The Aircraft Engineer October 31, 1930

7/27/2019 The Aircraft Engineer October 31, 1930

1/6


2/6

October 31, 1930 Supplement to FLIGHT

ENGINEERINGSECTION

Edited by C. M. POULSENOctober 31, 1930

CONTENTSPAGE

Technical Features of theAirMail. ByFrank Radcliffe, B.Sc, A.R.Ae.S. 73Floats for Racing Craft. ByWilliam Munro, A.M.I.Ae.E 74In th eDrawing Office ... ... ... ... ... ... ... 76Technical Literature ... ... ... ... ... ... ... ... 77

TECHNICAL FEATURES OF THE AIR HAIL.By FRANK RADCLIFFE, B.SC., A.R.Ae.S.

(Continued from p. 70.)Now let us look at some of the requirements for a modernmail land 'plane. The following is suggested as being reason-able : * *Cruising speed = 150 m.p.h. .-. top speed = 178 m.p.h.Range in still air = 750miles.Pay load in the form of mail bags = ] 000 lb.The following indicates what would be essential as regards

equipment:Crew : two pilots, one to act as navigator, postal operator

and wireless operator.Wireless equipment (transmission and receiving).Navigation equipment.Night-flying apparatus.Postal apparatus for collecting and dropping mails.(The weight of all the above equipment could be assumedto be 800 lb.)It will be evident at once that our mail 'plane, whilst

having the performance of a modern day-bomber, will beof a very much bigger class, as the disposable load (ex fuel)will be approximately 1,800 lb. The actual size Mill be leftfor consideration in a subsequent article.(d) Aerodrome and Ground Equipment.If the technicaldevelopment of the aeroplane presents difficulties, then

the demands ol the air mail service from an operational pointof view seem overwhelming if 100 per cent, efficiency is to beobtained. For convenience, the problems will be mentionedin the following order:

(1) Meteorological.(2) Night flying,

and (3) " Blind " flying.rr.r (1) Meteorological

On our assumption that flying stages will be of a t leastfive hours' duration, it seems necessary that informationshould be available to the navigator at very frequent intervals

relating to the weather conditions. A fixed aerial appearsto have decided advantages over a trailing antenna inasmuchas it allows of continuous radio-communication when flyinglow. In the U.S. A. information on the weather is transmittedby the Chamber of Commerce at 15-minute intervals on along wave-length over sections of 150-200 miles' radius, andmany of the operating companies have a private networkof short-wave stations for carrying on a two-way communi-cation with their own aircraft.

(2) Night FlyingIn addition to the information regarding the weather,that wireless makes possible, the pilot needs guidance alonghis route when flying by night, and here the problems arisewhich still await solution. Aerial lighthouses and beaconsare suggested by some as a means of meeting the difficultiesof night flying, and these have been largely employed inAmerica. They are expensive to instal, and have the dis-advantages that fog and low clouds will obscure the lights,thus rendering the lights ineffective to the aircraft that flyabove the fog or clouds. It is of interest to note that theweather conditions, whilst being favourable to the lightsystem in the U.S.A. would be less so in Europe. Undoubtedlythere are many sections of our Empire routes that will needa lighting system, and, given the right meteorological con-ditions, the results should be very satisfactory.The aerodromes require suitable night landing equipment,and the following units appear necessary (5) :I. Location Beacon.Easily recognisable Neon tubesgiving out definite flashes.II . Landing Floodlight.For illuminating the surface ofthe aerodrome at night. As the pilot always desires to landagainst the wind, it is mounted and made as a mobile unit.The floodlight is required to have a large horizontal spreadwith a very small vertical divergence.III. Boundary Lights.Consistingof flashing red or amber

lights mounted on a low standard with a joint at the basefor safety in the event of an aircraft colliding with it .IV. Obstruction Lights.For denoting danger spots on ornear the aerodrome. The range of visibility of this class oflight, in normal conditions, is about 3 miles, and a portableacetylene red lamp is used for temporary obstructions ; whilstelectric lamps mark buildings and telegraph poles.V. Wind Indicators.Suitably illuminated, are necessaryfor guiding the pilot, and these take the form of a horizontalT.(VI) Ceiling Lights, which take the form of searchlights,are useful for finding the height of clouds. This information

1192a


3/6

74jStTPFLEMENT TO - - - jFLIGHT OCTOBER 31, 1930THE AIRCRAFT ENGINEER

-*.

