naca tm 676 towing tests of models as an aid in the design of seaplanes

7/27/2019 NACA TM 676 Towing Tests of Models as an Aid in the Design of Seaplanes

1/26

v TECHNICAZ MEMORANDUMS

,

NATIONAL ADVISORY COMMITTEE FOR AERONAUTICS eNo. 676

TOtiING TESTS OF MODELS AS AN AIDIN THE DESIGN OF SEJIJ%ANES ~

By P. Schr~der

lTerft-Reed.erei-Hafezi, Vol. II, No. 16* August 22, 1930

,,I ., .- . .

.)

j ., ,,. .

WashingtonJuly, 1932

;.,> /, -:.,, ,

,s


2/26

. .

~11111111111111111imlimlllll:=___3:76014373451NAT 10NAL MIVI SORY COMWZ?TXE FOR .AERONAiJT1(2S

TE CHN1 CAL JiEMORKNW?M NO. ~676 ,,.,TOWING TESTS, OF MODEL5 AS ~. AID IN THE DESIGN OF S3DLANES*

By P. Schr~der. ... .,Is INTRODUCTION

The seaplane, both before and during its take-offrun, is a water craft- As such, it must have a sufficientdisplacement and must fulfill the well-known requirementsof longitudinal and transverse stability while afloat,just as any s-nip= But in addition to these it must attaina speed which is altogether beyond the capacity of iaostwater craft. This presents no difficulty if there isavailable an extraordinarily large propelling power; as isusually the case -with small float seaplanes. If, however,we are concerned with avery large aircraft, one which isintended for long-distance flights, we must lift from thewater, when taking off, just as heavy a load as possible.It might easily happen that tile seaplane could carry itsdesigned load if it were once in the air, but that it ,.could not reach take-off speed on the.water- The problemthen arises: What form of float is best suited to carry arelatively heavy load to a high speed with a given propel-ling power? This problem has been studied ever sincethere have been seaplanes by the comparison of the take-~ff performance of seaplanes which have %een built and bytowing models in experimental model basins. As a resultof this work; there are available to-day a num%er of prov-en designs. ,.

,Nevertheless, in the design of a seaplane, one, alwaysmbets anew the.problem of determining the suitable float.Forit should benoted that the successful take-off ofaproveri.seaplane %s not dependent solely. on the floats, Ifthe flbat of a s~ccessful seaplane is assembled with an-other wingc63.1 and power plant it may fail to take off,..:.... .:-. . . .-...,;ll~er ~Jodellschieppvers~dch als EilfSmi.tt.91beiXZI hhUI$. .eines Seeflugzeuges. 17erft-Reederei-Hafen, August 2.2, ,1930, pp. 349-354. Pappr presented al the Ninth M6e+i+~of the Gesellschaft der Freunde und Forderer der Haml&g-ischen Schiffbau-Versuchsanstalt, June 11, 1930.


3/26

2 N~A.C.A. T~chni.cal Mernormyium lToC676

even with the same power loading. Moreover, the identi-cal float assembled, with the. same wing cell and powerplant may have en.tirely different take-off characteristics,acc~rding to lthe specifications underwhich%he assemblyis made- ...

These facts have not yet. received the considerationwhich they deserve in tho study of tho take-off phcnolmcnaof soaplanos. In part this is because tho results of testsare as a rule not made availablo for publication. Themost serious obstacle, however, is presented by the prac-tical difficulties encountered in carrying out tests insucha manner that they make it possible to perceive theeffects of changes in the design of the airplane on thetake-off performance. ,...:.,.

So far, the.difficulties uontionod have been avoidedin the following hanner~ To begin with, the design ofthe airplane, including the.float system, was prepared byreference to successful prototypes Then, the float sys-tem was tested in an experimental model basin, and inthese tests the model was subjected to tlie conditionswhich would be imposed upon it by the proposed wing celland-power plant- If the results of the towing tests wereunsatisfactory, thdform of the float was changed accord-ing to judgment until satisfactory results were obtained.Thismbthod of itself makes it impracticable to attemptchanges in the design because the scope of the experimer.-tal workfor each new seaplane would thereby become infi-nite.

The procedure doscri%ed gives good service when it isa question of obtaining a quick decision concerning a sPe.-cific design. It probably will be necessary to.resort tothis method ,in the future for the same purpose. However,it is not Suited to the ~yrther advancement of our task,which is to obtain the best take-off performance for the .,floats of large seaplanes. While being tested, the floatsystem must be subjected to the roquiremonts of a specificairplane design, consequently, such tests do not permitgeneral conclusions regarding the float system unde~ test.It is not possible, On the basis of such tests.? .,~Q.pr~.-..diet, how,the float: system would behave when assembled withanother, win-~cellor power plant. It isnot even possibleto predict .tlie,bffect ofs small change in the originaldosi&n 011 thetako,-off performance. .,.


