WHITCOMB winglets small nearly vertical winglike surfaces mounted wing tips provide reduction induced drag Winglets developed Dr Richard Whitcomb NASA Langley Research Center Most published data has concentrated improving efficiency subsonic jet transports however winglets have installed new Leatiet Longhorn business jet two Burt Rutans very efficient light planes IAI Arava STOL transport Therefore winglets have proven effec tive subsonic Mach numbers full-scale Reynolds numbers purpose article report preliminary investigation use winglets 2-meter class sailplane Winglet Theory complete single source winglet information contained NASA TN D8260 Design Approach Selected Wind Tunnel Results High Subsonic Speeds Wing-Tip Mounted Winglets Richard T Whitcomb Bob Meuser summarized report Free Flight column August 1977 issue ModelAviation highly recommend anyone seriously interested winglets beg buy borrow copy TN D8260 document sale National Technical Information Service Springfield VA 22161 costs $375 interested winglets $375 worth will try summarize portions applicable model air planes ttelacethtllce lifting surfaces produce higher pressure high pressure below wing tends curl air around wing tip low pressure above wing forming vortex continues downstream shown Fig 1 tip vortex responsible induced drag large part total drag method reducing its strength will reduce drag has long known tip plates will reduce induced drag however increased friction inter ference drag usually results increase total drag Winglets reduce induced drag generating Before after Left Daughters Jill Leigh Standard Spica Right Dad ar .. , ven effective subsonic Mach numbers full-scale Reynolds numbers low Reynolds numbers involve basic research 52 Model Aviation 762 FR FIGURE L WING TIP VORTEX outward flow above wing counteract tip vortex winglet must produce large side loads order generate sufficient outward flow have significant effect tip flow must efficient lift producers net drag reduction achieved Therefore winglets incorporate airfoils have relatively high aspect ratio efficiently generate side forces something cannot done low-aspectratio flat end plates Winglet Design Designing winglets just like designing regular wings allowances complex circular flow field operate Most NASA winglets have large upper winglet located wing trailing edge smaller lower winglet located near wing leading edge asshown Fig 2 lower winglet contributes very little total drag reduction fre quently omitted Therefore upper winglets will considered application model air planes leading edge winglet should located slightly forward wing airfoil maximum thickness order minimize flow separation intersection winglet wing upper surface winglet should 675 3 q TYPICAL WINGLET SECTION Upper Surface 27 Wingleti deg SECTION A-A designed lift coefficient approximately equal wing lift coefficient speed maximum drag reduction desired side load generated winglet should higher near root lower near tip elliptical lift distribution normally desired wings Therefore winglets normally incorporate lot taper Substantial washout would also required produce desired span loading free air however decreased inflow angle increased height tip vortex provides approximate aerodynamic twist required Thus no washout winglet normally required Drag reduction increases winglet height heights much excess wing tip chord usually presents structural design problems corresponding increase winglet per*-2lCt H Upper-4Upper Surface Lowerroaf7Spanhc Lawertip380150 / 65ct/ Upper Surface Span 023 CtDihedral ____70 H 36 SPICA STANDARD SPICA 2M FIGURE 2 NASA WINGLET Standard2M Wing Span 99 in78 WingArea965 sqin780sqin AspectRatio 10178 Weight SSoz55oz Wing Loading82oz/ftl02oz/ft formance effective inflow angle produced tip vortex usually greater desired winglet angle attack winglet has toed out producing apparent negative angle attack spite appearances winglet really does produce inward force Winglets also increase lift over outboard part wing further increase lift-to-drag ratio aircraft should perform lower wing loading increased lift over outer part wing combined inward load winglet results higher wing root bending moments same wing winglets less bending moment wing span increased sufficiently equal performance shorter wing winglets Model Winglets has no published data performance ofwinglets model air plane Reynolds numbers Model airplanes obey same aerodynamic laws full-scale aircraft however fly region small changes Reynolds number have large effect nature flow Therefore principal differences between model airplane fullscale winglets should airfoil selection winglet angle attack measurement winglet effects determination optimum winglet angle difficult job best Therefore low-aspect-ratio wing selected flight test model because FIGURE 3 SPICA May 1980 53 10% thickness turbulator spars promote early boundary layer transition Details winglet shown Fig 4 I~ 4 1 IWING LEADING EDGE K AIR0IL I00/o GoT96 i ADJUSTED BY SHIMS AT LEADING OR TRAILING EDGE 300 70 FIGURE4 SPICA 2MWINGLET effects ofwinglets stronger tip vortices should larger easier measure larger winglets required low-aspect-ratio wide-chord wing will increase Reynolds number thereby making airfoil selection less critical model selected winglet flight tests Standard Class design Spica named after star pronounced Spika Spica stable easy-to-fly model has proven competitive four Unlimited Class contests second place two third places pilot error Spica also has stronger wing Standard Class models unusual load distributions produced winglets should present no problems Spica converted 2-Meter Class model substituting winglets detachable tip panels give 78-inch span 78 aspect ratio general arrangement both configura tions shown Fig 3 winglets generally patterned after 850 10 recommended configuration TN D8260 Fig 2 tip chord increased avoid un necessarily low Reynolds numbers leading edge sweep really necessary drag reduc tion retained move winglet area aft avoid reducing directional stability think looks better too winglets also increase effective dihedral because added side area above wing tips Selection airfoil difficult because winglet tip chord 4 inches puts below critical Reynolds number popular sailplane airfoils normal flying speeds winglet close FAI free-flight model terms its Reynolds number range perhaps some popular free-flight airfoils might successful airfoil finally selected Go 796 thinned III WINGLET ANGLE 275 DEG 4LOW SPEEDHIGH SPEED 28C 0 -0--0 Performance Measurement next step determine optimum winglet