can then be forwarded to meteorological stations for inclusionin the weather bulletin, thereby indicating the maximumheight at which flying must be carried out if the aerodromeor route lights are to be seen.(3) " Blind " Flying

This term is intended to cover flying in fog, in conditionsof poor visibili ty due to clouds, and along a night route notmarked by light beacons. This section of practical aero-nautical development is the one that is requiring the mostcareful investigation at the present time, for on its successfulaccom plishment depends very largely the future of commercialnight flying. Its problems are concerned with the devisingof instruments that will indicate immediately :

(i) The course an aircraft is taking relative to the earth;(ii) The a lti tude at which the aircraft is above theI ground ; and(iii) The at ti tu de of the aircraft relative to th e tr uehorizon.I. A few note s on the problem s will prob ably be of inte res t.(Th e enquiring re ade r is referred to recent issues of " AeroDigest " for information on wha t ha s been achieved inAmerica in this direction.)

(1) The present method of communicating direction withaircraft in this country, and in operation on the Europeanroutes is worked from ground stations by means of wirelesstelegraphy (6). There are at present three stations, viz.,Croydon, Lympne, and Pulham, and the resul ts during1929 have continued to be entirely satisfactory."' (2) At present no entirely satisfactory m ethod of determin"ing the alti tude of an aircraft above the ground exists, sothat a knowledge of the country over which flying is con-tem pla ted would be presum ed (or else, flying would need tobe carried out at high alti tudes ). There is need, therefore,for extensive research in this field.(3) The problems covered by this group have to deal with

the flying atti tude of the aircraft in three dimensions, andwhilst of supreme importance, finality has by no means beenreached ; in fact, their solution appe ars sti l l to be a longway off. :;RECA PITU LA TIO NLooking broadly at the various problems which arise inconnection with both the aircraft and their operation, i twou ld appear tha t :

(1) The solution of the problems de aling with th e designof the aircraft themselves are more advanced than thosedealing with the more difficult phases of flying, such as1* blind " flying.' (2) The future of comme rcial flying seems to be in th ehands of the research wireless engineers.

(3) Means of collecting or dropping mails whilst in flightwill create very great difficulties if 100 per cent, efficiency isdesired.(4) Speed alone will ha\ re to be the essence of the airmail service if i t is to be an everyd ay economic fact.(5) There will have to be a generous co-operation on thepart of all sections which make up an air mail service if pro-gress is to be accelerated to its fullest extent.

l:;*"^ r ; " (To he continual)Seferemvi: _ - r "', - . . . -(5) Information has been supplied by Chance Bros, and C o., Ltd.,Marine and Aerial Lighthouse Engineers, Birmingham.(6) Fuller details on these matters will be found in the following paj)ers:" Practical Navigation of Aircraft," Capt. F. Tym ms, M.C. E.Ao.S.Journal, May, 1925." Landing Aircraft in Fog ," Flight-Lieut. H. Cooch. R.Ae.8. Journal,June, 1926.

FLOATS FOR RACING CRAFTB y W I L L I A M M U N B O , A.M.I.Ae.B.

Ii may be recollected that some years ago Mr. Munro, whowas at that time on the Technical Staff of the Gloster AircraftCo., contributed some articles on seaplane stability calculations.He then went to the U nited States, where he took up the post aschief engineer to the Towle Aircraft Com pany, and while theredid a considerable amount of work on flying boats andamphibians. Mr. Munro has now returned to England andjoined the Technical Staff of the Supermarine Aviation Worksat Southampton.

In view of the fact that i t has been decided to build a newseaplane testing tank at an early date in U.S.A., and thatsimilar activity has alread}' commenced in Canada, the writerhopes the following notes may be of interest and some value.The subject of float or pontoon design may be consideredas having three m ain divisions :(1) Design suitable for commercial machines.(2) Design for fast ships of the racing type.(3) Design of wing-tip floats or sponsons.Having designed for any one of these types, a tank, testis the final criterion of the likelihood of success which justifiesactual building.In many cases, however, where a sufficiency of data isavailable regarding previous successful ships, tank tests aredispensed with by the manufacturer on the score of cost.It cannot be emphasised too highly that where any radicaldeparture from well-known type is contemplated, tank testsare not merely instructive but essential for clean running,good performance, absence of porpoising tendencies and actualsafety.It is proposed then to deal briefly with the characteristicsand construction of the flotation gear suitable for high-speedcraft, capable of travelling at speeds of over 300 m.p.h. withcorrespondingly high land ing and take-off speeds. These

