwing and a canard configuration:

• In both, lift is upward.

• The can ard foreplane and the SFwing 's outboard areas mus t bothstall first .

• The aft wing of a canard and theinb oard portions of a SF wing mustarrive at their angles of zero liftbefore that of the foreplane or out­board panel.

Canard design technology is thusapplicable to SFtailless design , withone major difference: the innerportions of the SF wing are no taffected by dow nwash fromthe outer portions. In canarddesign , downwash from the fore­plane significantly affects the aft­plane and is a design consideration.

~A~

~~-_.-

Figure 3.The 1935 Faurel A.V. 10tailless lightairplane.

PLAIN TAILLESS AIRCRAFTFigure 2 is a three-view drawing ofthe Arnoux "Simplex"-a 1922 rac­ing monoplane, which was pow­ered by a 320hp Hispano-Suizaengine . Its top speed was 236mphan d its landing speed a brisk84mph. It crashed during a testflight before the Coupe Deutsch.

Flight controls were elevons andrudd er, and the airfoil was a sym­metrical Goettingen 411. The veryshort tail-moment arm from the CGto the elevons must have made lon­gitudinal control and CG locationvery sensitive; stops restricted thedow nwa rd movement of theelevons. Roll and yaw control wassatisfactory, and the structure was

~ =>

~. .

o i

Figure 4.Hoffman disk-type alrpane.

good . To obtain the correct CG, atractor engine and propeller werethe only choices . The major disad­vantage, longitudinally, of the plainwing is the short tail-moment arm .

Obviously, lower aspect ratioswith the resulting longer chordswo uld be an improvement.Coupling low AR with heavy taperresults in even longer centralmoment arms.

Figure 3 illustrates the concept­the Fauvel A.V. 10 of 1935. Poweredby a 75hp Pobjoy engine, it had asharply tapered wing with an AR of5.4. Its airfoil was heavily reflexed,without washout, and uniformacross the span . Inboard trailing­edge elevators provided pitch con­trol; outboard ailerons provided rollcontrol; and a rudder controlled yaw.

The AV 10 performed well andwas granted a French certificate ofairworthiness, but no further devel­opments occurred. Structurally, the

wide, thick win g was light. A trac­tor engine and prop were th e onlychoices.

The low-AR, wide-cho rd con figu­ration was develop ed in to theHoffman disk-typ e airplane shownin Figure 4. The airfoil was a stable ,reflexed M-section; th e ailero nswere the wingtip, floati ng variety;the elevators were inset at th e semi­circular trailing edge, and a largevertica l surface was provided. An85hp tractor engine and prop wereused. It flew well, but no furtherdevelopments took place.

Low-AR wings do not stall untilthey reach high angles of attack;and th e danger of spins is remote.Slow, safe, landings at high anglesof attack are possible. TheHoffman 's lon g main landing gearreflects thi s capability.

In RIC model terms, th e taillessplain wing con cept is alive and wellin Bill Evans' "Scimitar" series.

SWEPTBACK AIRCRAFTSweepback (SB) favors higher aspectratios. For a given angle of SB (mea­sured on th e Y4 cho rd line) higherARs result in longer tail momentarms for better lon gitudinal con­trol. Higher SB angles have thesame effect but result in lower lift.

High ARs demand greaterstrength and higher weight. Also,sweepback induces twist underflight loads, and that tends toreduce th e Wingtip'Sangle of att ack.Good , torsional stiffness is requiredto remedy this.

During th e '30s, the Germa nHorten brothers developed a seriesof flying wings as shown in Figures

Figure 5.The Harten brothers' lirst "Ilying wing"sailplane01 1933.

THE BASICSOF RIC MODEL AIRCRAFT DESIGN 113

CHAPTER 23 ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

~

~ ..~~~

Figure 11.Taillessairplane of F. Hadley Page andG. V.Lachmann.

Figure 10.Wenk-Peschkes "Weltensegler" sailplane(1921 type).

Figure 9.The Wenk-Peshkes "Weltensegler"sailplane at the 1921 Rhiin Competition.

plane. This design illustrates thecombined plain and sweptbackwing plan form, with a rectangular,dihedralled center section andanhedralled, sweptback, outer pan­els. The outer panels are set at lowerangles of attack to provide thedownload to balance the forwardCG. Controls were on the trailingedge of the outer panels.

Th ese outer panels, like aninverted V-tail, provided both hor­izontal and vertical surfaces. Theelevons acted, in concert, as eleva­tors: but differentially as ailerons.The downswept controls alsoacted as rudders into the elevon­induced turn, thus overcomingany adverse yaw.

As Figures 9 and 10 illustrate, thewing was externally braced, it hadan AR of 11, and it weighed a low93 pounds for a span of 53 feet andan area of 195 square feet. It flewsuccessfully, but later broke up inflight , causing the pilot's death.

Figure 11 portrays a British pro-

Figure 8.The Davis Wing.

chord line. Controls consist of split­drag rudders outboard and elevonsinboard. Wisely, the narrow tips areequipped with fixed leading-edgeslots to delay wingtip stalling.Obviously, the pusher engine andprop are best. No dihedral is neededon sweptback wings.