angle mea sure its performance see winglets really work couple hours low-speed wind tunnel would quickly produce data re quired out question monetary reasons Blame Rawdon presented simple cheap method measuring sailplane perform ance reasonable accuracy August 1979 issue Model Aviation unfortunately requires people entire active sailplane group area However absolute values sink rate airspeed need measured requirement compare effects small variations specific model Thereforeflight time fixed altitude can substituted sink rate airspeed steady flight approximately constant fixed stabilator angle Consequently plotting flight time fixed altitude versus stabilator angle can provide required data evaluating effects small variations model winglets change strength tip vortices downwash stabilator will change winglet angle polars thus produced cannot compared directly Still polars provide good way averaging large number flights maximum flight times winglet angle can easily determined plotting against winglet angle give opti mum winglet angle minimum sink rate Therefore required evaluate winglets stopwatch still air lot patience performance evaluation flights made during last hour before sunset surface winds less 5 mph launches made using same winch turnaround pulley 250 meters winch launch flown identically possible data ignored deviation normal launch observed model floated off towline trimmed fly 500- 1000-foot diameter circles average Out lift sink major problem method data has lot scatter no matter much care taken fly flight same way 5 O 280 SPIC DARD p-AC SPICA 2M WITHOUT WINGLETS len Lii U Lii I T CD j 2" x 0 -16-120804004 STABILATOR TRI MANGLE FIGURE 5 TYPICAL FLIGHT DATA POLAR 54 Model Aviation 08 -2-A Winglet Angle Deg FIGURE 6 EFFECT OF WINGLET ANGLE ON MAXIMUM ENDURANCE -4 b U I. Left Outerwing components used test programwinglet Standard 2-meterwingtip Standard outerwing panel Right Winglet installed Spica 2M wing As drawings thoroughly detail arrangement must executed preciselyotherwise wasting time means large number flights must flown order obtain enough data average sufficiently high statistical confi dence level Flight Test Finally calculations designing construction daydreaming complete time acid test sure everything would go planned first flight conducted 7 pm weekday deserted field winglets set -5 degrees initial flight insure winglets would stall few hand launches showed model controllable no excuse further delay model went up tow rails released normally Almost immediately wing dropped recov ered before corrective rudder could applied model continued rock wings frequently showed no tendency drop off spiral dive couple turns revealed rudder overly sensitive perfectly flyable next task evaluate stall recovery sufficient altitude permit recovery unusual maneuvers nose dropped straight ahead stalls level flight well left right turns remainder first flight flown large circles trying out different stabilator trim settings first flight timed however model seemed float manner belied its over-lO-ounce wing loading rudder throw decreased next three flights flown different stabilator trim settings over 3-minute average duration wing rock continued evident low speeds winglet shimmed out trailing edge give winglet angle -425 degrees flight time jumped near 4 minutes almost no wing rock point knew winglets working remained flights optimize winglet angle obtain comparison data could also call friends out amazed weird looking model next six weeks over 150 flights made various winglet angles establish data base Over 125 made late evening evaluate performance re mainder flown under thermal conditions evaluate handling qualities Another 50 flights made under same controlled conditions obtain comparison data 78-inch wing winglets 100-inch wing typical polar obtained winglet angle -275 degrees shown Fig 5 flagged symbols indicate flights considered invalid because poor launches winch malfunc tions obvious lift sink etc Polars obtained five winglet angles well model winglets Standard Class Spica maximum endurance obtained polars plotted against winglet angle determine winglet angle minimum sink rate data presented Fig 6 along maximum duration 2-Meter model winglets Standard Class model start project goal get 2-Meter winglet version perform well Standard Class version As can seen Fig 6 goal exceeded importantly adding winglets 2-Meter wing gave 15% improvement still air duration contest flier knows still air performance does win contests Therefore flights also made evaluate handling qualities thermals windy weather As might expected large side area winglets makes model sensitive gusty crosswinds same sensitivity lateral dis turbances also causes wings rock flying vicinity lift Its almost model waving saying dummy winglet model will also turn very tight difficult trim tight thermal circle Most contests area have separate classes 2-Meter models contest entered Spica 2M date has LSF Last Chance contest held Plainfield IL July 22 1979 As frequently happens model performance superior pilot settle 4th place two points out third Still performance good seriously considered flying LSF International Championships two days later Speed-distance tasks flown found time practice event few practice flights SOAR practice field revealed roll rate high speed too slow allow good lap times fact combined forecast high winds made decide fly Unlimited Class decision soon regretted problems breaking towlines tow release malfunction eliminated chance respectable showing before first days flying over As soon returned home went out club field worked high-speed roll rate problem appears high speeds winglets loaded sufficiently give very high roll rate Simultaneous application full rudder slight amount up elevator pro duces tight chandelle turn rapid recovery airspeed dive end turn extended slightly Conclusions winglet investigation far completed enough data has obtained show winglets can provide significant improvement sailplane performance however winglet angle must adjusted optimum performance given airspeed difficult since winglets should removable transportation anyway measured seems likely skin friction interference wing-winglet intersection causes some reduction L/D high speeds increased side area winglets produces wing rock flying gusty winds near thermals rudder very effective inducing roll normal flying speeds less effective high speeds wing operating low lift coefficients winglet model very stable difficult trim tight haA ff turn May 1980 55 j
Edition: Model Aviation - 1980/05
