The Aircraft Engineer 29 March 1928

7/27/2019 The Aircraft Engineer 29 March 1928

1/8

March 29, 1928 Supplement to FLIGHT

FLIGHTENGINEERINGSECTIONEdited by C. M. POULSEN

March 29, 1928CONTENTS

M e t a l C o n s t r u c t i o n D e v e l o p m e n t . By H. T- P o l l a r d , W h . E x . ,A . F . R . A e . S ... ... ... ' ... ',.'

S e a p l a n e S t a b i l i t y C a l c u l a t i o n s . By W i l l i a m MunniT e c h n i c a l L i t e r a t u r e ... ... ... ... ... ... ...

EDITORIAL VIEWSIn his previous article on " Metal Construction Develop-ment , " Mr.Pollard dealt with the subject of fuselage membersmade from strip material . In the present issue he turns tothe subject of wing spars made from strip, and gives illus-trations of various types, such as the Boulton and PaulArmstrong-Whitworth, and " Bristol."We had rather expected that some of our readers wouldhave offered certain mild criticism of one or two things inMr. Pollard's article in the February number of T H E A I R -

C R AFT E NGINE E R , which appear to be a little open to doubt,or at any ra te to discussion. As no one has done so, perhapswe may be permitted to call attention to these points, sotha t , if t hey are unimportant we may be informed to tha teffect, while if they are of importance it is essential thatthey should not be overlooked.In Table 1 on p. 14 of the February 23 issue. Mr. Pollardwave, in column 9. the thickness of material for longeronsand st ruts of a strip frame, some of the latter being only0-006 in. thick. We have not. perhaps, the same accessto materials specifications as have aircraft firms, but to usit seems that such strip might easily vary at least 0-001 in.either way. If it does, it would appear that the strengthof the member made of such thin stuff might vary by.perhaps. 30 per cent., quite apart from any question of theeffect of corrosion on material of this "' fag paper thickness."It would, we think, be of interest if Mr. Pollard would informus whether his experience has indicated that any risk of thisnature is likely to be met with in practice.

The second point to which we should like to refer is theload of 800 lb. assumed to be applied at 0 in Fig. 1, on p. 14.On p. 17 Mr. Pollard states that the dimensions and loadsare such as might apply to the structure of a machine weigh-ing 4.500 lb. gross weight. In view of the fact that in mostmachines of normal proportions the tail skid load is some-where in the neighbourhood of 10 per cent, of the weight,is not 800 lb. too small a load to assume ? In other wordsis not the factor of safety too small ?

METAL CONSTRUCTION DEVELOPMENTBy H. .].POLLARD. Wh.Ex.. A.F., R.Ae.S.

(Continued front Page 17.)Further questions of detail design may now be considered.We will consequently turn our attention to wing spars.The reader will probably be aware that the design of parts ofaeroplane structures built up from metal strip began to bevery seriously considered late in the year 1918. when the

rapidly dwindling supplies of long lengths of timber suitablefor spars made it urgently necessary that these membersshould, if possible, be made from other materials. With theend of the war the intensive study of the possibilities ceased,but sufficient work had been done to prove that metal sparssuitable for normal single-seater and two-seater machinescould bemade which would show subs tantial savings in weightover wooden spars of the same stre ngt h. Very efficientshallow spars. 2 to 4 ins. deep, were than made and tested,bu t the early promise of rapid development did not materialise.Sufficient data for the successful design of deeper spars wasnot available and the problem of making metal ribs of equalweight and stiffness compared with timber ribs was foundmost difficult to solve. Speaking generally, it is not difficultfor a spar designer to produce a section from 2 to 5 ins. deep,giving a stress of 60 to 70 tons per sq. in., but between, say,5 and 12 ins. this result is not easy of attainment. Above12 ins. the ordinary box spar, and probably the single webspar, is scarcely practicable, and n girder arrangement islikely to prove more efficient. In particular cases, by theintroduction of numerous and varied stabilising members,exceptions could be made to the above general statements,but special cases of spar design need not at the momentconcern us.