Richard Engel's "Winglet" (ModelAirplane News, March 1994),powered by a pusher .40 and with awing area of 900 square inches,is a good example of a flying ­wing design.

~~~

==)::1==.~:=.:::t:I : = =-L ...L -L

A more recent flying-wing design,the Davis Wing, is shown in Figure8.It incorporates the design features ofthe ill-fated Northrop flying-wingbombers of the '40s. It also bears aclose resemblance to the Hortendesigns.

The engine is a 65hp, water­cooled Rotax 532, in a well-stream­lined pusher installation.

This wing had an AR of 6.67, a sur­prisingly large wing area of 240square feet and a gross weight of 975pounds for a wing loading of 4.06pounds per square foot (low for apowered full-scale light airplane). ACessna 172 weighs 2,300 pounds,has 174 square feet of wing area andwing loading of 13.2 pounds persquare foot.

The Davis's top speed was a brisk150mph---excellent, on 65hpi stallspeed was a modest 42mph, thanksto its low wing loading. Its emptyweight was 565 pounds, so it carried73 percent of its weight as usefulload.

The wing is sharply tapered andswept back 28 degrees on the 1;4

COMBINED PLAIN ANDSWEPTBACK AIRCRAFTFigures 9 and 10 show the 1921Wenk-Peschkes IWeItensegler" sail-

Figure 7.The Buxton gliderof 1938.

5 and 6.The Horten flying wings had:

Figure 6.A60hp pusher propona Horten glider.

• Elevators inboa rd and aileronsoutboard on th e trailing edges.

• Thick , sharply tapered planformsof symmetrical airfoil sections.

• A cabin arrangement th at, inlater models, requ ired th at the pilotlie in a prone position, completelyenclosed in the wing.

One version had an enclosed 60hpengine driving a pusher prop on anextension shaft (Figure 6). For RICmodels, an electric motor enclosedin the wing, with an extension shaft,driving a pusher prop at the wing'strailing edge would be practical.

Figure 7 illustrates th e Buxtonglider of 1938. This interestingdesign had a thin, high-AR wing,symmetrical airfoils washed out tothe wingtips, and vertical fins andrudders at the wingtips. Outboardelevons provided pitch and roll con­trol. The pilot was hou sed in a podbelow the Wing. Small split flapswere used at the wing roots.

• Washout toward th e wingtips.

• Dihedral on the lower wingsurface.

• Yaw control was provided by airbrakes placed outboard on both thetop and bottom surfaces, flush withthose surfaces when not being used.No vert ical surfaces were used.

114 THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Tailless Airpane Design ... CHAPTER 23

shows a swept-forward, tailless,free-flight model. Note the heavilycambered airfoil sections and thelarge vertical surface.

AILERONS AND ELEVONSAdverse yaw is an important con­sideration when dealing with high­aspect-ratio (AR) wings of plain,swept-back or swept-forward con­figurations-particularly for ail­erons or elevons located near or atthe wingtips. On this au thor'sdesigns , the modified frise aileron(see Figure lA in Chapter 10, "RollControl Design") with heavy differ­ential has been proven to provideroll control without adverse yaw.However, if they're used as elevonsfor elevator control, they shouldhave equal up and down action. Atwo-servo arrangement, where theelevator servo moves the aileronservo back and forth , will providethe elevons with equal up anddown action as elevators, and withdifferential action as ailerons.

On plain or delta wings of low AR,the need for anti-yaw differential isgreatly reduced. On swept-forwardwings (without high-lift devices),modified frise ailerons located at thewingtips and with anti-yaw differen­tial are suggested. Elevators are thenlocated at the inboard trailing edgeswhere their moment arm from th eCG is the greatest .

For swept-forward wings with

Aileron and elevator chordIncreased 100% oc:::a=-

A-A

T20

Notch- l ,P1L..!::~ili~

m~leading- -~

ldg droop

Increased vertiul talllf'!11nct'll$Id moment arm

~ Incn.~d rudder ...~

Cross-section {t: C III~~3a~lp tins

Figure 15.Delta RPV configuration modifications.

Figure 14.Delta RPV; three-viewsketch ofbase-lineconfiguration.

{cQ. <1!

tractor power unit isrequired; a pusher instal­lation would presentserious problems in cor­rectly positioning theCG.

SWEPT·FORWARDWINGSFew swept-forward tail ­less airplanes havebeen developed. Figure16 shows one suchdesign-the Landwerlin­Berreur racing mono­plane of 1922. This"Buzzard"-type aircraftfeatured separate eleva ­tors and ailerons and alow -aspect-ratio tailfin . It was powered by a700hp engine.

Figure 17 (from anAeromodeler annual)

Figure 14 illustrates the originalconfiguration of a Delta RPV(remotely piloted vehicle), whichunderwent wind-tunnel and flighttests at the Langley Research Centerin Virginia.

Figure IS shows the modificationsresulting from wind-tunnel tests,confirmed by subsequent flight tests.Note the NASA leading-edge droop(Model Airplane News, June 1990­NASA Safewing) and RAO slots onthe outboard wing panels to improvestall resistance. An RIC model basedon the modified design would be aninteresting project. The low AR, widechord, and thick airfoil result in alight, strong structure. Obviously, a

Figure 12.The Tscheranowsky-Gruhon "Parabola. ..

Figure 13.The 1930 Abrial A-Viii light delta-wingairplane.

ject: the Handley Page-Lachma nntwin-pusher-engine tailless. Thiscraft had the combined plain andswept planform, but with large ver­tical surfaces at the wingtips. Thiscompensated for the fuselage andcountered an "engine-out" situa­tion.

The tab on the floating airfoil infront of the main plane is coupledwith the landing flaps to counteractthe nose heaviness caused by thedeflected landing flaps. The adventof WW 11 probably stopped furtherdevelopment of this in terestingdesign.

DELTA WINGSThe delta plan form has the advan­tage of flying to very high ang les ofattack before stalling. High-liftdevices are neither practical norneeded on this type of wing .

Over the years, many delta-wingdesigns have evolved. Figures 12 and13 illustrate two such planes. Figure12 is of the Tscheranowsky-Gruhon"Parabola," which was built by theZ.A.H.I. in 1931. Its wing section hada thickness of 7.7 percent. Figure 13shows a design that might raise prob­lems with lateral stability-the 1930Abrial A-Viii light airplane. It waspowered by a 9Shp engine; it had a22.4-foot span and 173 square feet ofwing area; and it weighed 1,320pounds. Note the reflexed airfoil.

THE BASICS OFRIC MODELAIRCRAFT DESIGN 115

CHAPTER 23 ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

"':=&.25Chon!

Open

C_~L:~/O'\) 30'

Closed

Figure20.Split -dragrudder design.

cussed. On swept-forward wings,becau se of th e directional ins tabil­ity of this planform, large centralvertical surfaces are man dato ry.

This author's Plover glider (seeCha pter 26, "Constru ctionDesign s") had a vertica l ta il­moment arm of twice the wing'sMAC and an area 10 percent of theWi ng's . A large vertical surfacecould result in spiral instability

SPLIT·DRAGRUDDERS AND SPOILERSNorthrop and Davis flying wings

employed split-dragrudders at the wingtipsas in Figure 20. Openedon one wing panel , theadded drag acted likerudders . Engel's"Winglet" also has split­drag rudders .

Spoilers may be usedfor both glide controland direc tional control,but they may alsoreplace ailerons for rollcontrol when usedon the Wing's upper

\ TYPlcal Wlnglel Section4'

Upper surface1--. _..(..

cal tail area isdescribed in Chap-ter9, "Vertica l TailDesign and SpiralStability.")

Note that thesidewa ys -p ro jectedareas are proportionalto the angle at whichthese outer panel s areanhe-draled; and theirplan -view area isinversely proportionalto this angle.

