O VER VIEW OF STALL/SPIN Characteristics Of General Aviation Aircraft /. INTRODUCTION

From man's earliest attempts at flight, stall/spin accidents have plagued the development of virtually all types of aircraft. Even today, stall/spin accidents account for more fatal and serious injuries than any other kind of accident. The classic stall/spin accident is one in which the pilot stalls the aircraft at too low an altitude to effect recovery. This situation usually arises insidiously - the pilot not expecting to stall and out of practice for good stall recovery techniques. Two examples of stall/spin are the following: First, after experiencing an engine failure on takeoff, the pilot attempts to turn back to the runway quickly before using up available altitude. Although most pilots know that a steeper bank angle results in less altitude loss in a 180 degree turn, many do not appreciate the need for increased airspeed to avoid reaching the angle of attack for stall. Further, the amount of altitude required for a safe power-off 180 degree turn is not known to most pilots. Altitude loss can range from 125 to 275 ft. (depending on the tightness of the turn) for a Piper PA-18A aircraft to 400 to 800 ft. for a Cessna 185 aircraft. A second example of a classic stall/spin accident is one that can occur when turning onto final approach in a crosswind. The pilot perceives that, for his normal bank angle, the turn rate is such that the aircraft will not be lined up with the runway. As a consequence, the aircraft is banked more steeply, using rudder to bring the nose around faster; simultaneously, increased back pressure is applied to avoid dropping below the intended glide path. Suddenly, the low-wing drops sharply and, given an unaccustomed closeup view of the ground from a steep banked, nose down attitude, the pilot instinctively pulls back harder on the elevator control - which, unfortunately, is the wrong direction for recovery. Satisfactory stall/spin behavior has been a design goal for virtually all aircraft, but difficult to attain. In the early days it was not known which aerodynamic parameters were responsible for loss of control at high angles of attack (ADA). In contrast, today, not only is there a good understanding of the factors involved, but also analytical prediction techniques and large scale wind tunnels are available to provide design criteria for adequate stall/spin

by SETH B. ANDERSON (EAA 73687)

Ames Research Center, NASA Moffett Field, CA 94035

resistance. Of the many factors related to the stall/spin problem, primary attention is given to aerodynamic considerations.

although it is recognized that human factors and pilot training are also very important aspects. Attention is directed only to the parameters which cause departure from controlled flight - not those associated with the spin and spin recovery. The purpose of this paper is to improve flight safety by the following: 1 - Identify the key factors affecting stall/spin flight behavior. 2 - Obtain a clearer understanding of operational limitations at high ADA. 3 - Focus attention on possible solutions to the stall/spin problem. //. DISCUSSION

For a normal category general aviation aircraft, the Federal Air Regulations (FAR Part 23) require the aircraft to exhibit clear and distinctive stall warning and for the aircraft to respond to normal use of the controls in such a way that neither excessive altitude losses nor dangerous attitudes are encountered during stalls and recoveries. Al-

though some aircraft fully meet the FAR requirement, many do not. We need to examine the effect of certain aerodynamic parameters on aircraft behavior near the stall to fully appreciate the difficulty of the problem.

Aerodynamic Considerations

Why has it been so difficult to design general aviation-type aircraft with the aerodynamics that provide good low-speed handling qualities so that even the novice pilot would not get into a stall/spin problem? In particular, from the manufacturer's viewpoint, can an aircraft be made inherently spinresistant without special gadgetry and without undue compromises in maneuverability, overall performance, appearance and cost? This has been a difficult question with no unique answer because many interrelated factors affect stall/spin behavior. Low speed behavior of an aircraft is influenced by the following: slope of lift curve top, CL versus alpha; rolling moment; roll damping; roll control power; aileron adverse yaw; directional stability; yaw damping; dihedral effect; side force; yawing moment; yawing moment due to rolling; pitch stability; pitch control power. Obviously, this is a formidable list to review and it will be possible to discuss only the key parameters. The variation of lift coefficient with angle of attack near maximum lift (shape of the lift curve top) is often used as an indication of potential seriousness of the stall/spin problem. A sharp lift curve top, that is, one where lift decreases rapidly with increasing AOA usually results in a large bank angle (roll-off) at the stall. The down-going wing experiences a larger AOA with reduced lift and also a very significant higher drag, while the opposite occurs for the up-going wing; thus setting up the autorotative forces which sustain the spin motion. When flow separation starts outboard, lateral controllability deteriorates, roll damping is reduced, and large excursions from level flight can occur. An obvious improvement is to design the wing for initial airflow breakdown disposed symmetrically at the wing center section. This lessens the tendency to roll-off at the stall, maintains adequate roll damping, and minimizes directional divergence. For

an aft tail configuration, the reduction of downwash tends to pitch the aircraft SPORT AVIATION 19

