JOURNAL OF THE AERONAUTICAL SCIENCES

VOLUME 8 DECEMBER, 1940 NUMBER 2

Wing Loading, Icing and Associated Aspects of Modern Transport Design

CLARENCE L. JOHNSON Lockheed Corporation

INTRODUCTION be very little cause to consider limitations of wing load­ ing such as are now considered unsafe. HE PURPOSE of this paper is to investigate the The various steps that are currently being taken to effect of wing loading on transport airplane de­ T make the use of higher wing loadings safe are: sign. Wing loading affects the flight characteristics of the airplane in that the higher the unit weight to be car­ (a) Improvements in airports and airway facilities ried by each square foot of wing, the higher the speed will allow higher speeds. at which the airplane must fly to sustain that load or (b) The use of improved landing gears which will the greater must be the at which the allow faster with better landing control wing meets the air. Values of wing loading vary from and better braking. 3 lbs./sq.ft. in gliders and to 60 lbs./sq.ft. in racing (c) Improvements in engines and propellers are type aircraft. At the present time, typical transport being made available to give better power- airplane wing loadings vary from 20 to 35 lbs./sq.ft. of weight ratios which improve take-off and emer­ area. gency safety. It has been unfortunate that many airplanes in the (d) Detailed studies of stability, control factors and higher wing loading brackets have had poor flying char­ wing stalling. acteristics for reasons not at all connected with wing (e) Improved wing flaps are, being used and studies loading. The most usual of these faults has concerned made on boundary layer control by the use of stalling characteristics. There is no fundamental rea­ engine power. son why an airplane with high wing loading should have (f) Studies are being undertaken on icing and re­ worse characteristics than those with low wing load­ lated problems toward a solution for keeping air­ ing. In fact, due to the higher stalling speed actual plane efficiency high under all possible condi­ control should be improved. tions. The fact that the stalling speed varies slowly with It is perhaps in the last field that development has wing loading is of utmost importance in determining been most tardy. This subject will be discussed at the economic value of the airplane for transport use, as quite some length in this paper as it is one of the critical comparatively small changes of wing loading result in factors affecting wing loading. a great change of pay load. For every condition of flight except stalling there is FACTORS AFFECTING WING LOADING some combination of power and wing loading that will The problems related to wing loading which now face give a constant performance. In a short time, power the aviation industry, come under two broad groups. will be used for the landing and stalling speed condi­ These are: tions to control the airflow around the wing in such a FLIGHT PROBLEMS manner that even the minimum speed factor will be a 1. Handling characteristics— 2. Take-off and climb perfor- function of power. From this point onward there will landing speed, stability, mance maneuverability 3. Icing effects Received March 5, 1940. 4. Pilot technique 43 44 JOURNAL OF THE AERONAUTICAL SCIENCES

AIRPLANE 'Cs

\j SPAN 9SF7. [ LENGTH 7SFT ^d WING LOADING-40LS/S$.FT GROSS WT. 44 000 LBS.

A AIRPLANE D*

84 FT. 52.3 rr. WINGLOADING 2i./L£/soFr W/A/G LOADING-JO lO/SO FT CROSS WT: 19,000 L as. GROSS WT 4SJO00L3S

nic

FIG. 1. Comparative planform drawings of airplanes "A" and FIG. 2. Comparative planform drawings of four-engined air­ planes "C" and "D."

