| Tutorials home | Safety tour | Emergencies | Meteorology | Flight Theory | Navigation | Communications | Builders guide |


Fly safe! home page

Decreasing your exposure to aerodynamic risk

3. Don't stall and spin in from a turn


Rev.13 — page content was last changed 23 June 2012

Google logo



Loss of control in low-level manoeuvring is a major cause of serious accidents. This makes it vital that the reasons for loss of control in those situations be understood. Some studies indicate that 80–90% of stall/spin accidents involve turning either in the circuit or in other low-level (i.e. below 1500 feet agl) flight — even when just sight-seeing. Proximity to the ground appears to sometimes lure pilots into fatal reactions, though low-power descending turns seem to be more frequently involved than level or high-power climbing turns.

"The weather was sultry with temperature in the mid-30s, a gusty wind was starting to pick up and thunderstorms were expected to develop. The visiting pilot and passenger decided to head for home taking-off towards the north-west. Their hosts, preparing to wave goodbye, expected them to follow their usual practice of flying over the homestead before heading south. The aircraft was seen to climb to an estimated 250 feet then commence a turn to the right but seemed to be still climbing. When the aircraft was nearly beam on to the witnesses the bank increased suddenly while the nose dropped and the aircraft swung right around towards the south-west. The onlookers watched in disbelief as the aircraft plunged into the paddock, starboard wing first with the engine at full power."

      Content



3.1 A remarkably easy way to top yourself!

How much height might be lost in a stall/spin incident?
The pilot in that intended fly-by failed to recognise the stall in the turn and had no time to recover from the consequent first stage spin; he had perhaps only reached 300 feet agl when everything turned sour. The height lost during a normal stall and recovery incident in a very light aircraft is probably between 50 and 250 feet; depending on atmospheric turbulence, the aircraft type, the aircraft attitude at stall, the docility of stall onset and the pilot's awareness of the incipient (beginning or initial stage) stall. However, loss of height in a stall/spin event, as described above, is very much greater, perhaps 100–300 feet during the incipient spin stage, 200–400 feet to stop the autorotation and 300–500 feet during the recovery; a total of 600–1200 feet if the incipient spin is allowed to develop into autorotation.
Do YOU make sure you know the accelerated stall characteristics of the aircraft you are flying?
Unfortunately many pilots are not wary of the stall onset when the wings are loaded up, because they are used to benign stalls and have never explored an aircraft's accelerated stall characteristics; which will be different — and in some aircraft quite viciously so — to the normal 1g stall characteristics. No pilot can escape from a stall/spin event if there is insufficient height to do so, but prompt recognition of the incipient stall and fast corrective action can save the day.

All of which is why low-level stall/spin events are so absolutely deadly and why the only real solution to a stall/spin event is absolute avoidance; all the spin recovery training you may undertake is not going to help once the aircraft is spinning below the minimum recovery height. Stay within the aerodynamic limits and never place the aircraft in any situation which would make such an event possible; fly the aeroplane or, more to the point, make sure the wing and tailplane always keep flying! Never, never indulge your self-supposed ability to produce fast pull-ups on take-off or a wing-over and beat-up.

Developed spin recovery training is not included in the RA-Aus Pilot Certificate or the GA Private Pilot Licence syllabus, though stall and incipient spin awareness and recovery are normal parts of both syllabi. However, please read Autorotation — the fully developed spin and Spin recovery confidence building in the flight theory section.

Stall/spin near impact
This is near to last in an award winning photo sequence of a stall/spin accident. Taken by Ian Ward, published in the 'Wimmera Mail Times' and subsequently in the AUF magazine. In this image autorotation seems to be established with the nose yawing downward and the aircraft rolling to the right. Note the pilot's totally natural ground rush aversion reaction: full back stick (evidenced by the elevator position in all photos) thus holding the aircraft in the stalled condition. Also the pilot is endeavouring to roll the aircraft upright with aileron (starboard aileron down), but consequently deepening the stall, and thus increasing the lift loss and the drag on the down-going and non-flying starboard wing. The pilot had no escape from the inevitable.