In the design office well supplied with drawings of sections*of spars and with information regarding the stresses that aredeveloped by various kinds of loading, the task of selectingan old or designing a new spar is simple, providing no radicaldepartures are made from previous practice.The purpose of this article is to bring to the notice of thetechnician who is inexperienced in these matters, considera-tions which should help him to succeed in designing spars orother structural members in thin metal. An investigation ofa simple concrete example brings out some of the chief points,bu t we will in the first instance review two important casesof the effects of stress which have been investigated mathe-matically.In his treatise on Mathematical Theory of ElasticityProf. A. E. H. Love studies the case of a rectangular plate212a


2/8

Bxreruatsm TOFLIGHT22

MARCH 29, 1928THE AIRCRAFT ENGINEERhaving sides of length 2a and 26, and thickness 2h, securedat the edges and acted on by edge thrust Pj per unit lengthin^the direction of side of length 2a,and P2 per unit length onthe side of length 2b.

The conditions of loading and the dimensions of the plateare shown in the following diagram.

P, PERUN IT OF LENGTH

P2 PERUNIT OFLENGTH

The form of the solution isnnz(xw = W sin

provided that 2aa) . nr.(y ^ b)-Sm~2b~- (1 )

where w is the displacement of a point on the plate at(.r(/)from the originW is a constant.

m and n integers (giving the number of corrugations or" waves " parallel to the sides of the plate)and D = - - - this term is called byProf. Love the " flexuralo 1 >) *> 0"0h ,, ,,It has consistently been found that the ratios of t/R and

90

80

70

SO*DWI5"10

JO

20

10

tR

001 -002 003 -01";

>

STRIP THICKNESSRADIUS OF TUBE

0+ -005 -006 0(

RELATION BETWEEN CRITIC/!INTENSITY OFSTRESS & RATKOF SEMI THICKNESS TOPLATWIDTH FOB FLAT PLATE UND

/y

\y*

01 015hI

y/

/

/

d.EER _

rf

-

/ t

A

)7 -00S 009 -0

\

-J-

/

t1

/

SIMILAR RATIO FOR TUBE.ABSCISSAE BEING WALTHICKNESS TORADIUSRATIO

02 03STRIP SEMI THICKNESS

BREADTH OF STRIP04 (

3

5

Graph 2.2126


3/8

M A B C H 29, 1928 23

THE AIRCRAFT ENGINEER SUPPLEMENT TOFLIGHT are too low, the theoretical value of t/R for tubes showinga much greater discrepancy with experiment than does thevalue of h/B for flat plates.

For the most carefully prepared and conducted tests onthe most carefully made tubes the writer would expect tofind increases of at least 50 per cent, in the value of f/Rover the theoretical values for the same stresses.Unfortunately, similar mathematically established formulashave not been established for complete corrugated sections, butwith the following notes and the above equations as a guidethe trained technician should not experience great difficultyin inventing his own formulae for finding at what stressesthe sections he designs will become elastically unstable,providing he has the opportunity of observing the behaviourof actual spars under load and has also the data to enablehim to compare the results of a fair number of tests. Without

of a section receives from an adjoining section, the numericalvalue of which is dependent almost entirely on experience.It is obvious, then, why it is inadvisable to attempt to publishformulae at the present time for general use on built-upstruc ture members. Such formulae as have been deducedmay be established ma thematically a t some time in the future ;then th at will be the occasion for publication . Failingmath ema tical suppo rt, it is probable tha t formula* for p ar-ticular types of spar only will be available, the limitationsof each type being strictly defined.Empirical formula? containing terms whose values inparticular cases are dependent purely on the designer'sexperience are scarcely suitable for publication, since theprobability of misapplication by the inexperienced is toogreat.It should, finally, be noted that, whatever expression isderived for determining the stress at which instability sets in,it must be capable of application to a complete section or

