CHAPTER 6 Reducing vortex-induced drag: 2 Planform, twist, wing tips and winglets 6.1 Planform of wings Aspect ratio is by far the most important factor in reducing vortex drag, but some model aircraft designers throw away part of the advantage sained from high aspect ratio by carelessness in detail. As suggested in Figure 5.2, the aerody- samic or effective span of any wing is always slightly less than the physical length of the sur- face, because the tip vortex leaves the wing slightly in-board of the tip. A bad choice of tip shape, ora bad planform, or a wing with too much or too little twist (variation of rigging angle from place to place), reduces the effective wing span: the wing will behave as if it were smaller in both area and aspect ratio, The factor Kin the induced drag equation will enlarge. 6.2 The rectangular wing The easiest type of wing to build is one with rec- tangular plan and the same cross section from tip to tip. All the ribs or templates are identical and there are no awkward joints in leading or trailing edge members. Such a planform is not the best zerodynamically, the basic reason being that some parts of such a wing are underemployed, not car- rying their fair share of the model's weight. The circulation of air and vortex strength over one part of a wing influences the direction of flow over the adjacent parts and changes the local angle of attack. With a rectangular wing the tip vortex is strong and hence the downwash near the tip is large. The closer a segment lies to the tip, the more A lively sports model of straightforward design. The rectangu- Jar wing plan ensures safe stalling characteristics and the generous tail volume ensures stability, yet with full span ailerons the mode! is capable of a full aerobatic schedule. (Photo: Model Flight) 55"MODEL AIRCRAFT AERODYNAMICS Rectangular wing plan Maximum possible section tit coetficient : — ntl Ltn Tive refaiaty ide . oat Xe, seo - aan Strongly tapered wing Stal! Maximum possible section if coeticient Dangerous tip stal Arrows show local lit coofficionts Moderately tapered wit TTL ‘Arrows show local ft coefficients Mex poss, Elliptical wing plan Maximum possible section iftcoeftoiont TMT Simultaneous stall Arrows show local it coefficients Figure 6.1. Planforms, load distribution and stalling. it is influenced by the vortex. The section angle of attack near the tip, and hence the section c,, is reduced almost to zero. Thus, although the wing chord is constant, the load carried by each chordwise section falls off sharply towards the tips. The wing, even with no geometric twist, works at reducing aerodynamic angles of attack over the whole outer span, with the result that the load distribution resembles that sketched in Figure 56 ‘Load distribution yea distabuton| Load aistribution 6.1A. (In this diagram the rectangular distribution of area is shown, and the actual load carried by each part of the wing is shaded with a faint outline of the area marked in.) ‘Assuming the wing is of constant aerofoil section throughout, the maximum possible section cp at each spanwise point is the same. This is indicated by the heavy black line in the left-hand diagram. Since, however, the tips are at a lower» . REDUCING VORTEX-INDUCED DRAG: 2 PLANFORM, TWIST, WING TIPS AND WINGLETS seaic of attack, they are still well below the stall- angle when the roots reach it (as indicated by Se vertical arrows}, Because of the vortex-induced Sewnwash, the angle of attack at the tips is always tess than at the root, so the outer parts of the wing © approach their maximum possible lift coef- Scent. The rectangular wing thus has an inher- safe stalling character, the wing in the centre while the tips are still well below the stall. » is no tendency for a tip to drop first. If the sedel is radio controlled, the ailerons remain effective over the outer sections even when the is on the verge of stalling (arrow heads ‘hing the black line). Such wings need no geo- setric washout, but this is at a cost in terms of =Seetive wing loading. If some of the area were seen from the tips and distributed over the central eetions of the wing, the wing as a whole would | later, at a higher total G,, since the total C, is sede up of the average of all the section ¢ across span. When a model wing is drawn out with a certain area, itis too easily assumed that each part of ‘ll carry a share of load which is in proportion he area at each point. This is encouraged by the simple expression (see Chapter 2) for wing load- ing, WIS, which implies that weight divided by total area gives a true standard of comparison between models of various types. If, however, because of bad planform, some parts of a wing are relatively idle, the rest has to take on an extra burden. This implios that while one portion of the wing is lifting very little, working at a low G,, the other parts are at high C, to compensate. As the equation for induced drag shows (C,, = [C,2/3.142 x A] A) increasing section lift coefficient at a point along the span causes vortex drag to rise propor- tionally to the square of ¢, so the drag of the whole wing is higher. At the same time, the idle part of the wing still contributes some skin drag and profile drag. Such a bad distribution of lift is reflected in the planform correction factor, k. Noticeable losses of performance can result. 