Failure to recognise the aircraft condition is not confined to relatively inexperienced pilots. In June 2009, the three-pilot crew of an Airbus 330 operating as Air France flight 447 (and following an airspeed indication aberration) failed to recognise that they were holding the aircraft in a fully stalled condition with a pitch attitude of around 16° nose-up and an angle of attack above 35°, while the aircraft was sinking at 10 000 ft/min. The engines were responding normally to pilot inputs. That stalled descent continued for 3.5 minutes from 38 000 feet to the surface of the Atlantic, where all 228 persons on board perished. Similar disastrous failures, on the part of the flight deck crew, to recognise the stalled condition have occurred previously; for example, in 2005 a West Caribbean Airways MD-82 crashed in Venezuela after sinking from 36 000 feet in 3.5 minutes — 160 persons perished.

Some refresher notes on the pertinent aspects of aerodynamics
Before proceeding further, the following are some refresher notes on a few pertinent aspects of aerodynamics — for those who have forgotten the theory or just can't immediately recall it.

  •   The airspeed at which an aircraft stalls depends in part on the basic wing loading multiplied by an aerodynamic load factor. For convenience the load factor used is the non-dimensional ratio of lift force being generated to aircraft all-up weight but expressed in terms of 'g' acceleration units. So if the lift force generated is 50% greater than the aircraft's gross weight (while at rest on the ground) the wing structure load factor is expressed as '1.5g'. If a wing reaches the critical angle of attack of 15° or 16° when loaded up — i.e. an aerodynamic load higher than 1g — the stalling speed will be higher than the normal 1g stall speed at that particular mass and wing configuration. The effects of that accelerated stall are usually more pronounced than a 1g stall. An accelerated stall is not a 'high speed' stall — the latter is one form of accelerated stall.

 V-n diagram A V-n diagram has been modified to show the stall speed positions at a 2g load factor (Vs2g) and a 3g load factor (Vs3g). The manoeuvring speed Va in this particular diagram is coincident with Vs4g. This means that if the pilot, or the atmosphere, attempt to apply a load greater than 4g, at or below this airspeed the aircraft will stall. Thus at speeds at or below Va it is probably not possible to reach the positive limit load factor of 4.4g.

  •   Uncoordinated or cross-controlled flight: applying pressure to the rudder in one direction with opposite aileron applied is cross-controlling. (The cross-controlled situation can also be brought about if the airframe is improperly rigged.) This is normally a rather sloppy way to fly but also a condition that can lead to an uncommanded roll and incipient spin if you inadvertently exceed the critical angle of attack [aoa]; particularly in uncoordinated climbing or lower speed descending turns, such as that made in the approach to landing. A planned and properly executed cross-controlled steady sideslip during final approach IS a normal and safe height loss manoeuvre for non-flapped aircraft. (Some afficionados use sideslip in addition to full flap for a really steep, high drag approach — if it is appropriate and beneficial and if the aircraft designer allows it.)

  •   Once established in a coordinated level turn the lower inner wing has slightly lesser airspeed and thus less lift than the outer wing, which produces a tendency for the outer wing to rise and the bank angle to increase. This requires the pilot to apply a slight opposite pressure to the control column which is known as 'holding-off bank'; this is quite normal and probably the pilot may not notice doing so because it should be just part of maintaining the chosen bank angle throughout the turn. In a climbing turn the outer wing has a slightly greater effective aoa than the inner wing and thus additional lift. Combined with its faster speed this reinforces the tendency for the bank angle to increase and the need to hold-off bank.

  •   However, in a descending turn the steeper path of the inner wing means that it will have a slightly larger effective aoa than the outer; this may compensate, or over-compensate, for the faster velocity of the outer wing. In order then to maintain the required bank angle it may be necessary to apply a slight inward pressure to the control column; i.e. in a coordinated descending turn the bank may be 'held on'. Holding-off bank in a gliding turn can lead to a stall/spin condition.

  •   If an aircraft is inadvertently stalled in a coordinated turn, where the ailerons are in the neutral position, both wings usually display the same progressive stall pattern; thus there should be no pronounced wing drop in a well-designed aircraft. In a coordinated climbing turn you would expect the outer wing to stall first, but propeller effects could negate or reinforce this tendency.

  •   When flying at speeds below 1.5 times Vs — generally regarded as the minimum safe speed near the ground, as long as Vs is calibrated airspeed rather than indicated airspeed; see the following Note 1 — the aileron moments are increasingly less effective with diminishing airspeed, so larger aileron deflections are needed to bank the aircraft. There is always a tendency to be more forceful than necessary, thus overbanking the aircraft at a critical stage. The same applies to rudder effectiveness particularly at low power settings.