-13-25

W

28

FIG. 2

FIG. 3THIC KNE SS OF MAT ERIAL IN EACH CASE OIS

DEVELOPED WIDTH 2 014RADIUS Of GYRATION 0786 1123ABOUT AXIS TH ROC .G.& PARALLEL TO FLATS

1774

the opportunity of actual observation any formulae he maydeduce would probably have little value.In this summary of the factors that must have their placein the desired formulae, obviously one of the most importantwill be K , th e rad ius of gj-ration of the section. Com ponentcurves (arcs of circles are usually taken) of the cross-sectionmust have a radius to thickness ratio proportional to thestress tha t is to be developed. The desired general largeradius of gyration m ust not be obtained by the use of extendedflats. The am ount tha t can be allowed depends on theboundary conditions, but the ratio of flat width to thicknessma y rarely be perm itted to exceed 20. Equa tion (4) aboveshows that the breadth of the section must be included, andaho th e thickness, as shown in equations (4) and (5). A termor terms involving the size and number of corrugations andtheir relative position must also be included, and a veryimportant factor relates to the degree of fixation one portion212c

to any portion of a section, however small, because theinstability may concern a section as a whole, as, for instance,a complete spar flange or a portion of such a flange ; in thelatter case, for example, a portion of a flange might consistof an arc having an excessive R/ t ratio at an early load,a dent or short-pitched wave might form along this part,which would subsequently develop and spread, causingthe prem ature collapse of an otherw ise perfectly sta blesection.Elasticians will doubtless solve these problems in timein spite of their complexity, but, as in the caae of the flatplate and tube formulae, it is quite certain that the resultsof the mathe matic al investigations will need considerablemodification on being brought in to line with practice ; henceis emphasised the need for large numbers of carefullyconducted, recorded, and co-ordinated experiments.Consideration of a selected few of the large number which

E 2


4/8

24TO MAKCH 29, 1928

FLIGHT THE AIRCRAFT ENGINEERhave been carried out for the Bristol Aeroplane Co. willfurnish further useful information.Fig . 1 is a section of a spar used ex tensively in " Bristoldesigns . A series of tests have been mad e with this spar invarious gauges : the dimensions of the test spar centres areshown in Fig. 2, and the ratio of end load to lateral load wasP = 24 W.

P being the thrust and W half the lateral load.In Grap h 2 the relation between the stress developed and thegauge of the flange is shown.The results of three other experiments relating to flangesnumbered 3a, 36 and 3c are shown, but in this case stressesdeveloped are plotted against depth of section (see Graph 2).As an exercise the student of the problem should try to forman expression from the data given which, when evaluated,gives approximately the stress developed on the tests.Flange 3a was not designed for use in a spar, but for acompression rib ; it was convenient, however, on one occasion

Tak e the bending moment across this section to be 31inch /tons. Then, assuming the web section to be adequatelysupported (by ribs or other meansa matter that will bebriefly mentioned later), the section A should withsta nd themaximu m stress of approxim ately 65 tons/s q. in. thatwould be induced. Portion B has to withs tand an averagestress of about 40 tons per square inch, and consequently thethickness would be much less, as shown, while the stress acrossthe web varies from a small negative to a small positiveamount, and consequently the web thickness could have avalue as shown : as dimensioned this web would probab lywithstand a compressive stress of about 22 tons per square inchor a thrust of 1.600 lbs.The student should work out many such examples forhimself, tak ing different eases of bending m om ents, deflec-tions, and end loads. The manuf actur ing side of the questionshould be kept in view ; for instanc e, it would be foolishat this stage to call for strip tapering in thickness from, say,

to use the section in a spar. This particular spar was thebottom rear member in the metal wings of the Bristol Luciferpreliminary training machine.The stress in an ideally designed spar would be constantalong the length and across the breadth of the section at allpoints, and for the material used would be the maximumpermissible for the max imu m loading. This sta te, of course,is impossible of attainment, usually load factors at points ofsupport and points of maximum bending between supportsreach the lowest permissible figure. In spa rs ma de of timb er,but particularly in those made from oval tube, there is littlescope for making variations in thickness of material, pafticu-larlv across sections where the stress variation is generallyvery wide.From Fig. 4 the ease with which variations can be madewith strip spars is apparent.I of the section = 1 2 "")Z , , = 0-48 >app rox.A = 0-34 J