Page Numbers: 52, 53, 54, 55
WHITCOMB winglets small nearly vertical winglike surfaces mounted wing tips provide reduction induced drag Winglets developed Dr Richard Whitcomb NASA Langley Research Center Most published data has concentrated improving efficiency subsonic jet transports however winglets have installed new Leatiet Longhorn business jet two Burt Rutans very efficient light planes IAI Arava STOL transport Therefore winglets have proven effec tive subsonic Mach numbers full-scale Reynolds numbers purpose article report preliminary investigation use winglets 2-meter class sailplane Winglet Theory complete single source winglet information contained NASA TN D8260 Design Approach Selected Wind Tunnel Results High Subsonic Speeds Wing-Tip Mounted Winglets Richard T Whitcomb Bob Meuser summarized report Free Flight column August 1977 issue ModelAviation highly recommend anyone seriously interested winglets beg buy borrow copy TN D8260 document sale National Technical Information Service Springfield VA 22161 costs $375 interested winglets $375 worth will try summarize portions applicable model air planes ttelacethtllce lifting surfaces produce higher pressure high pressure below wing tends curl air around wing tip low pressure above wing forming vortex continues downstream shown Fig 1 tip vortex responsible induced drag large part total drag method reducing its strength will reduce drag has long known tip plates will reduce induced drag however increased friction inter ference drag usually results increase total drag Winglets reduce induced drag generating Before after Left Daughters Jill Leigh Standard Spica Right Dad ar .. , ven effective subsonic Mach numbers full-scale Reynolds numbers low Reynolds numbers involve basic research 52 Model Aviation 762 FR FIGURE L WING TIP VORTEX outward flow above wing counteract tip vortex winglet must produce large side loads order generate sufficient outward flow have significant effect tip flow must efficient lift producers net drag reduction achieved Therefore winglets incorporate airfoils have relatively high aspect ratio efficiently generate side forces something cannot done low-aspectratio flat end plates Winglet Design Designing winglets just like designing regular wings allowances complex circular flow field operate Most NASA winglets have large upper winglet located wing trailing edge smaller lower winglet located near wing leading edge asshown Fig 2 lower winglet contributes very little total drag reduction fre quently omitted Therefore upper winglets will considered application model air planes leading edge winglet should located slightly forward wing airfoil maximum thickness order minimize flow separation intersection winglet wing upper surface winglet should 675 3 q TYPICAL WINGLET SECTION Upper Surface 27 Wingleti deg SECTION A-A designed lift coefficient approximately equal wing lift coefficient speed maximum drag reduction desired side load generated winglet should higher near root lower near tip elliptical lift distribution normally desired wings Therefore winglets normally incorporate lot taper Substantial washout would also required produce desired span loading free air however decreased inflow angle increased height tip vortex provides approximate aerodynamic twist required Thus no washout winglet normally required Drag reduction increases winglet height heights much excess wing tip chord usually presents structural design problems corresponding increase winglet per*-2lCt H Upper-4Upper Surface Lowerroaf7Spanhc Lawertip380150 / 65ct/ Upper Surface Span 023 CtDihedral ____70 H 36 SPICA STANDARD SPICA 2M FIGURE 2 NASA WINGLET Standard2M Wing Span 99 in78 WingArea965 sqin780sqin AspectRatio 10178 Weight SSoz55oz Wing Loading82oz/ftl02oz/ft formance effective inflow angle produced tip vortex usually greater desired winglet angle attack winglet has toed out producing apparent negative angle attack spite appearances winglet really does produce inward force Winglets also increase lift over outboard part wing further increase lift-to-drag ratio aircraft should perform lower wing loading increased lift over outer part wing combined inward load winglet results higher wing root bending moments same wing winglets less bending moment wing span increased sufficiently equal performance shorter wing winglets Model Winglets has no published data performance ofwinglets model air plane Reynolds numbers Model airplanes obey same aerodynamic laws full-scale aircraft however fly region small changes Reynolds number have large effect nature flow Therefore principal differences between model airplane fullscale winglets should airfoil selection winglet angle attack measurement winglet effects determination optimum winglet angle difficult job best Therefore low-aspect-ratio wing selected flight test model because FIGURE 3 SPICA May 1980 53 10% thickness turbulator spars promote early boundary layer transition Details winglet shown Fig 4 I~ 4 1 IWING LEADING EDGE K AIR0IL I00/o GoT96 i ADJUSTED BY SHIMS AT LEADING OR TRAILING EDGE 300 70 FIGURE4 SPICA 2MWINGLET effects ofwinglets stronger tip vortices should larger easier measure larger winglets required low-aspect-ratio wide-chord wing will increase Reynolds number thereby making airfoil selection less critical model selected winglet flight tests Standard Class design Spica named after star pronounced Spika Spica stable easy-to-fly model has proven competitive four Unlimited Class contests second place two third places pilot error Spica also has stronger wing Standard Class models unusual load distributions produced winglets should present no problems Spica converted 2-Meter Class model substituting winglets detachable tip panels give 78-inch span 78 aspect ratio general arrangement both configura tions shown Fig 3 winglets generally patterned after 850 10 recommended configuration TN D8260 Fig 2 tip chord increased avoid un necessarily low Reynolds numbers leading edge sweep really necessary drag reduc tion retained move winglet area aft avoid reducing directional stability think looks better too winglets also increase effective dihedral because added side area above wing tips Selection airfoil difficult because winglet tip chord 4 inches puts below critical Reynolds number popular sailplane airfoils normal flying speeds winglet close FAI free-flight model terms its Reynolds number range perhaps some popular free-flight airfoils might successful airfoil finally selected Go 796 thinned III WINGLET ANGLE 275 DEG 4LOW SPEEDHIGH SPEED 28C 0 -0--0 Performance Measurement next step determine