    Note 1: many — possibly most — airspeed indicator systems underread or overread considerably at high angles of attack. If the necessary position error correction to IAS to provide the calibrated airspeed [CAS] has not been supplied by the manufacturer or determined by the homebuilder, there is potential for serious misjudgement. For example, if the indicated Vs is 30 knots and there is a position error of minus 8 knots at that aoa, then the corrected stall speed is 38 knots CAS. Consequently if the pilot calculates the minimum safe speed as 1.5 times 30 = 45 knots then the expected 50% safety margin might be only 20%; i.e. 45/38=1.2 times Vs [CAS]. Of course 45 knots IAS will probably not be 45 knots CAS so there is another adjustment in the calculation, but you get the picture.

  •   In the following text 'top/bottom rudder' refers to the relative position of the rudder pedals when turning; 'top' being the rudder pedal opposite the lower wing, thus if the aircraft is banked and turning to the left then pressure on the right rudder pedal will apply top (or outside) rudder, and pressure on the left rudder pedal will apply bottom (or inside) rudder. An excess of bottom rudder produces a skidding turn; too much top rudder produces a slipping turn or may even halt the turn, so producing a full sideslip. (In a coordinated turn there is just sufficient bottom rudder applied to keep the slip ball centred.)

  •   All of the following is applicable to three-axis controlled aircraft but some parts may not be generally applicable to weight-shift controlled trikes.



3.2 Do you know how much the angle of attack increases in a turn?

As a consequence of providing the centripetal force for a sustained turn, the wing loading (i.e. lift force) must be increased as angle of bank increases. The loading increases rather slowly up to a bank angle of 30° — where it is 15% greater than normal level flight loading — after which it increases rapidly, where it is 41% greater at a 45° bank angle, and so the load factor will be 1.41g. The right-hand column in Table 3 shows the increase in stall speed, which is proportional to the square root of the load factor. You can see that the percentage increase in stall speed is about half the percentage increase in wing loading.

Table 3: wing loading and Vs increase
Bank angleCosineLoad factor needed
[wing loading increase]
Vs multiplier
[increase]
10° 0.98 1.02g [+2%] 1.01 [+1%]
15° 0.965 1.04g [+4%] 1.02 [+2%]
20° 0.94 1.06g [+6%] 1.03 [+3%]
30° 0.87 1.15g [+15%] 1.07 [+7%]
40° 0.77 1.30g [+30%] 1.14 [+14%]
45° 0.71 1.41g [+41%] 1.19 [+19%]
50° 0.64 1.56g [+56%] 1.25 [+25%]
54° 0.59 1.70g [+70%] 1.3 [+30%]
60° 0.50 2.00g [+100%] 1.41 [+41%]


The lift force increase in the constant-speed turn is provided by an increase in the lift coefficient [CL], which in itself is brought about by increasing aoa. Increasing aoa while maintaining constant speed produces an exponential increase in induced drag (this is related to the CL²; perhaps doubled at 45°, trebled at 60°) thus resulting in loss of height or change in rate of climb/descent unless power is substantially increased.

A rule of thumb for light aircraft with normally cambered wings is that each 1 degree aoa change — starting from 2° and continuing to about 14° — approximates to a 0.1 CL change and each 0.1 CL increase/decrease at a constant airspeed represents a wing loading change of roughly 8%.

So, from the table above, a 30° bank angle in a sustained turn adds 2° to the basic aoa for the airspeed, a 45° bank angle adds 5° and a 60° angle adds 12°. The basic aoa for normal descending and climbing speeds in the circuit are probably in the 6–8° and 6–10° regions respectively so anything more than a moderate 30° banked turn makes severe inroads into the safety margin between the effective aoa of some sections of the wing and critical aoa. As well as that, down-aileron increases and up-aileron decreases the aoa of the outer wing sections.

Something to be borne in mind is that wing loading must also change with the payload carried, as do the stall speeds and the performance speeds. If a two-seat recreational aircraft is normally flown with just the pilot on board, the aoa associated with a particular calibrated airspeed is significantly less than when flying at the same airspeed with a heavy passenger and perhaps a full fuel load. For example if the aircraft is normally flown with only the pilot on board with an all-up weight of 400 kg but when flown with a heavy passenger, your gear on board and full fuel then all-up weight increases to 540 kg and the wing loading is increased by 35%. Thus, CL and the aoa for any particular IAS/CAS will be greater than that to which the pilot is accustomed; maybe 2–3° at low airspeeds and much less at high airspeeds. All of this means that the low-speed bank angles you use safely at low weight may well be deadly when heavy.