ECONOMIC PROBLEMS 1. Payload carried 5. Hangar size, servicing prob- 2. Performance gained lems (b) Both airplanes meet C.A.B. requirements struc­ 3. Airplane initial cost 6. Crew required turally. 4. Operating costs for given (c) Both airplanes have the same stability. The speed airplane with the larger wing must have a cor­ The preceding items will be discussed and enlarged respondingly large vertical and horizontal tail. upon. The best way to make various points clear is by The length must also be greater, but due use of examples which cover the present design trend to the smaller relative e.g. travel on the lightly for the coming series of transport airplanes and also loaded airplane, the fuselage length does not cover the type of equipment now being used. The vary directly with the wing span. problem will be attacked in the following manner: (d) The wing airfoil section and aspect ratio must 1. Two bi-motor airplanes will be designed to study be the same for both airplanes. current transport types. The first has a wing loading (e) The fuselage cross-sectional area is a constant. of 31.8 lbs./sq.ft. and a gross weight of 17,500 lbs. This (f) Crew and passenger facilities are the same for airplane, designated as "Airplane A," has already been both airplanes. flying for quite some time so its performance and weight (g) Both airplanes have the same type of high characteristics are known exactly. For the low wing wing and . loading class, a second bi-motor, "Airplane B," is de­ 2. To study projected transport designs, two four- signed, with a wing loading of 21.1 lbs./sq.ft. Both engined airplanes will also be considered. Four-engine airplanes, to be comparable, must fulfill the following airplane "C" has a take-off wing loading of 40 lbs./sq.ft. conditions: Airplane "D" "is designed to have a 30 lbs./sq.ft. wing (a) The airplanes must carry identical payloads the loading. Both four-engine airplanes have Wright same distance when equipped with a given power G205A two speed supercharger engines, and are in­ plant. For the example considered, the payload tended to carry approximately 8500 lbs. of payload is 3775 lbs., the distance 950 air miles and the 1100 air miles at 250 m.p.h. cruising speed. The crew engines Pratt & Whitney Hornets. consists of 5 persons. Tricycle type landing gears are WING LOADING ASPECTS OF TRANSPORT DESIGN 45 used and the Lockheed-Fowler flap as a high lift de­ TABLE 1 vice. The four-engine aircraft are designed on a com­ Physical Characteristics of Four Airplanes Studied parative basis to meet the requirements of Item (1 b-g) * Two speed superchargers. f Tail area ratios shown, expressed in terms of the basic tail above. areas of Airplanes "A" and "C," are required to keep stability- Figs. 1 and 2 show the comparative sizes of all four factors constant. airplanes discussed, while Table 1 summarizes various physical and performance characteristics. ! ITEM TWO ENGINE AIRPLANE FOUR ENGINE AIRPLANE Airplane A B c D It is of utmost importance to make careful weight analysis for all airplanes involved in order to keep the No. of Engines 2 2 4 4 1 Engine Type P& W P& W W.A.C. W.A.C. comparison valid. The airplanes with light wing load­ Hornets Hornets 205A * 205A* ings have naturally higher wing and tail weights than 1000 at 1000 at 4500 ft. 4500 ft. Rated BHP and Alt. 800 at 800 at and and those of higher wing loading, although the weight per 5000 ft. 5000 ft. 900 at 900 at sq.ft. of area is considerably less. To evaluate landing 14,000 ft. 14,000 ft. Gross weight—Lbs. 17,500 19,000 40,000 45,000 gear weight, care must be taken to use tire and wheel Wing Area—Sq. ft. 551 900 1,000 1,500 sizes currently available. In some cases, this will Wing Loading—Lbs./sq. ft. 31.8 21.1 40.0 30.0 Power Loading—Lbs./BHP 10.94 11.90 10.0 11.26 penalize one aircraft more than its competitor. Empty Weight—Lbs. 11,000 12,340 25,330 29,100 Useful load—Lbs. 6,500 6,660 14,670 15,900 For a given cruising speed, the airplane with the Pay load—Lbs. 3,775 3,775 8,597 8,504 light wing loading requires more engine power and Range with above payload 950 950 1,100 1,100 therefore more fuel than one with a higher wing loading. Speed for above range—MPH 215 198 250 250 Wing span—ft. 65.5 83.7 95 116.2 This subtracts from the net useful load. Table 2 Wing area ratios 100% 163% 100% 150% gives a complete study of various factors involved for tHorizontal tail area ratio 100% 163% 100% 150% tVertical tail area ratio 100% 163% 100% 150% two different operating problems of the four-engine Aspect ratio 7.78 7.78 9V03 9.03 Type landing gear Normal Normal Tricycle Tricycle airplanes. Overall length The ranges taken for a comparative study are ac­ Change in wing weight—Lbs. 800 1,860 tual air miles flown against zero wind with no reserve Change in tail weight—Lbs. — 205 — 540 Change in fuselage wt.—Lbs. — 230 — 850 fuel provided. Actual operating air-line range would Change in gear weight—Lbs. — 85 _ 270 therefore be about 50 per cent of the distances indicated Change in Misc. weight—Lbs. — 20 — 250 Change in empty weight—Lbs. , _ 1,340 — 3,770 in all cases. Approximate airplane cost $80,000 $89,700 $253,000 $291,000 Considering the above, it will be seen that the two- engine airplane with light wing loading is 1500 lbs. TABLE 2 Comparison of Range and Payload Data on Four-Engined heavier than the one with the high wing loading while Airplanes "C" and "D" the corresponding figure for the four-engine airplanes Wing Loading vs. Payload: is 5000 lbs. A study of various similar airplanes will convince anyone interested in the accuracy of the pre­ Airplane "C" Airplane "D" ceding weight study. Gross Weight 40,000 lbs. 45,000 lbs. It is well to note the comparative dimensions of the Empty Weight 25,300 29,100 airplanes involved from a point of view of hangar Useful Load 14,670 15.900 space required and also the selling prices. Current two- CREW WEIGHTS 2 pilots 340 lbs. 340 lbs. engine transports sell for approximately $7.30 per pound Radio operator 170 170 Steward 150 150 of empty weight, while the less numerous four-engine Stewardess 130 130 airplanes are available for about $10.00 per pound. Due to the development of low drag, high lift wing RANGE DATA ' Cruising Speed—MPH 250 250 flaps, the problem of maneuverability at slow speeds and Cruising Altitude—Ft. 12,000 12,000 Power required per engine—BHP 650 790 the actual landing speed problem has been kept well in Fuel consumption—specific .435 .45 Case A < Fuel used per hour—gallons 189 237 hand. Most of the present types of transport airplanes Miles per gallon of fuel 1.32 1.055 Fuel required for 1100 miles—Gal. 833 1,041 are equipped with split type wing flaps which are not Oil carried for 1100 miles—Gal. 38 48 of great benefit in improving maneuverability at ap­ V Weight of fuel & oil for 1100 miles 5,283 6,606 proach speed conditions. Transport airplanes now be­ 1 I Cruising Speed—MPH 200 200 Cruising Altitude—Ft. 8,000 8,000 ing designed will make definite use of flaps for maneuv­ Power required per engine—BHP 455 520 Fuel consumption—specific .42 .425 ering, so that it is in order to give an example of equiva­ Case B < Fuel used per hour—gallons 127.5 147.5 Miles per gallon of fuel 1.57 1.35 lent wing loadings for equal maneuverability. Con­ Fuel required for 1100 miles—Gal. 700 815 Oil carried for 1100 miles—Gal. 32 37 sider Fig. 3, which shows the lift obtainable at various V Weight of fuel & oil for 1100 miles 4,440 5,168 angles of attack for an airplane with flaps retracted Payload for 1100 miles range at 250 and a 27.8 lbs./sq.ft. wing loading, compared to the MPH at 12,000 ft.—lbs. 8,597 8,504 same curve for an airplane with modern flaps set for Payload for 1100 miles range at 200 approach maneuvering condition. When flying at 120 MPH at 8,000 ft—lbs. • 9,440 9,942 Payload for 1900 miles range at 200 m.p.h.i the unflapped airplane has an angle of attack of MPH at 8,000 ft—lbs. 6,200 0,163 slightly over 5° and it has reserve lift to make a 2G pull- 46 JOURNAL OF THE AERONAUTICAL SCIENCES