down out of the stall region. In addition, natural stall warning is usually obtained by buffeting of the aircraft and controls. The stall pattern depends on many factors including wing position, the amount of sweep, judicious selection of airfoil sections, proper combinations of wing thickness, twist, taper ratio, etc. To provide a docile, straight-ahead stall for all combinations of flap, gear, engine power and CG travel without unduly degrading maneuverability and high speed performance is admittedly not an easy task. Although a gentle stall is certainly a desirable feature, it in itself has not completely solved the stall/spin accident problem. As noted later, the Piper J3 Cub has a gentle stall, but a poor stall/spin accident record. Other aerodynamic factors which promote adverse stall/spin behavior must also be considered. Of the various aerodynamic parameters listed, one of the most important is directional stability since a reduction in its stabilizing function at high AOA allows yaw excursions to build up to a rate high enough that autorotative (spin) forces can predominate. Providing strong (positive) directional stability out to stall AOA for many configurations is difficult, particularly when the tail is located in the stalled wing wake. Finally, and equally important, are the lateral stability and control characteristics. Of the various aerodynamic parameters listed, one of the most important is directional stability since a reduction in its stabilizing function at high AOA allows yaw excursions to build up to a rate high enough that autorotative (spin) forces can predominate. Providing strong (positive) directional stability out to stall AOA for many configurations is difficult, particularly when the tail is located in the stalled wing wake. Finally, and equally important, are the lateral stability and control characteristics. When roll-off occurs at the stall, adequate roll control power is needed to quickly level the wings to minimize altitude loss. Dihedral effect is also significant, since at high AOA the selfrighting tendency to raise the low wing by sideslipping may be nonexistent. In addition, roll/yaw cross coupling must be minimized because adverse yaw due to aileron deflection tends to promote spin entry when ailerons are used to prevent roll-off at the stall. In spite of continued emphasis given in pilot training to use only rudder to raise the low wing at the stall, most pilots will instinctively use aileron in a stressful situation. This point is so important that in advanced fighter aircraft, the control sys20 MAY 1989

tem is programmed to ignore the pilot's lateral control input if aileron is applied to the direction of roll-off. The need for stall warning has been recognized as an essential element to help prove safe operation at low speeds. Increased buffeting and shaking of the aircraft and controls as the stall is approached is the most desirable form of stall warning since it is relatively difficult to ignore and gives the pilot an indication if he is progressing into or away from the stall. Artificial cues (oral tones and panel lights) are less effective and tend to be ignored under high

stress situations. Good stall warning can be particularly helpful to prevent an impending stall in operations close to the ground such as immediately after take-off or during landing flare. The relationship between wing flow breakdown and pitch control at the stall is also important and is different depending on tail placement. For a conventional (aft) tail location, wing stall usually occurs before pitch control effectiveness is lost. For canard configurations, pitch control must lose effectiveness before appreciable wing flow separation occurs to provide satisfactory stall behavior. These important design considerations will be discussed in more detail later for individual aircraft. To briefly summarize at this point improving stall/spin resistance is relatively complex. Stable contributions of several aerodynamic parameters, elimination of roll/yaw cross coupling, and a favorable wing flow separation pattern are required. Next, the development of several historically important aircraft and stall/spin research programs are reviewed to reflect on the aforementioned factors which influence stall/spin behavior. Historical Overview In man's earliest attempts at flight, the need for satisfactory stability and control at low speeds was underrated. Otto Lilienthal in Germany in late 1800 recognized the desirability of having

pitch (static) stability in manned gliders, but lacked sufficient pitch control power for safe low speed operation. Lilienthal made many successful glides before he was killed in an accident when his glider was upset by a gust and stalled. The Wright brothers, on the other hand, appreciated the need for adequate control, but initially never understood the need to provide positive stability. As noted in accounts of their first flights, their canard configuration was unstable, resulting in over controlling pitch attitude during the flights. The use of wing warping for lateral control produced large values of adverse yaw which contributed to spin tendencies when stalls were encountered. Because of stability and control problems, frequent inadvertent upsets occurred. In one case, Wilbur Wright allowed the aircraft to pitch up to the stall during a moment of confusion when he inadvertently stopped the engine. The stall occurred at low altitude, resulting in a nose down impact with the ground. The 1911 modified "B" Flyer used trailing edge ailerons instead of wing warping to improve lateral control and an aft tail for better pitch stability. The 1909 Bleriot, first of the monoplane series, used wing warp for lateral control and was judged to have only 10% of the aileron control power of modern aircraft. The stall was deceptive, occurring with no warning by dropping a wing and forcing the pilot to turn into the low wing for recovery at about 25 mph. The Curtiss Pusher of 1910 illustrated a common stall problem because high drag and small power available adversely delayed stall recovery. The stall was almost instantaneous with little warning; if speed dropped too low, full power was not enough to keep from mushing into the ground. In less than a decade after Kitty Hawk, stall/spin accidents tended to stagnate the development of aviation. Poor stability, low lateral control power, and roll/yaw cross coupling common among early aircraft had an adverse effect on safety in low speed flight. The stall/spin problem was a mystery in the early aircraft. Flying was conducted close to the ground because of the frightening aspects of high altitude flight. It was many years before pilots recognized that some kind of rotation was involved in loss of control. Low altitude operation precluded fully developed spins before ground impact. The nose dive attitude was evidence in itself to pull back on the stick to attempt spin recovery. It was Orville Wright who discovered that many of the nosedown