4/26

N.A. C.A. Technical. i!ernorandum 3T0. 676b 3.. .,Hence,, it ,~as. nece.s,s.aryto develop a method ofre-search which would makeitpossible to conduct the testsin such a manner that they would be independent of the re-quirements of .a special airplane design. Such tests cou~dbe made before the, completion of the ,work of designing theaircraft and would then be just asmuch an aid to the air-plane designer in the design of the seaplane as would theaerodynamic model tests, which no designer to-day would -forego.The making and evaluation of a towing test of a sea-plane float system, which answers the requirements justmentioned, is the subject of my present remarksi The ex-

perimental data presented have been taken from the testsof the transoceanic airplane designed by Dr. Rumpler,whfchwere made by the. E.S.V.A. only a short while ago.

II. THE MAKING .OF THE MODEL TESTS AND. . THE PRESENTATIOtiOF TH3! RESULTS ..

Because of practical difficulties in obtaining meas-urements, we cannot simulate the accelerated take-off runsInstead .we s,elect a number of definite, speeds and tow themodel at each with constant speed. The modelmust betowed at each speed at several trim angles bebause the re-sistance in general depends very much ~n the angle of trim.Besides the speedand angle of trim, the lift which is de-veloped by the water must also be prescribed. Then onlydoes the question of resist&ice havea single meaning.As the load which must .be carried by the water is stronglyinfluenced by the wing cell -d power plant, even for agiven definite flyi,ngweight, runs must be made at differ-ent loads for each s?eed and tria~

~ In each case, besides t~e resistance.,, the, trimmingmoment necessary to produce the trimangle must he measured.In the later computations these moment measurements formthe point of departure for the d-etermination of the trimposition which the floatsystem wou,ldassume under a givenaircraft- .,.. .,.

Tigure 1 shows in,grapfiic form apart of t!he scheduleof t,~st runs for the study of the ltu.mpler model., In thismanner the model is towed at all odd angles of trim up to11O. For each of tile parallel abscissas indicated, we ob-


5/26

. . .

4 N.A. G.A. Tech:rilcalemorandu& 5?00676

tain. a resi stance cur~ye and a rnom,entcurve which, corre-spond. to a given trim angle and.a prescribed load. .

For example, Figure 2 shows the results of the re-sistance measurements for a trim angle of u = 5...An exainple of the moment measurements is shown in Fig-ur8 3. The reference axis for the moments shown is thetransverse axis through the point at which the center of

gravity is to be, according to the first design of the air-planOm,-

The ftrst presentation of the results is derived di-rectly from the p~actice in making towing tests. We caneasily derive from these curves those rosistabco and ao-ment curves which correspond to a given speed and a pro-scribed anglo of trim for any arbitrary change in the lift.In Figure 4 we hav~ theresistancos, in Figureo5 themoments in the new ,presentation, meay~red at a = 1 .The results are placed at the disposal of the air-plane designer, in this form two curve sheets for each trimangle investigated. These experimental data include allthe properties of the float system which we need for theinvestigation. of the take-off with any power plant andwing celluleoIII* THE APPLICATION OF T3E TEST RESULTS IN

THE DESIGN OF THE AIRPLANE.. 1. The Influence of the Wing Cell and Power Plant onthe Take-Off Performance

For any. form of float system, ev,ea the oest imagina-ble, there aielimits to the take-off performance which isobtainable, fixed by the wing loading G/F and the poticrloading G/N of the aircraft. In these, G is tho grossweight,, F the wing area, ,and N tho engine horsepower.However, in the investigation of the take-off, G; F, an,dN are to be considered as constant becauso they priricirnpally dotermino the flight c~larac,~cri.sties. .Furthe.r.moro,tho .wirigform,solocted must be de:te.rrnine,drom%h6aerody-. ..riami.cst.add-poiritalone-i ... ....:, .,.,.,,. .. .

,.. ., ... .::.,..,. ,.,,.: ... ...... .,.


6/26

N~A. C.tA. TecQnjcal Kemorandum. lTo..6.76 5

Tho arrangement of the wing CO1l and powor plant rol -ativoto the float system can.,influbnco the t.ako-off re-sistance very. munch. The question of which arrangernontwould give tho best take-off perfor,rnancos which aro.possi-ble with tho prescribed. float system ,cti be determined bymeans of such towing tests as have.just been described.It is in this sense that model tests with the float sys-tem are to ,he regarded as an aid in the designing of theseaplane. .,The effect of the wing cell,and power plant on thefloat system can be expressed a-s a lifting force and a mo-ment. Changes in arrangement which will change this force

and this moment without impairing the flight performanceare: . .,.(a) ChaVging the distance from the step to thevertical through the center of gravity of the sea-plane. This can be done without changing the formof the float gystem by shifting the float system lon-gitudinally relative to the wing cell or by shiftingequipment and cargo longitudinally.(b) Changing the distance H from the centerof gravity to the line of thrust of the propellers.This can be done practically by moving equipment andcargo vertically or by moving the shaftsti