notes will indicate the general trend of successful float designfor this particular type of craft, but as each particular shipproves in hard experience to have its own particular vices,it is emphasised again at the risk of being tedious that nomatter what data a designer may possess, a tank test is thesafest and often the cheapest proof of the correctness of hisdeductions. DesignIt will be readily agreed t ha t when building th e class ofseaplane under consideration, the ordinarily vital mattersof detail design suitable for production, such as the eliminationof panel b eating or " bum pin g," are of necessity forced totake a second place, and the question of design approachedfrom a muc h different viewpoint. Aerod ynam ic considera-tions may be said to outweigh the cost problem and refine-ment of line is even allowed to impair to some extent thehydrodynamic performance, particularly in regard to stepdepth . Eve ry effort is made t o reduce the step size to aminimum, and yet have a step depth suitable for the gettingoff speed of the machine.In certain tank tests made for floats for high-speed crafta dep th of ste p of 1 in. was tried ou t. The tes ts showedth at this de pth was insufficient. Howe ver, up to a certainpoint reduction of step depth is in l ine with the requirementsaimed at, as a relatively deep step is necessary for low gettingoff speeds and a shallower one for higher taking off speeds.In view of the small wing area of the typical racing machin


4/6

OCTOBER 31, 1930THE AIRCRAFT ENGINEER StTFFLEMENT TOFLIGHT

2 . Coefficient of Fineness of Water-plane, which is theratio of the water-plane area of the float to the rectanglewhich encloses it.This coefficient would be 0 6.3. Block Coefficient, which is the ratio of the volume ofthe float to the volume of a block having the same overalllength, same extreme breadth and same extreme depth.This coefficient would be 0-435.*

Reserve BuoyancyAs would be expected, the frontal area of floats for racingcraft is made as small as possible, and the amount of reservebuoyancy is cut considerably from the 90 per cent, to 100 percent, generally used for comm ercial craft. This figure maybe as low as 60 per cent. In prod uction float work the wall-sided float with cambered deck is making strong headwayand the essential difference in line for speed ships is bestillustrated b y the body plan shown in Fig. 1, and the stream-line ty pe of profile indi cated in Fig . 2 . These are n ot scaledrawings and it would be misleading to accept them as such,but they give a very good idea of the characteristics found

necessary in prac tice. As is well known to stude nts of floatdesign, the bow shape and angle of vee-bottom are of primaryimportance.Metacentric Height

Another important point in which the make of float underconsideration varies from the normal is the amount oftransverse metacentric height deemed necessary. The

G.M. or distance between centre of gravity of machine andthe metacentre in ordinary cases ranges in the neighbourhood3 _of 1-1 JW ft., where W equals displacement of shipin lb. As the position of the metacentre is determined b ythe ratio ^., where I equals the moment of inertia, andV equals the displacement, it will be clear that starting offwith a narrow beam pontoon for the reasons already given,and maintaining the usual ratio of lengths to breadth theneither the track between floats has to be excessive, resultingin a heavy undesirable undercarriage and increased resistance,or th e ratio will give a v aiue smaller th an usual and lessstatic stability in a transverse direction. The latter is thecondition generally accepted and the values of transverseand longitudinal metacentric height are approximatelyindicated in Fig. 5. The figures below show app roxim atelythe results which might be expected from wind-tunnel testsof clean floats suitable for racing seaplanes of about 3,000 lb.weight :

2-in.1-in.

StepDepthLiftat 100ft. /sec.

. 3-2. 3-38

Dragat 100ft. /sec.6 -05-42

Construction

Dragin lb./sq. ft.1- 51-31

DragCo-efficient0-060-056

See also " Seaplane Stability Calculations " by Win. Munro, February23 , 1928, in FLIGHT SrppLEMENT AIRCRAFT ESHINEEK. an d March 29, 1928.

The following notes on construction may be of interest.The pontoons are all-metalAlcladand finish givenby the anodic process. A centre line bulkhe ad s uit abl y

FEMALE DIE

MALE DIEFINISHED HOLE

PUNCH & DIE FOR FLANGEDLIGHTNING HOLESFIG.6.

2S

20

X 10ID

**

/

r ~_ p

r1:IC

1I* -

3.E .i

I *-

"" i

1000 2000 3000 4 000GROSS W EIGHT OF PLANE IN POUNDS FIG. 7.