6.3 The strongly tapered wing The converse of a rectangular wing is one such as that shown in Figure 6.1B, strongly tapered with tips almost pointed. This is not only very ineffi- The RhOnadler was a very suacessful (full-scale) sailplane in the 1930s. It had a strongly tapered wing which was efficient at lew speed, but roquirad very strong washout and change of wing profile at the tips to prevent dangerous lip stalling and spin- At moderately high airspeeds the tips were at negative angles of attack and ‘lifted’ downwards, bending the wing down ruining the penetration.MODEL AIRCRAFT AERODYNAMICS cient but dangerous. The strength of downwash over the various parts of such a wing is such that the local anglo of attack increases towards the tips, where the area is smaller. There is an aerodynamic wash-in, the tips are overloaded and stall first ~ indeed, with a wing like that sketched, they would be almost permanently stalled (arrows rising above the maximum ¢, ine). The narrow outer panels are called upon to provide far more lift than their sec- tion c,max. permits, while the roots contribute lit- tle, The model would fly better if its outer wing panels were cut off altogether. Shown in Figure 6.1C is a compromise. The wing is tapered but only moderately, achieving some worthwhile improvement in efficiency by matching the load distribution more closely to the area. The stall is likely to occur somewhere in the middle rogion of the semi-span. This, in many cir- cumstances, will cause a relatively mild tip stall effect but much less than the strongly tapered wing. 6.4 Washout ‘The strongly tapered wing does possess one advan- tage: because it has a very broad and thick root, it may be very lightly built without loss of strength. For this reason some early full-sized sailplanes such as the Rhénadler of 1932 adopted this type of wing. The tip-stalling problems were avoided by giving the whole wing a marked twist or washout to reduce the geometric angle of attack over the outer panels by approximately the amount needed to equalise the downwash across the span. This tended to distribute the load more in proportion to the area and so reduced the vortex drag, The result was an efficient wing, but at only one airspeed. At the designed speed, the whole wing was working at roughly constant aerodynamic angle of attack, butat any other speed the distribution changed. In particular, as the speed increased, the wing tips reached their local zero angle of attack quite soon as the average angle of attack of the whole wing decreased. At any higher speed than this, the outer 58 wing panels actually operated at negative angles of attack to the local airflow, and began to ‘lift’ down- wards, Although this lift force was directed down, the resolved drag component was still directed aft. ‘Thus, not only did the tips of these highly twisted wings throw extra down loads onto the rest of the wing, but they added vortex drag. More impor- tantly, as the speed increased, the profile drag at the tips, operating at negative angles of attack, rose rapidly. From the cockpit the wing tips could be seon to bend downwards at some quite moderate airspeed, and the penetration suffered accordingly. The same effect may be observed on many model sailplanes with too much washout. Too: much wing twist, introduced to cure abad choice of planform, results ina ‘one speed’ wing. This may be exactly what is desired for an F1A (A2) sailplane, but not for any type that needs to fly at varying speeds. Even for an A2, wing twist renders the model more sensitive to slight trimming errors. A small depar- ture from the ideal airspeed causes a dispropor- tionate rise of drag. For reasons to do with high profile drag and promature stalling at low Reynolds numbers, cal- culations show that almost all the aerodynamic advantages of the tapered wing are lost on small models. For this reason rectangular wings are pre- ferred for all these. By careful use of washout, the tip vortices may be reduced in power even on such awing, at the single trim for minimum rate of sink. ‘Washout often proves very useful in preventing dangerous tip stalling on all models, particularly for scale types where the wing of the prototype is strongly tapered. Washout also aids aileron control at low speeds (see Chapter 7). It does not, however, produce the most efficient wing possible. 6.5 Wash-in Wash-in, the twisting of a wing to a higher angle of incidence at the tip, promotes tip stalling, in- creases the strength of the tip vortex and should be avoided. If, in order to trim a model for turning flight or to balance out torque reactions from the~ REDUCING VORTEX-INDUCED DRAG: 2 PLANFORM, TWIST, WING TIPS AND WINGLETS. propeller, it is necessary to rig one wing at a differ- ent angle from the other, this should normally be done by warping one tip toa smaller angle, wash- out, rather than washing-in the opposite tip. Wash-in has the same effects as drooping an aileron, Differential ailerons (see Chapter 13) are ared so as to mitigate the ill effects of this by raising (washing out) the aileron on one side much more than drooping the other. Adverse aileron, drag, causing an undesired yaw, is caused by the difference in vortex drag between the downgoing and the upgoing sides, 6.6 The elliptical wing Mathematical analysis and experiment show that the only type of wing that will produce, at all speeds, constant downwash and a load distribu- m exactly matching the area is one with ellipti- cal planform distribution (Figure 6.1D). The effective angle of attack everywhere is the same and the C, maximum is reached simultaneously slong the entire span (arrows all touch the black line together). This follows from the equal distri- bution of load, area for area, of the wing. Every part of the total area contributes its fair share of the lift- ing work. The perfectly elliptical wing plan exemplified in this 1942 design, the Hayseed by Robert Munn, buill now as.a ‘vintage’ model by John French. (Photo: Don Howio) In practice, such perfection, with the simultane- ous stall, is rarely achieved, since the wing is usu- ally slightly yawed or ‘one wing low’ prior to the stall, and the elliptical planform will then appear to cause tip stalling of a mild kind. Tip stalling is also encouraged by the lower Re of the outer wing. Itisalso important to note that in turns, especially the tight turns required often for thermalling in very light and small hand launched sailplanes, the lift distribution across the span of a wing is changed by the fact that the inner wing tip in the turn is moving through the air at a lower speed than the outer tip. The loading is changed