(c) Changing,the angles of inclinationa ofthe wing chord to the C.W.L. (designed water line).,.(d) Changing the angle, of inclination ~ of the

propeller shafts to th,e C,,W.L.Methods. (a) and (b) are the most effective. They de-termine the t~imming momentapplied, to the float system

and t%,e.rebythe trim positions which occur during the take-off. Methods (c) and (d) mainly c,hange the lifting force.They have little influence on t~e moment.In order to obtain that arrangement which will leadto the ,most favorable teke-off resistance which is.possi---,-,...-ble with, the-ffioa~-s~stsfi being considered, and takingi,nto account all of the secondary requirements, the follow-ing met~nod ,seems most .sui-table: For some arrangement which-is practica~le., .~~ which seems plausible when comparedwith proven con~{,ructions, the take-off performance is de-termined from the tests. Then each ofmethods (a), (b),


7/26

6 N.A.C.A. Technical Memorandum No. 676

(c), ad (d) iS tried to det=mine in which direction achange must be made to improve the performance of tl~efirstdesign. O,filywhen none of the practicable arrangementsleads to satisfactory results isthe float system underconsidera.tion .to be rejected as uilsuitable for the pro-..posed aircraft, and.a chango in the form of the float sz~.s-tem to be considered. ..,:. . .,....,~.,, .:,,; ,, .

In this the fundamental requirement is that the~formof the float system shall be derived primarily from thehydrodyha:miepiht of view. In assembling:i.%.ith.tkeo wing oelltid- power plant it is necessary to imsure %y asuitablb adfi,angement that the float system is working und-er conditions that approach as closely aspo.ssible thdseunder w-hichit has its minimum water resistances ,,

Be-sides the changes in arrangement which influencethe ~conditionsader which the float system .mustwork,we I%voin tha altitud-o controls an additional. important

~~iidfor influencing the take-off of a seaplane. The grcat-erthe speed which has been reached, the greater the lim-its within which one can influence the trim by pulling upor nosing down with the elevators. Consequently, if athigh speeds the position of equilibrium free to trim isun~avorabl e, the float system can De brougilt into a favor-able, position by using the elevators.. .

From this we derive the following divisions for theinvestigation of the t~e=off characteristics of a p~o- .,posed seaplane: . ... .,,-, .,1], The most favorable resistance curve whichcan be obtained i.nthe free-to-trim condition bychanges in arrangement is derived. The elevator haslitt.ltieffect at low speeds and this curtieis decis-

ive four the first part of the take~off..,.21 The resistance in tile free-to-trim condi-tion. during the second part of the take-off will Ce.n-erally exceed the most favorable resistance of the

. . . . ., .

float- syktemby a considerable amount. Informationas. to. t.htilevator movement which is required .to u ;b~ing the...lo.+t~.sy-stem into its best attitude is tobe obtained from the tests. .The best attitude-whichcan be attained by suitable elevator movements is decisive for the course of .thb resistance curvs. diox--ing the second part of the.:take-off. .. .,,,.. ,., . . ............ ,.,. ,...


8/26

N,A. C,A. Technical Memorandum Ho. 676 7. .:,... . . ....:,.In the past we have leen obliged to leave to the pf -lot the task of determining this requirement for each

se~plane by making anumber of trial take-offs. This iSa thankless task, as the requirement changes with everychange in the loadirigand the distribution of this load inthe aircraft. Furthermore, with heavily loaded seaplanesit is not without :dariger~ forwith unsuitable trims, whichhavo a high rQsist&tice, other inconveniences are iriTOlV~dkThe float system shows a tendency to pitch and to leavethe water (porpoise). In addition, the spray becomesheavier, the farther the ,float system departs from its mostfavorable trim. This can be very disagreeable in connec-tion with the propeller. The greater the total weight oftho seaplano becorms, the moro importantit is to.avgidsuch dangers asfai aspossiblo by resorting to modoltests.

~~ The Determination of the Resistance Curve forthe Take-Off Free to Trim. ..

The working of a definite example is much more in-structive than a purely abstract discussion, so we willnow imagine that we have been giv,en the problem of carry-ing out the study of the take-off characteristics of thedesign for a transoceanic airplane proposed by Dr. ,Rumpleron the basis of model towing tests- In accordance withthe considerations just discussed the first step is to de-termine the resistance for take-off freo to trim underthe conditions of the proposed design. According to thisdesign it is specified that:G = 115,000 kg (253,530 lbo)F = 1,000 mz .( 10,754 sq.fto)o = 2.5 -c 4 ., H = 1.14 m ( 3.74 ft. )

Static thqg..sl..op.r..p.ellersrca~a~ .23,,,5Q.O,,kg,51,809 lb.) de-creasing to about 16,900 kg (37,258lb@) at 150 km/h (93.2mi./hrO). The center of gravity is 2P9 m (9.51 ft.) for-ward of the step and 4.75 m (15.58, ft.) above the designedwater line (C.W*L.) q .,.. 1?