crashes were due to stalls and not

structural failures. He advocated pushing the elevator control forward rather than back for stall recovery - a concept not too popular with pilots who saw the ground from a steep nosedown attitude. The World War I aircraft became more maneuverable, higher powered, and structurally capable of being spun without falling apart on recovery. In fact, stall spin maneuvers were often used to escape from an aerial opponent. The spinning aircraft could lose altitude more rapidly (due to low L/D with separated airflow) and more safely with the low airspeeds involved, compared to his pursuer who risked structural failure in high-speed dives. The maneuverability advantage of low wing loading promoted biplanes and triplanes of compact design such as the Sopwith Camel, Sopwith Triplane, the Spad VII and Fokker DR1 - with straightforward stall characteristics. The rectangular wing planforms tended to promote flow separation near the wing center with a flat lift curve top and reduced rolling tendencies at the stall. With the multiwing arrangements (triplanes) in particular, the leading wing would stall first, due to increased upwash resulting in gentle nosedown stall recovery. However, low directional stability and large aileron adverse yaw made it easy to inadvertently spin these aircraft and many were lost if stalled at low altitude. The first systematic flight tests to provide a better understanding of the effects of stability and control at high lift were conducted by NACA at Langley Field, VA in the summer of 1919. Tests were conducted on a Curtiss JN4H "Jenny" advanced trainer. This biplane was judged to have a straightforward stall; however, because of the high drag associated with multiple struts and bracing wires, high rates of sink could develop quite easily, requiring generous application of power to recover. The commentary of flight behavior near maximum lift included, "The airplane can thus be flown level in a very badly stalled condition, the action of the longitudinal control being reversed (i.e., if the machine is losing altitude, it is necessary to decrease the angle of attack by pushing the stick forward in order to ascend). Furthermore, the airplane is very unstable laterally at angles in excess of 12 degrees, and it is prone to fall off into sideslip." Most pilots could not fly the aircraft beyond Ci_max because the ailerons were ineffective and aileron adverse yaw was large enough to trigger autorotation. Following World War I, the surplus

military aircraft became the first general aviation vehicles used by barnstorming, stunt pilots and in charter operations.

Primarily because these aircraft had low wing loading, they could be stalled and spun close to the ground with little airspace needed for recovery. In fact, several World War I aircraft were deliberately crashed in spins for World War I movies, such as "Lilac Time", with the pilot walking away unhurt. The technique was to ballast the aircraft with aft CG to promote a flat spin and provide only enough fuel to get to altitude, thus avoiding a chance of fire. With the low

vertical velocity and flat attitude on ground impact, the cockpit area sustained relatively low accelerations. In the late 1920's, the stall/spin problem received formal attention when the Daniel Guggenheim Fund sponsored a contest to promote performance and safety of flight. The Guggenheim Safe Aircraft Competition required that the test aircraft be able to fly from a minimum speed of 35 mph or less to 100 mph or more "hands-off controls at any throttle setting." Of the several aircraft entered in the contest, the Curtiss Tanager was the winner. It featured a floating-tip aileron on the lower wing and full span leading edge slats; features which were found to be superior in maintaining control at and beyond the stall. Delivery of the aircraft to the Guggenheim Flight Test Section was made on October 29, 1929. The aircraft had superb low speed performance due to the use of full span slats and flaps.

The flight tests got off to a bad start however, when the Curtiss test pilot bent the landing gear fittings in a demonstration of short field technique when he landed out of a sem-whip stall close to the ground. This relatively safe flying aircraft, with a wing loading of 8.5, never became popular. It is ironic that it was severely damaged, when flying in gusty air during a slow flight demonstration, by mushing into the ground. In the 1930's, increased performance resulted from structural and aerodynamic refinements including tapered

wings and advanced NACA airfoil sections, which, unfortunately, adversely affected stall/spin behavior. In an effort to improve this situation, studies were conducted on means of controlling wing tip stall by use of leading edge slots and slats. Although these devices were found to reduce roll off and improve roll damping, the large drag penalty in cruise detracted from their general acceptance. Another approach to improve stall/ spin safety was pioneered by Fred Weick in the mid-1930's. Good low speed behavior was obtained by promoting wing center section flow-breakdown and limiting elevator travel such that the entire wing could not be stalled. In addition, trim changes due to engine power were minimized and the use of slot-lip ailerons resulted in less adverse yawing moment. The airplane designated the W-1A was purchased by the Bureau of Air Commerce and at their request tested by NACA. This unique concept which used two-control operation (no rudders), stemmed from a 1932 NACA test of a Fairchild 22 monoplane. By providing adequate dihedral effect and sufficient directional stability, two-control operation was found to be feasible. By eliminating the possibility of crossing the controls at the stall, spin tendencies were nonexistent. The two-control concept was employed by Weick in the Ercoupe aircraft which continuously received a degree of popularity in various production versions. It is interesting to note that although the Ercoupe has never had a spin accident, it has not come along entirely unscathed. NTSB accident records show 15 crashes of the Ercoupe - the large majority due to stall/mush behavior. Because lateral control remained effective, ground impact generally occurred in a flat, level attitude with relatively minor crew injuries. This stall/ mush type accident would have been alleviated to some degree by increased engine power. The Ercoupe was eventually made available as a three-control aircraft. Presumably this was done to increase its popularity with pilots who desired greater maneuverability and improved control in crosswind landings. It would be of interest to know if the addition of yaw control has resulted in stall/ spin accidents. Prior to World War II, additional technical information pertaining to stall/spin became available. The British at the Royal Aeronautical Establishment (RAE) examined the aerodynamic design features contributing to the wingdropping problem at the stall. In this country, NACA examined the stalling

