The Aircraft Engineer June 19, 1931

7/27/2019 The Aircraft Engineer June 19, 1931

1/7

Junr 19, 1931 Supplement to FLIGHT

ENGINEERINGSECTIONEdi t ed by C. M. P O U L S E N

June 19,1931CONTENTS

PAGEVariable Lift Wings. By F.Duncanson, B.Sc, Wh.Ex. 41Metal Construction Development. By H. J. Pollard, Wh. Ex.,A.F.R.Ae. Soc 45Technical Literature 47

VARIABLE LIFT WINGS.B Y F. DTTNCANSON, B.SC., Wh.Ex.

Mr. Duncanson, who is on the Technical Staff of theGloster Aircraft Com pany, Ltd., is no stranger toreaders ofTHEAIHCKAFT ENGINEER, as he has previouslycontributed articles on cantilever wings and on theinfluence of size on structure weight. In this issueMr. Duncanson takes up the subject of variable camberwings, and comes to the conclusion that with modernefficient aircraft the advantages to be derived from theuse of variable camber are greater than they were inthe older types of machine. Mr. Duncanson estimatesthe weights and performance of two types of machinedesigned to do the same work, onewith fixed wings andone with variable camber wings. He arrives at theresult that the variable camber-wing machine will havea top speed some 12 m.p.h. greater than that of thefixed-wing machine, while the rate of climb is also verymaterially better, as are also service ceiling and abso-lute ceiling. In addition, Mr. Duncanson points out,the use of variable camber wings enables smaller overalldimensions to be attained, which in turn means im-proved manaiuvrability, better view, and a reduction infuel consumption. Mr. Duncanson does not regardvariable camber gear as a means to reducing landing>peed, but as a means to better performance and greatermviauvrability, and it is from this new point of viewth-i t he examines the subject.

J HE primary object of employing a device whereby theu.t!1On of a w i n m a v b e v a r i e d is to obtain a wing"ing aerodynamic quali t ies conducive to high speedc f : e d w i th qua l i t i es tha t are necessary for goodc - " b and slow landing speed. The aerodynamic pro-Ptrties sought are, firstly, the best possible L ift / D rif tra-.os at lowvalues of K ; secondly, good L/D valuest n aerate values of KL; t h i rd l y , as high a maxi mum* as possible. These qu alities are best visual ised by ae> resen tation of the aerofoil characterist ics on thete .i * L c h a r t -^ 'cs of

n Fig. 1 are plotted the charac-a typical medium-lift aerofoil, shown by

curve A, while the charac ter i s t ics tha t may be obtainedfrom this aerofoil when fitted with a t rai l ing edge flapare shown by curves B and C, B indica t ing the effectof set t ing the flap at a sl ightly negative angle relat iveto the main port ion of the wi ng , and C indica t ing theeffect of se t t ing the flap at a large posi t ive angle.

:-4 -y x x ENVELOPE

8 9 i-O

The envelope of all the wing charac ter i s t ics obta inablebetween the ext remes i l lus t ra ted in Fig. 1 is indica tedby curve D. In some of the early researches on vari -able camber wings the envelopes indicated that muchgrea ter improvements could be obtained over the char-acterist ics of the original sect ion than are shown inF i g . 1, but thi s must be ascribed to the selection ofprimary aerofoi ls which nowadays would be regarded asvery inefficient. The proport iona te ga ins in wingcharac ter i s t ics indica ted by Fig. 1 are, however, quiteenough in certain designs to just ify the adopt ion ofvariable camber .556a E


2/7

ISUPPLEMENT TOF L I G H T42 JU N E 19, 1931THE AIRCRAFT ENGINEER

Results of experiments on actual machines are alwaysmore convincing than those obtained in the wind tunnel .It is , therefore, very inter est ing to com pare theL / D X K,_ curves of a complete aeroplane for thenormal wing and for various flap settings, as shown onFig . 2. These curves are construc ted from d ata given inR. and M., No. 1085.*

This point is illustrated by Fig. 3, which is replottedfrom t he d iagr am published in " Fl ig ht " of March 71930, page 270. Assum ing th at a wing incidence of 17'may be obtained by a machine of normal proportionsduring a three-point landing, a slot ted wing would haveits K L increased from 0.54 to 0.61 at this incidencei.e., a 13 per cent, increase, whereas an increase to 0.87'

, , , t

7//

,'// /

fj /V/1 1Ij 7s

s

XF L A P S N O R M A L

^FLAPS DCS!