. .


9/26

~ ,; N~A~C.A;. Technical Memorandum No. 676

-! :,Ttid model:is 1/16 full size. .The displacement of themodel. a.trest .is..therefore,...2 .. .,,. .,. . ...... ....,. D.=.G ~-s=.28=08 kg -(61.91 lb,) (1)., :. .,. ., .. f...in which ~ =-16, d-gnotes the model scale. The modelspeed v corresponds to the airplane speed V.. .,.,.,

P = v ~-1/2 (2).,,,

The towing-carriage speed of 10 m/s (32.8 ft./see.) underthese- conditions suffices to.sirnulate a Speed of 144 km/h(89.5 m.i./hr.) of t.he.airplane. The weight on the watera, which inust be supported by the water at any speed..ofthe model v, is derived from:.a = D-e (3),.

Iilthis, e is the air lift supplied by the wing celland power plant reduced to model scale. .... e =E ~-3 ., (4),,

E consists of the wing lift Ef and a contributionEt which is derived from the thrust of the propellers:..

E= Ef-tEt (5)The wing lift Ef is calculated from . .,

Ef = Ca q l?, .(6)

The lift coefficient, Ca is given in the polar diagram(fig. 6) as a function of the angle of attack of the wingchord. This is the result of aerodynamic model testsiIa addition to the value of the lift coefficient c~Jwhich we need to begin with, we also obtain the drag coef-ficient cw for the air resistance of the seaplane. .Theangleof the win-g 8 is . ,.>., ...,,. ,.,. ,: 6,=; U:+0 .,. .. :@,..,.,... . ,......:....,According to the specifications of thedestgh:, :a wi~g


10/26

N.A. C.A. Technical Memorandum No. 676 9

of 8,5 corresponds to a trim angle of 6. The part ofthe---liftderived from the propellers Et, is

Et = S sin frl (9)as a result of the inclination of the propeller axis atan anglo T to the horizontal. The angle T is accord-ingly,

n= c%+~ (lo)In this case T is. always 4 greater than the angle oftrim. Compared to Ef, Et is of secondary importancebut generally is not so small as to be negligible.

The results of this computation we can see in Figure?. The scale of abscissa in this case is set off propor-tional to the squares. By this means we get a straightline for the weight on the water at each trim angle. Con-sequently, we need to carry out the calculation for onlytwo speeds at each trim angle.We now determine the resistances and moments for

a= 1. With the weight oa the water now known we cantake the resistances and moments which correspond toa= 1 from Figures 4 and 5.

In this manner similar resistance and moment curvesare obtained for each trim angle investigated. The securves fulfill the simultaneous requirement that at everypointA+E=G; (11)

that is, tho weight on the water and the lift from theproposod collule and propeller thrust always total thegross weight. Yigures 8 and 9 give us those curves fortho tr.ansocoanic flying boat of Dr. Rurnplor.

The next problem is to determine the trim angleswhich the seaplane will assume during the take-off- Theseare dependent upon, the moment loads M. which the floatsystem receives from wings and tail and from propellers-This constraining moment M. consists mainly of two parts

M. = Ml i-Mt (12)Of these Ml is the part derived from the aerodynamic

___


11/26

io N. A.-C:A*Technical~emorafid-um $-.:676

force sand Mt that from the power plant. .,1,11i.:d~eteti-mined %y aerodynamic measurements and in thi:s case i:t:rnaybe con.side.redas known, jus:t as is the polar diagram ofthe wings. Mt is computed from

.,.. . . =:slf .,.. .. .. ..:Mt .~ ,, ,: ~~ .;:. . .. . :in which it must tie remembered that S is dependent -uponthe, speed. For the model the imposed moment is then

(14)

The moment m. thus computed is shown in ,our example asthe cudve m. of Figure 9a The moments -m:,aising fromthe action of the water are drawn .positive when they liftthe bow; the constraining moment m. is drawn a$ positivewhen it depresses the bow, so that positions of equiliti-rium are indicated by ~

(15)+ m. O ;,T~lgrefore, the various points at which the curve no iri-tersects the curves cc= constant are such pointsof equi-librium. In Figure 9 we can now read the variouschangesin trim angle which may he expected in a take-off free totrim under the conditions of the design. Figure 10 showsthis curve.

With this knowledge of the changes in trim we are ina position to determine the corresponding resistance curvein Figure 8, With this the first step of our problem issolved. We see at once that left free to trim the sea-plane has a much g~eater_resistance at high speeds thanat trims of from 3 to 5.