On sweptback

0.16CT

WingleIIncidence -I"Upper lip, upper wlnglel ·4'Lowerrool, upper wlnglel -7'Lower wlngle!·11'

J2yplcal WingleI Secllon

T.:I"

Figure 18.Whitcomb Winglet

ta illess wings, the loca­tion that pro vid es thegreatest ve rtica l tail ­mom en t arm is at thewingt ips (con t ro l sur ­faces with greatermoment arm s need lessarea for equal effective­ness) . If symmetrical air­foil sections are used inthe dual-wingtip vertical Figure 19.surfaces, "toeing -in " Grantz Winglet.their chord lines by 2 or3 deg rees is suggested .

Two forms of winglet s-theWhitcomb and th e Grantz- maybe used as win gtip vertical sur faces(see Figures 18 and 19). Thedimensions of bot h are related tothe wingtip chord and will providevertical areas that mayor may notbe adequate. Determine the areasnee ded and, maintaining the sam eproportions, size the winglets tothe de sired area. Rudder are ashould be 30 percent of the area ofan y of the vertical surfaces dis-

VERTICAL SURFACESFor plain, delta and swept-forwardta illess planforms, a single vertica lsurface on the cen terline is opti­mum. Placing the rudder-hingeline at or behind the wing trailingedge provides a hea lthy mom en tarm. Positioning 1;4 to 113 of thevertical tail area below the wingwill improve its effectiveness atwing-high angles of attack whe rethe above-wing portion may beblan keted by the Wing's turbu­lence. The anhe-draled and swept­back outer pane ls of the combinedplain and sweptback tailless con­figura tion present side areas thatact as vertical surfaces. (The verti-

Figure 17.M-tailless (with negativesweepbackjbyK. Ginalski o( Poland.

Figure 16.1922 Landwerlin-Berreur.

inboard, high-lift devices, slotte delevators/elevons (similar to theslotted flap shown in Cha pter 14,"Design for Flaps") are suggested.These provide additiona l lift to bal­ance that of th e high-lift devices.

It's suggested that elevators thatare separate from ailerons be usedwhere possib le. The top-hingedvariety (see Figure l C in Chapter10) with equal up/d own actio n issugges ted.

116 THE BASICS OFRIC MODEL AIRCRAFT DESIGN

Ta illess Airpane Design ... CHAPTER Z3

o .<, .4 6 8 10 I ~ 1.4 1.6u tr coefficie nt. C,C..... - M .,.SJ

Figure 23.AC andMACof mult/-tapered wings.

MAC 2MACl

0.?6~ l~T.JIIJ=i~+r:~~ .~_.:=ft :e> ,,+,r

~ 0.?4tL s I s·t+- . I I . I

~022 ' ,tJ8I

I -;"~!I I I j " I I I I¥: I . i! ,

r.:.020I ,

I I e«; I I I I

I I I

lO/D,

~ .OI4 I I ,~.?

I

0 ,ii.olo,j, I

~OO8I ,

a. 006 I

I' I ,I

u .004I I~

.GO.?, ,

I , I I.. - .-./ I ~ I,

~ -z, Ie Airfoi l: 2R, -/5 -8 50

1: - .3R.N. (~/fecfi vt!J 8.24a OOOOar~: 10 -3 -34 Tt!$t: V.O.T.1/80

FormulasDistance Y(ACIDeation) = (AreaAI Xl ) + (Area 8 I X2)

(AreaA+ Area 8)

Wing MAC . (Area AI MAC 11+ (Area 8 I MAC 2)(Area A+ Area 8)

2 .0 4

o

.?Z .4 4

2.0 .40

s 8.. 4I

s ,~ 0

,

• I

8 I, I I ,

Figure 23.TaperedNACA 2RI-15-8.50airfoil.

LEADING·EDGE FIXED SLOTSDesp ite washout, swept-back,highly tapered win gs are prone totip-stall ing at high angles ofattack. This resul ts in loss of lon­gitudin al control. Fixed LE slots,as shown in Figure 22, delay thestall about 9 degrees and increasethe max CL substan tially, but ha vevery low drag. Both Northrop andDavis used th em at th e wingtips,extendin g for 2S percent of thewing's semi-span.

The basic dimension s for the slotshown in Figure 22 may be appliedto any airfoil sectio n.

Figure 21.Spoil-flapdesign.

~ Chord-----..

1_ H.4lI ChO~fi~

Closed OpenPlyol

SPOIL FLAPSSpoil flaps are shown in Figure 21.They were used on th is author's"Dove"-a powered glider. The spoilflaps were used for glide control andproved to be successful. Their com­bined areas were 7 percent of theDove's wing area. Extended, theydidn't change the Dove's in-flightattitude, but the y did cause a greatersink rate. They were used for slow,steep descents from height and forshort, no-float landings. Used sepa­rately, they could act as drag rudders.

surface only.Placing the spoiler's LE beyond

70 percent of the wing chordavoids the lag between controlaction and response, which is char­acteristic of spoilers located fartherforward on the wing chord.

Spoilers create desirable into-the­turn yaw, because only the spoileron the inside of the turn is raised; itsmate remains flush with the wing.

The Hortens used spoilers onboth upper and lower wingtip sur­faces for direction al control. Whennot in use, both split-d rag rudd ersand spoilers lie flush with the wingsurface and cause no drag .

i~

.2 0 .2 .4 .6 .8 10 12 1.4u « co effic ie nt, CI.

C. . ... - M ....$1

6 U ' ..1_ I , , 1 ,f-~ ,

4 :.....684r..,: Oe<4 f ' I

,'Tf' . , , +1

2 , I ,( I . 1 -r-10

. I ! II : : ! ' I ' ~\~

I , I I . 'i l l ' ~

8+ Hi - ~-+-!-t=tE.\1::6

I , . . .\--f-I ,

4, , , :

I ._ --++ -+-,I I ~

I I 'I ,

I , ,I , ,

I , ,I I I I,

I , I ..!

, ,I Co I

II I

I ,

H=Alrfo,f: 2R, -15- 0R.N (e'fUfio,e! 8.090.000Dol/!: ' -2 7-34 tesr -VO.T.1177

-

oC.

I

A,rfo il:2R, -15 ' O Vt![ ( f l /s ec .j : 70..3

~Ze~~5fn~o;;:;,t;o.~· I.';;~:J.(!to.°Ooo - .2W~re f ("$fed: L.M.AL T("sf : V.DT. 1/77Correc ted for tVI'Vl t! /- w olf effeet. - .

LI D I

i\., ,I , I I

I ,I

16..,2 8 4 0 4 8 12 /6 20 24 28 32M 9 /(" o ( o tr oc» , a (rif!treesJ. ref~rr~ to r oof chord

' • • z .o,

. -8! I I

8 j6

~1 -~~.. -- i

O.SltJ"~,.;nq I• .. P'~ I ,-Q! "7" 32

: I i '/1.28- " II , !

2I I l"~ f j'--... I .2_i ! /Y I A ,....I I I ;/,1 I I I

0I I ! , ; , I I I,

8I I I ! ,.I

~';f, I, , I Ie /2

I I II c.

08- It /i-j;../ , i

.CWb't'

~-'i , I I I ; !:' 048 1r 16ar~

Anq'. 0'a"~, a-<lIUWt.-t:Kial " CWt; T __ wit .. '-' . .... oIoiC

Figure 22.Fixedleading-edge slotat Rn600,000.

Figure 24.Tapered NACA 2RI-15-D airfoil.

THE BASICS OF RIC MODELAIRCRAFT DESiGN 117

CHAPTER 23 A THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Figure 26. TaperedNACA 00-15-3.45 (4 to 1) airfoil

STATIC MARGINAs previously discussed, the AC andNP of tailless airplanes coincide . Forstability, the CG must be ahead ofthe AC/NP. This produces a "forcecouple"-lift upward and CG down­ward-that must be balanced by arear download.

The larger the static margin (thedistance between the CG andAC/NP), the greater the aft down ­load necessary. Centrifugal force cre­ated during maneuvers requires anincrease in all three : lift, weight atthe CG and balancing force.

Large static margins, however, aremore stable longitudinally; smallmargins promote maneuverability,but reduce stability. A safety margin(SM) of 5 to 10 percent of the wing'sMAC is suggested.

The swept-forward wing obtainsequilibrium by increased lift createdtoward its tips. This permits the useof cambered, high-Cj-rnax airfoils,healthy stability margins and high­lift devices.

WEIGHT DISTRIBUTIONThis is important, longitudinally, fortailless airplanes, because of theirlimited longitudinal control whencompared with "tailed" airplanes(Chapter II , "Weight Distribution inDesign"). Massing the fixed weightsof power and control units as closeto the CG as possible is recommend­ed for tailless designs. Positioningthe fuel tank on the model 's CG willavoid a possibly destabilizing shift ofthe CG as fuel is consumed and thetank becomes lighter.

LOCATING THE AC AND MACIn Chapter 1, "Airfoil Selection,"graphic methods for locating theAC and MAC of straight, taperedand sweptback wings are explained.

For multi-tapered wings-such asthe one shown in Figure 23--obtainthe Y4 MACs of each panel (A and B)using the methods shown in theaforementioned article. Calculatethe area of each panel (in squareinches) and , using the simple formu­las that accompany Figure 23, obtainthe wing's AC and its MAC. A