optimum winglet angle mea sure its performance see winglets really work couple hours low-speed wind tunnel would quickly produce data re quired out question monetary reasons Blame Rawdon presented simple cheap method measuring sailplane perform ance reasonable accuracy August 1979 issue Model Aviation unfortunately requires people entire active sailplane group area However absolute values sink rate airspeed need measured requirement compare effects small variations specific model Thereforeflight time fixed altitude can substituted sink rate airspeed steady flight approximately constant fixed stabilator angle Consequently plotting flight time fixed altitude versus stabilator angle can provide required data evaluating effects small variations model winglets change strength tip vortices downwash stabilator will change winglet angle polars thus produced cannot compared directly Still polars provide good way averaging large number flights maximum flight times winglet angle can easily determined plotting against winglet angle give opti mum winglet angle minimum sink rate Therefore required evaluate winglets stopwatch still air lot patience performance evaluation flights made during last hour before sunset surface winds less 5 mph launches made using same winch turnaround pulley 250 meters winch launch flown identically possible data ignored deviation normal launch observed model floated off towline trimmed fly 500- 1000-foot diameter circles average Out lift sink major problem method data has lot scatter no matter much care taken fly flight same way 5 O 280 SPIC DARD p-AC SPICA 2M WITHOUT WINGLETS len Lii U Lii I T CD j 2" x 0 -16-120804004 STABILATOR TRI MANGLE FIGURE 5 TYPICAL FLIGHT DATA POLAR 54 Model Aviation 08 -2-A Winglet Angle Deg FIGURE 6 EFFECT OF WINGLET ANGLE ON MAXIMUM ENDURANCE -4 b U I. Left Outerwing components used test programwinglet Standard 2-meterwingtip Standard outerwing panel Right Winglet installed Spica 2M wing As drawings thoroughly detail arrangement must executed preciselyotherwise wasting time means large number flights must flown order obtain enough data average sufficiently high statistical confi dence level Flight Test Finally calculations designing construction daydreaming complete time acid test sure everything would go planned first flight conducted 7 pm weekday deserted field winglets set -5 degrees initial flight insure winglets would stall few hand launches showed model controllable no excuse further delay model went up tow rails released normally Almost immediately wing dropped recov ered before corrective rudder could applied model continued rock wings frequently showed no tendency drop off spiral dive couple turns revealed rudder overly sensitive perfectly flyable next task evaluate stall recovery sufficient altitude permit recovery unusual maneuvers nose dropped straight ahead stalls level flight well left right turns remainder first flight flown large circles trying out different stabilator trim settings first flight timed however model seemed float manner belied its over-lO-ounce wing loading rudder throw decreased next three flights flown different stabilator trim settings over 3-minute average duration wing rock continued evident low speeds winglet shimmed out trailing edge give winglet angle -425 degrees flight time jumped near 4 minutes almost no wing rock point knew winglets working remained flights optimize winglet angle obtain comparison data could also call friends out amazed weird looking model next six weeks over 150 flights made various winglet angles establish data base Over 125 made late evening evaluate performance re mainder flown under thermal conditions evaluate handling qualities Another 50 flights made under same controlled conditions obtain comparison data 78-inch wing winglets 100-inch wing typical polar obtained winglet angle -275 degrees shown Fig 5 flagged symbols indicate flights considered invalid because poor launches winch malfunc tions obvious lift sink etc Polars obtained five winglet angles well model winglets Standard Class Spica maximum endurance obtained polars plotted against winglet angle determine winglet angle minimum sink rate data presented Fig 6 along maximum duration 2-Meter model winglets Standard Class model start project goal get 2-Meter winglet version perform well Standard Class version As can seen Fig 6 goal exceeded importantly adding winglets 2-Meter wing gave 15% improvement still air duration contest flier knows still air performance does win contests Therefore flights also made evaluate handling qualities thermals windy weather As might expected large side area winglets makes model sensitive gusty crosswinds same sensitivity lateral dis turbances also causes wings rock flying vicinity lift Its almost model waving saying dummy winglet model will also turn very tight difficult trim tight thermal circle Most contests area have separate classes 2-Meter models contest entered Spica 2M date has LSF Last Chance contest held Plainfield IL July 22 1979 As frequently happens model performance superior pilot settle 4th place two points out third Still performance good seriously considered flying LSF International Championships two days later Speed-distance tasks flown found time practice event few practice flights SOAR practice field revealed roll rate high speed too slow allow good lap times fact combined forecast high winds made decide fly Unlimited Class decision soon regretted problems breaking towlines tow release malfunction eliminated chance respectable showing before first days flying over As soon returned home went out club field worked high-speed roll rate problem appears high speeds winglets loaded sufficiently give very high roll rate Simultaneous application full rudder slight amount up elevator pro duces tight chandelle turn rapid recovery airspeed dive end turn extended slightly Conclusions winglet investigation far completed enough data has obtained show winglets can provide significant improvement sailplane performance however winglet angle must adjusted optimum performance given airspeed difficult since winglets should removable transportation anyway measured seems likely skin friction interference wing-winglet intersection causes some reduction L/D high speeds increased side area winglets produces wing rock flying gusty winds near thermals rudder very effective inducing roll normal flying speeds less effective high speeds wing operating low lift coefficients winglet model very stable difficult trim tight haA ff turn May 1980 55 j
Edition: Model Aviation - 1980/05