1192c


5/6

76SUPPLEMENT TOFLIGHT OCTOBKE 31, 1930THE AIRCRAFT ENGINEERlightened is fitted, which is continuous from bow to sternan d theskin issupported bytransverse frames andbulkheadsmade in twopieces and riveted to the centre-line bulkhead.I t is considered advisable to divide the pontoon length intoat least f ive water t ight compartments, the forward orcollision bulkhead being more robust than the othe r s .Fig. 6 shows a useful type of punch and die used for makinglightness holes. The planing bottom stiffeners and sidestiffeners are continuous fore and aft, and all rivets to theskin are made countersunk as shown in Fig. 7, to minimiseresistance. The thicknesses of material required for amachine of approximately 3,000 1b.weight would be :

Centre line bulkheadCentre line bulkhead stiffenersFrames andbulkheads ..Bulkhead stiffenersPlan ing botto m stiffenersSide stiffenersChine angleSide skinDeck pla tingPlaning bottom forward . .Planing bottom aft

0-0400-0280-040-0-0510-0280 0400-0400 510-0400 320-0510-040 Fig. 3 shows a sketch front view of a biplane racer inwhich thehorizontal member between the floats is a stream-line tie-rod and Fig. 4 indicates details of the a t t a c h me n t ofsame inside the float.On the planing bottom the average stiffener spacing is6 in. and on thesides 8 in.,where Z or bulb angle stiffenersare used, these stiffeners having half-inch or five-eighths inchflange to skin.The following tab le of riveting gives good results inpractice :

Thickness ofPla te .0-0200-0280-0320-0490- 051 0-0650-08 0

Diameter of Rive t .In .

128 *Laps.Thickness. S.R.Laps . D.R.Laps . T.R.Laps .I n . In. In.0-032-0-049 1 - If0-051-0-065 | 1$ If008 -0128 | if 2Butt Straps.Thickness. S.R. Strap. D.R. Strap. T.R. Strap.In. In. In.0032-0049 II If 2f0-051-0-065 If 2& 30-08 -0-128 l | 2j 3J

ChassisTo ensure safety it is recommended that thechassis shouldbe stressed for the following cases :Condition 1Landing onwater with noangle of bank in such an a t t i tudeth a t the resultant water reaction acts at a poin t on the linejoining the nose of float with the bottom of the step, anddistant one- third of the length of the line from the nosethe reaction being inclined backward from thenorma l to thisline at an angle tangent1. The resultant couple to bebalanced by the iner t ia of the machine.Facto r required7 5 (onweight of machine) .Condition 2The machine at rest on thewater with noangle of bank andinclined sot h a t themain planes are at their stalling angle.Factor required6 (onweight of machine less floats).Condition 3Machine rollingweight of machine acting through C.G.of machine andC.B. of one float.

Factor required2 (onweight of machine.)Condition 4 r~' r~" ':C.P.F. FlightFlying wire loads.Factor required9.

Condition 5C.P.A. FlightFlying wire loads.Factor required7.Condition 6Side load on each float equal toweight of machine lessweight of floatsdivided by number of floats, and applieda t a point in theplane of the topsurface of the float verticallyabove the float C.B.Factor required10.