0-018 to 0-01 in., but longitudinal variations in metaldisposition are very easily madefor example, as in the caseof Fig. 4. If part A were subjected to an extra compressivestress along a portion of its length, an extra lamination orlaminations could readily be slipped over and secured atC and C, nor would it be necessary to secure the flangestogether on their curved surfaces, securing of the flats beingsufficient.It is often convenient in the case of cantilever spars to havea number of laminations at the root, falling off to a singlethickness a t the tip , thus securing a fairly even d istribution ofstress. (For practical purposes it is not necessary to have theshape of the outside of one lamination exactly the same asthat of the inside of another, that is to say, a single set of toolswill often serve for all the laminations.)Because of the ease with which high stresses can bedeveloped in shallow members it often happens that a goodspar reckoned on a strength/weight basis can be designedwhich only occupies a portion of the depth of the aerofoil,

2\2d


5/8

MABOH 29, 1928TO

say, 55 or 60 per cent,, and w hich also may have the advantageof enabling external bracing fittings to be totally enclosed inthe aerofoil. Wh ile this latte r is very desirable it is betterto obtain this by other methods than cutting down the spardepth, as the shallower the spar the greater the deflection.The sight of a plane flexing up and down when a machine istaxying does not inspire the pilot with confidence, eventhou gh th e wing might be quite, stead y in the air. Fulladvantage should be taken of the aerofoil depth, and for thisreason Fig. 4 was chosen as an example of a type speciallysuited for deep aerofoils. This type of spar requires variousweb stiffeners, an d it is the writer's opinion th at development ofhigh stresses in it is a matter of more difficulty than in the boxspar type, but with the present trend of aeroplane design inthe direction of fewer external struts, it is a type which willin the end pay for the labour that is put into its investigationand development .

A com parison of a few cu rren t typ es of spar will give someinformation on matters of design.In Fig s. 5, 6 and 7 are shown typ es of spars developed byMessrs. Boulton and Paul, Armstrong Whitworth, and theBris tol Aero plane Comp any. The first and second are copiedfrom th e " Encyclopaedia Bri tann ica." Two very specialfeatures of the Bou lton and P aul spar are (1) the very neatarrangement at the riveting edges, where the 180 web bendmakes t he assembly of these spars previous to riveting a verysimple matte r, and (2) the very excellent web supp ort. TheArmstrong spar is particularly noteworthy because of theentire absence of flats, the flanges and webs being whollycurved. A special feature of the " Bristo l" spars is the ratherwide flat riveting edges, which enables the external fittingsto be secured after the long box has been riveted up along itsfour edges, as shown in Fig. 8. These fittings take the formof shallow channels secured to the spar edges after assembly.making possible a continuous process of assembly, in whichholes may be punched and rivets clinched simultaneouslyunder a gang press. Rivets are at the moment put in byhand between each stroke of the press, but the possibility ofan entirely automatic assembly process in which the rivetsmay be fed into place mechanically are evident.

These wide flats might easily be a source of weakness inthe spar, but they are kept as far from the points of maximumstress as possible and a very substantial curl terminates theweb edges.On tests on short spars where the end load has greatlyexceeded the lateral load there has been a tendency for theseflats to fail before the yield point of the material has beenreached ; b ut the simple manufacturing processes perm ittedby the design has made it worth while using a heavier gaugeflange or web in the one or two isolated cases that haveoccurred.The question of " developed stresses " in spars is obviouslyof an extr em ely complicated na tu re. Formulae deduced fromthe Euler Bernoulli theory are used for the simple reason thatthere are no others to replace them ; they are, moreover, easyof app licatio n. Any one who has seen the formation of long-pitched waves in the compression flange of a spar or seen dentsdeveloping round rivets or observed the spar visually con-tracting in depth or spreading in breadth must have felt thatthe proces s of multiply ing the end load an d deflection,dividing by the section modulus, etc, was yielding a resulthaving li t t le bearing on the actual stress intensity in the sparunder load. It m ight be argued that the above process resultsin very approximate figures of comparison, but even in thiscase consideration would have to be given to the methods ofloading.