\\ "\

iiiii

FIG.2

VN 8"

\iii

- PUHCM.HH

This rep or t gives figures of carefully-m easured liftand drag of the complete aeroplane with pro pellerstopped, at flap settings of 5.1, 0, 8 and 15.7.I t i s thou ght tha t an inte rmed ia te se t t ing be tween 5.1 and 0" may have given a curve in about theposition shown by ? ? P.It will be seen from the abo\*e that the effect of flapson the maximum lift of the original aerofoil is toincrease this by abou t 20 per cen t . For a comprehen-sive discussion of the present state of development ofvariable camber the reader is referred to Capt .Macmillan's very able and interest ing art icle which waspublished in THE AIRCRAFT ENGINEER of June 20, 1930.The object of the pre sent article is to consider th ead van tag es of variable camlier from different poin ts ofview, and to suggost new avenues for research anddevelopment work.In the case of modern designs of aircraft where ahigh degree of aerodynamic efficiency is being obtained,the wing drag is now a greater proport ion of the totaldra g of the machine than was formerly the case. Anyimprovement in the characterist ics of a wing wil l there-fore result in a gre ater propo rt ion ate imp rovemen t inthe performance of a modern aircraft than would beachieved in the case of an obsolete design . An otherimportant advantage of variable camber flaps in thecase of modern highly efficient aircraft is their effect incoirsening the extremely flat gl iding anglo, which ischaracterist ic of these machines, when coming in toland, so th at a pa rt from th eir effect in reducing la ndin gspeed, they give a useful effect as air brakes.A combination of variable camber wings with lead-ing edge slots gives very great advantages, inasmuchas the full benefit of slots in reducing landing speed isonly obtained at angles so large as to be unattainablein normally proport ioned aircraft , whereas the two incombination result in a high l i ft being obtained at thenormal a t t i tude the machine would take in an ordinarythree-po int landing . The l i ft obtained by the combina-t ion of flap and slot is , moreover, higher than thatobtainable with slots alone.

1-21-1

10 9

8

7

6

5

4

32

1

-

""VSLOT OPENAP DOWN 20' /

yi

5 10 15 0

/ FLAPNORMAL,R A F. 28

FIG 3

[X = 28 * /

3PENNORMA

AEROHSECTIO

I

31LN

20 25 30

Lift and Drag of the Bristol Fighter with Fatrejr variable camber wings.

representing a 61 per cent , improvement, is inherent inthe wing but cannot be used, the required wing inci-dence bein g 28 . W hen the slot and var iable camberflap are combined, however, the KT/ at 17 is increasedto 0.98, representing an 85 per cent . gain.It is obvious that no form of varinhle lift wing isworth a dopting unless lateral control is both light andeffective at and beyond stalling speed; in this connec-tion it is pleasing to note that in the case of variablecamber wings recently t ried i t was found that thelateral control was effective throughout the speed rangeof the machine and at all variations of wing section,and that the ailerons felt even lighter with full camberth an with flaps no rm al. I t mu st also be born e in mindthat whatever be the devices used to augment the l i ftof the wing, these must be so arranged that for high-speed conditions the profile of the wing is not interferedwith to any appreciable ex ten t . This object may beachieved by careful at tent ion to detai l design.In order to obtain the maximum possible benefit fromthe use of variable camber, this should be regarded notas a means for reducing the landing speed of exist ingmachines, but as a means for reducing their overall sizeand weight, and increasing performance and manoeuv-rability, while still retaining a reasonably slow landingspeed.To illus tra te th is new po int of view we will make acomparison between two aircraft, both designed foi tnesame purpose, one with fixed wings, which will herein-after be referred to as the F.W. machine, and the otherwith variable camber wings, which will be referred toas the V.C.W. machine, on the basis of the same land-ing speeds for both des ign s. The wing section selectedfor bo th cases will be R .A .F . 28. Fo r the purpo?" this ex amp le a hyp othe tical specification will " eadopted, the leading requirements of which are:

(1) Stal l ing speed, 55 m.p.h.(2) Ma xim um speed a t 10,000 ft. t o be not less tian160 m.p.h.(3) M ilitar y load, 1,200 lb. ,(4) Dura t ion, 6$ hours at a cruising speeo130 m.p.h.(5) Span not to exceed 44 ft .of

5566


3/7

JUNE 19. 193143

THE AIRCRAFT ENGINEER SUPPLEMENT TOFLIGHTWith regard to the F.W. machine, we arrive at anestimate of the all-up weight by first of all assuming thatthe total weight of a suitable water-cooled engine instal-lation of 500 to 600 b.h.p. is 1,350 lb., to which is

added the military load, fuel and oil, having a totalweight of 2,500 lb. The weight, less structure, istherefore 3,850 lb. From experience with similar

The V.C.W. wing weighs 1.84 lb. per sq. ft., BO thatthe wing weight is 1.84 x 401 = 737 lb.844 - 737 = 107 lb. = saving inwing weight.

The reduced chord of the V.C.W. machine results inreduced fuselage length, greater concentration of massesdesigns we know that the structural weight will be of and lighter stresses, and the consequent saving of fuse-

the order of 34 per cent. The total weight of theaircraft will therefore be: o 850W.F.W. = - ~ = 5,840 lb.

Theoretically, the KL max. of a R.A.F. 28 Biplane,with what little help ia obtained from the top wingslots, will be 0.57 ; but in practice we know that a KLat the stall of 0.63 will be realised. (Throughout thisinvestigation, in order to be on the safe side, thebenefits of any doubts, such as this, are given in favourof the F.W. machine.)The wing loading appropriate to this KL max. andtlie stalling speed of 55 m.p.h. is 9.7 lbs. per sq. ft.The wing area will therefore be 602 sq. ft. The F.W.aircraft is illustrated in side view by Fig. 4, in frontview by the left-hand side of Fig. 5 and in plan viewby Pig- 6. The average aspect ratio of the wings turns F.W. wing weighs 1.4 lb. per sq. ft., so tha t theweight is 1.4 x 602 = 844 1b.

lage weight, calculated on conservative assumptions, hasbeen found to be 96 lb. The V.C.W. machine willachieve the required duration on 137 gallons of fuel asagainst 155 gallons in the case of the F.W. machine.The saving in fuel alone will, therefore, account for afurther weight reduction of 137 lb. There will also beseveral small weight reductions in other components ofthe machine, such as the petrol tanks, piping, controls,tail unit, etc., but to be on the safe side these reduc-tions will be neglected. The weight estimate of theV.C.W. machine will therefore be:

W.vc w = 5,840 - (107 + 96 + 137) = 5,500 lb.The V.C.W. aircraft is illustrated in side view byFig. 7, in front view by the right-hand side of Fig. 5and in plan view by Fig. 8.An appropriate wing design is found to have thefollowing proportions:Span 41 ft. 6 in.; chord,5 ft. 3 in.; mean aspect ratio, 8.It is obvious that a two-bay wing structure is neces-sary in this case. This does not mean that any appre-ciable aerodynamic loss need be feared, provided that

care is exercised in streamlining. Evidence in supportof this statement may be found in the case of theGloster Multi Gun Fighter, whose performance isphenomenal in spite of the handicap of a radial air-cooled engine, large military load and additional dragof wing guns.The biplane wing characteristics of each design, cor-rected from the monoplane tunnel figures by means ofthe standard Prandtl methods, are shown plotted onFig. 9.The parasitic drags of the two alternatives have beensummed up in the usual way, and in this connection itshould be noted that in spite of the V.C.W. machinehaving a more favourable body fineness ratio and