SPORT AVIATION 21

problems peculiar to low-wing airplanes. These sources identified roll instability at the stall as the primary cause of the problem pointing out that the trend to increase aircraft performance by using high taper ratio wings lessened the warning of loss of control. Further, the low-wing aircraft inherently had less lateral and longitudinal stability, thus tending to increase the consequence of inadvertent stall. These studies concluded that complete elimination of the roll instability did not seem possible but that by providing increased stall warning, greater stability at the stall, and a means to limit elevator effectiveness, the severity of the problem would be reduced. Although aircraft designers knew in general the aerodynamic features that contributed to poor stall/ spin characteristics in the 1930's, little quantitative information was available to accurately describe the handling qualities of aircraft near the stall. A first step in defining flying quality requirements was started by NACA in the late 1930's. A Stinson SR-8E aircraft was instrumented, and the dynamic behavior at the stall was measured. This aircraft, with its highly tapered wing and large aileron adverse yaw, was noted to have an unstable roll oscillatory behavior and autorotative tendencies at the stall. In addition, in 19391940 flying qualities tests were conducted on five light airplanes which showed lateral instability at the stall. It was observed that certain stability and control parameters might possibly be modified to improve stall/spin safety. These modifications included: (1) Increasing wing washout to eliminate tip stall and provide improved roll damping; (2) increasing the area and aspect ratio of horizontal and vertical tails to improve pitch and directional stability; (3) moving the elevators out of the propeller slipstream to reduce effectiveness; (4) depressing the thrust axis to reduce the nose up trim change with power; and (5) reducing yaw control power by

limiting rudder travel. The results of these modifications applied to a Piper Cub J-3 showed the following improvements in stall/spin behavior: lateral instabilities were essentially eliminated, increased directional stability and reduced rudder control power made the aircraft unspinable, and stalls in straight or turning flight were eliminated by virtue of insufficient elevator control power. These modifications detracted very little from the overall performance of this particular aircraft and could reasonably be made without extensively modifying the external appearance. As 22 MAY 1989

far as known, these stall/spin proof features were not incorporated on any follow-on versions of this type aircraft. After World War II, the general aviation fleet was again expanded by surplus military aircraft and by improved performance four-place aircraft such as the Beech Bonanza and North American Navion. Again, stall/spin accidents continued to take their toll in take-off and landing, in part because the higher wing loading of these aircraft required more altitude for stall recovery and also because roll instability at stall was aggravated by higher-lift flap systems. In

addition, increased engine power adversely affected stall behavior. Because propeller slipstream rotation causes an increase in sidewash on the vertical tail, large values of left sideslip occur when the aircraft is kept at a constant heading during stall approach. Moreover, since the left wing receives an upwash due to propeller slipstream, most aircraft tend to roll to the left when stalled in the flap down, power approach landing configuration. Notorious in this regard were trainer aircraft such as the Vultee BT-13 and the North American AT-6. It is of interest to note that when the stall strip was removed from the left wing of the author's BT-13, the left roll-off at the stall was greatly reduced. Other well known aircraft which have a marked roll-off to the left in a power-on flapsdown stall are the Beech Bonanza and Douglas DC-3. Interest by NACA in improving the general aviation aircraft safety record picked up again in the 1950's. Although some improvements had been made in low speed handling qualities, more quantitative results were needed to guide new aircraft designs. Results of flight tests conducted on a Taylorcraft BC-12, which used a NACA 23012 airfoil (sharp lift curve top), indicated that satisfactory lateral characteristics at the stall required relatively large wing washout angles (up to 8 degrees). Limiting elevator travel to have sufficient power to accomplish a three-point landing, but insufficient to exceed the angle of attack

for satisfactory lateral control could be achieved only for power-off, forward CG conditions. Additional flight tests in this series were conducted on an Interstate S-1A, a Fairchild PT-19, a Piper AG-1, and an Ercoupe with a modified tail. By limiting the longitudinal control power such that the angle of attack obtainable was 2 degrees less than that for max lift, these aircraft could not be made to spin. It was recognized that limiting elevator travel would not be a completely satisfactory solution for all aircraft because the control required varies too much with changes in flap configuration, power and CG location. The British at the RAE were active also in the late 1950's in reviewing stall characteristics for a wide variety of general-aviationtype aircraft. Their conclusions were to employ airfoil sections of known good stall behavior (flat lift curve top) and to avoid high wing taper ratios. Emphasis was also given to locating the horizontal tail to provide adequate stall warning. Although this work gave an excellent account of the behavior of many popular aircraft at the stall, the quantitative information which would give the designer a useful way of incorporating desirable aerodynamic features was apparently lacking. The level of research activity devoted to stall/spin problems increased again in early 1970 chiefly as a result of NTSB Report AAS-72. This report emphasized that in spite of improvements made over the postwar years, stall-related accidents still accounted for the largest portion of fatalities and injuries in general aviation flying. The relative ratings of stall/mush for a wide variety of aircraft are given in Table 1. Of the 31 general aviation aircraft considered in the study, a wide difference in accident rates was indicated (as much as 20 times greater in some cases). Most of the aircraft with the poorer ratings are older types with the exception of the Cessna 177 and Grumman American Yankee. It is surprising to note that the unsophisticated (no flaps, constantchord wing platform, low power) singleengine types like the Aeronca, Cub, Luscombe and Taylorcraft which have good stall warning and docile stall behavior displayed the worst records. Evidently, the aerodynamic features that tend to promote spin entry, i.e., large adverse yaw due to aileron deflection, high elevator control power to allow high AOA penetration, and large rudder control power which are inherent in these aircraft give the unwary pilot enough opportunity for self destruction.