3. The Resistance Curveswith Elevator(at Fixed Trims)(

i-l,, . !Uie next problem is the determination

Control

of the resist-axicecurves in the second part of, the take-tiff when thes.eti~laneis pulled up, to a fixed, triin::We tit~l:,s an....ex~mple, consider trimrning to ~.=#o This isthe mostfavorable trim. for the d,esign under consideration..;;..:

First we must determine by computation whether the


12/26


13/26

12 N,A. C.A. Technical Nemorandum No. 676

.... . . To.,he water resistance thus computed, tbe air re-,sistance

(18)-is tb be added i- The drag coefficient cm is taken fromthe, polar diagram. The data regarding Cws, which in- -eludes the so-called parasite drag, Ifare also derivedfrom aerodynamic tests. ,:.. ,..,,

Tile difference between the sum of the resistahties.and the propellef thrust is available fo& deceleratingthe seaplaae, If the reciprocal of th~d difference is de-noted hy r; an elementary principle of dynamics givesthe formula for take-off time t and length of take-offrun s,

The integral can easily be solved graphically- In ourcase,t = 70 sec. and s = 1,430 m (4,690 ft.)

for a take-off in which the seaplane is left free to trimuntil it reaches 70 km/h (43.5 mi./hr.), aild then is hel,dat 5.5. AQ Example of the Investigation of Changes in Design

We have reached the objective of our investigation ofthe design under consideration, and are now confronted bythe question .~hether still better performances can be ob-tained from the float. system. The next step ofa complete.investi~ation of the take-off, according to our discussion,is the trial of all four changes in design (1.11-l-a,b,c,d)which may be used for the purposet TO coilsider all fourcases would lead to tiresome repetition. Accordingly, Iwould like to iimit myself, for purposes of illustration,to describing in,detail only the investigation of the ef-fect, of a change in the poeition of the center of gravity-

We. will assume, that as a result of a changiin the ~:distribution of t~e cargo, the center of gravity is movedaft8,oO mm (31.5 ino).. Since its original position wasalso _o.uraxis of moments, this shift means a stern-heavy

-.


14/26

N,A, C.A. Technical Memorandum No, 676 13

moment for the model ofm~ = 28.08 X 008 : 16= 1.404 m kg (10.1$5 lb. ft.) (2o)

The previously computed moment m., is decreased bythis amount as seen in Figure 9. The curve of the new -trim angles is obtained from this figure and is shown inFigure 10. The corresponding resistance curve for a take-off free to trim is found in Figure 8.For high speeds, a change in the position of the cen-ter of gravity is of no importance since we are then ableto assume any trim by means of the elevators. Under these

circumstances; a change in the angle of the wings relativeto the C.W.L. is effective. As o increases, the wingslift agreater part of the gross weight; consequently, thewater resistance decreases. The investigation requires arepetition of the whole process of analysis. Figure 12shows the results for an increase in the angle of thewings from a = 2.5 to o = 4.500 The expected reduc-tion in water resistance is obtained, but the greater partof the gain is lost because of increased air resistance.We must not conclude that the investigation was fruitless,however. In this case the original design already layvery near to the most favorable proportions. That factwas determined by the investigation.

IV. CLOSING REMJiRKS

The take-off which we have just discussed referred toa take-off without wind and at a prescribed gross weightsIt is easily possible to compute the take-off for any o.th-er desired gross weight from the experimental results pre-viously discussed and according to the method of, evalua-tion presented and also to take account of the winds Atthis time I can only mention this fact.

Even if it is decided to melee essential changes in the wirig ceil or power plant an estimation of the take-off performance is still possible without its being necess-ary to make new test runs. Such radical caanges are, forexample, the choice of another wing profile, another wingarea, or the fitting of more powerful engines.The cost of such a comprehensive investigation is nat-


15/26

14 N. A. C.A. Technical Memorandum No. 676........ .urally greater than that of the customary simpler proce-dur e. But if we keep inmind that the value of the en-tire seaplane depends on its ability to take off, %ecausethere canbb no flight without take-off, .and also considerthat the cost of a giant seaplane runs into millions ofmarks, the cost of the research isfully justified, If weundertake a complete investigation of a float system, asis here recommended, we.are assured that we will be pro-tected against disagreeable surprises, which will causemuch greater expense, if they are first encountered .on thefinished seaplane, than a complete model test would costs,. . .

A. shortened tegt which is limited to th.erunsabso-lutely necessary fbb the consideration of a.given. take-offcondition and which neglects the determination.of the oth-er properties of the float system does not offer,this pro-tection to the same extent. Experience has already showrithat in the region of maximum resistance, two differentresistances may occur. The more stingily the researchprogram is laid out, the more easily does the more unfavor-able case escape observation,

A more reliable way of preventing a costly increasein the scope of the tests :lies in the solutioilof the fol-lowing research pro~le~: ..Let the resistance WL and thetrimming moment ml of a high-speed flying boat-with aload al be measured.at a speed VI . The resistance W2and moment mz for the same flying boat with a load aaare to be computed from these results for a correspondingspeed. The H.S.V.A. has done this and will report soonon the solution of this problem.*

In conclusion, I express to Dr. Rumpler my bestthanks, because he has so kindly agreed to the publicationof-.the, experiaental data which I have given and therebyhas enatled ine to enliven-the presentation with actualtest results.