and the innor tip is likely to stall first. ‘To prevent this an increase of outer chord above that of the pure el- lipse is needed. The perfect load distribution is also upset to some extent by the presence of the fuselage which disturbs the flow and, with some wing positions, may reduce the load-carrying capacity of the cen- tre section of the wing to nothing. For these reasons the ellipse is best regarded as an ideal to be approached as closely as possible, rather than the best practical wing planform. For small models the bad effects of low Re at the tip may dictate a rectangular plan, as mentioned above. To aim for an elliptical area distribution is not quite the same thing as saying the wing should be a perfect ellipse. It may be so, but any other form which gives a chord at each point the same as the pure ellipse will have the same effect. In Figure 6.2 some of the possible variations are illustrated (something further is said later about the crescent wing). 6.7 Compromises with the ellipse Perfectly elliptical wings are not easy either to lay out on the drawing board, or to build accurately. Many good compromise shapes are available and these can be made lighter, because of their simplic- ity, so the small aerodynamic losses may safely be ignored. Several popular examples are shown in Figure 6.3. Of these, the first was commonly found 59MODEL AIRCRAFT AERODYNAMICS “Leading edge salon win sony elise Figure 6.2 Altemative forms of elliptic chord distribution. on full-sized sailplanes during the 1920s and 1930s, and still appears on many models. If used, it might be slightly modified by employing a squared off tip, without serious loss of efficiency. The curved trailing edge is not altogether easy to build, Ifa built-up structure is used, ribs over this section of the wing cannot be made by the sand- wich method of construction. Each rib has to be plotted and cut individually. 60 “The second planform in Figure 6.3 is very good from a structural point of view, giving ample root depth for spars, and can be safe if the tip chord is nol reduced too far. The stall will occur first some- where about halfway out along the wing (Fig 6.1C) Construction of a foam-cored wing of this type is straightforward. There will often be a mild wing drop at the stall. This may count as an advantage for aerobatic models, assisting entry to spins, flick rolls and other manoeuvres of the kind. The uncontrollable tip stall of the too-strongly tapered wing, howover, is never desirable. Tip stalling can be avoided altogether by careful choice of aerofoil section, as will be discussed in Chapter 7. The third planform in Figure 6.3 is good in all respects except that the root tends to be somewhat cian cane panels BQ onew pari Hoome to | “Senuoman pantorn) —— Wess tan ha wasoutrtpostain ws aawwes Figure 6.3 Compromises with the ellipse. Wherever the out- line wing has a grey edge, this part of the wing will be over- Joaded in flight and will tend to stall first~ REDUCING VORTEX-INDUCED DRAG: 2 PLANFORM, TWIST, WING TIPS AND WINGLETS ‘hin. It is easy to build and the departure from the ellipse is quite small. A convenient place for a Sthedral joint is provided by the start of the tapered panel. Calculations by F X Wortmann suggest that such a wing is built with three degrees of wash- ut (i.e. reduction of incidence towards the tip), its performance is improved over a wide range of speeds. The washout should be progressive along she entire semi-span, rather than confined to the ‘“spered panels only, as modellers usually do it. ‘The fourth planform in Figure 6.3 is that cur- sently preferred by most designers of full-sized sailplanes. The approximation to the ellipse is very close and the wing root is deeper than for the previous type. To build such a wing requires alittle more work as an extra template is needed for the semi-span position. With a built-up structure it is possible to space the ribs more widely with lighter spars and covering over the outer panels, to save mass. Again, a convenient dihedral joint is provided, and on larger models the wing may be Scale models of high sonic and supersonic jot aircraft present many difficulties because the wings are designed for speeds at high Mach numbers. Different types of wing profile are necessary and the balance and stabilily equations alter markedly. At low airspeeds swept wings tend to stall at the ‘ips, but such models, with appropriate adjustments, caz fly very well and very fast. This is John Hull's F 15 Bagle, pon ered by a ducted fan, Note the downturned wing tips. (Photo: Don Howie} designed to part at the mid-span position for trans- port. It is argued by some aerodynamicists that a wing will produce a trailing vortex at every point where the trailing edge departs froma straight line. This is probably true where there is a marked change of dihedral, as on many model aircraft, or if there is a sharp break of taper. Following this argument some designers have adopted the last planform shown in Figure 6.3. The trailing edge of the wing is straight and is drawn at about 90 degrees to the aircraft centre line. All the taper, still approximating the elliptical chord distribution, is on the leading edge. The result is a slightly swept back wing with respect to the quarter chord line so the aerodynamic centre of the whole wing is aft of the root quarter chord, which must be taken into account when positioning the centre of gravity. Whether such a planform produces any detect- able gain in vortex drag is hard to discover in prac- tice, 6.8 Bad planforms Some models have appeared with the planform shown in Figure 6.4. This is no doubt intended to reduce the interference drag where the wing root meets the fuselage, but the effect is wholly bad. Vortices are created which reduce the effective span and aspect ratio of the wing. The result is somewhat similar to opening a gap between wing and fuselage at the root, in effect making the mono- plane wing into two separate surfaces. On some scale models of very early aeroplanes and gliders, this may be inescapable but it has drastic influence on performance. j= gn veto ag Ne Figure 6.4 Planform to be avoided. 