FIG. 3. FIG. 4. up without stalling, for the wing loading of 27.8 lbs./ loading on controllability will not be discussed sq.ft. The flapped airplane flying with one-third flap here, but has been thoroughly studied.* The conclu­ extension at 120 m.p.h. is carrying 38 lbs./sq.ft. of sions resulting from the study of the rudder control­ wing area and has exactly the same maneuverability lability problems do not show that any great difficulties in that it can make a 2G pull-up without stalling also. will be encountered in obtaining rudder control after A 2G pull-up is equivalent to a 60° bank. This ex­ engine failure. ample shows the value of the proper flap type for ma­ TAKE-OFF AND CLIMB PERFORMANCE neuvering. The values of wing loading shown corre­ spond roughly to the landing values for the four-engined In order to evaluate an effect of wing loading on per­ airplanes having 40 and 30 lbs./sq.ft. take-off wing formance and climb it is necessary to make a complete loading, respectively. There is often considerable performance analysis of the four airplanes referred to comment that when the flaps are extended, the air­ above. Without discussing the details, the basic air­ plane will have a very low rate of climb but this is not plane characteristics are presented on the following true for the example shown. With one engine inopera­ two curves. The data shown applies to the full scale tive, the four-engined airplane with the 38 lbs./sq.ft. airplanes and it is based on a large number of wind- wing loading will have a rate of climb of over 600 feet tunnel and flight tests. While the airplanes with the per minute at normal (not take-off) power. lower wing loading have lower drag coefficients than those with higher loading, their total drag is consider­ The conclusion can be drawn that as far as handling ably greater than the drag of the latter airplanes for al­ * characteristics and stability are concerned, higher wing most every flight condition. The curves labeled "with ice loadings offer no serious problems. It might also be formation" will be discussed later. (See Figs. 4 and 5.) said that present maneuvering speeds can still be maintained on a well designed airplane by proper use A comparative performance of four airplanes is sum­ of wing flaps although the limiting landing speed will marized in Figs. 6, 7, 8 and 9, as well as in tabular increase slightly over values now in use. The latter form as shown on page 53 (Table 3). factor is not considered serious when good stall char­ The outstanding advantage of the heavily loaded acteristics are available, as they certainly must be, and airplanes in speed and climb at normal altitudes is very when ample power is used to give good climb in emer­ * Rudder Control Problems on Four-Engine Airplanes, C. L„ gency. The problem dealing in the effect of wing Johnson, S.A.E. Journal, June, 1940. WING LOADING ASPECTS OF TRANSPORT DESIGN 47