556c


4/7

SUPPLEMENT TOF L I G H T44

THE AIRCRAFT ENGINEER JU N E 19, 1931

22

20IS161412108642

\ rr

1

LFIXEC

\

*\ \

WIN(

VARIV. w\ABLE

V\

2 3 -4 !* L

: A M B (

N\ \N> >

R WIK

\

a

FIG.9\

7 -8 -9

60 0

50 0

400

300

IrtA

H P0

1

\

FIXED

^ ^ w ir

WINS AlRCF

kRIABLE CAM>I6 AIRCRAF^

f//VFIG.IO

/

V

) SO 80 100 120 140 160 180V M.P.H

greater clearance between the wheels and bottom plane,to be on the safe side, no advantage has been taken ofthese indications of reduced drag coefficients in favourof the V.C.W. a l te rna t ive .The performances at sea level of the al ter nat ivemachines are shown superposed on Fig. 10, while theperformance part iculars with respect to height areshown graphical ly by Figs. 11 and 12.The fol lowing table summarises the leading par-t iculars of these a l te rna t iv e a i rc raf t :

Span (both planes)Overall lengthOverall heightChordMain plane areaTotal weight fully loadedWing loadingSpan!*WAirscrew diameterSpeed at 6.1.Speed at 5,000 ft.Speed at 10,000 ft.Speed at 16,000 ft.Rate of climb at s.I.Itate of climb, 10,000 ft.Time to 10,000 ft.Service cellingAbsolute ceiling .,Alighting speed

Fixed WitvAircraft.. 44 It. 86 ft. 6 In. ... 12 ft. B in.. 7 ft. Ota.. 802 BO. ft.. 5,8601b.. 9-71b./Bq. ft.. 0-881 ... lift. 165 in .ph.. 104 m.p.h.. 162 m.p.h.. 165-8 m.p.h. ... 1,180 ft. per min.. 560 ft. per min.. 12-6 min.. 19,000 ft.. 21,000 ft.. 56 m.p.h.

Variable CamberWinff Aircraft.41 ft;e m.32 ft .3 In.11 ft .6 in.5 ft. 3 lu.401 sq. ft.5,5001b.13-7 lb . / sq . f t .0-81810 ft. 6 in.177-5 m.p.h.176-7 m.p.h.174 m.p.h.168 m.p.h.1,250 ft. per min.640 ft. per min.11-2 min.20,250 ft .22.200 ft .66 m.p.h. Th e span loa ding in more favourable to th e F.W. than to th e V.C.W.aircraft.

Apart from the higher performance of the V.C.W.machine, the fol lowing advantages are natural lyobta ined from this des ig n: (1) Smaller overal l dimensions.(2) Improved manoeuvrability on account of the de-creased mom ents of ine rt ia of th e airc raft about all

three axes .(3) Pi lo t 's view gre atly imp rov ed, n ot only becauseof the narrower top and bottom wing chords, butbecause the sm aller gap enables the top wing to beplaced at such a position relative to the cockpit thatthe edge view only is visible to the pilot.(4) Redu ced fuel consu mp tion. (In th e examplechosen this amounts to 11.5 per cent .)As time goes on the need for the adoption of variablelift devices becomes more and more apparent, sinceincreased perform ances a nd g rea ter useful loads arecontinu ally being dem anded . Su bst antia l advantagesmay be obtained in practically every type of aircraft,and th e applicat ion of variable camb er, already in anadvanced state of development, is susceptible of stillfur ther improvements reg arding such mat te rs aseffect ive lateral control and simplici ty of design.The author wishes to acknowledge his indebtedness tothe Co ntroller of H .M . S tati on ery Office and to theSecr etary of the Aero nautical R esearch Committee inconnect ion wi th the prepara t ion of thi s a r t ic le .