61MODEL AIRCRAFT AERODYNAMICS 6.9 Swept wings Swept wings, back or forward, are used on full- sized military and commercial aeroplanes because they are beneficial for flight approaching the speed of sound. Low speed aircraft, particularly two seat sailplanes, often use swept wings to assist with balance and cockpit layout. The rear pilot of a sail- plane like the Blanik or Ka 13 needs a view to the sides and upwards. Ifthe wing is straight, balance requires the rear seat to be close to the main spar of the wing and the view is then seriously restricted. Sweeping the wing forward improves the outlook. As Figure 6.5 shows, the aerodynamic centre of a swept forward wing is often ahead of the wing root leading edge and the rear seat can be located here. Tailless aircraft commonly have swept back wings for reasons of stability (see Chapter 12). Small amounts of sweep have very little effect on vortex drag. Sweep forward actually aids con- trol at low speeds, delaying wing tip stall, but back sweep has the reverse effect and. control at low speeds may require the use of special devices such as slots, boundary layer fences, etc. Modellers using pronounced sweep back, perhaps on a scale model of a jet aircraft, may find similar devices essential. Sweep back also has a slight positive ‘Sweep back useful for high ‘Mach number fight. (ie. near speed of sound) ‘Sweep back promotes tip staling ‘Aerodynamic centre, For construction used to locate the ‘aerodynamic centre, see Appendix + Sweep forward delays tip stalling ‘Sweep forward often used to improve pilot's view from the cockpit. ‘Sweep forward also useful for high Mach number Figure 6.5 Swept wings. 62» . REDUGING VORTEX-INDUGED DRAG: 2 PLANFORM, TWIST, WING TIPS AND WINGLETS. edral effect and may be justified on aerobatic sodels which, in the absence of normal dihedral, d this effect useful for steadier flight both ‘averted and right side up. Sweep forward works in the opposite sense, lightly destabilising the model in the lateral direc- tion. 5.10 Points of detail With all tapered wings on models, the tip chord should err on the generous side to avoid tip stall- ing caused by scale effects and laminar separation. The aerofoil section at the tip should normally be thinner than at the root, for the same reason Whatever the planform of the model, serious osses occur if there are gaps through the wing at ny point. These can appear in flight, particularly n large sailplanes, if the wings flex and work slightly apart al rigging joints. Through such gaps the air flows from the high to the low pressure side of the wing, creating turbulence and reducing lift. Control gaps have similar effects and all such leak- ages should be caréfully sealed. 6.11 Tandems and canards 1 the total lift load required to support the weight of the aircraft is shared between two main lifting surfaces disposed fore and aft, as in a true tandem, layout, then as with the biplane there will be four instead of two tips and four tip vortices. As men- tioned before (Section 2.5) an aircraft which has a load-carrying tail, or forewing in the case of a canard, is in a strict sense a tandem and there will be some vortex drag created by the lift of the sec- ond surface. Only if the stabiliser is rigged to carry zero load in normal flight does the difference in pressure above and below the surface disappear. With it goes the tip vortex and the associated drag, Neither the tandem nor the canard permits this arrangement. (See further explanation in Chapter 12.) 6.12 Wing tips In terms of practical experience in flight, itis very difficult if not impossible to show that any particu- lar type of wing tip has‘a large advantage. This is the case with full-sized aircraft — it is even more so with models. Mere fashion sometimes seems to be quite influential. There is a tendency for manufac- turers and designers to introduce new shapes as soon as there is the slightest hint of experimental evidence to support them ~ the changes draw attention to their products. The promoters stress the latest research findings, hoping thereby to make more sales. Further changes, again with some slender scientific support, may follow after a few years. Compared with aspect ratio and the general planform and twist of the wing, wing tips are of small importance although, if a wing has a very bad tip, the resulting disturbance of airflow may promote tip stalling. Modifications to a wing tip may sometimes make a difference to control responses and stability which may be more signifi- cant than any drag saving. In the case of a radio controlled aircraft, aileron control may be affected. In general, however, the difference in performance between a model with a good tip shape and a poor one will be barely detectable in flight. There may be something to be gained by trial and experiment, but probably not very much. Practical aircraft ‘wings must end in tips of some kind, and wherever there is a difference of air pressure above and below a wing (or tailplane, fin, forewing etc.) a vortex will form at or near the tip. There will be a drag penalty - the greater the relative difference of pressure (i.e. the higher the G,) the more severe the penalty will be. 6.13 Tips for fast flying models Ithas already been emphasised that tip vortex drag is most important for aircraft flying at high lift coefficients, i.e. at high angles of attack, slowly. It follows that little vortex drag is to be saved on high 63MODEL AIRCRAFT AERODYNAMICS speed aircraft by modification to the wing tips. Parasite and profile drag are much more important for such models and the wing tip should be designed for savings here. The plain squared-off tip probably does lose something and should be rounded off in both front and plan views. What is important is that the wing tips should be simple. Fancy appendages