7/27/2019 The Aircraft Engineer January 30, 1931

1/4

J an u ary 39,1931 Suppl em ent to FLIGHT

ENGINEERINGSECTIONEdited by C. M. POULSEN

January 30, 1931CONTENTS

PAGETKe Paris Aero Show , 1930. By H. J. Pollard, Wk.Ex., A.F.R.Ae.Soc. 1The Progress of Aero Engine Design at the Paris Show. By CaptainAndrew Swan, B.Sc,A.M.Inst.C.E. ... ... ... ... ... 4Technical Li ter atu re ... . .. . .. .. . ... ... ... ... 8

THE PARIS AERO SHOW, 1930.By H. J. POLLARD, W h.Ex., A.F.R.Ae.Soc.

After a preliminary tour of an aeronautical exhibition,one naturally en dea vou rs to decide what are the chief featuresof development. Opinions freq uen tly differ as to whetheran alteration is an improvement, and consequently the termdevelopment is not necessarily synonymous with improve-ment. The results of development of ideas on details ofdesign are often neutral or negative, but the ground hasbeen cleared by the work done, allowing a clearer vision ofprospects, and not infrequently developments of ideas leadto definite imp rove me nt in design.In therecent Paris Aero Show the most noteworthy featuresof general development appeared to be the increased adoptionof monoplane structures, together with the increased use of

metal covering for both wings and bodies. A fair percentageof the fuselages were of the metal monocogtie type, and byway of a beginning some observations on these will beappropriate.The combination of covering, lateral, longitudinal, andin some cases diagonal members, may be effected in a varietyof ways ; the chief methods available may be t abula tedas follows :Class. Description.1 Transverse and lateral members at right angles, theshear being taken up by the thin flat sheet.2 Transverse and lateral members at right angles, withdiagonals added, the shear being taken up by theverticals and diagonals jointly with the thin sheetcovering.3 Deeply corrugated sheet with the corrugations runninglongitudinally for the purpose of resisting the bendingforces, the shear being taken partly across the corru-gations and part ly on transverse formers.4 Corruga ted panels reinforced by lateral and longitudinalmembers.5 As class 4, but with diagonal members added.

96a

7/27/2019 The Aircraft Engineer January 30, 1931

2/4

SUPPLEMENT TOFLIGHT JAN UA RY 30, 1931THE AIRCRAFT ENGINEERIn the foregoing, all reinforcements are placed inside thebody . Ano ther classification of structu res in which externalreinforcements are used might be tabulated, but the aboveclearly shows what variations are possible, and in fact,examples of all these classes were on view in the Salon deVAviation. In some cases more than one, class of construc-tion was used in the same aircraft.Class 1 seems to be the mo st rational construction, as acomplete structure is obtained and anything additional wouldapp ear to be superfluous. This would be th e case if the inne rstructure were stressed to carry the loads independently ofth e rigid skin (as in a fuselage cited below). Th e skin, how-ever, may be an essential part of the structure as a stabiliserand stifiener to the individu al mem bers of th e frame. Dia-gonal members in a rectangular panel mean smaller areasof unstifl'ened sheet, and are effective in improving externalappearance and also give a more robust structure, but as ama tte r of fact, Class 1 was the c onstru ction most in evidencein the aircraft exhibited, notable among such constructionsbeing Dt-woitine D.27, Dcw oitine D .35, Liore, Am iot, C.A.M.S.80 an d N ieuport-Delage. In oth er cases, whilst a largeproportion of the complete, body was of the above construction,som e variations were made ; for example, in th e L atecoerehydroplane, the last quarter of the body length was fabric-covered.The D ornier hull cons tructio n was of Class 1 over the rearhalf and Class 2 in front. Similarly, with the Caudron,except that as a third variant external stiffeners were usedon both upper and lower surfaces.Class 2 type of construction was exhibited by S.P.C.A.and Wibault.The Junkers .Junior had Class 3 to itself, whilst as far asone remembers, the same remark applies to Nieuport-DelageType 541 as regards Class 4.As would be expected, the exponents of Class 5 were few.Potez had sheets having shallow corrugations f-in. deepor so, probably at about 6-in. pitch ; the structure carryingthese was fully braced.The Hanriot H.431 had corrugated detachable panels,held by turnbuttons to a complete strut and wire structure,square tubes suitably joined at the panel joints being used.Also on the Hanriot 215 corrugated panels were used on abraced structure over part of the fuselage, the after thirdbeing fabric covered.The most notable construction having the skin sandwichedbetween internal and external stiffeners was in the A.N.F.machine.Examination of the external appearance of machines inClass 1 proved particularly interesting. In some cases theappearance was most pleasing from the point of view offreedom from local buckles, in other cases such buckles werevery pronounced.The futility of attempting to give reasons why metal-covered bodies should be produced in such a variety of waysis fairly obvious. W hy have E nglish designers who aredeveloping steel aircraft produced wing spars of such diverseforms ? Simply because it is no t usual for two people tothin k alike ; yet it would not be possible to say tha t anyEnglish interpretation of what a steel wing spar should belike is definitely superior to any other spar. Similarly, withthe French monocoque bodies : criticism is possible, bu t