25000

20000

15000

VARIABLE CAMBER W ING AIRCRAFT\

10000

5000

RATE OF CLIM B V I I N 20 0TIME TO HEIGHT MIN K> 4 0 020 60 030 80 04 0 10005 0 12006 0 100 120 140SPEED M.P.H

556d


5/7

JUNE 19, 1931 THE AIRCRAFT ENGINEERSUPPLEMENT TOFLIG H T

METAL CONSTRUCTION DEVELOPMENT.B Y H . J . POLLABD, Wh.Ex. , A.F.R.Ae. Soc .

" Developable Surfaces."Apart from designs of foreign importation, metalconstruction development in this country has in themain been concerned w ith framed str uc tur es , chieflygirders of the strut and wire type, occasionally withgirders rigidly brace d, the desired exte rna l surfacesbeing obtained by fairing, general ly with non-structuralmaterial.We need not concern ourselves here with th e r eason swhy the long established pr incip le of gir de r c onstr uc-tion has been retained; suffice it to say that the problemof the substi tut ion of structural components made fromhigh-tensile steel presented enough problems in itselfwithout the additional complication of innovations inoverall str uct ur al design. The re is now, however, aninsistent demand for the utilisation of surface materialfor bearing parts of the structure stresses, and aircraftstructural engineers are study ing the problem. I t isnot intende d, a t this sta ge, to discuss th e possiblemerits or dem erits of " monocoque " str uc tur es , bu trather to place before the reader certain general con-siderations relating to the shapes of surfaces intendedfor " rigid " covering.

Tt is obviously very de sirable t ha t t he conto ur of asurface to be covered* by metal sheets should, whereverpossible, be such that a flat sheet of the coveringmaterial may, on being laid on the surface , conformthereto without any stretching or without the necessityot cutting up into smaller pieces in order to el iminate"' us, crinkles or buckles in th e finished cov erin g. Ae-rved surface which can be " opened out " and lid


6/7

SUPPLEMENT TOFLIGHT JU NE 19, 193:THE AIRCRAFT ENGINEERplane. Similarly, with the other pairs of lines. There-fore a surface is obtained composed of a series of tri-angular planes aab, bbc, etc. The first of these planesurfaces bounded by lines aa and bb may be bent roundline bb until it lies in plane bcc. This larger flat sur-face may be bent round line cc , and so forth until thepolygonal surface is developed into one single plane.

the planes are intersecting straight lines forming adevelopable surface as described in connection withFig. 1.A short consideration of the second class of ruledsurfaces will help the reader to a clearer understandingof the whole problem.In Fig. 2 we have a representation of a series of lines

FIG 4Definitions of the terms used are, perhaps, made clear

by puttin g the above statemen t into mathematicallanguage. Since two consecutive generating lines inter-sect, they lie in one plane, and a surface such as theabove may be produced by the ultimate intersections ofa series of planes, and since any two consecutive planesintersect on a line on the surface, the equation repre-senting any one of the series of planes can involve onlyone arb itrary co nstant. To make this clearLet the equation of any surface be/ (x, y, z, a) = 0 (1)

where a is a constant.Let a be changed to a, then the equation is/ (x, y, z, a,) = 0 (2)

This simply means that equations (1) and (2) representsurfaces of the same shape, but differing in size orposition, or both.If the surfaces intersect, then all parts on the curveof intersection are satisfied by the above equa tions. Ifa, is made to approach very near to a then the curveapproaches some limiting position, and the locus of allsuch limiting positions for different values of " a " is asurface which is called the envelope of the family ofsurfaces (1).If either x, y or 2 in the above equation be made zero,