undoubtedly add parasitic drag and should be avoided as should excrescences A Geometric span Etfectve span such as tip wheels, skids, etc. There is not much else to be done. 6.14 Wing tips for slow flight Sailplanes and other models trimmed for flight at high angles of attack will gain something from a tip design which reduces the strength of the tip vortex B Geometric span — Effective span Effective span tow pesae octal bt —— aaa Probable resut @== Slight upsweep of walrg op cased ‘yaverad user D Downturned tip ‘Thooretical basis cE raoiniase @ ———— Figure 6.6 Wing tip shapes 64- REDUCING VORTEX-INDUCED DRAG: 2 PLANFORM, TWIST, WING TIPS AND WINGLETS of this can be done, compels it to form further eut in the spanwise sense, Aerodynamically, the effective span of a wing is not determined merely the geometric span. If the vortex forms some- sshere inboard of the tip, which it nearly always, * not always, does, the wing loses efficiency in proportion to the inboard migration of the vortex. ‘This is often represented by a span efficiency fig- ure and very few real wings are better than 959% span efficient in this sense, Attempts to improve she wings of sailplanes and gliding models have ded to concentrate on devices to prevent the vortex forming inboard, thus seeking to increase she span efficiency as much as possible. It has not, swever, boon established that these methods suc d. Barly full-sized aircraft and many models ‘Twin fins as end plates for tai A Wing wake passes below taille ‘Severe loads on tusolage inad landing Tailplane as end plate for fin and rudder Wing wake may strike taiiplane Figure 6.7 Tail unit endplates. have wing tip shapes which encourage premature formation of the wing tip vortex. If, for example, the tip is raked forward (Fig 6.6A) or generously curved at the trailing edge (Fig 6.68), the vortex may be expected to form close to the point where the trailing edge curvature begins. It is generally accepted now that the trailing edge should run as straight as possible to the extreme tip. Evidence from wind tunnel tests and flight tests on full-sized sailplanes and some powered aircraft has been gathered over recent years to show that the airflow near the tip of a wing can be improved slightly by adopting a generally upturned form. ‘The first support for this came from the German aerodynamicist Sighard F. Hérner, whose book Aerodynamic Drag was published in 1951. A development of the Hérner tip was widely adopted for full-sized sailplanes, and is illustrated in Figure 6.6C. The tip is essentially square but the leading Fin vulnarablo to damage 'T Tail estrains vortex ‘only at top of fn Tailplane clear of fuselage ana wing wake Fuselage wake reduces. ficiency of tailplane sad Talplane restrains vortex at ‘base of fin only Taifplane may be given slight ‘inedral to reduce wuinerabity in landing 65MODEL AIRCRAFT AERODYNAMICS of 7 Camber inwards —_¥) Vortex: ‘Hin pressure Front sein Cambor outwards Cs vortex Upper winglet Lower winglet 3 : Resolution of forces on winglets Few rectonin ___—} sideways ‘main wing vertex ‘component = Forward component Figure 6.8 The Whitcomb winglet. edge is curved back to meet the trailing edge approximately at right angles. On the underside the wing tapers in thickness upwards to a crisp edge, rather than a rounded form. The purpose of this is to allow the high pressure air below the wing to sweep easily to the tip where its energy may serve to carry the vortex to the extreme limit of the span. In practice the full effect does not seem to occurbut the extension of the trailing edge to the extremity probably does carry the vortex slightly further out than with a rounded tip. The Hérner tip is easy to make, quite elegant in appearance, prac- tical in service, and at least has not been shown to be disadvantageous Tips that turn downwards, as shown in Figure 6.6D, have also frequently been used on sailplanes. ‘These should not be confused with the upswept 66 Horner tip described above. Their purpose is mainly to protect the wing tip and ailerons from damaging contact with the ground. Aerodynami- cally, the downturned tip might have some good effect, tending to confine the high pressure air and restrict its movement round to the upper side, the opposite of the Hémner tip. In an exaggerated form, if the camber of the wing is carried round all the way, the result resembles the lower part of a Whitcomb winglet (Fig 6.8). As before, there may be some benefits for aileron control and tip stall- ing, although little is known of this. More recently, sailplanes with distinctly upcurved and back swept tips have appeared, as illustrated in Figure 6.12. The outermost panel of the wing, usually made detachable, sweeps back slightly and has aslightly increased dihedral anglewhich blends to a Hérner tip, and at the trailing edge of this tip a small winglet is usually added {see Section 6.17). Because the winglets have the effect of changing the general lift distribution, the main wing is often slightly loss tapered than usual. for the Schempp Hirth Ventus 2 sailplane a reduc- Son in the minimum rate of sink of 6% has been claimed. (This amounts to a matter of 3cm per second or 7 inches per minute, to measure which in actual flight is extremely difficult.) What is probably of more importance to the model flier is shat the slight additional dihedral and the smoother flow over the tips improves the handling of the aircraft at low speeds and gives better aileron control. Not all aerodynamicists are fully convinced of the benofits and it is always difficult to distinguish gains in performance made by improved wing section design, turbulators and the introduction of sew structural materials, allowing wings to be shiner and stiffer, from the wing tip effects. 