It is interesting to note that the best rated aircraft (e.g., Cessna 182 and 210) have similar pro stall/spin features

except for one subtle difference. The pilot has to work a lot harder to stall these aircraft because high inherent longitudinal stability and large elevator hinge moments at stall require very

large pull forces to reach high AOA. In the mid-1970's, Burt Rutan introduced several canard configured aircraft (VariViggen, VariEze, Long-EZ, etc.) which possessed inherent stall/ spin resistance. By virtue of operating in the upwash of the wing, the highly loaded canard loses effectiveness at moderate AOA before appreciable wing stall occurs resulting in a gentle nosedown stall recovery. Other noteworthy spin resistance features include: (1) Relatively low rudder control power at high AOA (limiting the ability to develop yaw rate for spin entry) and (2) inboard located ailerons which minimize adverse yaw. In the late 1970's, large scale wind tunnel tests were conducted to improve spin resistance on a Beech Musketeer aircraft at the NASA Ames Research Center. A wing leading edge modification was developed to alter the stall pattern so that the onset of separation was localized at the mid semi-span. A glove shaped to resemble the lower surface of a 17% GAW airfoil (leading edge droop) was attached to the entire span of the wing leading edge except for a short (10") gap at the mid semi-span which retained the original 63 415 airfoil. The change in pressure at this junction produced strong steamwise vorticity which reduced the local induced AOA outboard. Results showed a relatively flat lift curve top out to 40 degrees. AOA, with greatly improved lateral control effectiveness and less roll off at high AOA. Subsequent flight tests verified the improved spin resistance characteristics of this type stall modification. In the 1980's, NASA-Langley continued wind tunnel and flight research concentrating on refining outboard wing stall protection methods for a broader variety of aircraft configurations. In addition, the effect of wing leading edge modifications on spin recovery was examined. These tests identified that an optimum leading edge "fix" was configuration dependent - a low aspect ratio wing requiring a somewhat modified approach compared to a high aspect ratio wing. Further, it was shown that a particular mod which improved spin resistance may result in a more difficult spin recovery. Homebuilt aircraft - What about the

stall/spin record for the more popular experimental aircraft? Does the Glasair, Lancair, Mustang II, etc. have good stall/spin resistance? Unfortunately, not much technical information is available from which to make an accurate assessment. Many of these aircraft have not been exposed to realistic operational use long enough to establish a firm statistical base. NTSB accident records are generally not available on the later homebuilt aircraft and most manufacturers understandably do not volunteer all the facts on stall/spin behavior. In addition, few pilots of homebuilts

have the expertise needed to properly evaluate and interpret stall/spin behavior for the general public. In addition, they do not have the motivation to explore the high AOA range in the depth required. Understandably, departure from controlled flight can be intimidating, particularly for the low time pilot. I have questioned several pilots at Oshkosh about the stall behavior of their aircraft, and the usual reply is that the stall is mild without a bad roll off. When asked whether they have examined the stall behavior with aft CG flaps down, climb power, with appreciable left or right sideslip, and stick held full back in a delayed recovery - they say, "Oh, no, that's too dangerous." Unfortunately, the 'loo dangerous" area may someday strike back and with the lack of exposure, the correct recovery technique may not be applied in a high stress situation. Next, two popular homebuilt aircraft are examined in detail to rate their stall/ spin characteristics based on the three factors which primarily influence stall/ spin resistance: (1) aerodynamic flow behavior, (2) pitch control limiting, and (3) cross coupling. The BD-5 is a good example of an aircraft that may appear to have docile stall behavior. There are valid reasons, however, why it can be dangerous and less forgiving, particularly when improper stall recovery techniques are used. It's had a relatively poor accident record considering the small number of hours

flown. Over 80% of the accidents involve stall/spin. The BD-5 and other popular homebuilt aircraft use a NACA 6 series airfoil (64212) designed to promote laminar flow over 40% of the chord. Use of these airfoils may have a serious disadvantage because of the small radius leading edge (sharp nose). When operated at low Reynolds number (small chord wing), max lift is relatively low because extensive flow separation tends to occur from the leading edge resulting in an abrupt decrease in lift and low roll damping. The BD-5 stall is characterized by lateral (rolling) oscillations with no "G" break. With the stick held full aft, the rolling oscillations become larger and more difficult to control. Departure usually occurs after 3 to 5 roll cycles with a rapid roll off to an inverted nose low position. Recovery is accomplished by relaxing back pressure and rolling to the nearest horizon. In accelerated flight stalls (2-4 G's), the roll off is very abrupt with essentially no warning. I found it "easy" to snap roll through 360 degrees if release of back pressure was delayed. Stall recovery techniques are somewhat different for the BD-5 because of the stall flow behavior peculiar to the 64 series airfoil. Characteristically, flow reattachment will occur only when an appreciably lower AOA is achieved compared to the AOA for initial flow separation. This flow hysteresis effect is more pronounced if the builder was careless and allowed kinks or creases to occur in the leading edge wing skin during construction. Stall recovery is complicated in that a much larger than expected increase in airspeed is needed to lower AOA for unstalled flight. In addition, if the pilot raises the nose too abruptly, a secondary or accelerated stall will easily occur with more dire consequences. There is a pronounced tendency to enter a "snap" departure roll if ailerons are misapplied during pull out. Holding aileron neutral is difficult because of low stick centering forces and a normal tendency to pull the side arm controller towards the right rear quadrant during pull ups. The accident records (NTSB investigations made in the 1970's) list loss of flight patch control (stall/spin) for a large majority of the fatalities. Was the BD-5 prone to establish a poor stall/spin record based on its inherent characteristics? Looking at the key design features associated with stall/spin resistance: (1) Aerodynamics - The wing uses a 64212 airfoil at the root and 64218 at the tip, no washout, and moderate taper SPORT AVIATION 23