Page Numbers: 52, 53, 54, 55
WHITCOMB winglets small nearly vertical winglike surfaces mounted wing tips provide reduction induced drag Winglets developed Dr Richard Whitcomb NASA Langley Research Center Most published data has concentrated improving efficiency subsonic jet transports however winglets have installed new Leatiet Longhorn business jet two Burt Rutans very efficient light planes IAI Arava STOL transport Therefore winglets have proven effec tive subsonic Mach numbers full-scale Reynolds numbers purpose article report preliminary investigation use winglets 2-meter class sailplane Winglet Theory complete single source winglet information contained NASA TN D8260 Design Approach Selected Wind Tunnel Results High Subsonic Speeds Wing-Tip Mounted Winglets Richard T Whitcomb Bob Meuser summarized report Free Flight column August 1977 issue ModelAviation highly recommend anyone seriously interested winglets beg buy borrow copy TN D8260 document sale National Technical Information Service Springfield VA 22161 costs $375 interested winglets $375 worth will try summarize portions applicable model air planes ttelacethtllce lifting surfaces produce higher pressure high pressure below wing tends curl air around wing tip low pressure above wing forming vortex continues downstream shown Fig 1 tip vortex responsible induced drag large part total drag method reducing its strength will reduce drag has long known tip plates will reduce induced drag however increased friction inter ference drag usually results increase total drag Winglets reduce induced drag generating Before after Left Daughters Jill Leigh Standard Spica Right Dad ar .. , ven effective subsonic Mach numbers full-scale Reynolds numbers low Reynolds numbers involve basic research 52 Model Aviation 762 FR FIGURE L WING TIP VORTEX outward flow above wing counteract tip vortex winglet must produce large side loads order generate sufficient outward flow have significant effect tip flow must efficient lift producers net drag reduction achieved Therefore winglets incorporate airfoils have relatively high aspect ratio efficiently generate side forces something cannot done low-aspectratio flat end plates Winglet Design Designing winglets just like designing regular wings allowances complex circular flow field operate Most NASA winglets have large upper winglet located wing trailing edge smaller lower winglet located near wing leading edge asshown Fig 2 lower winglet contributes very little total drag reduction fre quently omitted Therefore upper winglets will considered application model air planes leading edge winglet should located slightly forward wing airfoil maximum thickness order minimize flow separation intersection winglet wing upper surface winglet should 675 3 q TYPICAL WINGLET SECTION Upper Surface 27 Wingleti deg SECTION A-A designed lift coefficient approximately equal wing lift coefficient speed maximum drag reduction desired side load generated winglet should higher near root lower near tip elliptical lift distribution normally desired wings Therefore winglets normally incorporate lot taper Substantial washout would also required produce desired span loading free air however decreased inflow angle increased height tip vortex provides approximate aerodynamic twist required Thus no washout winglet normally required Drag reduction increases winglet height heights much excess wing tip chord usually presents structural design problems corresponding increase winglet per*-2lCt H Upper-4Upper Surface Lowerroaf7Spanhc Lawertip380150 / 65ct/ Upper Surface Span 023 CtDihedral ____70 H 36 SPICA STANDARD SPICA 2M FIGURE 2 NASA WINGLET Standard2M Wing Span 99 in78 WingArea965 sqin780sqin AspectRatio 10178 Weight SSoz55oz Wing Loading82oz/ftl02oz/ft formance effective inflow angle produced tip vortex usually greater desired winglet angle attack winglet has toed out producing apparent negative angle attack spite appearances winglet really does produce inward force Winglets also increase lift over outboard part wing further increase lift-to-drag ratio aircraft should perform lower wing loading increased lift over outer part wing combined inward load winglet results higher wing root bending moments same wing winglets less bending moment wing span increased sufficiently equal performance shorter wing winglets Model Winglets has no published data performance ofwinglets model air plane Reynolds numbers Model airplanes obey same aerodynamic laws full-scale aircraft however fly region small changes Reynolds number have large effect nature flow Therefore principal differences between model airplane fullscale winglets should airfoil selection winglet angle attack measurement winglet effects determination optimum winglet angle difficult job best Therefore low-aspect-ratio wing selected flight test model because FIGURE 3 SPICA May 1980 53 10% thickness turbulator spars promote early boundary layer transition Details winglet shown Fig 4 I~ 4 1 IWING LEADING EDGE K AIR0IL I00/o GoT96 i ADJUSTED BY SHIMS AT LEADING OR TRAILING EDGE 300 70 FIGURE4 SPICA 2MWINGLET effects ofwinglets stronger tip vortices should larger easier measure larger winglets required low-aspect-ratio wide-chord wing will increase Reynolds number thereby making airfoil selection less critical model selected winglet flight tests Standard Class design Spica named after star pronounced Spika Spica stable easy-to-fly model has proven competitive four Unlimited Class contests second place two third places pilot error Spica also has stronger wing Standard Class models unusual load distributions produced winglets should present no problems Spica converted 2-Meter Class model substituting winglets detachable tip panels give 78-inch span 78 aspect ratio general arrangement both configura tions shown Fig 3 winglets generally patterned after 850 10 recommended configuration TN D8260 Fig 2 tip chord increased avoid un necessarily low Reynolds numbers leading edge sweep really necessary drag reduc tion retained move winglet area aft avoid reducing directional stability think looks better too winglets also increase effective dihedral because added side area above wing tips Selection airfoil difficult because winglet tip chord 4 inches puts below critical Reynolds number popular sailplane airfoils normal flying speeds winglet close FAI free-flight model terms its Reynolds number range perhaps some popular free-flight airfoils might successful airfoil finally selected Go 796 thinned III WINGLET ANGLE 275 DEG 4LOW SPEEDHIGH SPEED 28C 0 -0--0 Performance Measurement