6.15 Tip plates and tip bodies When a wing is mounted so that it completely bridges the walls of a wind tunnel, no wing tip vortices form, as mentioned above. The bound vortex extends from wall to wall and is cut off cleanly. The wing then behaves as if it had infinite espect ratio and vortex-induced drag is nil. Attempts have been made many timés in the past to get the benefits of an infinite aspect ratio by fitting endplates or large, streamlined tip bodies which, on some full-sized aircraft, have also been used as fuel tanks. The plates or bodies are intended to act like the tunnel walls and prevent the formation of tip vortices. They do succeed in this to some extent, but to be completely success- fal they have to be very large. (They also steepen the slope of the lift curve, see Figure 5.8.) This is not achieved without penalty. The tip plates themselves cause form and skin friction «drag, and this parasitic total may easily be larger than the saving. This depends very much on the C, REDUCING VORTEX-INDUCED DRAG: 2 PLANFORM, TWIST, WING TIPS AND WINGLETS. at which the aircraft flies. Since at high speeds vortex drag is small in any case, while form and parasite drag are large, tip plates and bodies have a very bad effect and their use cannot be justified atall for aircraft that commonly fly faster than their speed for best 1/D (see Fig 4.10). At lower speeds, a model with a low aspect ratio wing may be improved by fitting endplates. According to tests carried out by A Raspet of Mississippi State Col- lege many years ago, the best size is about twice the wing root chord in length. Such plates would be a considerable nuisance on any practical model air craft. If the wing is already of high aspect ratio, large endplates of this type will have little effect since the aspect ratio is already high, the gain in drag from the endplate is proportionately smaller, and parasite drag no less. Tip vulnerability is greater. Plates, or tip bodies, smaller than the rec- ommended size do not inhibit the tip vortices enough to make much difference, and still add their quota of parasite drag. They should be avoided, The few full-sized aircraft and sailplanes which do have tip bodies usually do so for non-aerodynamic reasons: for example, asailplane wing and aileron end may be protected by a small Anighly effective, electric powered ducted fan model, based ‘on the Grumman F9¥ Panther, by Mal Pring, The span is 0.93 metres. Speeds in level flight exceed 145 km/h. The wing tip bodies may improve the airflow over the ailerons, but they add a certain drag penalty. (Photo: M. Pring) 67MODEI, AIRCRAFT AERODYNAMICS lip body such as that of the Blanik two-seater, or a powered aircraft may lack internal space for large capacity fuel tanks. The wing tips may be the best place for mounting external ones. 6.16 Tailplanes and fins as endplates ‘Tip endplates can be useful to increase the effec: tive aspect ratio of a tailplane or fin, with possibly good effects on stability and control response (Fig 6.7). A tailplane or canard forewing fitted with end fins will have a steeper lift curve slope and become more powerful. The end fins can serve as fins for the whole model so their drag will be hardly any groater than that of a simple central fin of equal, or slightly less, area. However, the twin fins will have four tips and their aspect ratio will be very low reducing their effectiveness, and this, with their structural vulnerability, makes their use of doubt- ful value. They may cause more trouble than they are worth in practice. However, the tailplane itself may act as an endplate to a fin, if itis mounted on top of the fin, in “T’ configuration, or if the fin is mounted entirely above it, This restrains the vor- tex at one end of the fin and increases its effective aspoct ratio. The ‘T-tail arrangement also carries the tailplane out of any possible airflow disturbance caused by the wing- root-to-fuselage junction, Both fin and tailplane may then be slightly reduced in area, which helps to compensate for the increase in structural weight caused by the necessity to stiffen the fin to carry extra loads. A high mounted tailplane is also less likely to blanket the fin dur~ ing the towline launches or spinning, It is more vulnerable in ground loops and heavy landings but Jess easily damaged in normal landings because it is higher off the ground than a low tailplane. 6.17 Winglets ‘Wing tip plates of the kind just described should be distinguished from winglets and tip sails, which are different in principle. A tip plate or body 68 The Pelican, a canard flying boat by Andy Lennon, with (pusher propeller. Special problems with operating from the water make it necessary fo accept some extra drag from the ‘hull and stabilising sponsons. The fins on the wings act to some extent as endplates for the lifting surfaces.(Photo: A, Lennon} is intended to restrict or prevent the tip vortex. Winglets and tip sails are designed to use the vortex by extracting some of its energy. This not only weakens the vortex but, if the energy can be turned into a force in the right direction, there is a further small gain. Winglets of the type sketched in Figure 6.8 were first developed by R T Whitcomb. ‘As shown in Figure 5.1, the vortex airflow round a wing tip is inclined outwards on the underside, upwards just beyond the tip, and inwards above. ‘The precise angle of the flow to the direction of flight changes as the strength of the vortex varies at different angles of attack and flight speed. Ifan aircraft such as a commercial jet transport operates most of its time at one steady speed, it is possible to design a set of winglets which project into the vortex flow al such angles that they can, like small wings, extract some ‘lift’ force. If the winglets are set correctly this force will have a forward-acting component which can appear in the general force diagram for the whole aeroplane, as an addition to the thrust. The bulk of the winglet’s lift will, however, be directed laterally and this will not only tend to bend the winglets themselves but will increase the bending loads on the wing main structure, Since the winglets generate lift, each winglet will havea vortexal its end main wing vor- tex without