I, I I •

I I ,, ,

C" ,

1 I,I

I , , ,I I , I

I , , ,

Airfoil : 00 -/5 -3.45R,N. (e ffec t;.,.) 8, / 3O,(}(XJDoTe: 9~28·3" Te3f : V OT- IIlS

o .Z . ., .6 .8 1.0 U l 4Lilt roe" ;".· '" t'i

C... _.. .. .~,~

fo (J -./

-:e- .2

provide root and tip airfoil ordi­nates and aerodynamic center loca­tion. "S" is wing area and "b" isspan. Although tested at high Rns,these wings are a useful guide forswept-back designs.

DIHEDRALSweptback and delta wings need nodihedral. The plain and swept­forward types should have the dihe­dral angles that are suggested inChapter 9. Combined plain andsweptback wings need a healthyamount of dihedral in the plain sec­tion to compensate for theanhedraled tips.

, I

c.

L/.

~~ ~ ~ 0 4 8 ~ ~ . N M RM g' . of af1 aclf , ct (d~(r~e3J r e t err ea fa r oof chOrd

28

2. I

320

go /s{) /2 , ,J1

Figure 27. Tapered NACA 00-15-3.45 airfoil

WASHOUT AND SWEEPBACKFigures 24, 25, 26 and 27 reflectwind-tunnel tests performed byNACA on four different wings. Allwere stable at the stall (pitchingmoment becomes nega tive). Thewing shown in Figure 24 has areflexed airfoil and 8.5 degrees ofwashout. The wing in Figure 25 alsohas a reflexed airfoil but no washout.The wings shown in Figures 26 and27 have 3.45 degrees of washout.

In Figures 24, 25 and 27, thetaper ratios are 2 to 1 from root totip. In Figure 26, the wing's 4-to-ltaper invited early tip-stall, alongwith reduced CL max. These figures

118 THE BASICSOF RIC MODEL AIRCRAFT DESIGN

Chapter 24

Afterbody

Two hull or float des igns wereselected for th is chapter. The

• Optimum CG location s, relative tothe step.

• Scale-wing lift forces were includedin the tests.

• Spray, porpoising and skippingtests were conducted duringsimulated takeoffs and landings.

• Water resistance, with a range ofloads.

• Trim angles, "free to trim " underhydrodynamic forces in the displace­ment range, i.e., up to the hump andat various controlled trim angles atplan ing speeds in excess of humpspeed.

HULL DEVELOPMENTThe hull or floats described herewere developed by NACA scientistsand tested in 2,OOO-foot-long towingbasins. Recorded were:

oI=- Spraystrips

CG~ - 100Forebody

"=~k:ri"'=4::::::=Ea-- Sternpostangle-""-'-----~

Keel

Chine

Bow

Top contour

Tumblehome --- - - - - "

'">".,"'". J~Q n

Water rudder

e~triP~_,,<_ - _~ nY1- Maximumbeam

Section A-A Section B-B $echonC-C

PLANING ACTIONAND THE STEPFigure 2 illustrates the step's func­tion. Planing at speed, the forebodycreates a trough in which the after­body planes . With adequate stepdepth, the hull or float rides on twoareas, and porpois ing, or skipping, isminimized.

-There must be adequate buoyancywith substantial reserve while afloat.- Planing surfaces should have awetted area that's large enough topermit the model to accelerate toflying speed quickly.-The hull's (or float's) trim angle atthe hump should not cause thewing's airfoil to exceed its stallingangle of attack.-Spray should be well-controlled;in particular, it should be preventedfrom hitting the propeller.-There should be no porpoising ontakeoff, and no skipping on landing.-The model should weathercock toface into the wind when at rest, orwhen taxiing on water at low speeds.

• Float and hull basics. Figure 1shows views of a float , or hull,with three cross-sections. Note thefollowing key points:-The "step" separates the fore­body from the afterbody.-The "keel flat" is the referenceline for the "trim angle" shown inFigure 2.-The "sternpost angle" governsthe hull's (or float 's) trim angle atthe "hump."-The "beam " is a criticaldimension.-The "step depth" is also a criticaldimension.-The "angle of deadrise" bears onthe hull's planing performance.-The "deck" is only a referenceline. The top contour is the design­er's choice.

Hull and Float

Design

F ew events give greater satis­faction than the successfulfirst flight of a model air­

plane that one has conceived,designed and built. Ensuring thesuccess of that first flight and ofsubsequent flights is what th isseries is all about.

Flying off water adds two newelem ents: hydrostatics (buoyancy)and hydrodynamics (planing lift).

Flying boat or floatplane flyingis, if anything, mo re fun than fly­ing off land. There are few treesover water to reach up and grabyour model, and water is more for­giving than terra firma.

• Float a n d hull factors. Forsuccessful water flying, the follow­ing conditions must be met:

Figure 1.Hullor ffoatbasics.

THE BASICSOFRIC MODEL AIRCRAFTDESIGN 119

CHAPTER Z4 ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

degrees has hum p trim of 12.5degrees--well above the Wing'sstalling angle.

A properly designed forebody bot­tom and spray strips will run verycleanly. Spray hitting the wings, tail,or propeller can slow takeoff, not tomention damage the prop. At prop­tip speeds of close to 300mph, wateris pretty "soli d."

---------------Wakearolile----- Sternpost---

--' 8 -l'---..-CG ------- C

A --- <, ----------.10'

-, - - TnmangleI

I

Figure 2.Forces ona hull In two-step ptanlng.

dimensions of both are comparableto those of R/C mode l water planes.

The first design has a short after­body that's suitable for floatplanes.The second, with a long afterbody, issuitable for flying boats. Bothdesigns were tested with stern postangles of 6 , 8 and 10 degrees.

THE "HUMP'"Figures 3 and 4 provide resistanceand trim angles for the short andlong afterbody hull/floats. Both fig­ures merit close scrutiny.

Note the high points in the resis­tance curves--known, for obviousreasons, as the "hump." Not surpris­ingly, the maximum trim anglescoincide with the hump. Beyondhump speed, trim and resistance falloff as the hull accelerates to planelion the step."

Up to the hump, trim is controlledby both hydrostatic and hydro­dynamic forces with little effectiveelevator action. Beyond the hump,trim is progressively elevator­controlled as speed increases toliftoff velocity. Notable is the influ­ence that sternpost angles have ontrim angles at the hump for bothafterbod y lengths. By judicious selec-

tion of the stern post angle, one cancontrol hump trim angles within afairly wide range.

There are two causes of humpresistance:

• The hull is transitioning frombeing a floating object supported byhydrostatic buo yancy to being aplaning object supported by hydro­dynamic forces that act mainly onthe forebody bottom, but with buoy­ancy still having some effect.

• The hu ll/float must rise from fulldisplacement depth, floating, to itsplaning depth aided by wing lift as itaccelerates.

If the wing's AoAis above its stallingangle at hump trim, the wing willstall, and its contribution to raisingthe aircraft will be largely lost .Stalled, the wing will lose roll damp­ing and aileron control, and thewing floats may dig in and causewater looping.

A model wing's stall angle-at lowRn, in ground effect, and with slot­ted flaps extended-may be as low as10 degrees. A short afterbody hull /float with a sternpost angle of 10

BEAM AND CG LOCATIONThe hull/float maximu m width, orbeam, is critical for good water per­formance. Too milch beam addsweight and air drag and makes themodel hydrodynamically ready tolift off before the wing provides ade­quate lift. Skipping and wing stallmay result .

With too little beam, the model sitslow in the water and has higherhump resistance and heavier spray.Takeoff runs are lon ger. Too muchbeam is better than too little.