the designer in each case could probably give a decidedanswer to our questions.The body of the smaller Nieuport-Delage was entirelyfree from buckles and had a beautiful finish : this, however,was obtained at the expense of much riveting, 22 longitu-dinally double rows of rivets being counted. The internalreinforcement consisted of large numbers of end-flangedchannel sections, the transverse loops being semi-circularflanged channels laid over the longitud inals. The skin,lateral and longitudinal members were secured with cheese-headed bolts and nuts, as shown in Fig. 1, the shanks of thebolts passing throug h the longitudinal channel. It was no tpossible to see whether distance pieces were used or not.While the necessity for cutting either of the stiffeners wasavoided, this arrangemen t did no t inspire too much confidence,unless the hoops are called upon to do very little work, or if

a little relative movement between these parts is of littleconsequence. Other meth ods adop ted for reinforcing themetal skin are given later.A construction which appeared to be superior to the Nieu-port w as to be found on th e Dewoitine D.27 : here onhtwelve dou ble rows of rivets were used. Th e panels wereapproximately 18 in. by 9 in. and the whole was very freefrom buckles. The body curv ature was about the same ason th e Nie up ort, hence closely-spaced stiffeners were notnecessary for the purpose of bolstering up the plating againstlocal deformation.The question of external appearance might be consideredto be so closely bound up with external shape that, forexample, local buckles in the thin flat plating would be absentin bodies of fairly sharp curvature, and that in " slab-sided "bodies th e avoidance of su ch buckles would be difficult. Asa matter of fact, some of the flat-sided exhibits were excellentin this respect, notably the Dornier, while the Liore boatwas another example of excellence in the matter of freedomfrom local buckles, unsupported panels being up to Q in, by1 ft. On the other hand, th e Latecoere hydroplane plates,which were approximately 8 in., in width, were very badlybuckled in places.Of course, thickn ess of platin g h as m uch to do wit h freedomfrom buckles, but in the three cases cited the plate thicknessappeared to be about the same.No useful purpose will be served in describing in detailevery difference between the numerous monocoque bodiesexh ibited . Possibly at future ex hibition s we ma y get someindication of which constructions have proved the mostsuccessful.As stated above, the external appearance of the exhibit?varied con siderably. The imp ortanc e of this from a sellingpoin t of view cann ot be ignored, and th e desire to k eep platingfree from buckles may have led to what we have termed theClass 2 and oth er classes of constru ction, inste ad of theextrem ely simple constru ction w e have called Class 1. Thedetails of the reinforcements of the other classes are worthme ntio nin g: the usu al stiffener w as roun d or rectangularsectioned chan nel h aving o utw ard lv e xte nd ing flanges : inoth er case s, simple an gles, or angles in w hich th e edges werestiffened by ben ding . Oth er section s were used for thispurpose, but the angles and channels were the most widelyused.In constructions where the whole of the reinforcement liesinside th e skin, the principal q uestion arising is what shallbe the scheme where the lateral and transverse memberscross each other.In some cases it might be desirable to make the transversemembers discontinuous at the hoops, or vice versa, havingsome simple fitting or joint at the point of juncture. Anexcellent example of this was seen in the Amiot three-enginedmachin e. The constru ction can easily be seen from Fig- '- This method of using a separate pressing was also foundin some cases where a long ridge was raised in a sheet forpurpo ses of stiffening (see Fig . 