next step determine optimum winglet angle mea sure its performance see winglets really work couple hours low-speed wind tunnel would quickly produce data re quired out question monetary reasons Blame Rawdon presented simple cheap method measuring sailplane perform ance reasonable accuracy August 1979 issue Model Aviation unfortunately requires people entire active sailplane group area However absolute values sink rate airspeed need measured requirement compare effects small variations specific model Thereforeflight time fixed altitude can substituted sink rate airspeed steady flight approximately constant fixed stabilator angle Consequently plotting flight time fixed altitude versus stabilator angle can provide required data evaluating effects small variations model winglets change strength tip vortices downwash stabilator will change winglet angle polars thus produced cannot compared directly Still polars provide good way averaging large number flights maximum flight times winglet angle can easily determined plotting against winglet angle give opti mum winglet angle minimum sink rate Therefore required evaluate winglets stopwatch still air lot patience performance evaluation flights made during last hour before sunset surface winds less 5 mph launches made using same winch turnaround pulley 250 meters winch launch flown identically possible data ignored deviation normal launch observed model floated off towline trimmed fly 500- 1000-foot diameter circles average Out lift sink major problem method data has lot scatter no matter much care taken fly flight same way 5 O 280 SPIC DARD p-AC SPICA 2M WITHOUT WINGLETS len Lii U Lii I T CD j 2" x 0 -16-120804004 STABILATOR TRI MANGLE FIGURE 5 TYPICAL FLIGHT DATA POLAR 54 Model Aviation 08 -2-A Winglet Angle Deg FIGURE 6 EFFECT OF WINGLET ANGLE ON MAXIMUM ENDURANCE -4 b U I. Left Outerwing components used test programwinglet Standard 2-meterwingtip Standard outerwing panel Right Winglet installed Spica 2M wing As drawings thoroughly detail arrangement must executed preciselyotherwise wasting time means large number flights must flown order obtain enough data average sufficiently high statistical confi dence level Flight Test Finally calculations designing construction daydreaming complete time acid test sure everything would go planned first flight conducted 7 pm weekday deserted field winglets set -5 degrees initial flight insure winglets would stall few hand launches showed model controllable no excuse further delay model went up tow rails released normally Almost immediately wing dropped recov ered before corrective rudder could applied model continued rock wings frequently showed no tendency drop off spiral dive couple turns revealed rudder overly sensitive perfectly flyable next task evaluate stall recovery sufficient altitude permit recovery unusual maneuvers nose dropped straight ahead stalls level flight well left right turns remainder first flight flown large circles trying out different stabilator trim settings first flight timed however model seemed float manner belied its over-lO-ounce wing loading rudder throw decreased next three flights flown different stabilator trim settings over 3-minute average duration wing rock continued evident low speeds winglet shimmed out trailing edge give winglet angle -425 degrees flight time jumped near 4 minutes almost no wing rock point knew winglets working remained flights optimize winglet angle obtain comparison data could also call friends out amazed weird looking model next six weeks over 150 flights made various winglet angles establish data base Over 125 made late evening evaluate performance re mainder flown under thermal conditions evaluate handling qualities Another 50 flights made under same controlled conditions obtain comparison data 78-inch wing winglets 100-inch wing typical polar obtained winglet angle -275 degrees shown Fig 5 flagged symbols indicate flights considered invalid because poor launches winch malfunc tions obvious lift sink etc Polars obtained five winglet angles well model winglets Standard Class Spica maximum endurance obtained polars plotted against winglet angle determine winglet angle minimum sink rate data presented Fig 6 along maximum duration 2-Meter model winglets Standard Class model start project goal get 2-Meter winglet version perform well Standard Class version As can seen Fig 6 goal exceeded importantly adding winglets 2-Meter wing gave 15% improvement still air duration contest flier knows still air performance does win contests Therefore flights also made evaluate handling qualities thermals windy weather As might expected large side area winglets makes model sensitive gusty crosswinds same sensitivity lateral dis turbances also causes wings rock flying vicinity lift Its almost model waving saying dummy winglet model will also turn very tight difficult trim tight thermal circle Most contests area have separate classes 2-Meter models contest entered Spica 2M date has LSF Last Chance contest held Plainfield IL July 22 1979 As frequently happens model performance superior pilot settle 4th place two points out third Still performance good seriously considered flying LSF International Championships two days later Speed-distance tasks flown found time practice event few practice flights SOAR practice field revealed roll rate high speed too slow allow good lap times fact combined forecast high winds made decide fly Unlimited Class decision soon regretted problems breaking towlines tow release malfunction eliminated chance respectable showing before first days flying over As soon returned home went out club field worked high-speed roll rate problem appears high speeds winglets loaded sufficiently give very high roll rate Simultaneous application full rudder slight amount up elevator pro duces tight chandelle turn rapid recovery airspeed dive end turn extended slightly Conclusions winglet investigation far completed enough data has obtained show winglets can provide significant improvement sailplane performance however winglet angle must adjusted optimum performance given airspeed difficult since winglets should removable transportation anyway measured seems likely skin friction interference wing-winglet intersection causes some reduction L/D high speeds increased side area winglets produces wing rock flying gusty winds near thermals rudder very effective inducing roll normal flying speeds less effective high speeds wing operating low lift coefficients winglet model very stable difficult trim tight haA ff turn May 1980 55 j