The student should be on his guard against being misled bywh at are po pula rly and erron eously called " figures of m erit."AN OMISSIONThe attention of readers is drawn to an omission, due toan over sight, of a radiu s dime nsion on F ig. 1 of Mr. Pollard sillustr ation s. The Figure will be found in the top left-handcorner of the set on p. 23, and the missing radius dimensionrefers to the radius arrow shown in the left-hand lobe of thetop spar flange. Rea ders desiring this figure to be completeare asked to insert at that point the dimension 0-05" O.R.

212*

SEAPLANE STABILITY CALCULATIONS.By WILL IAM MUJJIRO.

(Concluded from page 20.)5. Displacement to Load Water Line.

This is tabulated in similar fashion to the calculation oftotal displacement, but in this case the areas of half-sectionsentered in column 2 are taken below water-lines, as indicatedon body-plan Fig. 5. Thu s we get :Area ofStation. Half-section.121 U1110987i;

543o1i0

Displacement

0 - 00 - 00-040-601 1 11-381-541-701 -311-140-910-690-460-290 - 0

to L.W.L.

Simpsons'Multiplier.A2

H424i4o4o4Ho

i

33 11 1-725~ 1 3

Funct ionsof Areas.0 - 00 00 0 62-402-225-523-086-802-624-561-822-760-690-580 0

33-112 64

A ~ X -2,427 lbs .The displacement to load-water line is the total function ofareas multiplied by one-third the interv (al between stationsand mu ltiplied by two for both sides of the float : then brou ghtto pounds by multiplying by 64.(j. Position of Centre of Buoyancy.This has to be found(a ) vertically :(b ) fore and aft.(a ) Is not usually worked ou t in detail , but is taken approxi -mately as a known ratio between the distance of the C.B.below the load water l ine and the maximum draught of thefloat to the same waterline.This ratio is given by Major R. E. P enny, A.F.R.A e.S.,as 0-2 of the draught.*If calculated, the method employed would be to make outa " Curve of Displacement " at varying water-lines, and thenapply the same rule as used in ship-design, i.e..

Area of Curve of displaceme ntLoad displacement Dis tanc e of C.K. below

L.W.L.' Sec paper on "Sea plan e D evelopment " hv Major R. K. Penny.A.l'.K.Ac.S., read before lioyal Aeronautical Society, April 28th, 1927.

Fig. 7.


6/8

S C P F L K M B N T T OFLIGHT26

THE AIRCRAFT ENGINEER MAR C H 2 9 , 1 9 28Each displacement at the different water-linescalculated as for displacement to load water line.(b ) Fore and aft position of centre of buoyancy.To obtain this pos i t ion we tabulate as under :

Half-area Simpsons" Functionsof Section Mu ltiplier of Areas

w oul d be

S ta t ion .121 H1110987

0 00 - 00 040-601 1 11-381-54

543o

0

1-7(11-311 1 40-910 46

0 ( 1

H4242

2424

0-00 -0

-06-40

Lever6

Mom ents

0-2- 543

0-0-0-

az086 80

003011-606 6611 043-OS3( I 6 *

2-624 561-822- 760 600 58

1>3455 A

2-0211-12r> 4K1 1 - 0 43 - 4 53 - l i )

0 - 0 34 SS33 11

E x c e s s a f t = 3 4 - 8 8 3 0 - 6 8 - ^ 4 - 2 .Ce nt r e of buo va nr v a f t of s t a t ion (i 4 -2 1 -723 5 M

ins. (0-208 ft.) .The m idship statio n. N o. (i. of th e f loat is taken as th es tar t i ng p oint : the m oments forward of No. H s ta t ion areadded together : similarly, the moments aft of No. ( i stationare added together and the excess of moments aft is multipliedby the interval between s ta t ions and divided by the tota l offunctions of areas.We now have a l l the information necessary to determinewha t tran sver se me tacentr ic height our f lotation system willp r ov ide .