. ..- -.... Schr$der, P.: Determination of Resistance and Trimmingmoment of Planing Water Craft. T,li. Noa 619, N.A,C.A.,

1931, . .....,

. . ,,.. . . . . ... .. - :::.1.


16/26

.. .

., ...N. A. C.A. Technical Memorandum ITo.676.,

)..ISCUSSION

15

:.. .. . . ,.

In the .diqcussi on of the .pr~ceding pa~er, Oberbaurat ,11 can oniy con-,r. Ing. !~eitbre~ht, .,qrlinf..rernark~.d:gratulat e the H6S iV~A.- and D,n~ Schroder. on thi s work whichthey have described he~ee .QY.quest ioned.o not relate tothe results of- the tests bu-t to the methods used.l!Uegtion 1.-,L Aside from the various loadings of theseaplane in the initial condition, the variables in theinvestigation are: . : ,.

(a) speed(b) moment ,!(c) trimIn the work under discussion the speed and trim were cho-sen as the constant quantities in the test. In our oPiil-ion it is simpler not to change the speed and moment wit3.-in one test and to plot trim angles as functions of themoment, I believe the consideration of the possibility ofaffecting the speed by means of the elevators or by chailgesin weight distribution is made simpler. Naturally, this isconditioned upon the. fitting of a trim-con,trolled lift.(One wilich simulates the wing lift, varying with-the angleof attack of the wing.- Translator.) ..

Q!A!@M!.9._a- 1s there any hesitation .at attemptingthe quantitative measurements of accelerated runs if dy-namically similar models are used? An obstacle to. carry-ing out these tests up to now has been the difficulty oftaking ca,re of the chaoge in wing lift on the model cor-responding to changing trim angles. This problem has al-ready been -solved hy us and I believe also b~ othors, sotilat in this respect there.aced be no further hesitation.

Question. 3.- Are there any figures wh~ich compare the..time of take-off andta~e-off run computed acc.or,ding toequation (19) for mod~l aiid.full size?li, .,.Mr. Dieaer (Engineem), Friedr.i.chshafen, remarked asfollflws: ..ll.1-.,wo&dike t~..add .to the presentation-.-by Dr.Schroder a point which he has not referred to in this dis-cussion of the balance nf forces and moments on the air-craft while taking off. This is the moment which is de-veloped by the air stream of the propeller ~n the flying

surfaces. The lift of the wings is more or less strongly

. ... . . .--.


17/26

16 ~T.A.c..A. Technical llernorandu.mNo.676

affected by the slipstream according to the relative posi-tions of propeller and wings as is also the induced amgleof attack, over that area of the wings which is swept bythe slipstream., Because of tilis, the direction of thewake behind t-hewing i,s also bhanged. and like.wisej the di-rection of the blast on. the-control sur~aces, an,d theirmoments. It is further changed by tQe increase in airspeed in~the slipstr.earnas compared to the flying speed.This moment deyeloped by the effect of the slipstream oilthe wings and control surfaces is greater than the momeatof the thrust about the center of gravity of the systemand coilsequentl~ cannot be neglected in working up the re-sults of towing testsa Unfortunately, tho desigiler is ina difficult position because at present tl.is moment gen-erally cannot be calculated and can bo estimated only veryapproximately from wind-tunnel tests. Consequently, thiscircumstance causes an undesirable uncertainty in the .a:?-plicat~otiof tank tOsts to large seaplanes...

1t17ithregard to the question raised by the previousspeaker concerning the agreement betwooa tho take-off por-formanco obtqined by analysis of towing tests and thetake-off performance of actual airplanes, .1 might referto the results on the flying boat Dornier Val, the modelof which was towed last year at various trim angles inthe manner described by the lecturer. The computation ofthe model tests for a gross load of 6,200 kg (13,669 lb,)gave a take-off time of 27 seconds, while the actual rieas-ured take-off time was about 23 secoild.s, showing a rela-tively good agreement.ll

Mr. H. Herrmann (Engineer), Bremen, made the followingreioarzs:, .llThelecturer has shewn US hQW mUC-hfurther re-search m,ethods have, d,eveloped.at the Haa3urg tank in thelast few years. He even goes so far as, to work out di-rsctiogs for t~e pilot .to follow during take-off. Ofcourse, asapilot one c,an abide by such ,direc.tions to acertain bxtent. But th,ere are many considbra,tions againstit. According to experience, the trim angle is measuredtoo large on small uodelsi and consequeiltly m importantpiece of fundamental dat,a is wrong., We have no neas,urewhatever of the ,magnitu.deof this error. Co:ls.equen,tly,.it appears most advisable.,to check thewho,le pro.cetdurg.,ejgainby t~,e-off,tpsts. 11 ,.,.. . ,..,,.. ..., . ... .,, ,In,his conclusionj,Drj Schr~der replied: llOberb&aratDr. ~eitbrech$has expressed the opinion that it would bebetter to bringinthe trim moment as the independent vari-