ratio. Although there are valid engineering considerations for this design, the pre-stall wing flow behavior does not provide adequate roll control and desirable values of roll damping. In addition, stall warning in the form of buffet is not intense enough and occurs too close to departure to be effective. Total rating points - on the "poor to bad" side. (2) Stall limiting - This factor, as with most aircraft, was not "built-in" as a basic design feature. As noted previously, improvements in stall behavior can be obtained by limiting pitch control travel so that the wing cannot be completely stalled. However, for the BD-5 and other similar configured aircraft, this will sacrifice takeoff performance since a reduction in pitch control effectiveness adversely increases take-off speed. Obtaining a desirably low rotation speed is more difficult for the BD-5 because the thrust line is appreciably above the vertical CG producing a relatively large nosedown moment during take-off. Unfortunately, if the recommended pitch control limits are used, too much pitch control power is available and the wing AOA range can be penetrated too deeply into the stall region. A suggested compromise is the following - with a mid CG position, set the nose up pitch limit to allow rotation at speeds no less than 70 mph. Rating of this factor - poor to bad. (3) Cross coupling - Adverse yaw due to aileron deflection is not large on the BD-5 and is only noticeable to low speeds. In addition, rudder control power is relatively low resulting in less tendency to inadvertently develop large yaw rates required for spin. However, directional stability is relatively low at high AOA and even "small" amounts of adverse yaw can trigger spin tendencies particularly during accelerated flight recoveries - overall rating, good to poor. In summary, one might expect the BD-5 to generate a poor stall/spin record, based primarily on the wing flow aerodynamics. The adverse flow behavior on the BD-5 wing can be a remedied by adding an increased leading edge radius to the airfoil and increase camber over the outboard wing. This modification which protects the outer wing panel from stalling, improves roll damping and roll control at high AOA and greatly increases stall/spin resistance. In addition, several BD-5 builders have used a NASA GAW airfoil section

which provides greater lift and a more gentle stall. These modifications, unfortunately, reduce laminar flow with slightly less top speed performance. As mentioned previously, canard air24 MAY 1989

craft can be designed to provide passive stall limiting. A closer look at this feature is provided by examining another popular homebuilt, the VariEze. This aircraft had only a few stability and control problems over its developmental history. Although there are no publicized stall/spin accidents, this type aircraft must be designed correctly to avoid stability and control problems at low speeds. A more serious stall departure can occur when CG location is too far aft for either conventional or canard configurations; however, the stall behavior may

be quite different for the canard. An aft CG position for early models of the VariEze allowed greater penetration into stalled flight than desired. Several landing accidents were caused by divergent wing rocking or roll-off as speed was reduced in the landing approach. As would be expected, when AOA is increased to large values on the rear swept wing, outboard flow of the boundary layer will induce flow separation at the wing tips causing roll-off or pitch-up. An early "fix" was obtained by protecting the outboard wing area with leading edge droop. The placement of the canard, airfoil section employed, and canard geometry are key factors in providing good lowspeed behavior. The pitch stability characteristics determined from fullscale wind tunnel tests show three areas in the AOA range where important flow effects occur. The first change occurred at relatively low AOA (approximately 4 degrees) where outboard flow of the boundary layer slightly degraded wing lift and stability. The second change in stability occurred near 14 degrees AOA where a significant increase in stability resulted from canard stall and the associated reduction in down-

wash over the inboard wing area. This increase in nosedown pitching moment provides the desired passive stall limiting. A third change in stability at 22 degrees AOA is destabilizing (nose up), resulting from outboard flow separation on the wing.

Canard pitch effectiveness which is primarily a function of geometry (aspect ratio) and airfoil section must be pfcperly tailored for good stall behavior. A gradual trailing edge flow separation pattern occurs on the VariEze canard at an AOA sufficiently below appreciable wing stall. The effect of airfoil section on canard stall lift characteristics is important. For example, with a NACA 0012 airfoil section, a gradual increase in lift beyond CLmax could cause a poststall pitch-up tendency. With rearward CG, a high AOA trim (deep stall) condition could occur from which recovery may be impossible. With the usual canardwing planform geometry typified by the VarEze aircraft, one might expect reduced directional stability and damping because of the short moment arm from the CG to the vertical tail. In fact, a marked reduction in directional stability occurs in the AOA range of 10 to 20 degrees which could contribute to spin tendencies if this higher AOA range was penetrated. Dihedral effect increased by a factor of 4 in the AOA range from 0 to 20 degrees. The combination of high dihedral effect and low directional stability could result in a high roll to yaw, lightly damped Dutch Roll behavior. This has been manifested in wing rock control problems at low approach speeds for early models. Aileron effectiveness deteriorates markedly in the higher AOA range (above 10 degrees); this would be expected with the inboard location of the ailerons as a result of the outboard boundary layer flow near the wing trailing edge inherent in swept planforms. Rudder control effectiveness is relatively low compared to a typical conventional configuration and also decreases markedly in the higher AOA range. The low speed, high AOA flight behavior of the VariEze is relatively docile with no G break and the minimum speed (full-back stick) characterized by a porpoising motion and no lateral or directional departure tendencies. With power on and full aft stick, increases in altitude occur. Even with abrupt control inputs, no departures from controlled flight are apparent, although some unfamiliar motions can be obtained in a rapid pull up with full rudder and aileron inputs applied to intentionally cause upsets. Although the potential exists for this configuration to depart and spin if high AOA could be reached, the aircraft's passive stall prevention method is sound and straight forward. How does the VariEze rate in the three critical categories for stall/spin re-

sistance? (1) Aerodynamics - The flow behavior on the swept wing planform at moderate to high AOA is inherently poor in terms of providing good roll damping and pitch stability. Wing AOA must be limited to relatively low values for satisfactory low speed behavior. Rating good to poor. (2) Stall limiting - The passive stall limiting feature of the properly designed canard pitch control effectively restricts AOA range of the rear wing to minimize flow separation. Accurate CG control is mandatory for this concept. Rating good to excellent. (3) Cross coupling - Inboard location of ailerons results in small adverse yaw. Low rudder control power limits ability to generate large yaw rates at stall. Rating - good. Although it has not been possible to examine the stall/spin resistance of more recent homebuilt aircraft in the same detail as the foregoing, the writer would welcome inputs to establish a broader data base and in the end improve flight safety for all pilots.