3.3 Do you know why loss of control is more likely if the controls are not coordinated in a turn?

If an aircraft is being held in a level turn at a particular bank angle with constant power and excess bottom rudder is applied and held, the aircraft will rotate about the normal axis (yaw) in the direction of rudder deflection. Airspeed over the outer wing increases slightly while airspeed over the inner wing decreases, producing a lift differential; so there will be a secondary roll effect that increases the bank angle. At the same time, the yaw increases fuselage drag — which decreases airspeed and thus lift, and the nose drops a little. This is an uncoordinated skidding turn, which often happens when the pilot tries to 'hurry' the turn with bottom rudder instead of increasing bank and we have a situation where the aircraft is overbanking with the nose yawing inward and downward.

If the pilot reacts by applying and holding opposite aileron to restore the required bank angle — i.e. holding-off bank — then, due to the downward deflection of the inner aileron, the outer 30% or so of the lower wing is flying at a much higher aoa than the corresponding section of the higher wing. (If equipped with flaperons the whole lower wing would be flying at a higher aoa.) The lower wing will also be producing more aileron drag — mainly because of the increase in induced drag — so the inward and downward yaw will be increased and there will be a tendency for the pilot to raise the nose by increasing control column back-pressure, thereby increasing aoa overall while, at the same time, speed will continue to decrease because of the increased drag — unless power is increased.

The pilot is now 'pushing the aerodynamic limits of the flight envelope'. Any consequent tightening of back-pressure on the control column to raise the nose (or any inadvertent back-pressure applied when, for instance, looking at something of interest below you, looking over your shoulder, being distracted by something in the cockpit, using the radio or even any encountered atmospheric turbulence, wake turbulence from preceding aircraft or gust shear) may take the aoa of the inner wing past the critical angle. The aircraft loses its lateral stability (positive roll damping) and it is most likely that the lower wing will drop in an uncommanded roll, and thus become increasingly more deeply stalled than the upgoing wing — which may not be stalled or just partly stalled.

Here is a condensed RA-Aus accident report:

"The two seat cabin ultralight stalled and spun just as the aircraft was starting the turn onto base. The pilot halted the autorotation and was very close to complete recovery from the descent with wings level when the aircraft contacted the ground. The very fortunate pilot's later explanation was that just as he was about to turn base he heard another aircraft give a base call. While his attention was diverted into searching for that aircraft speed bled off, control inputs were miscoordinated and the aircraft stalled and started to spin."

(The pilot's prompt recovery action also demonstrates it is far better to crash in a nearly level attitude rather than in a nose-down attitude.)

If that initial roll is not promptly recognised as an incipient stall or partial stall and allowed to continue — or perhaps incorrectly countered with opposite aileron without first unstalling the wing(s) by easing forward on the control column — the increasing aoa of the lower wing deepens the stall and causes greatly increased asymmetrical drag. Additional yawing forces in the same direction as the lower wing come into play, the nose-down pitching moment increases and the nose drops further. This is the incipient spin condition, where autorotation is about to commence; autorotation will happen quickly, and in some aircraft very quickly indeed. The result is the stall/spin fatality you hear about when an unwary pilot allows a spin to develop without sufficient height to recover; and of course you say 'How sad it is for the family' — while thinking (perhaps falsely) — 'but I'm too wary to get caught by such a simple misjudgement!

A similar situation may eventuate if the pilot picks up a dropped wing with rudder without first unstalling the wing (see Picking up a dropping wing with rudder) or if an aircraft taking-off exhibits a wing-rocking tendency (because its airspeed is too low) use of rudder could activate an incipient spin.

If the cg is aft of the rearward limit (thus closer to the centre of lift) the amount of elevator deflection or control force needed to rotate the aircraft to the critical aoa is reduced; i.e. just a relatively small rearward movement of the control column may rotate the aircraft to the critical aoa. If MTOW exceeds the design limit and/or the cg is aft of the rearward limit then recovery from the initial stall may be impossible. See the stick force gradient.

Apart from the weight and balance aspects, the rule to avoid such situations is in proper energy management — always maintain a safe speed near the ground consistent with the bank angle employed, continually envisage the wing aoa, i.e. keep the wing flying and keep the slip ball centred; and never apply an excess of bottom rudder in an attempt to tighten any turn if height is below the safe recovery height (3000 feet agl perhaps) for a fully developed spin.
How often have YOU come within a hair's breadth of eternity while being blissfully unaware of it?
Pilots need to be particularly careful when sightseeing. There is always a tendency to overbank the aircraft and pull back too much on the stick ('bank and yank' — perhaps also without adding power) so you or the passenger can get a good view of something on the surface directly below.