18/26

N. A.C.&. Technical. Mernaranduii No. 676 17

able instead of the, trim angle. This procedure is, ofcourseg tobe--.pr~fer,red ,when:we ara coneeraed with test-ing a definite se,aplangdesign, for ml~ich we know ?CCU-rat,e.ly,the moments prod,uced hy.t:he wings and power plaatcBut.Engineer Diemer has also pointed out that these vei.ymoments are very.difficult to determine,. ,1.can only agrciewith hima This circurn.stancohas been sottlod,inmy mi,ndwith t-no decision to choose thoangl-o at willarid tomoa&In this manner, tho moasuroments thom-ro tho mornonts.solves remain free from unroliablo assumptions regardingtho magnitudo of tho oxtornal momonts which are intro-.ducodi In tho working up of the results oi *Ko drawingboard, according to my method, one can; dotormino tho of-foct of any dosirod moment. Ono can obtain tho samo re-sult if ono runs at fixed, proscribed momonts, but cannotrely upon a dofinito variation of momont with spood andtrim in carrying out tho tests. Honco, tho advantagowhicil tho choico of tho mononts as tho indepondont varia-blo should givo is lost, But , if ono must carry out aninvestigation of tti.osamo oxtont in both cases, I beliovoit more advantageous to fix the .qngle. In addition onemust always be careful to carry out model tests with dif-ferent forms of floats in such a manner ti~at they are com-parable with one anotlier. Comparison at equal trim an-gles is directly apparent. it is not at equal moments.

The trim-controlled lift of the float is ail advant-age only if one is investigating a specific design. Incomplete investigations of the float system, which are in-tended to give information concerning the take-off per-=formance of any aircraft which may be equipped with thefloat system being studied, the measurements in generalmust not be sub.jectod to the requirements of a particularaircraft, as I have established thoroughly in my lectura,IIror carrying out the measurements in accelerated runs~

all instruments must record automatically andwork withoutlag. As yet we have no such instruments. Furthermore,the value of measurements in accelerated runs should notbe over-estimated, The manner in which the acceleratii~gforce varies with the speed is unknown before the run andconsequently must he selected arbitrarily. In my opinion,a necessity f-or-busying ourselveswith the new difficul-ties which are found in this does not arise until it ap-pears that in actual fact accelerations produce seriouschanges in the resistance of planing water craft. Tho in-vestigation of this question hy special tests is trulymuch to be desired~

, ,,-,, ,.,-. ,,- , , ,- -,...,, - ,,,....,.,, ... . . ...


19/26

I

18 N.A. C .A. Technical Meinorandum,,3.d... 676. . .. . .....lihe further chec~ing of the results ofmodel .tes%&

by take-off measureuents made on actual ~ircraf t; a$jj~l-giilccr Hcrrmann has just pointed out, is certainly nbces-sar~; Tho earlier mgasuromefits of take-off. timoy unfor-tuilately, could not be drawn Jpon for. this purposo, siilcotho method of research and evaluation which has j~~st boonprosonted has beeil devolopod but recently.[l

Translation byThe Staff, N.A.C,A~.Tank-,.

. . ..,

,., .. . ... ,... ..... . .

.,.. 1..,, ,. ,.

., . . .,.. 4

...,...,,...

.:.

,

.

.. . .., ,... . . .. . .i.

..,. . . . .,. .. .,

:..:.

,. -.. .,,, ,... ...:,,... ,,..- ...-,?.,:.,. . .,.,.. .. . . . ...


20/26

li.A.tA. Technical Memorandum No.676

8$i.*

.

Figs.1,2

0246 8 10/peed of model,v, in m sec.

Fig.1 Program of test runs for model 802 at afixed trim of a = 1

6~

43

210

..

----& -

Y o-e-_

Q ~.-1

1 2 3 4 5 6 7 8 9 10Speed of model,v, in m/see.

Fig.2 Model 802, Displacement of twin floats, D = 28.08 kg.A=16. Resistance curves at trim, u = 5 for variousweights on water, a.


21/26

N.A.C.A. Technical Memorandum No.676

l?ig.3angles

.

c1%+!s.

q-lo

q a= 10 ~a=go ba = ,50Oa= 3 ~a = 79- t.a -> .*_./76 ..5 / />/4 q ,/ \ ~.32 / . / \ \ \ .\1 ,/ / -Q, \\ \o / />/ ,

-1 - /./ ,./d +,. -- ./-2 / 1 /-4~#+-q--#1234567 8910

Speed of model, v,in m/see.Model 802, Displacement of twin floats,D=28.08 kg. A=16 .Moment curves at weight on water,a=20 kg for various trima.