A study of NACA reports on hulldesign indicated that a hull , plan ingat the wing's stall speed, should gen­erate enough hydrodynamic lift tosupport the model's gross weight.Further, at this speed, the "wetted"length of th e forebody bottomwould roughly equal the beam. Thewetted area would then be the beammultiplied by the beam (beam-),

The stall speed of a model dependson two factors: the wing's CL maxand its wing loading in ounces persquare foot of wing area.

Model airfoils have a broad aver­age CL max of 1.00, so wing loadin gis the major factor govern ing amodel's sta ll speed. It was con­cluded th at a plani ng area (bearn-)relationship to wing loadin g couldbe used for floa t/hull-bea mdetermination.

16'2.4 6

14'

ui 2.0 DISPLAC EMENT5

~'2°CD / w1 1.6 , 4 ~1 D'wUJ

~ 80

~ 1.2 LIFT 3

~ OFF ~ 6'~ 0.8 8° 2 <en " . ~ 4'UJ <,a: 0.4 , ...

STERN POST ANGLE S 20

Omph 14mph 27mph 0'14mph 27mphDmph

Figure 3.Resistance andtrim angles; short afterbody andsternpost angles of6, 8 and10degrees; beam2 loading at 2.5 oz. persquare inch.

120 THE BASICS OF RIC MODELAIRC RAFTDESIGN

Hull and Float Design ... CHAPTER 24

LIFT 3OFF

27mph14mph

2'

0'Omph

14'

16'STERN POST

ANGLES

/7

E//~~"._ """"~ 4° .:»>-

2

4

5

6

27mph

DISPLACEMENT. :HUMP

STERN POSTANGLES

14mphOmph

2.4

ui 2.0CD

T 1.6UJ

~ 1.2s!!! 0.8C/)UJa: 0.4

Figure 4.Resistance andtrim angles; long afterbody andsternpost anglesof 6, 8 and10degrees; beam2 loading at 2.5 oz. per sq. in.

Beam= ~~g!!UQU 0Beam2 loading

0

<.:> @~ 4

9:. 0 0

~ 3 00 0o 0

0 0 0 0

~o 00

00 00

So z 100.1 15 0 % 200 .1 250.1 300 .1

WINGLOADING- WEIGHT INOUNCESPERSQUAREFOOT.

An empirica l solu tion to thebeam problem was developed by ananalysis of the wing load ing s versusbeam- load ing of some 25 modelflying boats and floatplanes, asshown in Figure 5.

The curve in Figure 5 averages thevarious points and may be used todetermine your model's beam as fol­lows:

• Estimate your design 's gross weight(Figure 6 will help).

• Divide gross weight in ounces bythe model's wing area in square feetto provide its wing loading in ouncesper square foot.

• Refer to Figure 5, and select th ebeam- loading th at correspo nds toth e wing loading. For example, awing loading of 20 ounces persquare foot (horizon tal) calls for abea m- loading of 2.6 ounces persquare inch of beam (vertical).

• Divide gross weight by the beam­loading. The result is th e forebody'swetted area in square inc hes.A gross weight of 93.6 ounces,divided by a beam - load ing of 2.6ounces per square inch gives a wet­ted area of 36 square inches.

• The beam is the square root of th ewetted area. For 36 square inches,th e beam would be th e square rootof 36, or 6 inches .

• For a twin-float plane, divide thebeam in half for each float, i.e., 6divided by 2, or 3 inches per beam foreach float. Step depth should bebased on the total beam (6 inches, inth is example) and would be 8.5

percent of 6 inches, or 0.5 inch foreach float.

Figures 1 and 2 show the best CGlocation: along a line at 10 degrees toth e vertical, ahead of th e step/forebody bottom comer.

The wing's optimum location iswith its center of lift (Yo! of MAC) ver­tically in line with th e CG.

PORPOIS NG AND SKIPPINGPorpoising is th e up-and-down oscil­lation of the bow that occurs beyondhump speed. Skipping occurson land ing when the plane touchesdown several times. Landing too fastcontribu tes to skipping, butadequate step depth (8 to 9 percentof the beam) avoids both of theseundesirable cha racteristics.

PLANINGTAIL HULLSDuring th e1940s, in searcho f improvedp erforman ce ,NACA co n ti n­ued its towing­basin tests, bu ton a new hullform.

This hull fea­tured a deeppointed ste pand a CG posi­tioned at orbehind the step.The aim was toha ve the afte r- Oo z

body contributemore to thehull 's hydro-dy na mic lift- Figure 5.h ence, th e Beam chart.

name: "plani ng ta il hull. "This author designed, built and

flew a model with this hull-theFlamingo (see Chapter 26, "Con­struction Designs"). Powered by aTorpedo 0.15cid engine and con­trolled by a Babcock receiver andescapements, it flew well; the hullwas efficient.

Some years later, it was modern­ized with an 0 .5.Max O.35cid engineand a 4-channel radio that providedrudder, elevator aileron and enginecon trol.

One very und esirable trait sur­faced: th e Flamingo always weather­cocked pointing downwind-notgood for takeoffs ! This was becauseof its narrow afterbody, rearwardCG and deep step, all of whichcombined to make the model'sstern sink low in the water.

35 0 .1

THE BASICS OF RIC MODEL AIRCRAFTDESIGN 121

CHAPTER 24 .& THE BASICS OF RIC MODEL AIRCRAFT DESIGN

' 30

140

D. Flared E. "Edo" F. Cathedraldouble flared

5.SP~G.Suggested bollom

Figure 8.Hull andlIoat forebody bottoms andspraycontrol.

No spray control·LH With spray control-RH

lrf1 ~ a~adrlse

A . Flat B. V·bollom C.Dornler

FLOAT OR HULLPROPORTIONSFigure 10 provides proportions ofboth short- and long-afterbody hullsor floats. The short version, if usedfor a flying boat , would require anextension to provide an adequatetail-moment arm (TMA) for longitu­dinal stability. The long version pro­vides such a TMA.

BUOYANCYA cubic inch of water weighs 0.58ounce. A model weighing 100 ounceswould require a displacement of 100divided by 0.58, or 173 cubic inches,plus 100 percent reservebuoyancy, fora total of 346 cubic inches.

The NACA models on which Figure10 was based were designed with 100percent reservesfor a 94-ounce model(at the hull's lowest load). Adequatebuoyancy is not a problem.

For twin floats, a maximum depththa t's equal to the maxim um beamand a length that's 60 to 70 percentof the airplane's length provide ade­qua te buoyancy and reserves.

Hull wake profi leruns betweenbooms-- - ------- -- - --=

o

Hull

I Trimangle

Water line

BOW CONTOURSBow contours for full-scale aircraft

depend on theaircraft's func­ti o n. Fly ingboats forheavy sea dutywould haveboat-likebows;for more mod­erate dut y,bows mayhave a morestreamlinedsha pe. Thetype illustratedin Figure 10has proveni tself formodel hullsand floats, andit 's not diffi­cult to make.

Figure 7.Sea Loon II-planing action of hull andtwinboom afterbodies.

tive hydrodynamically, but it planeswith heavy spray. V-bottoms (type B)absorb landing shock, but reduceeffectiveness and have heavy spray.Types C, D and E are designed toreduce "pounding" on takeoff andlanding. Type F "cathedral" is popu­lar for motorboats; spray is well­controlled without external spraystrips, which are fragile and causehigh air drag.

Type G "suggested" combines theefficiency of the flat bottom with thespray control of the flared and cathe­dral types. Above all, its constructionis both simple and rugged (as shownin Figure 9) and applies to both hullsand floats.

Afterbodies do not require spraystrips; otherwise, construction is thesame as that shown in Figure 9 andbased on the principles in Chapter13, "Stressed Skin Design."

18001 2100z 24001 2700z

0

~ 0

00 0CD 0

00

00

00

00, 300, 600' 900, 12001 15001

GROSS WEIGHT INOUNCES.

FOREBODYFigure 8 provides typical forebodycross-sections of full-scale water air­craft. Type A "flat" is the most effec-

Z 1 10

3 1 00U• JO

g 80

~ 70

u 50~a 50

~ -"0i3~ 30

' 0

Figure 6.Engine displacemenlvs. gross weight.

' 20

Above-water side areas were wellforward; below-water side areas werewell aft. Wind striking the sidecaused the model to weathervane­but poi n ting downwind. Water­and air-rudder control tried hard tocorrec t this condition , but thedownwind win gtip float's waterdrag rendered these controlsineffective.