FIG. 7.

mm ii ipra BBKHimitffl WW BIBB IHFIIS HlHi ^^S FHn jjjijjjjp §Wi§ fflf^iliUilliil •Wt fflsfPwwpl wt ili PPiiePftl |ff»IINp4™wni •im- III jiiMiSfli w-N?mplfplliiiil f ;t?|h h; jt^ip Km[fH^|jif£i||jJI||Jil ^ Bffliff iillilli fill iilliBi PUP wmm WW l|MS gsafp 9 ii| Wiji™ MMfli^in K Hajprnmnjlji SHIS Iff m\\\ iifiiiiffli lill i#;i!§pipi i BPfflif fl^iwilli^BPII fteftiii W ii iniiP P i • »^jPI M Mat™ F# 11•Wf 1; «1I1 liiPi§-PP= H dfeii I^^^fc$MW$ittitMMf MMmmwkT ii r BIBPlMPiilPWiiiM M £#|gf#Pf e sWW Wl|Mik#d ^^MMf^mwk P#H4i4 Ssipi MMfetfe w ifiiiisyssWmB ijilpt iU ffiliib!a!ili!i!lii!ll4li |3J$3iH=j,%t SIIIIPI^^^B iji liii IWmkaiifPiM ?1 SgffipHli l •fl&ffli'iiW;!- RfraFi2F'' aMfrlsHisH^W' #PP llll! wm S§3iB<]M SSlBftilT1!!! Mm fflPBpf!|j{)lHi fffir ifi m ^^^P^Bi^K IKMBUBT lllllli MM asM Wi fiSfJIl ffBIBSSWBi ill Hfl^B pjjpggshj ASfflrfj|^fflH?^v,m ^|!ffi^||||||i|MH|t J1J Kmij i {H "'i| f|tjt h - •ixtep •' i llffii mil fflgigpik 81TOl«Bi^»» IMIfS *fflfflBPHHSIiffll I jpjsfijl i£pn FIG. 5. Hip li| mpipHipiKffli^ Si PHSiliJw'^™ H1 ™ jiffl Ipjpfg fSlpyi PwPppMnflMl Jjijl jjtijjlijjjgjjm j§§§j 5 1 HSHIisBilil B I Us BE jlll B^B strikingly shown. In both cases the smaller airplanes jgiliigafBBjjjej H^S^SH mlMwmMmMM\Mw\ym PB S llii&lffl with higher wing loadings are approximately 26 m.p.h. FIG. 8. faster than their comparable designs with lighter wing loading. There is a slight difference in absolute ceiling in favor of latter designs, but the low altitude climb is will use 22 per cent more engine power and a correspond-' better for the higher wing loading. The great difference ingly greater amount of fuel than that used by Airplane in speed for a given power is an important factor in C. The fuel cost alone for Airplane D will run over the economic operation of the airplane, particularly three cents per mile more than that of Airplane C. when it is considered that less fuel need be carried to go When all factors are considered, including first cost and a given distance, and therefore the payload can be in­ the effect of size on maintenance, the probable added creased. Airplane D, at a cruising speed of 250 m.p.h., operating cost per mile would give a differential of

111? ippf ^SSK^^BHBSW^9^&9S 1pppppi l P! Tpp JSjlppl Pll llllip |M Jis WIP ip ipip illj pifiil iff |M pi ||ii liii ii;jii;p|ii||| 1 ijlj ||jl ijijij||^ iipii ^^BH ij jjji liiipiil Si gjl ft

jujljjflpJRllMlik;:!;•;;•!; pk "*4^ijKt^i;l#if 4ijn;|.;:iji

; ^TipT^P3 ;|::;l;;;,b;;;;;:i:!;:;j';i:b; ipfemlps& 11 Syppfapl iiilp^^ lii^^B ^n^Mp ifl l'ql::fe||ipp|ijpp ^^^^^^^^^^^K li:j Ijii ill HI! 1 piiip^liillllllii ' ft ,1 iijpj ijii jjijii ijfy ^^^ 1 I^JUi^^^ B ^^^K

! Illillllllllgligllliilllililllllllilllllll P mm ilfeliraiiii m ppp P| j |!ife#"p iifpir! PPpI &;: :|jl;p iAfeM : Umm MM \wM:m fm.' lii•'r-i l^lv;|^,{gji. "isiyi•|.:..l«fl'.:.p i: WM;i Wi%- Mm FIG. 6. FIG. 9. 48 JOURNAL OF THE AERONAUTICAL SCIENCES

FIG. 12. Ice formation on propeller spinner. Cowl ice FIG. 10. Ice formation on Electra 10-A airplane. Note has fallen off. ice formation on top deicer attaching strips and rivets.