3). Pr ob ab ly due to breakage?at the ends of the ridges, these pressings offered themselvesas a convenien t solution. In oth er cases, similar ridges wen1made without adopting this artifice at the ends.In some cases, a small sectional cha nn el longitudinalpassed through a larger dimensioned hoop, also of circularchannel section, angle pressings being riveted to the mainmem bers at th e poin t of crossing (Fig . 4). In o ther case*.small chan nels, actin g as lo ngitu dinals, were attached tolarger channels by means of fish plates and small angin-as in Fig. 5. Again , in the Caudro n C.180, th e skin reinforcement was carried out by means of a number of channelsecured together, a nd t o th e skin, in the m anne r shown 'Fig. 6, while some of the Hanriot coverings were stiffen^by means of corrugations in the sheets, reinforced *'flanged circular channels riveted to the sheets at right'angles to the corrugations, as shown in Fig. 7. There i-obviously scope for considerable ingenuity in this direction-In o ne case, I believe on one of th e Latecoere exhibits- ^transverse formers appeared to be left out at the tailthe machine. The bunching together of the longitude* *may have been sufficient for vertical stiffness in this case.

96 6

7/27/2019 The Aircraft Engineer January 30, 1931

3/4

JANUARY 30, 1931 THE AIRCRAFT ENGINEER SUPPLEMENT TOFLI GHTIn considering curved bodies, the choice of a body havingis nearly a developable surface as possible is an obvious andfundamental adva ntage . There appeared to be very littlepanel beating on the bodies show n. The largemonocoques

were nearly all slab-sided ; the probable loss in body drag,due to this, must be considerable. Am ends were mad e, inmost cases, by fairings of wheels, tail skid, &c.Internal observations from the cockpits enabled one toobtain a good idea of the details of construction in manyinstances, but details of the internal construction of thewings were not so readily obtain ed. Only the Junke rs firmhad a stripped all-metal wing, and the details of constructionof this, which were of special interest, are dealt with later.Generally, the wings appeared to be of two-spar construc-tion, although the existence of two e xtern al lift m embersis not necessarily conclusive evidence th at there are onlytwo spars in the wing. The princip al difference in the wingsexhibited was the use and disuse of external corrugationsor other mea ns of reinforcem ent.Why has Wibault turned all the seams inside the wing,leaving a fiat outer surface ? Certainly no t for th e sa keof appearance, because th at is definitely w orse tha n onprevious exh ibits. N ot th at t he exis tence of sma ll localbuckles can affect the aerodynamic efficiency of a wing toany very great ext ent. The sta tem ent ha s definitely beenmade th at ex tern al ridges do affect th e airflow adve rselyand appreciably; on the other h and , experime nts made inAmerica with vertical fins placed on the wings near thetips have indicated an a erodynam ic ad van tage , a lthoughnot sufficient t o m ake th e u se of su ch fins w orth while.Of the metal-covered w ings show n, abou t half w ere plainand half corrugate d or ridge d. Th e use of very s hallowcorrugations can only be for appearance in the matter ofremoving or concealing local buckles in the shee ts. The