Extracts from three RA-Aus fatal accident reports:

1. " The pilot was conducting the flight for the passenger to take photographs of the property. Witnesses saw the Drifter fly over the farm buildings at an estimated 300 feet agl and then turn and fly back. The aircraft was seen to do a steep left turn during which the nose lifted. The aircraft entered an incipient spin from which there was insufficient height to recover."

2."The Sapphire was seen to be flying straight and level at about 150 feet agl. The aircraft entered a tight right hand turn, the nose was seen to drop and it appears the aircraft entered an incipient spin subsequently striking the ground in a nose-down attitude."

3."The pilot and his friend intended to do a short photographic flight ... for airport publicity purposes ... the passenger side door of the Skyfox had been removed to allow an unimpeded view for the photographer. No-one witnessed the accident however one pilot ... backtracking the same runway saw the aircraft take-off and start a normal climb but did not see the aircraft from that point as he continued to backtrack. It was only after line-up that he saw the smoke from the crash site. The impact site was 200 metres to the left of the upwind threshold and 900 metres from the start of take-off ... the wings hit almost simultaneously with the aircraft in a near vertical attitude, then bouncing 1.5 metres and coming to rest with the tail against a small tree, 60° nose down, right side up ... facing away from the take-off point ... indicates that the aircraft ... had turned back towards the runway, stalled and spun in with 180° rotation."

3.4 Popular precursors to a stall/spin: use rudder to hasten the turn or hold-off bank in a descending turn

The precursors to a stall/spin event in a low-power descending turn are the same as those for such an event in a level turn: if an excess of bottom rudder is applied the aircraft will be skidding and, unless some other factor is dominant, whenever an aircraft is slipping or skidding in a turn the wing on the side to which the rudder is deflected will usually stall before the other, with a consequent instantaneous roll in that direction. At descent speeds the aircraft is usually flying at a higher CL, and thus higher aoa, than when on the downwind leg for example. So a reduction in available aoa margin exists before allowing for the additional aoa required for the turn.

The descending turn from base leg onto the final approach to landing is the most obvious place for a pilot to attempt to hurry a turn with rudder, because of the need to align with the runway. A tailwind component on base leg to a crosswind landing will increase the tendency to hurry the turn with rudder as may other crosswind situations. If skidding, the excess bottom rudder is yawing the nose down, the rotation about the normal axis reduces lift from the inner wing and increases lift from the outer wing and the tendency is to use elevator to keep the nose up — which is going to bring aoa towards critical. Also because of illusory ground reference cues, there may be a tendency to increase the rate of turn by applying additional bottom rudder whilst maintaining the bank angle with opposite aileron — "holding-off bank". You should never hold-off bank in a descending turn (but see 'Note 2: holding-off bank'). If control column back-pressure is purposely or inadvertently applied the aircraft may enter a cross-controlled stall where it is going to roll further into the bank and enter an incipient spin.

In some aircraft it is quite possible that the pilot doesn't recognise that initial roll as an incipient or partial stall and allows it to continue, accepting it as part of the planned turn. The pilot will realise when he/she finds that applying corrective aileron increases the roll rather than reducing the bank. In similar situations there have been cases where the pilot has no doubt wondered why the elevators are completely ineffective when the control column is pulled right back to get the nose up.

Apart from the turn from base to final, such stalls might occur on final when avoiding bird strike or attempting a late correction to an out-of-line crosswind approach, or any time when you try to hurry a turn with bottom rudder. Stalls on the final approach, caused by failing to increase power when raising the nose to stretch the approach or reduce a high sink rate, will be exacerbated if the aircraft is also slipping. Possibly the most deadly low-level descending turn is the turn-back following engine failure after take-off.

Here is an extract from an RA-Aus serious injury incident report:

"The 8000 hour instructor (and student) had just returned from a flight and were over the top of the airfield when he thought the engine hesitated but then continued running. Considerable sink was experienced ... The instructor used rudder to yaw the aircraft toward the short runway then used rudder again to yaw the aircraft more to the right so that a landing could be made on the longest runway. The aircraft stalled and contacted the ground right wing first."