Edition: Model Aviation - 1980/05
Page Numbers: 52, 53, 54, 55
WHITCOMB winglets small nearly vertical winglike surfaces mounted wing tips provide reduction induced drag Winglets developed Dr Richard Whitcomb NASA Langley Research Center Most published data has concentrated improving efficiency subsonic jet transports however winglets have installed new Leatiet Longhorn business jet two Burt Rutans very efficient light planes IAI Arava STOL transport Therefore winglets have proven effec tive subsonic Mach numbers full-scale Reynolds numbers purpose article report preliminary investigation use winglets 2-meter class sailplane Winglet Theory complete single source winglet information contained NASA TN D8260 Design Approach Selected Wind Tunnel Results High Subsonic Speeds Wing-Tip Mounted Winglets Richard T Whitcomb Bob Meuser summarized report Free Flight column August 1977 issue ModelAviation highly recommend anyone seriously interested winglets beg buy borrow copy TN D8260 document sale National Technical Information Service Springfield VA 22161 costs $375 interested winglets $375 worth will try summarize portions applicable model air planes ttelacethtllce lifting surfaces produce higher pressure high pressure below wing tends curl air around wing tip low pressure above wing forming vortex continues downstream shown Fig 1 tip vortex responsible induced drag large part total drag method reducing its strength will reduce drag has long known tip plates will reduce induced drag however increased friction inter ference drag usually results increase total drag Winglets reduce induced drag generating Before after Left Daughters Jill Leigh Standard Spica Right Dad ar .. , ven effective subsonic Mach numbers full-scale Reynolds numbers low Reynolds numbers involve basic research 52 Model Aviation 762 FR FIGURE L WING TIP VORTEX outward flow above wing counteract tip vortex winglet must produce large side loads order generate sufficient outward flow have significant effect tip flow must efficient lift producers net drag reduction achieved Therefore winglets incorporate airfoils have relatively high aspect ratio efficiently generate side forces something cannot done low-aspectratio flat end plates Winglet Design Designing winglets just like designing regular wings allowances complex circular flow field operate Most NASA winglets have large upper winglet located wing trailing edge smaller lower winglet located near wing leading edge asshown Fig 2 lower winglet contributes very little total drag reduction fre quently omitted Therefore upper winglets will considered application model air planes leading edge winglet should located slightly forward wing airfoil maximum thickness order minimize flow separation intersection winglet wing upper surface winglet should 675 3 q TYPICAL WINGLET SECTION Upper Surface 27 Wingleti deg SECTION A-A designed lift coefficient approximately equal wing lift coefficient speed maximum drag reduction desired side load generated winglet should higher near root lower near tip elliptical lift distribution normally desired wings Therefore winglets normally incorporate lot taper Substantial washout would also required produce desired span loading free air however decreased inflow angle increased height tip vortex provides approximate aerodynamic twist required Thus no washout winglet normally required Drag reduction increases winglet height heights much excess wing tip chord usually presents structural design problems corresponding increase winglet per*-2lCt H Upper-4Upper Surface Lowerroaf7Spanhc Lawertip380150 / 65ct/ Upper Surface Span 023 CtDihedral ____70 H 36 SPICA STANDARD SPICA 2M FIGURE 2 NASA WINGLET Standard2M Wing Span 99 in78 WingArea965 sqin780sqin AspectRatio 10178 Weight SSoz55oz Wing Loading82oz/ftl02oz/ft formance effective inflow angle produced tip vortex usually greater desired winglet angle attack winglet has toed out producing apparent negative angle attack spite appearances winglet really does produce inward force Winglets also increase lift over outboard part wing further increase lift-to-drag ratio aircraft should perform lower wing loading increased lift over outer part wing combined inward load winglet results higher wing root bending moments same wing winglets less bending moment wing span increased sufficiently equal performance shorter wing winglets Model Winglets has no published data performance ofwinglets model air plane Reynolds numbers Model airplanes obey same aerodynamic laws full-scale aircraft however fly region small changes Reynolds number have large effect nature flow Therefore principal differences between model airplane fullscale winglets should airfoil selection winglet angle attack measurement winglet effects determination optimum winglet angle difficult job best Therefore low-aspect-ratio wing selected flight test model because FIGURE 3 SPICA May 1980 53 10% thickness turbulator spars promote early boundary layer transition Details winglet shown Fig 4 I~ 4 1 IWING LEADING EDGE K AIR0IL I00/o GoT96 i ADJUSTED BY SHIMS AT LEADING OR TRAILING EDGE 300 70 FIGURE4 SPICA 2MWINGLET effects ofwinglets stronger tip vortices should larger easier measure larger winglets required low-aspect-ratio wide-chord wing will increase Reynolds number thereby making airfoil selection less critical model selected winglet flight tests Standard Class design Spica named after star pronounced Spika Spica stable easy-to-fly model has proven competitive four Unlimited Class contests second place two third places pilot error Spica also has stronger wing Standard Class models unusual load distributions produced winglets should present no problems Spica converted 2-Meter Class model substituting winglets detachable tip panels give 78-inch span 78 aspect ratio general arrangement both configura tions shown Fig 3 winglets generally patterned after 850 10 recommended configuration TN D8260 Fig 2 tip chord increased avoid un necessarily low Reynolds numbers leading edge sweep really necessary drag reduc tion retained move winglet area aft avoid reducing directional stability think looks better too winglets also increase effective dihedral because added side area above wing tips Selection airfoil difficult because winglet tip chord 4 inches puts below critical Reynolds number popular sailplane airfoils normal flying speeds winglet close FAI free-flight model terms its Reynolds number range perhaps some popular free-flight airfoils might successful airfoil finally selected Go 796 thinned III WINGLET ANGLE 275 DEG 4LOW SPEEDHIGH SPEED 28C 0 -0--0 Performance