unduly compromising lift performance is currently too much a black art. This will improve with the help of new computational fluid dynamics methods which predict complex boundary layer flow with greater accuracy than ever before possible. In particular, three dimensional (spanwise) viscous flow effects including wing/body interference are currently being examined using the latest computational equipment. The end product of this advanced technology will give the designer the wing geometry information needed for improved high AOA operation for any airSTALUSPIN/MUSH ACCIDENT RANKING

Rank

Best

5

6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

Prospects for the Future Looking back over 80 years of flight, the stall/spin problem has not been eliminated. It is still a serious factor in current general aviation operation, and there is no clear understanding why large differences in stall/spin accident rates occur for various aircraft. Although the aerodynamic factors for good stall/ spin behavior are reasonably well known, incorporating them into current aircraft designs is not done, in part due to adverse effects on performance and maneuverability. For example, the flight characteristics of one recent four-place, high-performance, single-engine aircraft were evaluated with the following comments . . . "the exciting part of our test was the stalling . . . though both audio and buffet warning were strong and the aircraft flew impeccably down to the break, it displayed an invariable attraction to autorotation after a stall in virtually any configuration." In the future, necessary improvements in stall/spin behavior will occur only if the following well known basic features are designed into the aircraft preferably by natural methods or, if necessary, by artificial means: (1) Favorable stall progression on the wing to provide good post-stall roll damping and satisfactory stall warning; (2) tailoring pitch control effectiveness to prevent reaching angles of attack for adverse wing flow separation; and (3) minimization of adverse cross-coupling effects. A short discussion of each of these points follows. The first point, designing favorable stall progression on the wing for a wide variety of aircraft configurations without

1

2 3 4

Worst

31

Short Name

Accident Rates*

C-182 C-210 C-175

0.12

C-185 C-180 C-206 C-172

Cher-6

Cherokee C-150

Bellanca

Comanche Bonanza PA- 12

Tri-pacer Mooney

Ercoupe C-170 C-140 B-23 Stinson

Navion

C-177

Citabria Taylorcr PA- 18

Luscombe Cub Yankee Aeron. 1 1 Swift

0.15 0.08 0.21 0.19 0.27 0.23 0.25 0.25 0.26 0.31 0.31 0.40 0.22 0.40 0.53 0.44 0.42 0.52 0.59 0.50 0.79 0.94 1.24 1.19 1.50 1.44 1.52 1.82 1.67 3.05

'Fatalities per 100,000 flight hours — single engine GA aircraft (1965-1 973)

craft configuration. The second point, limiting maximum attainable AOA to values less than that which causes undesirable wing flow separation can now be applied successfully only on less complex systems aircraft (small CG travel, no flaps, etc.). Application of AOA limiting to higher performance aircraft is rapidly becoming possible with progress in advanced control systems employing computer logic to automatically tailor the flight envelope for increased safety. However, some compromise in performance and maneuverability must be accepted. All modern fighter aircraft employ AOA limiting and application to GA aircraft could follow pending cost reduction and acceptable systems integration.

Finally, means for elimination of adverse cross coupling and tailoring of key aerodynamic rotary derivatives to eliminate departure leading to spins is now available. Current fighters (F-18A and F-16A) have automatic spin prevention systems which essentially eliminate cross coupling and spin tendencies. At high AOA, these systems sense a yaw rate build up and at a designated threshold value, automatically apply rudder against and aileron with the incipient spin direction. As mentioned previously, if the pilot uses aileron opposite the direction of roll off, that input is ignored and although the stick moves the control surface does not. The application of this technology to GA aircraft may be the best and only effective solution to the stall/spin problem. ///. Concluding Remarks

Since the beginning of powered flight, stall/spin problems have slowed the progress of aviation development. Although the primary aerodynamic parameters that cause the stall/spin problem have been identified for many years, incorporation of the right combination of these factors has been an insurmountable challenge and persistently remains an elusive goal. Research has identified three key factors affecting high AOA flight behavior: (1) Stall progression on the wing, (2) pitch control effectiveness, and (3) roll/yaw cross coupling. In spite of good stall warning and docile stall behavior, experience has shown that if the pilot is given enough pitch control effectiveness to stall appreciable portions of the wing, and if moderate amounts of roll/yaw cross coupling exist - eventually a fatal stall/spin accident will occur. Hopefully, a better understanding of the cause of stall/spin accidents and a stronger motivation to examine high AOA flight characteristics at altitude will improve the accident record. Have a safe flight! About the Author Seth Anderson (13051 La Paloma Ave., Los Altos Hills, CA 94022) is an aeronautical engineer at NASA-Ames Research Center, Moffett Field, CA. He has worked with NASA and its predecessor, NACA, for over 45 years. He has authored over 90 reports and papers covering a wide range of aeronautical disciplines including aerodynamic performance, stability and control, handling qualities, flight safety, and operating problems of aircraft in low speed, transonic, and supersonic flight. Seth holds a Commercial Pilot's license, A & P license and a U. S. Hang (Continued on Page 90) SPORT AVIATION 25