NACA tested further variationsof this hull and arrived at a config­uration with no afterbody, just aver y de ep po inted ste p. Twoboom s extending back from twinengine nacelles replaced th e after­body and carried horizontal andtwin vertical surfaces at their aftends. This concept is reflected inth e author 's Sea Loon (Figure 7). Itflew well.

But the booms, which also pro­vided lateral stability on the water,did not sink into the forebody'swake as in Figure 2, but rod e on orjust under the undisturbed wateron either side of the forebody, asin Figure 7.

Figures 3 and 4 do not appl y tothis con figuration. Hump trim forth e Sea Loon was established bycarefully selecting the vertical-stepdepth to provide a 9-degree stern­post an gle. The objective was toavoid wing stall at hump trim.Once past the hump, the twinbooms were clear of the water.

122 THE BASICS OF RIC MODELAIRCRAFT DESIGN

Hull and Float Design .A CHAPTER 24

9

-- -

wind float or even capsize the model.These wing floats may be located

anywhere from the wing's tip to itsroot. Mounte d close to the root, thefloats must be larger to provide thegreater buoyancy needed; fartherout, the y may be smaller and lighterand have less drag.

The planing surfaces of thesewing floats must be of adequate areaand set at a great eno ugh angle tothe hull's keel flat to cause the floatto recover quickly while planingwhen disturbing forces cause th emodel to heel, lowering one wingtipfloat to the water surface.

WlNGnp FLOAT DESIGNRefer to Figure 11. When the modelheels to submerge one float, th e CGis displaced a distance "X ." This dis­tance, in inches, multiplied by themode l's weight in ounces, gives theunb alan cin g mom ent in inch­ounces. The corrective force is thebuoyancy of the submerged float inounces, multiplied by the distance

Five equal spaces

------

Lon anerbod 10 59.6 74.7 88.8 97 100 99 96 87 68 37.3

Beam widths In pen:enl 0' maximum beam al sla tlon 4

Station 0 .5 1 2 3 4 5a-b 6 7 B 9

Short anerbody 10 59.6 74.7 88.8 97 100 99 93 75 38

I--- - - Short afterbody65%01beam

Step=8-9%01max beam

::::::04.=l~l=JJ~~J::::ls;;:hodrt~a;fte:rtbo:dy;:sSite;;rn;-;po~st~a~n~g,:es~6;:JO-~103.8% 01 lorebody Stern post depth

8.5%01lorebodylength

FLYING BOAT LATERALSTABILITY AFLOATFlying boats and single-floatseaplanes need wing floats toprevent them from tippingover. These must provide suf­ficient buoyancy to cover asituation in which the modelis slowly taxiing crosswindwith the hull (or single float)on the crest of a wave and thedownwin d float in a nearbytrough. The upwind wingpanel is elevated at a consider­able angle to the wind, tend­ing to submerge the down-

Five equal spaces

Deck to keel 8.9 17.5 21.6 24.8 24.8Deck10 chine 8.9 11 15.3 22.5 24.8

Afterbody sections

Station 0 0.5 1 2 3 10 5

Topview

FOl8body deplhs In pen:enl 0' forebody lenglh

'--- - - - long afterbody -----..l

0.5 1 2 3 _ .4_ Topcontour 5A

F'"::'''CD QJp:'y strIP,mm []' IoP.65 beam

5~ -DJm m DJ1l CO--- ~erbOdY

58 6 7 8

,...-- - Forebody100% - - ---,..--- - ­Deckh~::::;s=::::::::;::::=;::==1====¢::=:;:==~;;;;;;;;;~__I_l

Figure 10.Hullor float proportions.

degrees. With a wing stall at 14.5degrees and hump trim of 10deg rees, the re is a good safetymargin- and wing stall at hu mptrim is avoided.

Beyond the hump, th e elevatorstake control of the model's trim,

and at liftoff speed, moder­ate up-elevator causes th emodel to become airborne .

~" balsa bulkhead

/(I I

(f

Knowing the hull's (or float's)total length and having arrived atthe beam, the dimensions of eitherversion are easily calculated. Notethat hull or float depths are based onthe forebody length, and widths arein percentages of the beam.

For twin -float planes, the calcu­lat ed beam is divided by 2 to provideeach float's beam. Overall floatlength is 60 to 70 percent of theplane's length . The step depth isbased on th e total beam and isapplied to each float.

WING ANGLE OF INODENCEChapter 18, "Propeller Selection andEstimati ng Level Flight Speeds," pro­vides the basis for calculating theangle of incidence necessary to pro­vide adequate lift at the model's esti­mated level cruise speed. For theSeagull Ill, th is was 0.5 degree.

r---;;----- 8eam: 6" - --- --,3/16" balsa Corner radII : 1"sheet .L

Figure 9.Typicalhull or floatconstruction.

WING'S STALUNG ANGLEAND HUMP TRIMChapter 16, "Landing Gear Design,"details the calculations necessary toarrive at the wing's stalling angle (atlanding-speed Rns, in ground effectand with flaps extended) .

The Seagull Ill's net stalling angleduring the takeoff run is 15 degrees.Since the wing is set at 0.5 degree inlevel flight, the stall would occur14.5 degrees later.

The Seagull III's hu ll is the long­afterbody type with a stern post angleof 10 degrees. Hump trim for th ishull is 12 degrees; but because theforebody keel flat is set at plus 2degrees for level flight, this model'shump trim angle is reduced to 10

THEBASICSOF RIC MODEL AIRCRAFTDESIGN 123

CHAPTER 24 .& THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Formula lor wingllp 1I0at "X" (CG til h ) i ht ( )volume (cubic inches) = movemen n nc es x gross weg ounces x 3.5Distance · Y" (Inches) x 0.58

Angle 01 heel-IIoatsubmerged

--I e\ CG movement · X" __ -/-----_ ....-- -- - Seagull III in a flaps-down landing. Note the

well-controlled spray from the forebody bot­tom andthe plane's "at the hump" an/tude.

3 degrees to the hull 's keel flat, asshown. Viewed from the front, thefloat bottom should para llel th ewater surface at con tact for maxi­mum recovery action when plan ing.

Figure 11.Wingtip-float-volume calculation.

Wing Wing

Figure 13.Development of "Thurston" float from basic bfock.

75% of length and widthat botlom

Wingtip 1I0at volume (cL)Block length= --.:.....:..------'---',---,

Beam [ln.] x etleclive depth (In.)

Totaldepth

Front

1I0at volume (cuIn.) x 0.58

Hull beam2 10ading

"Fishtail"

Beam formula =

Top

between the float and hull center­lines. The corrective buoyancy inounces has to be converted to cubicinches and increased for the reservebuoyancy. The formula in Figure 11for float volume does all this andincludes a 2S0-percent reserve.

To design a float that has low dragand the required volume is not diffi­cult. Layout a block that will providethe volume in cubic inches that pro­vides the calculated buoyancy (Figure12). The width is the float beambased on the hull beam- loading; itslength will be roughly four times thatof the beam. Both depth and beamare calculated using the formulas inFigure 12. Draw the 3-views of yourfloat in and around this block asshown. The float bottoms should beflat with sharp chine corners.

The float bottom should be set at

(note lIat botlom)

Beam fOrmula =V 1I0at volume (cu.!n.) x 0.58 Float depth= Float volume (cu.in.)

FloatIJH",~~.m:--2 O;:~~~~:~~:_=_=_=_=_=_~_i====F=lo=:at::b::ea~m>(-<ln:::::::J''ffi:depth ~

L_- rFloatoulline Side Front

1-- 4 x beam _Top

THE THURSTON FLOATThe Seagull III incorporates th eThurs ton float at its wingtips. Theseare light and rugged, easily madeusing sheet balsa and have low drag.Figure 13 provides their design basis.

WATER RUDDERSWater planes should have water rud­ders for directional con trol becausethe air rudder is ineffective when theplane taxies at low speed. TheSeagull III has a water rudde r at thebase of the air rudder. The Ospreyand Seahawk have water ruddersoperated by separate servos twinnedto the receiver's rudder channel. Allhave good water control. .&