Measurement next step determine optimum winglet angle mea sure its performance see winglets really work couple hours low-speed wind tunnel would quickly produce data re quired out question monetary reasons Blame Rawdon presented simple cheap method measuring sailplane perform ance reasonable accuracy August 1979 issue Model Aviation unfortunately requires people entire active sailplane group area However absolute values sink rate airspeed need measured requirement compare effects small variations specific model Thereforeflight time fixed altitude can substituted sink rate airspeed steady flight approximately constant fixed stabilator angle Consequently plotting flight time fixed altitude versus stabilator angle can provide required data evaluating effects small variations model winglets change strength tip vortices downwash stabilator will change winglet angle polars thus produced cannot compared directly Still polars provide good way averaging large number flights maximum flight times winglet angle can easily determined plotting against winglet angle give opti mum winglet angle minimum sink rate Therefore required evaluate winglets stopwatch still air lot patience performance evaluation flights made during last hour before sunset surface winds less 5 mph launches made using same winch turnaround pulley 250 meters winch launch flown identically possible data ignored deviation normal launch observed model floated off towline trimmed fly 500- 1000-foot diameter circles average Out lift sink major problem method data has lot scatter no matter much care taken fly flight same way 5 O 280 SPIC DARD p-AC SPICA 2M WITHOUT WINGLETS len Lii U Lii I T CD j 2" x 0 -16-120804004 STABILATOR TRI MANGLE FIGURE 5 TYPICAL FLIGHT DATA POLAR 54 Model Aviation 08 -2-A Winglet Angle Deg FIGURE 6 EFFECT OF WINGLET ANGLE ON MAXIMUM ENDURANCE -4 b U I. Left Outerwing components used test programwinglet Standard 2-meterwingtip Standard outerwing panel Right Winglet installed Spica 2M wing As drawings thoroughly detail arrangement must executed preciselyotherwise wasting time means large number flights must flown order obtain enough data average sufficiently high statistical confi dence level Flight Test Finally calculations designing construction daydreaming complete time acid test sure everything would go planned first flight conducted 7 pm weekday deserted field winglets set -5 degrees initial flight insure winglets would stall few hand launches showed model controllable no excuse further delay model went up tow rails released normally Almost immediately wing dropped recov ered before corrective rudder could applied model continued rock wings frequently showed no tendency drop off spiral dive couple turns revealed rudder overly sensitive perfectly flyable next task evaluate stall recovery sufficient altitude permit recovery unusual maneuvers nose dropped straight ahead stalls level flight well left right turns remainder first flight flown large circles trying out different stabilator trim settings first flight timed however model seemed float manner belied its over-lO-ounce wing loading rudder throw decreased next three flights flown different stabilator trim settings over 3-minute average duration wing rock continued evident low speeds winglet shimmed out trailing edge give winglet angle -425 degrees flight time jumped near 4 minutes almost no wing rock point knew winglets working remained flights optimize winglet angle obtain comparison data could also call friends out amazed weird looking model next six weeks over 150 flights made various winglet angles establish data base Over 125 made late evening evaluate performance re mainder flown under thermal conditions evaluate handling qualities Another 50 flights made under same controlled conditions obtain comparison data 78-inch wing winglets 100-inch wing typical polar obtained winglet angle -275 degrees shown Fig 5 flagged symbols indicate flights considered invalid because poor launches winch malfunc tions obvious lift sink etc Polars obtained five winglet angles well model winglets Standard Class Spica maximum endurance obtained polars plotted against winglet angle determine winglet angle minimum sink rate data presented Fig 6 along maximum duration 2-Meter model winglets Standard Class model start project goal get 2-Meter winglet version perform well Standard Class version As can seen Fig 6 goal exceeded importantly adding winglets 2-Meter wing gave 15% improvement still air duration contest flier knows still air performance does win contests Therefore flights also made evaluate handling qualities thermals windy weather As might expected large side area winglets makes model sensitive gusty crosswinds same sensitivity lateral dis turbances also causes wings rock flying vicinity lift Its almost model waving saying dummy winglet model will also turn very tight difficult trim tight thermal circle Most contests area have separate classes 2-Meter models contest entered Spica 2M date has LSF Last Chance contest held Plainfield IL July 22 1979 As frequently happens model performance superior pilot settle 4th place two points out third Still performance good seriously considered flying LSF International Championships two days later Speed-distance tasks flown found time practice event few practice flights SOAR practice field revealed roll rate high speed too slow allow good lap times fact combined forecast high winds made decide fly Unlimited Class decision soon regretted problems breaking towlines tow release malfunction eliminated chance respectable showing before first days flying over As soon returned home went out club field worked high-speed roll rate problem appears high speeds winglets loaded sufficiently give very high roll rate Simultaneous application full rudder slight amount up elevator pro duces tight chandelle turn rapid recovery airspeed dive end turn extended slightly Conclusions winglet investigation far completed enough data has obtained show winglets can provide significant improvement sailplane performance however winglet angle must adjusted optimum performance given airspeed difficult since winglets should removable transportation anyway measured seems likely skin friction interference wing-winglet intersection causes some reduction L/D high speeds increased side area winglets produces wing rock flying gusty winds near thermals rudder very effective inducing roll normal flying speeds less effective high speeds wing operating low lift coefficients winglet model very stable difficult trim tight haA ff turn May 1980 55 j