The formula used is B.M. = ~!where B.M. is distance from Metacentre M. to centre ofbuoyancy B, and l 0 is moment of inertia about centre line ofmach ine for two floats, and V is tota l disp laceme nt of twofloats (in cubic feet).This is shown b y referring to Fig . 7. a single f loat being usedas an i l lus tra t ion.Let G the angle of inclination be verv small.Y = W.S. or S .L.l\ area of wedge W.S.W j. . or L.S .L j .v = volume of W.S.W. for whole length of f loat.V = volume of displacement .

r ' : but LL , - S.L. Sin G practically)S.L.2 sin 0

Then r , =

But S/ij = 2/3 S.L. by properties of tr iangle..- . k Aj = 4/3 S.L. when 6 is very sm all.N o w v, / h hL=4 3 S . L . , I S . L . : s in= 2 3 sin 0 , S.L.:l

. - . v < h h,= 2 3 Y :; s in G dx .B u t > h hx-= X B . R . , a nd B . R . = B . M . si n G ;

. . V B.M . s in G = 2/3 j Y :! s in G dx ;

From Fig. 8 take a s t r ip P .Q. of length Y and breadth{indefinitely small) dx . Then, regarding P .Q. as a rectangleits moment of inertia about the base DC is :l/3(Y-rfar) Y* --- 1/3 Yda:(Ydx Ar ea)

Fig. 8.and the moment of inertia of the whole f igure about DC willbe the sum of all such expressions, giving us :

3 Y :i d.v.Both side-! of the fl:>at being sym me trica l we get :

Moment of inertia 1 --- 2/3 I X :i d.v.(Y = Semi-Ord. of L.W. Plane.) "Y :W.cAnd as B.M. is shown above to equal 2 3 ,

. . B.M. =- I .And as the f loats are situated at a distan ce from th e centr eline of mach ine equal to half the track : then i0 . the tota lmoment of inertia is equal to I (Area of W.P. I t rack- )as previously calculated.A nd

B.M. = II . .8 56"8

V =V

242775 84 = 22-5 f t .

-^ 75-8464The required Transverse Metacentr ic Height is thereforeB . M . - B . G .The position of G is known and B.G. we will assume is= 8'0".. . T ransv erse M etacentric ' Heig ht or G.M..

= 22- 5 f t . - 8 - 0 f t .= 14-5 ft.

Should the result arr ived at not lie in the locus of resultsgraphed for similar types of seaplanes, the easiest remedy i*to widen the f loat t rack commensurate ly.An example will make this clearer .Taking the f igures used throughout we will assume thatthe amount of metacentric height is not sufficient and thatG.M. should be 15' 0".Th e new B .M. therefo re will be 8' 0" -r- 15 ' ()" = 23 ' 0".

And as B.M. -- rT. . B.M. > V = I o. . 23 / 75-84 = I o for both floats

23 x 75-84 1 r. . - l0 tor one f loat872-16 =_ I o for one float.

And as Io -.-- I + (Ar ea of W .P . tr ac k2). . Area of W .P. x | track* - I 0 - l. . 4 1-2 5 X \ t rack- - 872-16 21-54

872 16 21-54. . 1 trac k- = T^ Z = 20-62 f t.i track = v 7 ^

. . Track required^ - 4-54 f t.

= 2 v 4-54 f t.= 9-08 f t.

212/"


7/8

MAKCH 29, 1928

THE AIRCRAFT ENGINEER SUPPLEMENT TOFLIGHTTurning now to the Longitudinal Metacentric Height, wefirst find the Moment of Inertia of the Waterplane" aboutthe midship station, No. 6 as follows :

03 on c1 2

121H111087

00-51-01-341-401-421-42

6 1-3354321

25120-960-710-460-33

S 33ifS

2H4247

2424H2

. 2 = 1104-6.

Longl. M.I.about C'r. of Flotation= 1104-6 - (41-25 . 0-982) = 1065.

Transverse M.I.1725 2= 56-2 X x - = 21-54.o >

Transverse Metacentre_ (21-54 > 2) -r (41-25 v 4-5

2; 2)_ 75.84

=- 22-5 ft,Longl. Metacentre

_ 2(1104-6 (41-25 x 0-982)}56-197 75-84 = 28-0 feet.

212^


8/8

S U P P L E M E N T TOF L I G H T28 M A R C H 29, 128THE AIR C R AFT ENGINEER

B ut v X ggx = V x BB, -= V X B.M. tan 0 (practically) ;. . V x B.M. tan 0 = 2 Yx- dx tan 6.