Figure 12.Method of developing float tines from basic block of wingtip float volume.

124 THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Chapter 25

Dinedral AnglesWing w/all . no all.High 2· 5·Mid 3· 6·l ow 4· 1·

/ "

50% 01 semi-span

25% "C"

, ...-'.

15% strip ailerons

/~Dihedral

Wing area=span xchord Y'I. M C

- - -+-- 2.5 10 3 xchord

Area1810 22%__~____ 01 wing area

AR 3105Elev. - 35%ollail area

A. t' 40%'

semi-spanailerons

I

~: '~ .i 10· '. /'Tricycle gear, _~...._. __ ._.. ····----·· l·

10· t _

~----------:c--:-lue;;;nOigl'h.h:::::::::::::::::::::;:---''I. MAC 1%01wing area-aileron

1

......- 1.5 10 2 !Chon! -~A 8o/.-Rudder onIY__-l~1-

.2· /f::::::::::::::::::=.._.!!R~ud!.llder 35%VT----I~~\.IJ:.--,---<:- 1"

.............. ··· · · · · · · · · · · · CG,~ · ~ · ··· ·

\.i8~

Semi span

Aspecl rallo 5 I 7

AR = Span IChord

y

Figure 1.Basic airplane proportions

Aircraft

Basic

Proportions for

• Figure 1. Basic proportions foreight models with engine sizes offrom .10 to .60.

RIC Model

M any modelers design theirmod els to reflect theirown ind ividuality. For

many reason s, they do not chooseto follow th e detailed and some­times complex suggestio ns present­ed by authors such as me.

The basic proporti ons presentedhere are for a range of models tohelp modelers exercise th eir urge tooriginate uni que, yet success ful,models. They are easy to follow andrequ ire a minimum of calculation ;and th ey're divided in to six cate­gories represented by:

• Figure 2. Basic twin-floatproportions.

• Figure 3. Basic flyin g boatproportion s.

• Figure 4. Basic glider proportions.

• Figure S. Proportion s for aerobat­ic models powered by .40 to .50engines.

• Figure 6. Airfoil layou t procedureand ord inates for six airfoils. Seeappendix for performanc e curves.

THE BASICS OFRIC MODEL AIRCRAFT DESIGN 125

CHAPTER 25 ... THE BASIC S OF RIC MODEL AIRCRAFT DESIGN

See Fig 2 lor hullBeam bonom design and

.25 I chord stern post depth

All • 40%semi-span

.25 chord

58% 01 length .-Iaherbody

Size to suit engine/1ank

.J.Jr-1=F: :;~:;:' Area 15% 01wing--.,rudder35% V.T.- -/-""Oo,l '''1

90'

Aspect ratio6

--- A-rea 20% 01 wing'......:-'""'U""--;:::!~....----.---, -:.-

Elevators 40% H.T.

75%01II mlspan

Chord

42% 01 lengthlorebody

Eng. disp. Maxtlnat Step(cid) beam (in.) depth (in.)

0 10 2375 7/16

0.15 2.5 151.32

0.25 2.5 151.32

035 2.825 1!.1

04 0 3.00 9/16

045-6 3.25 191.32

0.50 3.375 %0.60-1 3.5 11/ 16

Figure 2.Basic twin floatproportions.

~------- Length 5 x Chord - - - - - - .-1

Figure 3.Basic flyingboatproportions.

- --_·__··_-w ._ . ._J>.__

Depth8.5%ot lorebody

length

C

i_. ...L._.L j __ l....-.l .

i..- 50% 01 semi-span --..:

CG

m ,,L Step

Spray StrlpL-+::-~~::r-~;,,;:-,=:.e::!c~

Forebod; V

126 THE BASICS OF RIC MODEL AIRCRAFTDESIGN

Basic Proportions for RIC Model Aircraft A CHAPTER 25

\Rudder 35% VT

Aspect ratio 4.8sectionNACA 001area 20% of wingelevator 40% H.T.

Section NACA 0012

,-· 10·..............................J(

.25MAC

_.- 25C

I!...:-----~I-

• CGI

Optionalslottedflaps30% chord

Taper ratio0.6

Section NACA 0012

Aspect ratio 6

Figure 5.Basic aerobatic airplane proportions.

Semispan

Flapoptional

~Aileron25%chordand~%

semi­span

TNA

25% Mac

I•

1.5t 02 rMAC

Pivot

l==]::~=t===ti~~~~ ~5;C

Bto 10% 01 length

;"'1( - - - - - - Length------~Nosemoment arm Tail-moment arm _

Hor. tail Vert. tall% wingarea % wingarea

l~~ ~ ~13% 6% :q

V~~?JL.J.!::~~---===It==~,

Aspect ratio7 to 10

Taper Root 2 - 11Pl

Figure 4B.Basic gliderproportions.

AIRFOIL LAYOUT PROCEDUREEvery serious modeler should knowhow to develop an airfoil from itspublished ordinates.

These describe each airfoil bythree measurements:

• Chord length and stations alongthe chord.

• Depth (ordinates) above and belowth e chord line at eachstation .

• Leading-edge radius and locationof its center.

All measurements are percentage ofthe chord length. An exception isthe Clark Y, whose depth is mea­sured from its flat bottom, not itschord line. With the bottom level,the Clark Y is at an angle of attackof 2 degrees, measured on it schord line.

This author measures the sta­tions in l/ lO-inch intervals, along

th e ch ord lin e, from the leadingedg e. Some in terpolat ion isnecessary.

Depths abo ve and below thechord line are measured in 1/ 50­

inch intervals; some interpolatio nis needed . The necessary calcula­tions are simple.

StationsCho rd length x station percentage.Example: chord 7 in. x station 50 is3.5 inches from the leading edge.

THE BASICSOF RIC MODEL AIRC RAFT DESIGN 127

CHAPTER 25 .... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

for a high-aspect-ratio tapered wingwith many different ribs, this pro­cedure is both long and tedious.

Given chord lengths, airfoilsetion designat ions, skin thick­ness/spar location and sizes variouscompanies can provide very accu­rate computer-generated airfoil sec­tions at a reasonable cost.

Figure 6A illustrates a layout of a7-inch chord E193 section with ver-

tical line at each chord station. InFigure 68, the ordinate lengths,above and below the chord linehave been measured. Using Frenchcurves, the points are joinedsmoothly to outli ne the airfoil. ....

Ordinates (depths)Chord len gth (in.) x percent depth

2Example: a 7-inch chord with7.88% depth at station 50 is 7 + 2 x7.88 = 27.58 fiftieths above thechord line at station 50.

Most calcul ators have a "Constant"feature. Using it, th e cho rd len gthis entered once; the sta tion or ordi­nate percentages on ly are neededto complete the calculatio n.

Note th at ordinates below th echord line are negative, e.g., -2.5 . I

o .10 .20 .30

Chord 7 inches Stations~ ~

.40 .50 .60 .70 .80 .90

..I

1.00

Figure 6.Drawing £193 from ordinates.

Nose rad iusQu oted as a percentage of thechord's len gth, NACA airfoils, suchas NACA 2412, locate the center ofthe nose rad ius by "slope of radiusthrough the end of chord 2120."Simp ly measure 2 inch es from th echord leadi ng edge; erect a verticalline 0.2 inch h igh , above th e cho rdline. The diagon al, from th e cho rdline to th e top of the vertical line ,locates th e cen ter of th e noseradi us. On a lO-inch wing chord,this radius would be 0.158 inch.Laying out one airfoil section takes15 to 20 minutes. For an un taperedwing, thi s is no problem . However,

o .10

A- locating stationsand verticals

.20 .30 .40 .50 .60 .70

B- Measure ordinatesand draw curves

.80 .90 1.00

SEMISYMMETRICALFLAT BOTTOM

, CLARK Y

128 THE BASICS OF RIC MODEL AIRC RAFT DESIGN

E193 NACA 2412 , E197

Chapter 26

Construction

Designs

H ere are a few of the innov­ative RIC airplane modelsthat th e author has

design ed . The various sportplan es, canards, three-surface andamphibious design s and glidersincluded illustrate a variety of thedesign elements and approachesdescribed in thi s book.

Immm- - - -Type urp/liblollllllortGrollwelgllt 110 oz. (I.nd);

121 oz. (nlBI)Wing a~• ..........................666I11. I".Wing loading '4.3 oz••• tt.

(I.nd); tI.•oz••• It. (nlBI)881m2 loading 3.33oz.Engine ............................................•41Prop 11x6Power loading 239.9oz./rld(l.nd);

213oz./rltJ (11II"')

(MtltJeI AI"".", NIWtI, OCt. '92)

+3.50

t25"

1+--- NACA 441 5

4.5" chord

.... mmmm T 3-------1

mmI!LE!m- - - - - - ....Type t:lIlUlltJ

GrOll welgbl 7Soz.Wing ar• ..............................44411/. IR.Wing loading '4.3 oz. III. It.Engine , ,......... ...•3010.•35Prop 11Jx5 0I11JxB puhlrPower loading 215oz./cld

(Modll AI",,.,,, 1IIWtI, ..". '81)

)

51 "

THE BASICS OF RIC MODEL AIRCRAFTDESIGN 129