OVERVIEW... STALL/SPIN

(Continued from Page 25)

Glider Association Advanced (H-4) rating. His early experimental aircraft experience started shortly after WW-II with a Vultee BT-13 which was modified to improve performance and handling qualities. He has built and is currently flying a BD-5 aircraft incorporating several modifications to improve performance and flight safety. Active in EAA local Chapters giving lectures and advice to aircraft builders and designers, Seth was the first NASA participant at EAA Oshkosh in 1974, presenting technical forums. He has continued as a lecturer each year covering a wide range of topics including aerodynamics, stability and control, stall/spin, handling qualities and operating (safety) problems. Each year since 1975 he has conducted a forum on the BD-5 aircraft. In addition, he is the NASA-Ames technical representative for the NASA technology exhibit at the EAA International Convention.

FUGhfflCDM

$9995

HEADSET

plus »3°« S&H Regular price *169»5

Electrct Microphone Liquid Ear Seals Cloth Ear Covert Stereo Compatible Volume Control

24 Decibel Noise Reduction Non Gbre Surfaces I Year Warranty • Money Bock Guarantee

FREE BROCHURE: Call or write Kropp Enterprises, Inc.. P.O. Box 656,

Clermont, FL 32711 (904) 394-4727 Auto Fax: (904) 394-1718 "Serving Amenco's Auiotors Since 1980'

PUSH TO TALK SWITCH* 15" T(

ORDEREE

US

1-800-237-8400 EXT

LINES ONLY FLA 1-800-282-1469 X-29

Where The Sellers and Buyers Meet.

A new classified ad section in THE VINTAGE AIRPLANE 25c per word, 20 word minimum Send your ad and payment to ...

Wittman Regional Airport

Oshkosh, Wl 54903-2591

FASTENERS

&

HARDWARE

0 0 0 0 SIHVICI

0 0 0 0 • •lCt»

FREE PRICK LIST INFORMATION C A T O L O O 11.00 A P P L Y TO FINST OMDfN

BOH AERO 2 09 W

BACA

TRINIDAD, COLO 81 08 2

PARTICIPATE IN THESE SPECIAL EAA ACTIVITIES DEVELOPED EXCLUSIVELY FOR YOUNG PEOPLE! y

EAA YOUTH MEMBERSHIP

Full EAA Member benefits fen only SU annually

V EAA PROJECT SCHOOlfUGHT

Building real airplanes in schools and youth groups.

y

EAA SCHOLARSHIP PROGRAM

Providing support lor those seeking aviation related

educations y

EAA AIR ACADEMY

An Intensive hands-on summer aviation experience

If you use 80 octane avgas now, you could be using less expensive autogas with an EAA-STC. Get your STC from EAA—the organization that pioneered the first FAA approval for an alternative to expensive avgas.

Or write: EAA-STC, Wittman Regional Airport Oshkosh, Wl 54903-3065 For faster service, have your airplane's "N" number and serial number; your engine's make, model and serial number; and your credit card number ready.

FITTTINGS

A I R C R A F T

EAA AVIATION OPPORTUNITIES FOR YOUTH

AIRCRAR OWNERS SAVE MONEY ...FLY AUTOGAS

CALL TODAY FOR MORE INFORMATION 414-426-4800

VINTAGE TRADER

THE VINTAGE TRADER

July 28 - August 3

ATTENTION

Barney Oldfield Aircraft Company POBox228Depl E Needham. MA 02192

Champagne aerobalic performance on a

beer pocketbook Uses 65 to 125 hp NASAD approved Kits and components New

2 place Buddy Baby

now flying Info on all three $10.00

719/846-2542 EAA SCHOLARSHIPS

AERONAUTICAL ENGINEERING PILOT AND MECHANIC EAA ATT: CHUCK LARSEN WTTTMAN REGIONAL AIRPORT OSHKOSH, Wl 54903-2591 414/42W800

SPORT AVIATION

ANNUAL FILE . . . Set of 4 — Holds 4 Yrs.

$6.95 Post Paid

OutsM* U.S. and Canada add $3.00 EAA CH. 202, Box 202,

Panama City, Fla. 32402

at the EAA Aviation Center in Osntosh y

EAA AIR ADVENTURE DAYS

A one-day, hands-on aviation workshop tor young people presented at sites across the nation by EAA Chapters and clubs of the Academy of Model

VALVE COVER GASKETS

Aeronautics

[AA Air Academy programs are supported by ih» A/WCO Insurance Co. FOR INfORMATlON CONtACt

Chuck larsen. Education aiector EAA Avldlton Fourxjotlon

Wltmon Regional AJrport 0«*c«t\WIMv033066

lelepncoe (4U) 426-4800

90 MAY 1989

EAA

FAA-PMA • THE WORLDS BEST • TBO • RE USABLE

P.O. BOX 1366 1 -800-635-REAL

LAUREL, MS 39441-1366 (In MS 601 -649-0702)

4130CHROMOLYSTEEL

Round • Square • and Streamline Tubing • Sheet Material • Material Kits for many homebuilts Also Alum. Tubing Vt" to W - FREE PRICE LIST PHONE 303/665-5817

923 W.C.R. #7, Erie, CO 80518