'" x~udx. . B.M. = 2 | - -T h e n , as 2

T h e B r i s t o l F i g h t e r is to be s e n t to a s e r v i c e s qua d r onf o r e x t e nde d t r i a l s .P r e v i ous pub l i s he d r e po r t s on the s l o t and a i l e r on c on t r o la re as follows :R . & M. 968. Ful l sca le tes ts of a new s l o t and ai le ron*' control. By H. L. S t e v e n s .R . & M. 1051. S e c o n d r e p o r t on fu l l - sca le exper ience wi thJC ?Y dx can be s hown e qua l to the m o m e n t of t h e s i o t a n d a i i e r o n control fi t ted to a Br i s t o l F i gh t e r . Byi ne r t i a a bou t the c e n t r e of n o t a t i o n , we h a v e :

. . l o n g i t u d i n a l G.M. =. B.M. B.G.= 28-0 8-0 =- 20- 0 ft.Th e fore and aft s t a b i l i t y of mos t s e a p l a ne s is us ua l l y f u r t he rc he c ke d by c a l c u l a t i ng the l ong i t ud i na l G.M. (or m e t a e e n t r i cheight ) wi th the ma c h i ne t i l t e d ba c k at an angle of a b o u t8 .T h e m e t h o d a d o p t e d is e x a c t l y as de s c r i be d a l r e a dy but.of course , involves a new set of c a l c u l a t i ons for :(1 ) The new d i s p l a c e m e n t to the 8 w a t e r l i n e .(2 ) The new posi t ion of c e n t r e of b u o y a n c y .(3 ) The new a r e a of w a t e r p l a n e .(4 ) The new posi t ion of c e n t r e of n o t a t i o n .(5 ) The new m o m e n t of i ne r t i a of w a t e r p l a n e . a b o u t thene w c e n t r e of flotation.T he d i s t a nc e B.M. in this case will be, of course , much lesst h a n w h e n the ma c h i ne is at n o r m a l t r i m .I n c onc l ud i ng t he s e no t e s on the me t hods u s e d to d e t e r m i n et he s t a t i c a l s t a b i l i t y , a c o m p a c t t a b u l a r s u m m a r y is givenwh i c h is in ge ne r a l use, and is of c ou r se muc h ha n d i e r whe n onehas c lear ly in m i n d how to use S i mps on ' s mu l t i p l i e r s for eachs e pa r a t e c a l c u l a t i on .It will be s e e n t ha t in th is table , ca lcula t ions(2) Area of l oa d wa t e r p l a ne :(3) Cent re of flotation ;( 4 ) M ome n t of i ne r t i a of l oa d wa t e r p l a ne ;a r e wor ke d ou t side by s i de .1. T o t a l d i s p l a c e me n t of float is, of c ou r s e , known as soona s th e r e s e r ve buoya nc y is de c i de d on, but m u s t be c he c ke do u t on the float lines in the m a n n e r s h o w n .5. The d i s p l a c e m e n t to l oa d wa t e r p l a ne is s i mi l a r l y knownas soon as the we i gh t of m a c h i n e is d e t e r m i n e d , but m u s talso be t r i e d out on the l ines to fix the load water l ine pos i t iona nd c ons e que n t l y c a l c u l a t i ons 2, 3 and 4.I t m u s t be a dde d t ha t dyna mi c a l s t a b i l i t y r e qu i r e s f u r t he rc a l c u l a t i ons , and t h a t no a c c o u n t is he r e t a ke n of c ha nge s inwe i gh t of m a c h i n e , say, in a bombe r , l oa de d and r id ing l ight ,or of c ha nge s in t r i m due to f ree water sur faces in l e a kyf loa ts or wi th com pa r tm en ts f looded.

TECHNICAL LITERATURE.S U M M A R I E S OF A E R O N A U T I C A L R E S E A R C H

C O M M I T T E E R E P O R T S .FULL SCALE TESTS OF A BRISTOL FIGHTER WITH SLOT

AND AILERON CONTROL OPERATED BY A DIFFE-RENTIAL LINK MECHANISM.