CHAPTER 26 • THE BASICS OF RIC MODEL AIRCRAFT DESIGN

150

8.25

==zt:==

43.75"

(Mod,1 AI",I,,,,, 1IIwI, Jan. 'II)

+1.60

EIImlm- - - - - - - - - - -Type tllrltl-lllrlll. ".""GrOSS welgbl f11 oz.Wing area 226 III. In. (Iorsp");

4!iIJ III. In. (IIIp8); 11211I. Ill.(1ItJrlztJIIm1 mil)

Wing loading 21.7oz./ddEnglne 46SFProp APe 11116Power loading 210.8 oz./dd

~

l....---I---~ Slotted flaps

I, ,.,::r r-r-

:,.,"":en

"" ,............ ,'-..l:...... ""'-

,II>.... E168 -----<0co

E2.1!..... em<0u:i i--:<=II> l-

~f.-. 5" I.-

~,, ~m~-.

-. " '--

(Model AI",I."" NIIWB, SBpt. '93)

1mID- - - - · --.lftIe sportGrOSS welgbl 92OZ.

.Wing area 60811I. In.Wing loading 22 oz./rq. If.Engine ..48 ~/d

Prop APe1MPower loading 2tID oz./t:ld

r~H~t+~~ ,~~~J.~~O fd;X::r::S

15"

1!I!1!II- - - -Type pOWtlrsd glldITGross welgbl 55.375Wing area 60211I. In.Wing loadln 13.16oz./Iq. If.engine 16Prop APe 8X4Power loading 367oz./Cld

(1Iodel AI",I,,,,, NIIWB, NOli. 1994)

13 0 THE BASICS OF RIC MODEL AIRCRAFT DESIGN

Construction Designs ... CHAPTER 26

1+----- --- ---- 43.5"- - - - - - - - - - -

Inverted LE slot\ Pivot.1..

• E168

Stab llator section B-BMass balance

Slotted flap /

/rl --" -."':.1 Slot lip \

--..0;:::::-"...---':::" \:

Flap pivot -. +

(MtIII" AJIJI',n, Ntnn, Au,. '96)

mD---------- ---., JYpe ; ,. j,; SGrOSl weight : 81Wing area .,.Wing loading 25.4DZ./ItI. II.Engine 46Prop APC 11x1 IN 11xTPower loading 191.3 oz./fI1I

Pivot

Slot lip aileron

~" ~'~~7-'".

Fixed L.E. slot /Wing section A.A

14"

A

Slotted flaps(30% Chord)

Inverted L.E. slots

l'cl",",I--t--Slot lip aileron

, ,--_ _: ~:'C---

1.125"

Dihedral 30

eetA

Span57.75"

1lDII- - - - - - - - - --._ ",ortAIoBt plBn,GrOSl weight 113OZ.; 143oz. (wA/08I6)Wing..... .. :; 76B.,. In.Wing loldlng 211.112 DZ./BIt. 11.;

26.5oz. (wA/DBI6)EngIne 45Prep APC 1MPow8r loading 251oz./rlll;

317oz./CIII (wIl108I6)(llstlB11lIII1fIIt, JIm, '91)

THE BASICS OF RIC MODEL AIRCRAFTDESIGN 131

CHAPTER 26 ... THE BASICS OF RIC MODEL AIRCRAFT DESIGN

E214 @+3.5° E197 @+1.25°

I3.6"chord

6.46"average chord

Dihedral 5°

~CG

132 THE BASICS OF RIC MODEL AIRCRAFTDESIGN

15.8"

Ouler panel ACI

Construction Designs ... CHAPTER 26

1 75'

WL

38

6.63"

6375"

E193-

Bottom

_L35"beam

Si~"n 1--- - 17" +---+-+-- 17625"

ClarkY_ 2°

.--- -----51"------~NACA 0012 _1 0 - -.===-.

NACA OO15 -

WL 0" -- __:.~ ~~ .~80---- -

--,---'---20" 31"----

Wingspan:61"

mmD- - - - - --....... . . . . . .. . . ... . .•mpIJllJl_

IIy/IIf'"Gross weight 4IIoz.Wlig 11'8I 2511 If. III.WIng loading 23I1Z./1f. It.BeamZ loading 3.26 oz./lti. It.Englnl 16Prop .. 7x4 /IlIIIItI1Power loading 261.6 oz./CIII(fIDdB1 AmtItJn, Ot:t '87)

mm!I!J]]I- - - - ----,... ....................................ttyI",1JuI8r8a weight 112oz.Wing 1I'8I 1BfIII. In.Wing loading .. . 23.3oz./lll. fl.Btlmzloading 3.1111Z./1f. I".ElgI 4IddPrep 11Jl8 puB/IMPower IoadIH 243oz./rld

1\4+':< (Re ..,.,., 011. '12)

THE BASIC5 OF RIC MODEL AIRCRAFT DESIGN 133

Appendix..,

Cl 001

-."

~Eppler 211 is a foreplaneairfoil wIth a sharp stallat lowRn. Note thereduc­tionin angle ofattack ofzero lift asRn is reduced.

"Cl

"

~Eppler 197 is amoderately csm­bered airfoil with asoft, gentle stall.II has very lowdrag.

- ,

- ,

.. .. 12 ... ' 0 "-, co OUA

-, -, ..- 0 300

""",,CA C CI\

.~/--01 (/- O.

0 7 u 0 7 ... --I I - 0 1 ~

es os

1/ - 0.100 rr- Eppler airfoil 168is0' I: "

symmetrical with no, t ' pitching moment,o. I"

J\I- I< il-'o

except at the stall, duro- 0 1 0 .. 0 .. 0.10 0.12 0'4 ' 0 II ing which the airfoilcw OUA

- 0 3 ;;' becomes nose-down\ \ 01(10 andis stabilizing.

- os -05

\ \ O,1!lO

- 0 1 \\t~ ~: :

0.200

- 01 \~

"'- 0....

II "Cl c: ---............ Cl 001

- "'5

" 12 -,'-

~

Eppler 214 is anaft·loaded aIrfoil thathasgood lift. II starts to lift ata negative angle ofattackand has camber near thetrailing edge.

.12 ... .. II

- , co OUA

II " -.' -, -, .rsCl Cl 001

-."

" " - ,

'134 THE BASICS OF RIC MODEL AIRC RAFT DESIGN

o."J, .

E 105 110.5%1

,...........

Io400 ELllll I NOKANl l

UNI STUTTGART

.....Eppler 205Is moder­atelycambered. It hasgood 11ft andlowdragat lowRn andIs thin­nerthen Eppler 197.

~Eppler 222 Isalsomoder­atelycambered. Ithasgood 11ft andlowdrag atlowRn andIs thinner thanEppler 197.

E 222 110211

......· ......

MODELlWINOKANAlUNI STUTTGART

CL e M

-.35

12 •.3

c --==--_

10 1. 18

e -.1!l

-, 25

8 -,2

· 11 2

1 2 -.3

CL eM

·25

I

• '00000, 200000

• eoooo

. 12 .14.ce

.....Eppler 184Is areflexed airfoil with alow, nose-down pitch­Ingmoment.

02 ...

-e

. e

·2

' 2

I10

1 -25

~Eppler 230has a reflexedtrailing edge andhas anose-up pItching moment.

-e

-2

THE BASICSOF RIC MODEL AIRCRAFT DESIGN 135

.~...

- --- -

• CHOOSING AIRFOILS• WING LOADING• CG LOCATION• BASIC PROPORTIONS• AEROBATIC DESIGN

Have you considered customiz­ing one of your models toenhance its performance? ordesigning your own RIC modelairplane? If you have , this bookcontains a gold mine of practi­cal guidance, hints and tips thatwill guarantee your scratch­

building and model-customizing success. From aerodynamics to structuresand control surfaces, Andy Lennon offers practical solutions and an under­standing of why they work.

Which type of airfoil should be used? How should the weight and balance becalculated? How can a plane be designed so it will be stable and have very lit­tle drag? Should flaps be incorporated, and are they beneficial in reducinglanding speeds? With several decades of designing and flying successful modelaircraft, Andy answers these questions and many more in a practical, conciseway that will help you with nearly any project currently on your workbench.

Andy's book presents a thorough and comprehensive introduction to theintriguing world of model aerodynamics. It's jam-packed with graphs andcharts that are easy to understand and extremely helpful to the new or sea­soned designer. Airfoil selection , the all-important wing-loading calculationand finding the proper CG location are just some of the topics to be foundin the opening chapters.

Learn how to design efficient horizontal and vertical tails , determine horizon­tal tail incidence and estimate the downwash that affects that incidence. Andyexplains why these estimates are necessary and tells how to do it. Reducingdrag is a constant battle for the model designer; Andy shows how to do itby properly shaping fuselages, streamlining land ing-gear wires, and cor­rectly mounting the wing on the fuselage. If you 're seeking improvedaerobatic performance or a design that will perform well in a high-G turn,Andy again spells out the answers.

Interested in building unconventional models that utilize canards or threelifti ng surfaces? Andy clearly sets out the design principles. Secrets for suc ­cessful seaplanes and floatplanes are also covered. Andy tops off his bookwith a look at a few of his published designs, all of wh ich incorporate thedesign pr inciples presented in this unique volume.

Whatever your modeling background, this book will be a valuable refer ­ence source in your RIC library, and it will never be outdated. Filled withtimeless insights that range from the findings of early NACA reports toapproaches adapted in modern aircraft, this work will serve you well timeand time aga in.

Acomprehensive guide to designing radio control model airplanesBASICS OF

RIC MODELAIRCRAFT DESIGN

2023 12/05 2M HG

ISBN: 0-911295-40-2

II IIIII 90000>

9 78091' 295405 IIA'i --(AirAGEM E D I A

lIirDlaneNE\NS

modelairplanenews.com PRINTED IN THE USA $19.95