FIG. 11. Fuselage and pitot tube icing—Lockheed Electra. FIG. 13. Tail surface icing. five cents. The engines for the lightly loaded airplanes forms; a hard glaze ice has been reported as forming must run continually at higher power outputs than on an airplane at temperatures as low as — 40 °F. that required for the higher wing loaded airplanes, so Under this condition the ice was so smooth that no engine maintenance and reliability is actually affected detrimental effect could be noticed in the airplane per- directly. This gives an added safety margin to the formance. Perhaps the most general type of ice which airplanes with the heavier wing loading and to this forms on aircraft is a fairly soft rime ice which is very factor must be added, also, greater range, speed and rough and which can form in large amounts very climb. These additional safety margins go a long way quickly. There are little actual data available on the to compensate for a slight increase in landing speed. weight of ice that airplanes have picked up, but in ex­ The problem of take-off with heavy wing loadings treme cases it is estimated that this weight might be may be solved by the use of higher power and take-off as much as 6 per cent of the gross weight of the air­ flaps. For the airplanes considered, the present C.A.B. plane. take-off requirements will be met in all cases. The relation between wing loading and icing has not been carefully studied to date and the opinion exists AIRPLANE ICING among many pilots that an airplane with a higher wing The icing problem is one of the most important ones loading would be affected more detrimentally by ice facing the aviation industry today. There are a large than one with light wing loading. This is not an ac­ number of types of ice and conditions under which it tual fact. The airplane with the larger wing area WING LOADING ASPECTS OF TRANSPORT DESIGN 49

EFFECT- OF /JrE OX AtffPLANa Q&AG W/A/d Tl/WEL l£ST JIT itio<\mffim%nsm JLL \FLAP3ANO GE/t/P WETAACT&5 L4 tf"

wm NO YCE

FIG. 14. Wind-tunnel model prepared for simulating icing AtRkLWZ /C£ tests. y-cA^g CA$E B'

f

VRAG \COEfFAC/EA/T± C w 08 J% W 20 .24

FIG. 17.

the indicated high speed with all the power available to FIG. 15. Simulated cowl icing. the pilot with carburetor heat applied, was only 90 m.p.h. at 5000 feet. Practically full power was re­ quired for landing and only the pilot's skill prevented serious consequences in this particular case. The pic­ tures were taken after the airplane had been standing in a heated hangar for one hour in which time all of the cowling ice fell off the ship. Every rivet on the out­ board wing panel is accentuated by ice. As a result of this and other experiences, a series of wind-tunnel tests were undertaken to determine the possibility of repre­ senting ice effectively on a wind-tunnel model. It was found possible to obtain a very gooti representation by applying a mixture of plaster of paris and airplane dope to the various parts of the airplane as shown in the following figures. Complete tunnel tests were made and the effect of FIG. 16. Tail on wind-tunnel model showing preparation ice on the different parts of the airplane evaluated care­ for icing tests. fully. Changes of lift, drag and stability were meas­ ured under widely varying conditions. The ice con­ picks up more ice due to the larger area exposed. Fur­ ditions represented by the curve data included in this thermore it remains in the a longer report represent a very severe icing condition of a type period of time. In order to evaluate this problem, the not adequately removed by the present type of wing Lockheed Aircraft Corporation has in the past two deicer. In certain cases, ice has built up on the de- years been making a large number of wind-tunnel tests icer cap strips to such an extent no benefit has been on the effect of ice on airplane performance. The re­ derived from the operation of the deicing boot. Fig. 17 sults of these tests were checked in flight on a northern shows the results of ice formation on the complete air­ airline. Figs. 10, 11, 12 and 13 are pictures of a Lock­ plane as indicated by the photographs in Figs. 14, 15 heed Electra airplane which was iced to the extent that and 16. The most important result due to the presence 50 JOURNAL OF THE AERONAUTICAL SCIENCES

FIG. 18. FIG. 20.