IN THE D R A W I N G O F F I C E .THA T PULLEY PRO BLEM

By R. R O D G E R .In anarticle entitled " OnAngles " appear ing in THE AIR-

C R A F T E N G I N E E R , da ted May 30, 1930, Mr. H. Parkinsondefined an analytical solution to a pulley bracket probleminvolving a compound angle . Mr. Parkinson's conclusionswere criticised in Augus t 29 issue of THEA I R C R A F T E N G I -NEER by Mr. E. H. Atkin , the controversy apparently beingas towhich angles were actually required for thelayout of thebracket.B u t whyworry ? The problem is essentially one for theDrawing Office and is, therefore, due for solution by adraughtsman. In arr iving at that solution thedraughtsmanwill surely practis e hisa r t th e art of projection anddevelop-ment of surfacesand eliminate the possibility of maskederrors. As Mr.Atkin po in ts out, it is a comparatively easyma t t e r in analysis to assess the axes of reference incorrectlyeven when the angles themselves are known.Conjuring on a " guessing stick " with trigonometricalra t ios mayundoubted ly supply uswith thecorrect numericalvalues of certain required angles, but I'll warrant that ninedraughtsmen in every ten would prefer to see "t he cards onthe tab le. " Besides, chief and assistant designers aresome-times awkward people to deal with, and have a nasty habitof wanting to know why at inoppor tune moments . Thedraughtsman might endeavour to answer the query bywading through the analytica l proof, but I d o u b t if theresult ing assemblage of hieroglyphics would be very con-vincing to the chief.In any case, the graphical solution is so delightfullystraightforward, forms a record wherein one can seeeverystep and its consequences, and, finally, produce s a resulta finished drawingwhich can be issued directly to theshopsfor immediate use.Having , I hope, justified graphics to the detr iment ofanalysis , I submit below an idea of how " the man on theboard " would, or rath er should, reduce th is awkw ard pulleyproblem to cold facts in theplane of the paper .In the diagram the front elevation and plan are set outfrom the known apparent angles fi, 6 and y, the endelevation being a derivative of theothe r twoviews.Subsequent operations and ready reference in a problemof this kind are considerably simplified if an elementaryform of nota t ion be employed. Thus, all points of referencein their original positions are indicated by a le t ter , A, B, C.etc. Each subsequent posit ion is indicated by a numeralfollowing the le t ter , as Al , A2, A3, etc.,whilst additionallyeach operation should be i n d i c a t e d ' b y a distinctive colour.Thus , in theoriginal position all lines might be black, in thesecond position red, in the third position green, and so on.Finally, all reference points which lieabove, the plane, of thepape r are indicated by a plus sign, and all points below the-plane of thepape r by a minus sign. Poin ts in theplane of t tiepape r aregiven nosign at all.These last remarks may at first appear to be a little super-fluous, but they areprompted by thef ac t tha t I have oftenbeen amused by theantics of some draughtsmen, confront edby a maze of lines all in one colour anddevoid of notationsquinting down a piece of bent wire in an effort to discoverw h a t is happsning to their lines when they swing a systcfflfrom one posit ion to anoth er . This practice is to be dis-couraged as the observer 's eye and the piece of bent wire


6/6

OC T OB E R 31, 1930 11

THE AIRCRAFT ENGINEERSUPPLEMENT TOFLIGHT

B EN D D OWNT H R OU GH OFRONT ELEVATION

toBEND UP THROUGH y/" ( ^

held in the fingers areboth mobile relative to each other, andpractically any desired result may be attained whereas onlyon e is correct.A preliminary consideration of the problem leads us tothe conclusion that the fundamental data required is therelation of theplane WXYZ, which is theface of a structuralmember to which the pulley bracket is to be bolted, and thecommon plane of the cables OA and OB. Plane WXYZ isalready in the plane of the paper in the front elevation andnormal thereto in the remaining twoviews. Byro ta t ing thefront elevation about the centre 0 in an anticlockwise direc-tion, the twocables can be brought into al ignment in the endelevation, whilst still preserving the plane WXYZ normal tothe plane of thepaper in that view.To accomplish this, project A on to OB at C in endelevation.Again, project this intersection point Cacross on to OB infront elevation. Place a piece of tracing paper over thefront elevation andpivot thepaper at 0. Plot points A and Con this overlay and then, by t ria l and error, line these twopoints up with the axis XZ. This operation takes but a fewseconds. Prick throu gh points A andC in their new positions,remove overlay, and set in the new positions of the cablesOAl and OBI as indicated by the dotte d lines. Project Al(front elevation) on to line AC (endelevation) and throughintersection point Al draw dotted line 0 A 1 B1 .W e now have in the end elevation two required planesboth normal to theplane of thepaper. This ishalf the batt le .The next step is to throw a view normal to the plane OBI(end elevation). Th is gives the t rue run of the cables.In front elevation, swing axis WY to W1Y1 so that angleWO Wl equate angle A0 A1 . Project Wl and Yl on to WY(end elevation). Swing OWl and 0Y1 into plane OBI (endelevation), extended, through angles ip and O, respectively,giving OW2 and OY2, and project normal to plane. Thisgives us the developed centre lines about which the fixingbolts arepitched, OW2 for the upper plate and OY2 for thelower.The projection of theorigin 0 gives the bend line.W e now have five lines about which the cheek plates ofthe bracket and the pulley guard can be finally detailedready for issue to thesheet metal workers. ..-...--

The developed centre lines, OW2 and OY2, are onlyrequired when there is a restricted landing for the fixingbolts, e.g.,when the bracket is attached to a s t ru t or likemember where the bolts have to pass through or adjacentto the neutral axis of the cross section. In other cases wherethere is room to t u rn the bracket on theface of the structuralmember, these offset centre lines may bedispensed with, theline OE will become the centre line for pitching the fixingbolts, and common cheek plates may be used with differentbe nd a ng les

Documents

The Aircraft Engineer October 31, 1930