AA, BB, etc., in space. The line AA lies above BB,BB above CC, and CC above DD; none of the linesintersects the other, and xx 1, yy, etc., are the shortestdistance between these lines. The natur e of the surfacecan best be seen by imagining line BB rotated about x'unt il both lines (AA and BB) lie in one plane. If thisis taken to be thin me tal, th en, obviously, when BB isrotated back to the position shown in Fig. 2, every con-necting line must be stretched except the shortest linexx 1. This, therefore, is not a developable surface, butis known as a skew surface, or scro ll. We shall not dealfurther with this class of ruled surface.Returning to the definitions, the generating linesshown in Fig. 1 intersect in the polygon b, c, d, etc.,whose sides are in the direction be, cd. This polygonapproximates closer and closer to a continuous curveas the generating lines become nearer together, and inthe limit is a tru e curve. This curve is called theEdge of Regression or Cuspidal Curve. The curve isalways tortu ou s, i.e ., th e plane co ntainin g two sides ofthe original polygon does not in general contain tuenext side.This plane, which contains two sides of the polygonof which the tortuous curve (the edge of regression mour Fig. 1) is the limit, in its final position is knownas the osculating plane of the curve at the particularpo int . As two successive positions of it contain tiGth at is if the surface is a plane, the arbi trar y varia tion second side of the polygon, then clearly th e osculatingin the value of aj may be terme d giving th e plan e one plane passes from one position to the nex t by revolvingdegree of freedom, and the trace of the intersections of round the tan ge nt to the curve, and it is evident &

556/


7/7

19. 1931 47THE AIRCRAFT ENGINEER SUPPLEMENT TOFLIGHTexplained above that the envelope of the osculafAngplave to a twisted curve is a developable surface.Before demonstrating how the tangential property ofuhe tortuous curve may be utilised in laying out adevelopable surface, it may be helpful to make one ortwo further observations on such surfaces.

In the case of a cone it is obvious that there are twodevelopable surfaces, for the generating lines lying onone surface pass through the vertex forming a secondconical surface. Similarly with the tangents to an edgeof regression ; these pass either side of the curve, form-ing a second developable surface, or, as it is called, asurface in two sheets.The method of construction of such a developable sur-face is given in Thompson & Tait's Natural Philosophyand is as follows: " Lay one piece of perfectly flat, unwrinkled, smooth-cut paper on the top of another. Trace any curve onthe upper, and let it have no point of inflection, buteverywhere finite curva ture. Cut the paper quite awayon the concave side (see Fig. 3). If the curve tracedis closed, it must be cut open (see Fig. 3A). The limitsto the extent th at may be left uncut away are thetangents drawn outwards from the two ends, so that,in short, no portion of the paper through which a realtangent does not pass is to be lef t. "" Attach the two sheets together by very slight paperor muslin clamps gummed to them along the commoncurved edge. These must be so slight as not to inte r-fere sensibly with the flexure of the two sheets. Takehold of one corner of one sheet and lift the whole. Thetwo will open out into the two sheets of a developablesurface, of which the curve, bending into a curve of

double curvature, is the edge of regression. The tan-gent to the curve drawn in one direction from thepoint of contact will always lie in one of the sheet.-, andits continuation on the other side in the other sheet. Ofcourse a double-sheeted developable polyhedron can beconstructed by this process, by starting from a polygoninstead of a curve."As we have seen, a ruled surface may be developedinto a plane when all its generators are tangential tothe same curve. This is the fundamental fact in whatfollows.The basis of the method is to draw lines upon threeviews of the surface (required to be developed) inproper projection. The intersection of the lines formcuspidal curves, and these cuspidal curves must be inprojection, i.e., the points of tangency of the projec-tion of the generators with the cuspidal curve must be inprojection, or the curves or generators altered untilsuch agreement is obtained. I t then remains to pickeff, from the three views, suitable sections or formersover which the developable covering can be laid.The bounding lines or curves of a developable surfaceare known as the directrices. In demonstrating themethod of determining a developable surface we willchoose, for the sake of simplicity, a body in which thedirectrices are parallel or at right-angles to the planesof projection. In cases where the directrices are tor-tuous curves, the choice of suitable planes of projectionis often a difficult matter , and the process may be

tedious. The fact that the generators must, as well asbeing tangent to the cuspidal curve, also be tangent tothe directrices must not be overlooked, and, althoughus can be verified at once when the directrices areParallel to the planes of projection, yet when the direc-r ; 'ts are tortuous this may not be seen in the usualPa:i3s of projection, and additional planes are neces-sary. These have been chosen so th at proper contacty1]- ' J directrices is assured. Nothing more is in-o>"d than the ordinary rules of projection, althoughth(? use of several p1 - ' --'- ' arbitrarily.

the directrices'3 than the ordif use of several planes of projection will be necessary,chosen arbitrarily.

sh ~ S l m p! e examp le will take the form of the fuselagei n ' ^ m- ^ ^ ' 4- ^ s a v e s P a c e ^ e longitudinal scaleside view is smaller than for the other views.