EF/ ECT 0 c /CE ON Ah .£>?tf/V COA TROL Vim/ 'LARSt ?£7??/J CTFD FUL L 4IL ERON DEFLL "C770/ I

L4

A 1.2 <~~~ —*-—*• / I.O /-/U ?/C£ ( W Al ?/=£/?/ £ ¥\ ^ r qvc 8 > ^\ I -<^-^ k X / .6 // / 9IRPLA «f/KTi l/CE X .4 /I 1

4 f • 2 1 £*=± ?OU/A ^ -. 0 5 -0 4 ".0 3 0 2 -tf t t )

FIG. 19. FIG. 21. WING LOADING ASPECTS OF TRANSPORT DESIGN 51 of rime ice is a very substantial reduction in the maxi­ mum lift available. This reduction amounted to 32 per cent for the case with the flaps retracted and 23 per cent with the wing flaps extended. At the same time the airplane drag increased 47 per cent in the cruising condition. (See Fig. 18.) In a test with the ice re­ moved over the deicer, but with a P/2 in. ridge of ice on the cap strips of the deicing boot (dimensions shown for full scale), the lift was reduced and drag increased more than for the condition where the was iced completely. One of the most serious effects of icing was the great reduction in angle of attack for maximum lift when ice was present. (See Fig. 19.) Because of the fact that the normal stalling angle is reduced from 13° to 9° the pilot is confused because he has no means for knowing when he is approaching the stall. If the airplane without ice would give a stall warning due to tail buffeting at a speed of 5 m.p.h. before the stall, under icing conditions this would not be obtained, because the airplane stalls at a much lower angle of attack. Only an effective angle of attack in­ dicator can solve this problem unless the ice removal indicate the type of ice which can be accumulated on an is complete. It will be noted in the following figures' airplane. The ice accumulation between the that in spite of the decrease of lift, both with flaps and fuselage shown in Fig. 23 was obtained in five extended and retracted, the airplane characteristics minutes under heavy icing conditions. This is a good remain normal and flaps should be used for landing. example of insufficient deicer coverage. The other A test was run to determine whether icing of the gap figures show the effect of rivets in adding to the general between the wing and the flap would have any serious roughness. (See Figs. 24 and 25.) Inasmuch as the effect on the aerodynamic characteristics. Outside of a new transport airplanes will be very smooth, this effect reduction in maximum lift of approximately 10 per cent will not be present. Fig. 26 shows the effect of ice for­ there is nothing to fear from this problem. mation or of any spanwise protuberance on the maxi­ mum lift of an airfoil. Two effects are shown on the The effect of ice on longitudinal stability is normally curves. One is the fact the higher the protuberance the small (see Fig. 20), in fact for the condition tested there greater the disturbance and the other is the important was practically no change of stability. It should be effect of the chordwise location of the element. It will pointed out, however, that the longitudinal stability be seen that for a one-half in. high ice formation located will be affected if the icing conditions for wing and tail at the 5 per cent chord point on a wing with a chord of are not fairly similar. For instance, if the wing was 100 in., the maximum lift of the airfoil is only 54 per heated to remove all ice and no provision was made cent of the lift of the basic wing. If the protuberance for the tail, there might be an adverse effect on stability is moved back to the 15 per cent chord point the maxi­ and control. mum lift is 86 per cent of that of the basic wing. Most The presence of ice on the wing leading edge ahead of of our present day deicers extend only to the 5 per cent reduces the aileron control substantially as chord point on the wing. (See Fig. 27.) Therefore, shown in Fig. 21. The amount of rolling moment when any substantial ice deposit builds up behind the available with full aileron deflection is reduced to 64 per cent of its normal value for the ice conditions simulated. It was interesting to note that the stall character­ istics of the airplane tested improved when ice was on the wing. This effect has since been noted also by pilots flying the airplane in question. When fixed leading edge slots are used, ice will form on the lower lip of the slot and in some cases close the slot com­ pletely. (See Fig. 22.) This effect is easily counteracted by a small deicer shoe installed on the slot lip. When the preceding tests had been completed, flight measurements carried out in conjunction with a northern airline gave almost exact agreement with the results obtained from the FIG. 23. Four inches of ice accumulated in five minutes wind-tunnel test. Several photographs are shown which under heavy icing conditions. 52 JOURNAL OF THE AERONAUTICAL SCIENCES

FIG. 24. Wing ice accumulation.