winglets. Some saving in drag results. Winglets, as shown in the diagram (Fig 6.8) are cambered and twisted to meet the flow at each~ REDUCING VORTEX-INDUCED DRAG: 2 PLANFORM, TWIST, WING TIPS AND WINGLETS Professor Richard Eppler demonstrates the effect of winglets on a small glider. With one winglet turned up and one down, the model turned always to the right, indicating more drag on the downturned winglet. point at the most effective angle of attack. They are quite complicated to design and construct and are most efficient over a rather narrow range of flight speeds. If winglets are used it is most necessary to design them carefully and test them extensively before final adoption, They are not merely crude appendages stuck on the end of the wing, For them to work as intended, a great deal of work has to be done. 6.18 The commercial equation If an existing aeroplane is fitted with winglets, the increased bending loads compel some strengthen- ing of the mainplane, adding weight, and there is a reduction in load-carrying capacity, This may be compensated by the increased efficiency so that some fuel is saved. Clearly, whether the aircraft should orshould not have winglets is finally deter- mined not by aerodynamic considerations alone but by commercial factors such as the cost of the materials and the investment in design and the wind tunnel testing time, and the price of fuel. That the winglets do work as their inventors claim isnot doubted, but this does not imply they should nec- essarily be fitted to every commercial aeroplane. 6:19 Tip sails At about the same time as the Whitcomb winglets were being developed, J J Spillman was working on. tip sails of the kind shown in Figure 6.9. These were inspired by the wing tip feathers of some large soaring birds, which are spread, finger-like, to form a series of separate wing extensions with. slots betwoon. Essentially, these are intended to work in the same way as the Whitcomb winglets, but there may be three, four or five tip sails, ar- ranged radially and ‘en echelon’ round the tip. Each sail is adjusted to extract lift from the flow in its neighbourhood and, as with the winglet, some of this force is directed forwards, the rest adds bending load to the wing. The results are compa- rable and the same economic considerations apply. As before, an increase in aspect ratio has the same effect. AZ 7 Lt ores on sat Tip sals (may be 9, 4 or) ‘igh rection Loca tow cretion intip vortex ‘Resolved forces on sal Figure 6.9 Spillman (Cranfield) tip sails. 6.20 NASA tip sails Even more reminiscent of the bird wing, the NASA tip modification suggested in Figure 6.10 is intended to spread the tip vortex and reduce its strength, and this, too, reduces the vortex drag. Additional loads, as usual, must be borne by the mainplane structure and the slender tip ‘feathers’ are prone to flutter. 69MODEL AIRCRAFT AERODYNAMICS td Figure 6.10 NASA wing tip sails. 6.19 Disadvantages of winglets Winglets are intonded to reduce vortex drag which is most important for flight at relatively high angles of attack. They are usually set so that they will be most effective at one speed. For example, airliners cruise for long periods at steady airspeeds close to the best L/D trim. The winglets will be set to give their best effect in this situation. At any other speed they will work less effectively because the main tip vortex itself changes, requiring the wing let to be set at another angle. The winglets also contribute parasitic drag. At high speeds in par- ticular they will be a handicap because as airspeed rises vortex drag becomes less and less significant, while all forms of parasitic and form drag increase greatly (Fig 4.10). For example, if the winglet is set at-7 degrees at its root (Fig 6.11), it will be fixed at this anglo also when the aircraft is flying at an entirely different angle of attack, Even for sail- planes, which do need to fly fast between thermals, as well as slowly for soaring, winglets are of doubt- fal value from the drag point of view. The idea of making winglets adjustable to different angles in Figure 6.11 Winglets on a two-metre sailplane (Chuck Anderson). 8 oo serena sn we dnosan re gunn cocina -7°at rootyelative k~ to centre line nies, a ae | 4 bl 70~ . REDUCING VORTEX-INDUGED DRAG: 2 PLANFORM, TWIST, WING TIPS AND WINGLETS Aight, has been mooted but, at the time of writing, has not been attempted in practice. Winglets do, as a rule, improve aileron control and stability in circling, This, rather than any gain in performance, may justify their use, It has been shown in Chapter 5 that the most effective method of reducing vortex drag is by increasing the aspect ratio, ie. increasing the wing span for a given total area, It follows that whatever the gain from using winglets, a similar improve- ment could always be achieved by an increase in, pect ratio, This could be done by fitting a simple wing extension. Such a span extension would, of course, increase the bending loads on the main- plane and would add weight, so the question is ain decided by economics rather than aerody- namics. Nonetheless, whereas winglets require siderable research and, usually, wind tunnel ting to ensure they are of the most favourable ape and set at the best angle, to lengthen the wing is comparatively simple, Moreover, stretch- ing a wing in this way is guaranteed to reduce rtex drag at all airspeeds. A longer wing is more prone to flutter problems and slower in roll than a short wing, but adding winglets to a short wing also increases the danger of flutter and the addi- nal mass at the tip creates more rolling inertia. 