romigations in the Amiot wing covering, which mightliave been & in. deep an d -j in. pitc h d id no t, how ever,affect this purpose, as numerous shallow depressions wereto be seen on th e surface.Obviously, metal covering for wings has its advantagesand disadvantages, compared w ith fabric covering, a ndunless the former outweigh the latter, its use cannot endure.We have often pointe d ou t in the se pages th at unlessrigid covering can be used st ruc tura lly i n ad ditio n to itsfunction as a covering, it must compare most unfavourablywith fabric. The cost of the m ateria l, labour ch arges, etc.,certainly enter into the arguments, and Continental pricesfor the necessary m aterials, etc., ma y have an im port antbearing on the final choice of mate rial used th ere . A pa rtfrom such considerations, it is essential, from the point of

view of weight economy, that rigid covering should contri-bute something towards the structural strength of theaircraft. There are two ways of doing this : one way isto utilize the sheet coverings chiefly as a main load carryingmember, as is done in some types of Rohrbach wing. Tha tis , the coverings carry compressive and tensile forces directly,in addition to supplying the main torsional resistance of thewing. This method usually results in a heavy structure.The other method is to utilise the covering as a stabiliserand stiffener for th e intern al girder system ; economicaluse of rigid covering can be made in this way. This lat tersystem usually leads to a multi-spar structure.In concluding these remarks on rigid covering, we wouldsay that it does appear a desirable thing to explore fullythe relative merits of fuselages consisting of an inner, load-carrying structure, attached to which is an outer fairing,compared with bodies in which the two separate parts arecombined in one shell. Excellen t shapes can be obtain edby the first method, the manufacturing difficulties in theattainment of equally good shapes by the latter method areconsiderable. The former metho d lends itself to accu ratestrength computations ; the latter method does not.Once a suitable monocoque body has been developed,manufacturing costs should not be greater than for themanufacture of faired structures.The monveoque body gives more room for equipment,such as wireless, armament, etc., and the facilities for fittingthis should not present any difficulties. Greater damage ,without undue danger of collapse, may be permissible in a" shell " body tha n in a " frame " type of body, an d rep airsmay be more readily effected.As to weight, it would be a difficult task to make themonocoque body lighter tha n the faired rectangular frame.The shell type of body gives the impression of great rigiditycompared with the frame type, and incidentally one wonderswhether this has anything to do with the change over tomoiwcoquesin France, for in previous Shows the strut-and-wire type did not appear to be too strong torsionally.In one respect, the Show was very disappointing, and thatwas in the num ber of details of construc tion on view. TheJunkers Junior had the upper covering removed from thestarboard wing, together with one-half of the body covering.A Breguet was shown in the stripped state, and that wasthe lot, with the exception of the English exhibits. TheBristol Aeroplane Co., Ltd., Messrs. Armstrong Siddeley,and Messrs. Boulton and Paul definitely exposed to viewall details of construction, and in the former cases, wherecomplete aircraft were exhibited, the utility of the aircraftcould be readily estimated bv reason of the fact that the

7/27/2019 The Aircraft Engineer January 30, 1931

4/4

SUPPLEMENT TOFLIGHT JANU ARY 30, 1931THE AIRCRAFT ENGINEER

quantity, kind and accessibility of equipment could beclosely scrutinised.When these shows have greater commercial value, suchdisplay of the internal parts of aircraft will, of course, beessential for large sales, and although at this Show theconstructors mentioned had their products pushed well outof the way, if th e num ber of technical enquiries is any criterion,that factor did not affect the amount of interest displayed inthem.Needless to say, of the Continental exhibits, steel aircraftwere practically non -exi sten t: the Breguet was the, soleexception, although the C.A.M.S. had a steel spar on view(drawings of the spar sections are shown in Figs. 8 and 9,and another Continental effort at a metal spar is shown inFig. 10).The reinforcement of flat surfaces against secondarybuckling, as by means of channels riveted on is quite effec-t ive, but corrugation is a lighter and cheaper means of attain -ing the same end. (To he concluded.)