FIG. 25. Tail ice accumulation, showing curious forma­ tion obtained. Note insufficient deicer coverage. deicer or on the cap strip which holds it in place, we immediately obtain the maximum spoiling effect which it is possible to obtain. This is one of the main reasons for the poor performance of present aircraft under icing conditions. Two other factors having the same effect are insufficient deicer coverage and clinging of the ice to the deicer boot itself. Several steps have been taken to eliminate the last two mentioned factors and at the present time the Lockheed Aircraft Corporation is co­ operating with the deicer boot manufacturers in making wind-tunnel tests on an improved deicer which will extend back three times as far on the top surface of the wing as our present deicers and twice as far on the bottom. Icing affects not only the lift and drag of the air­ plane, but also engine operation due to carburetor FIG. 27. Developments in wing deicing methods, (a) Present deicing boot (wings not generally flush riveted), (b) icing and a number of various airplane accessories. Improved deicing boot (for next series of transport aircraft— Figs. 28 and 29 show icing wind-tunnel tests, made wings very smooth). Much study now being given to use of engine exhaust heat to determine the effect of the presence of guide vanes in for deicing. Definitely promising when mechanical problems an air scoop for supplying air to an engine or for ven­ are solved. tilation purposes. It will be noted that when the vanes were installed in the scoop, complete closure resulted planes A, B, C and D has been evaluated. The icing in a short time. condition assumed has been very severe as shown in Figs. 14 and 16 and this is particularly true for the four- ICING AND WING LOADING engined airplanes. The larger the airplane, the less Using as a basis the information derived in the wind- the relative effect of picking up a given amount of ice. tunnel tests indicated above, the performance of Air- This factor has not been included in the computations WING LOADING ASPECTS OF TRANSPORT DESIGN 53

FIG. 28. Air scoop with guide vanes tested in Goodrich FIG. 29. Same scoop as Fig. 29 without guide vanes. icing tunnel. Note complete closure. Scoop remains open. shown as it is counteracted by neglecting the weight of have as high three- or four-engined ceilings as the lightly the ice picked up. Performance curves for all airplanes loaded ones, they have better climb at normal alti­ are shown with and without ice and also for Airplane tudes. The stalling speeds are affected more than any C equipped with improved deicers and under icing other performance item and it is primarily in connec­ conditions. It is not expected that the new type deicer tion with this factor that the new type of deicer will be will greatly decrease the drag at high speed, but con­ beneficial, although the three-engined ceiling for the siderable improvement in rate of climb and ceiling re­ airplane with 40 lbs./sq.ft. wing loading will be in­ sults from its use. Tables 3, 4 and 5 show the com­ creased by 34 per cent. parison between the various airplanes under similar Higher wing loadings are fully justified on the basis ice conditions as well as for the case with no ice. The of the evidence shown below. airplane with heavier wing loadings exhibit the same performance advantage with ice that they do without PILOT TECHNIQUE ice, although the increments of speed and climb are not The use of wing flaps for maneuvering will require so great. While the heavily loaded airplanes do not very little change in the piloting technique now being

TABLE 3 TABLE 4 Performance Comparison Four-Engined Airplanes "C" and "D" Performance Comparison, Twin-Engined Airplanes "A" and Wing Loading—Airplane "C"—40 lbs./sq.ft. "B" Wing Loading—Airplane "D"—30 lbs./sq. ft. * It should be realized that the condition of icing discussed in ITEM ICE ON ICE OFF this paper is extremely severe and seldom encountered. Many Airplane A Airplane B Airplane A Airplane B types of ice are encountered which are completely or satisfac­ HIGH SPEED-5000 ft 194 pl78 245 218 RATE OF CLIMB torily removed by rudder deicing boots. sea level—ft./min 850 820 1360 1210 t Present small-coverage deicing boots. ABSOLUTE CEILING two engines—ft 14,200 16,300 23,000 24,400 STALLING SPEED Haps up—MPH 110 90.3 ICE ON * i ICE OFF | 90 74 ITEM STALLING SPEED Airplane C Airplane D Airplane C Airplane D flaps down—MPH 81 66 70 57 SPEED FOR BEST RATE OF CLIMB—MPH 133 124 135 124 HIGH SPEED SINGLE ENGINE CEILING at critical altitude—MPH 230 211 294 268 feet None None 960U 10,100 REDUCTION IN RANGE 23% RATE OF CLIMB DUE TO ICE 20% sea level—ft/min. 1160 1080 1640 1460 CRUISING SPEED 65% power—10,000 ft.—MPH 166 159 215 198 RANGE ABSOLUTE CEILING with 350 gal. fuel at 65% power—miles 733 702 950 875 4 engines 19,600 20,500 26,400 27,200

ABSOLUTE CEILING 3 engines 13,000 14,600 20,700 21,600 TABLE 5 STALLING SPEED flaps up 135 117.5 102 88.5 Estimated Effect of Improved Type Rubber Deicers on Air­ plane Performance RATE OF CLIMB on 3 engines at sea level Airplane "C"—Wing Loading, 40 lbs./sq.ft. •—ft/min. 520 490 940 840

STALLING SPEED ITEM ICE ON | flaps down 89.7 77.6 79.0 68.3 Present Deicer Improved Deicer