By H. M. G A R N E R , M.A. Presented by The Director ofScientific Research.R . & M. No. 1101 (Ae 279). (2 pages and 2 diagrams.)May, 1927. Price 4d. net.A defect of the slot and aileron controls used hitherto atthe Royal Aircraft Establishment has been a jerkiness inthe control at high speeds with the ailerons near the normalposition. It was thought that this was mainly due to thecam gear employed, and t ha t a differential link system wouldeliminate the defect.The link system was fitted to the upper wing of a Mark II IBristol Fighter, and the aeroplane was flown by a number ofpilots at the R . A . E .The jerkiness previously present in the control system wascompletely eliminated by the alteration in the system, whilethe effectiveness of the control at the stall was unaltered.

H . L. Stevens.R . & M. 1088. Preliminary report on the fitting of slots andflaps and slot and aileron control to a Bristol Fighter. ByH . L. Stevens.

AMERICAN NATIONAL ADVISORY COMMITTEE REPORT S.The National Advisory Committee for Aeronautics in theUnited States of America corresponds to our own Aero-nautical Research Committee. Twodistinct classes of reportsare issued, the first being know n as Technical Beports. TheseTechnical Reports are printed, and are i l lustrated by photo-graphs and/or drawings. The second class are known asTechnical Notes, and are issued in mimeographed form soas to enable them to be rapidly distributed to a somewhatsmaller, but directly interested, circle of readers. Copies ofthe Reports and Notes may be obtained from the Superin-tendent of Documents, Government Printing Office, Wash-ington, D.C., U.S .A.

AMERICAN NATIONAL ADVISORY COMMITTEETECHNICAL NOTES

The last list of these Notes waspublished inTHE AIRCRAFTENGINEER of March 31, 1927, to which readers are referredfor numbers previous to, and inchiding. No. 253.T.N. No. 254 :"M E T H O D OF C O R R E C T I N G W I N DT U N N E L D A T A FOR O M I T T E D P A R T S OF A I R P L A N EM O D E L S . "By R. H. Smith, Department of Construction and Repair .Washington Navy Yard. Thetitle of this Note isconsidered to be sufficient indicationof the subject-matter, and no summary will , therefore, begiven.T .N. 255 :"P R E C I S I O N OF W I N G S E C T I O N S ANDC O N S E Q U E N T A E R O D Y N A M I C E F F E C T S . "By Frank Rizzo, Langley Memorial Aeronautical Labora-tory.This investigation was carried out by the N.A.C.A. at theLangley Memorial Aeronautical Laboratory to determine theprecision of wing sections of wood-fabric construction usedon a number of aeroplanes. It was found that all wingsections deviated more or less from their respective prototypes.The mean thickness of the section was computed for thosewings with a noticeable sag. The aerodynamic effectsresulting from consideration of thickness variation are thenestimated from existing empirical information. The rib, sagand specified measurements of 14 sections investigated aregiven in Fig. 2.T.N. No.2 5 6: 'W A L L I N T E R F E R E N C E IN CLOSED-T Y P E W I N D T U N N E L S . " By George J. Higgins, Lang-ley Memorial Aeronautical Laboratory.A series of tests has been conducted by the N.A.C.A. inthe variable densitv wind tunnel on several aerofoils ofdifferent sizes and sections, to determine the effect of tunnelwall interference and to determine a correction which can beapplied to reduce the error caused thereby . The use of severalempirical corrections wasattempted with li t t le success. ThePrandtl theoretical correction gives the best results, and itsuse is recommended for correcting closed-type wind tunnelresults to conditions of free air.T . N . No. 257 :"T E C H N IC A L P R E P A R A T I O N OF THEA E R O P L A N E ' S P I R I T OF ST. L O U I S '. " Writ ten for theN.A.C.A. by Donald A. Hall, Chief Engineer, Ryan Airlines,Inc.In the main, this Note contains the same informationconcerning Col. Lindbergh's famous machine as t h a t pub-lished in FLIGHT of J u n e 9, 1927. Additional data are, how-ever, given.2\2h

The Aircraft Engineer 29 March 1928

Documents

Transcript of The Aircraft Engineer 29 March 1928