TH E PROGRESS OF AERO ENGINE DESIGN AT THEPARIS SHOWBy CAPTAI N ANDREW SWAN, B . SC, A.M.Inst.C.E.

Captain Andrew Swan, as many of our readers will probablyknow , is one of the Principal Technical Officers at the RoyalAircraft Establishmen t, Farnbo rough, where he is head of theEngine Department and in charge of the development andresearch work on engines. We count ourselves extremelyfortunate, in having been able to persuade Captain Swan togive us his impressions of the progress of aero-en gine designas exemp lified by the engine's exhibited at the Paris AeroShow which closed a few weeks ago.There was a large display of engines at the 12th Aero Showheld in Paris, particularly by French engine builders who, inmany cases, were showing examples of all their models.British constructors were well represented by ArmstrongSiddeley, Bristol and Rolls Royce engines of the latest type,whilst Alfa Romeo, Fiat and Isotta-Fraschini (Italy), Argus,B.M.W. and Siemens (Germany), Elizalde (Spain) and Walter(Czecko Slovakia), were showing a fairly full complement ofthei r products. No American engines were on exhibitionother tha n the Horn et and Wasp radial engines m anufacturedon licence by Siemens.Recent prog ress in engine design as typified by th e enginesexhibited is definitely along orthodox and well recognisedlines, and any changes noted are more those of detailedimprovement in the technique of design and construction orre-arrangement of accessories than actual departures of anyimpo rtance from present principles and practice. By this itis in nowise intended to decry the progress shown, as it isfully acknowledged that the perfection of even the smallest

detail, whether it be a point in design, or the adoption of amore suitable m aterial, is of imp ortan ce, an d it is the combina-tion of perfect detail from design to final construction whichdistinguishes the successful engine from its apparently similarbrother.In another category however, there are the compressionignition engines which are now making their appearance,namely, the Clerget nine-cylinder radial air-cooled enginesof 100 and 200 h.p., and the Junkers 600 h.p. water-cooledsix-cylinder opposed piston engine, of which some detailswill be given later.It is proposed in the first instance to make a general surveyof th e engine types sho wn an d to ind icate, with some criticism,the present trend of development, leaving to the end anymention of details worth y of note.Air-cooled and water-cooled engines were about equallyrepresented. Some mak ers, perha ps uncertain as to thefuture t ren d in thi s resp ect, make both typ es. Air-cooledengines, with few excep tions, were of th e rad ial ty pe , the single

The Iso tta-F rasc hin i " Asso Caccia " is an air-cooled Veetype of 450 h.p.row of cylinders being used for small and medium horse-pow ers, and th e dou ble row for th e larger horse-po wers. Twoexceptions were the Asso 80 R six-cylinder in-line enginedeveloping 100 h.p. and the twelve-cylinder Asso CaeoiaV-type developing 450 h.p., both made by Isotta Frascbim.who also make excellent water-cooled engines. Th e coolingair in the latter engine enters between the Vee of the cylinderblocks, and is divided and directed by baffles to pass acrosseach cylin der. Th is engine is no t a new typ e, but as fai as iskno wn ha s no t been pu t into series pro duc tion . The difficul-ties of adequately cooling the large cylinders of these engineby an air stream collected over a relatively small area an1 veryconsiderable, and it would seem necessary to provide a greaterfta*v of air by fan or ot he r me an s. Th e air-coole d vee engineof 400-500 h .p. is attrac tive in th at low frontal resiancewould be obtained in addition to the elimination of the water-cooling system, but as forced or induced air cooling wouldappear to be necessary, much of the simplicity of the s.V9teB1would unfortunately be lost.Another exception of special note was a small four-cy!in"er100-h.p. engine by t he firm of Chaise : In ste ad of the cylinder*being in line, each pair, that is the first and third and tnf