Parsmeter=v in m/see.8N7c.I-lk- 6.+$5~84

1711442 J .=! I -!-==

o 5 10 15 20 25 30 35Weight of model on water, a, in kg

l?ig,4 Model 802. Displacement of twin floats,D=28.08kg ,A=16 .Curves of resistance, W,at trim ofa=l for various weights onwater, a .


22/26

I?.A.C.A.Technical Menorandun No.676

.. .

=nrem T~eter= v in.

Figs.5,7

0 5 10 15 20 25 30 35Weight of ~odol ofiwater, a, in kg

Fig.5 Model 802. Displacement of twin floats,D =28.08 kg. h=16.Curves of moment, m, at trim ofu=l for various weightson water, a,

13456 7 8 9 10/peed of model, v, in a sec.

Fig.7 Ruxpler transoceanic flying lost. ~ = 16. Weight ofr+lodeln water, a,at 5 = 2.5 aznd ~= 4 forvarious speedsjv,andtrims,a.,


23/26

N.A.C.A. Technical Memorandum NO. 676 Figs. 6,8

a, 20.5 = 8b, 18.4c, 17.4 i, -0.3d, 14.4 j, -3.2e, 11.5 k, -6.1f, 8.5 1, -9.0~s 5.6 m, -12.0h, 2.6

Drag coefficient, cmFig. 6 Polar curve for Rumpler

transocea.nicflyi~ koct.

6

25c!V-l=4~d03E%o02c)j:1A2

01 2 : 4 5 6 7 8 ~ 10Speed of model, v, in m/see Curves of resistance,W, at fixed trim angles, a, forthe weight on water, a, indicated in Fig. io Resistance for take off free to trim. o Resistance for takeoff freo to trim after the centerof gravity has been shifted.

Fig. 8 Rumpler transoceanicflying boat. Curves of resistance ofmodel, W, at various speeds, v, for conditions indicated.


24/26

N.A.C.A. Technical Me::.randu,mNo. 676 rigs. 9,10

8 R -\/ .,., ...6 4

ga=. .0 . . . .. . -

.$ 4 ~?a=< J ~,a= 1 a= 3( /d / ~) -.. . ....$ 2 \~ o.a //*d 40gw 2 ~,.1A- ? ~,.-/ . .....b, m~ c, m~ d, m3-4 . ..-___._._ .

//1 2 3 4 5 6

78910peed of model, v, in m secCurves of moment, m, at fixed trim angles, a, for thewei:qhtsn water, a, indicated in Fig. 7o Constraining moment free to trim.o-. Constraining moment free to trim after the center ofgravity has been shifted.

Fig. 9 Rumpler transoceanicmodel, m, at various flying boat. Curves of moment ofspeeds, v, for conditions indicated.,

b, After shifting centerof gravity aft.c, Original desibg.

o 246810Speed ofmodel, v, in m/seeI 1 I I 1 Io :() ~fj 9KOY40Speed of full sized craft, V,

in km/hr.Fig. 10 Runpler transoceanicflying heat. Trim angle, a,running free to trim,at various speeds..


25/26

N.A.C.A. Technical Memorandum No. 676 Figs. 11,1,2

,.

24,000.. ,.~ :_P:opeile:: tkIUs_..2,000 i \ ~20,000 : %&8 ,000 - . //Free iotrim-:/ /- ..~-.~16,000 h ~/// \ /-%4, 000 - ~ // /~12,000- / F:.xe~ tr-m :--~ / _~ . x~ 10,000 \s. \.I-l f 0-. ___ __ .~ 8,000 z Aixsan~Lw ter resistzLncem 6,000 /

4,000 - - Wa$er :~esista]ce2,000 /.e , I* ,1I~ ~ -AiT resis-~ance \r \ \

0 26 43 60 80 100 120 140SpeedjVjinkm/hr.

Fig. 11 Rumpler transoceanic flyin@oat. Resistances of fullsized craft at various speeds and under various con-ditions indicated.

- Center of gravity 2.1 m forward of the step, 6 = 4.5 0--- Center of gravity 2.9m forward of the step, 6 = 2.5 0.20,000

jf18,000G 16,000:14,000~12,000~ 10,000% 8,000~ 6,000cd 4,000

2,0000

r- -#zAir &d Wa;er/ //.4.. . 4 . resi ta~.ce/ ..- - .=1/ / I)\ I \ h//a = jt. c\,.-Vfate \ \

y yiresi tar.C 1.i / ./I :1/ // \\ -/ \ \

/= - \I rebi~tar.ce \67 80 100 120 140Spqed,,V,in@!u?. ,.,,

Fig. 12 Rumpler transoceanicflyin~boat. !lheeffect ofa change in the design on the take- off resis-tance.


26/26

Illllllllllllllllwlmillllllllllllllll31176014373451 ~.\

naca tm 676 towing tests of models as an aid in the design of seaplanes

Documents