6.22 Winglets and tip sails for models As far as model aircraft are concerned, very few tess have been performed with winglets or tip sails. They are unlikely to produce benefits unless *hey are properly adjusted and very few modellers Save access to wind tunnels for the necessary test- “2g purposes. If there is no restriction on the wing- span of the model, it is safer to increase aspect ratio ®en to use winglets unless these have been cor- 'y designed. There are, however, occasions shen the wingspan is restricted by contest rules, er where an increase of aspect ratio (with a reduc- Son in mean wing chord) might take the wing wn to a low Reynolds number and so lose effi- sency. In such cases winglets, especially of the Whitcomb type, offer some prospect of worthwhile gains ‘The two metre sailplane class is a case in point. In 1980 tests of a model in this category were reported by Chuck Anderson (in Model Aviation, May 1980, pp. 52-5). On a wing with 25.4cm chord, of rectangular planform, aspect ratio 7.87, winglets as shown in Figure 6,11 were fitted. These seemed to improve the performance while remain- ing within the two metre restriction, After twenty years the model concerned and others of similar design, remain in use and still fly competitively in their class. These winglets also had some less desirable effects on lateral stability and control. It must also be pointed out that such additions to the tips are rather vulnerable to damage, especially in ground loops or landings which end with the model upside down, For small free-flight models and even for FIA (A2) sailplanes, as mentioned above (6.4), wing taper is not generally desirable but the addition of winglets or sails to a rectangular wing may prove worthwhile. The Reynolds number of the main- plane would be unaffected and the tip vortex, pro- viding the winglets were well designed, would be reduced. Anderson's two metre sailplane, very wisely, was made with the angle of the winglets adjustable so that by repeated test flying, the best setting could be discovered. Noel Falconer has used a refined winglet design on tailless sailplanes and electric powered models. Apart from saving drag, which is rather more severe on a swept back wing, the winglets also serve as fins, providing very necessary lateral sta- bility on the tailless aircraft. 6.23 Crescent-shaped wings Research into wing planforms in recent years suggests that some saving of vortex drag may be achieved by adopting a distinctly crescent-shaped wing plan, resembling that of the swallow, or the curved fin of a shark. The planform to be seen on some full-scale sailplanes such as the Discus and 71MODEL AIRCRAFT AERODYNAMICS Ventus 2 represents a move partway in this direc- tion but the full crescent wing is more extreme, with the trailing edge curving progressively more backwards towards the tip and a relatively sharp extremity. The basic distribution of chord across the span remains nearly elliptical but the ellipse is progressively sheared back to produce the shark fin form. ‘Advanced calculation methods were used in the first place to establish a theoretical basis for this study. Wind tunnel and flight tests lend some farther support to the idea but more work is being done. ‘A point that needs to be taken into account is that swept back wings are more prone to tip stall- ing and to wing flutter (see Sections 6.9 and 13.9). There may be some improvement in theoretical performance with such wings but they are more difficult to construct and if handling difficulties appear in flight, the gain will not be realised, [have built and flown three sailplanes with wings approximating this plan. All developed severe wing flutter at moderate airspeeds, despite correc- tive surgery. Whether any real saving in drag resulted remains uncertain. 6.24 Curved dihedral Since the early days of model flying, particularly with froe-flight models, polyhedral rather than straight dihedral has been very widely used and in at least one early text book for modellers (Model Gliders, by RH Warring, 1942), elliptically curved dihedral was recommended. The Fillon’s Cham- pion free-flight soarer of the 1950s adopted a smoothly curved dihedral form and much later, a very popular radio controlled design, the Hobie Hawk did the same. There was not much in the way of theoretical support other than the fairly obvious fact that the steeper dihedral of the outer wing panels, acting on a greater moment arm from the centre line of the aircraft, has a powerful stabilising effect so the total amount of dihedral is less than was necessary 72 Frontlavaton —- Pan vow Figure 6.12 The wing ofa moder full-sized sailplane using all the information gathered in recent decades about the best planform, wing tip shape, dihedral and winglets (based on the Schempp Hirth Ventus 2), with the straight form, There was a suspicion in many minds that the angular breaks of dihedral at the polyhedral joints created excess drag but there was little sure evidence for this. In more recent times, coming almost as an extension of the argument favouring winglets, it has been claimed, particularly by Professor Rich- ard Eppler, that the minimum vortex-induced drag fora wing of given span is obtained, not with small winglets standing more or less upright at the tips, but with dihedral following a circular or elliptical curve, flatat the centre and turning gradually more upwards towards the tips. The reasoning is sup- ported by mathematical treatment far beyond the compass of this book, but it is significant that some of the full-scale sailplane types which have entered production since the first publication of these ideas in 1995, have adopted a polyhedral form which looks remarkably familiar to model fliers. There are great difficulties also in fitting ailerons to such a wing, so a compromise is adopted. The Schempp Hirth Discus 2 is a promi- nent example (see also Fig 6.12). The effect on lateral stability is very marked but the main reason for this style is the need for vortex drag minimisa- tion. The performance of these aircraft is certainly no worse than before. ‘The difficulty of producing continuously curved wings remains. In flight, particularly in steep turns when the loading on the wing is high (see Fig 7.8) the wings bend upwards anyway and although the curve adopted is not truly elliptical, the effect is somewhat the same.