A, B, C, D, E, F and a, 6, c, d, e, f are the boundingplane surfaces. We will assume th at in the layout ithas been possible to make portions AHab a cone, CcDi/cylindrical and EFc/ a plane surface, but that BChrand Br/Ee can take the form of none of these simplesurfaces; 3'et they are required to be developable.In the first case, the surface is bounded by directricesBC and be and generators Tib, Cc. These latter lines(shown dotted) form the terminal tangents to the edgeof regression x. Other generators are drawn, and thesetogether with the edge of regression formed are pro-jected on to the other two planes, and adjustmentsmade until the projection is accurate. Similarly withsurface T>d Ee. In this latter case the plan view of theedge of regression has been omitted for clearness; thecuspidal curve is marked Y for this portion.

Finally the ordinates for as many additional sectionsas may be considered necessary are obtained. In theabove case only one intermediate section has beenchosen. In the case of the plane, conical and cylindricalportions of the whole surface, the ordinates or radii andangles are merely the mean of the corresponding dimen-sions of the extreme faces, but the ordinates for theother parts of the mid section are scaled from thepoints of intersection of the appropriate generators andthe plane of the section required.

It must be clearly understood that aerodynamic andoperational requirements of, say, a fuselage, may pre-clude the use of a developable covering. This is usuallyat once obvious, but if a doubt exists then the applica-tion of the principles explained in this article quicklydecides the matt er . Moreover, the necessary modifica-tions or alternative compromises between the aero-dynamic and constructional considerations are madeclear.Finally, it should be understood that considerablepractice by an expert draughtsman working on large-scale drawings is required before results can be obtainedwith rapidity and precision, but in view of the prac-tical advantage to be derived from application of theprocess, such expense in design as may be incurred isjustifiable.

TECHNICAL LITERATURESUMM ARIE S OF AERONA UTICAL RESEARCHC O M M I TTEE R EP O R TS

These Reports are published by His Majesty's StationeryOffice, London, and may be purchased directly from H.M.Stationery Office at the following addresses : Adastral House,Kingsway, W.C.2; 120, George Street, Edinburgh; YorkStreet, Manchester; 1, St. Andrew's Crescent, Cardiff; 15,Donegall Square West, Belfast; or through any Bookseller.

SPINNING EXPERIMENTS ON A SINGLE-SEATER FIGHTER.PART I. FURTHER MODEL EXPERIMENTS. By A. 8.Batson, B.Sc, and H. B. Irving. B.Sc. PART I I .FULL-SCALE SPINNING TESTS . By S. B. Gates, M.A.R. & M. No. 1278. (Ae. 424.) (10 pages and 12diagrams.) August, 1929. Price 9d. net.

The single Beater fighter which is the subject ol this report is a Btaggeredbiplane, which in its early forms gave difficulty in recovery from spins.One of these forms has already been the subject of model test and a report.*Later forms, In which modifications have been made to the body and tall,are here mainly dealt with. The model tests were made on a form rf themachine ln which, not only were fin and rudder areas increased, but thebody was lengthened and the tallplane raised from the middle to the topof the rear port ion of the body. These modifications resulted in preatlyenhanced damping moments due to body, fin and rudder while rolling,roughly as much of the increase belngtcaused by the lengthened and'deepenedbody as was due to the enlarged fin and rudder. The raising of the tallplaneoontrlbuted in no small measure to these Increases of both body and finand redder moments. a. & M. 1184. Experiments OQ a model of a Single Beater FighterAeroplane in connection with Spinning. Irving and Batson.

556 g

Documents

The Aircraft Engineer June 19, 1931