SPEED FOR BEST RATE 230 MPH 239 MPH OF CLIMB 150 135 155 140 HIGH SPEED at critical altitude RATE OF CLIMB—sea level 1160 ft./min. 1270 POWER REQUIRED ABSOLUTE CEILING—3 engines 13,000 ft. 17,400 ft. for 250 mph at 12,000 ft. RATE OF CLIMB altitude—BHP 4x650 4x790 on 3 engines at sea level—ft./mia 520 740 STALLING SPEED-flaps up—MPH 135.0 108.8.-" • POWER REQUIRED STALLING SPEED—flaps down—MPH 89.7 83.5 for 200 MPH at 8,000 ft. altitude—BHP 4 xVlO POWER REQUIRED 4 x870 4x455 4x520 for 200 MPH at 8,000 ft—BHP 4x710 4 x 670 . . . | 54 JOURNAL OF THE AERONAUTICAL SCIENCES used. There has been a habit among pilots to lower the landing gear before lowering the flaps in practically all cases. It is believed that this tendency has been the result of carrying over fear of forgetting landing gear extension from the time when retractable landing gears were rather undependable. Now that they have been improved so greatly and warning devices incor­ porated to remind the pilot to lower the landing gear, it can be expected that the pilots will often use a mod­ erate amount of wing flap when doing any maneuvering at low speeds. This should be particularly true on new aircraft because of the excellent climb with flaps ex­ tended. Fig. 30 shows the effect of various components of the airplane in changing the rate of climb at differ­ ent airspeeds. It will be seen that at 120 miles per hour an extended will reduce the rate of climb approximately 260 feet per minute while at 90 m.p.h. the reduction is 100 feet per minute. When the maneuvering flap setting is used the rate of climb is decreased by 200 feet per minute. Inasmuch as the 'FIG. 30. Loss of rate of climb, at sea level, due to various air­ plane parts—four-engined airplane "C." Weight 40,000 lbs. normal rate of climb with wing flaps retracted is 1640 feet per minute, this sacrifice of rate of climb is rela­ tively unimportant. both in the air and on the ground. Perhaps it would be well to point out that present day airplanes are land­ In regard to other changes in pilot technique, the ing as fast or faster than the values referred to in this most important will be in using the flaps for take-off. paper, primarily due to the fact present day aircraft No particular problems should arise here, although with fall far short of the desirable requirements in regard to normal type of landing gears there was a good deal of stalling characteristics. reluctance to use flaps for this purpose. This came about because the older type airplanes with a standard CONCLUSIONS landing gear .were subjected to a considerable amount of vertical tail blanketing when flaps were extended and From the preceding study, the following conclusions also in certain cases turning tendencies were developed may be drawn: due to this slipstream reaction on the wing flaps. Both 1. The use of increased wing loading is fully justified of these conditions can be largely eliminated by the use on a basis of improved performance, range and better of a tricycle landing$gear in which the vertical tail is economy. lifted clear of the air flow away from the flap during 2. Increased wing loading will lead to airplanes with take-off. Also the tricycle gear with its greater stability lower first cost and smaller size, both of which have can neutralize the turning tendency until the airspeed important effects on operating economy. over the vertical tail builds up. 3. The slight increase in landing speed will not be obtained at a reduction of landing safety due to im­ ECONOMIC FACTORS provements in stalling characteristics, control, sta­ All factors affected by increasing wing loading work bility and landing gear improvements. toward lower operating costs. The problem of hangar 4. The problem of icing is no more severe for the or storage space for future transports makes the abso­ airplane with heavy wing loading than it is for one with lute size of the airplane an important factor. One look light wing loading. The advantages of higher initial at Figs. 1 and 2 will indicate the advantage of higher rate of climb, higher speed and greater range still wing loadings from this point of view. remain with the heavily loaded airplane, even under severe icing conditions. PRESENT RESTRICTIONS 5. The icing problem is relatively less severe on large At the present time, the United States is the only airplanes than on small ones. Research now being country that has a definite restriction on landing undertaken by the various manufacturers and govern­ speed. To arbitrarily restrict designers from using ra­ ment testing agencies will lead to substantial improve­ tional rather than arbitrary requirements on airplane ment in performance under icing conditions. This will performance seriously hinders a healthy state of design further justify the use of increased wing loading. improvement to obtain many of the advantages given 6. Flight characteristics now common in large air­ above. The landing safety of an airplane is not deter­ planes should not be considered as necessarily apply­ mined solely by the speed at which it contacts the ing to future designs when changes in safety regulations ground, but rather by its controllability and stability are considered.