If flying cross-controlled when banked with an excess of top rudder — as in the sideslip manoeuvre or a slipping rather than skidding turn — then if it stalls the roll will probably be in the direction of the upper wing; i.e. towards an upright position, which is not quite so alarming and perhaps provides a little more time to react.

The following is an extract from an RA-Aus fatal accident investigation. A motor glider was returning to its home airfield after being airborne for about one hour 40 minutes; morning flying conditions were good with a five knot south-easterly. The accident occurred within a ground area considered quite safe for forced landings. The engine, propeller and pylon had been retracted and stowed within the fuselage; in such configuration the motor glider behaves as a pure glider and achieves its best glide ratio of 33:1 at 46 knots and minimum sink rate of 150 ft/min at 40 knots.

"The aircraft approached the airport from the west at approximately 400 feet agl, overflew the runway and continued straight ahead. It conducted a left hand turn back towards the runway before entering a stalled state and spiralling one and a half turns into the ground ... approximately 600 metres west of the western airport boundary and 400 metres north of the northern boundary ... it is the opinion of RA-Aus that the accident was attributed to pilot error and lack of situational awareness."

Note 2 — holding-off bank. Sailplane pilots probably spend more than 50% of flight time conducting small diameter circling turns within a lift source such as a thermal. The airspeed used to minimise diameter may only be 5–10% greater than the turning stall speed and under these conditions the outer wing tip will be flying at a significantly greater airspeed than the inner wing tip; for a 22-metre wing-span sailplane flying at 44 knots CAS in a 100-metre radius turn, the outer wingtip could be flying at 48 knots while the inner maintains 40 knots. Thus such aircraft develop more lift from the outer wing than from the inner even though the inner wing will have a higher aoa in the gliding turn, so there may be a need for sailplanes to hold-off bank.

3.5 What about loss of control in low-level climbing turns?

As we saw above, the increased lift force in the turn is provided by an increase in aoa. Now what will happen if you are climbing at Vx (the speed for maximum climb angle) using maximum power and decide — because of rising terrain or other obstruction (the latter particularly occurring in a delayed or misaligned go-around), an approaching aircraft or just natural exuberance — to make a quick 30° left turn using a 45° bank angle, while still maintaining the climb?

  • Coordinated climbing turn: if not keeping a close eye on the ASI and the airspeed has decayed just a little, the general aoa at Vx could be around 12°. To initiate a 45° bank turn, wing loading and thus aoa must increase by 41% which will take the aoa to 17°; i.e. past the critical stall aoa of 15° or 16°. Such full-power stalls in a coordinated climbing turn tend to result in the outer wing stalling first — because in a climbing turn the outer wing has a slightly higher aoa than the inner — with a fairly fast outer wing and nose drop. The roll towards the outside of the turn would initially level the wings but the increasing aoa of the down-going wing continues to accelerate the lift loss and increases the drag on that wing — a particularly rapid action if the propeller torque effect is such that it also reinforces the roll away from the original direction of turn. P-factor may also cause the aircraft to yaw when flying with high power at high angles of attack. Such stalls are likely to result in a stall/spin event if corrective action is not taken as soon as the initial loss of lateral stability — the uncommanded roll, or just a wing rocking warning — is apparent.

  • Cross-controlled climbing turn: if the turn is skidding (excessive bottom rudder applied) then the lower wing may stall first with the consequent roll into the turn — because only one wing is stalled — possibly being sufficiently pronounced to flick the aircraft onto its back. The propeller slipstream from a tractor engine will also be slightly asymmetric supplying more dynamic pressure and thus lift to one wing while reducing the effective aoa. We will discuss cross-controlled climbing turns further when we look at illusory ground reference cues.
Climbing at Vx, the best angle of climb speed, should always be regarded as a very short-term emergency procedure. Even a 30° banked climbing turn at Vx might produce an aoa of 14°, very close to the stall aoa and providing no margin for even minor turbulence, slight mishandling or inattention. Of course climb performance will be degraded unless extra power is available, which is unlikely because full power is normally used for the climb until a safe height is reached. The aoa margin that you should always have in hand to cope with such likely events is 3° or 4°, which indicates that, when climbing at Vx, turns should not be contemplated.

When climbing at Vy — the best rate of climb airspeed with aoa around 8° — then until a safe height has been gained turns should be limited to rate 1 (3° in azimuth per second or 180° per minute requiring perhaps 15° bank) to ensure an additional margin if wind/gust shear is encountered in the climb-out. When entering a turn during a full-power climb the aircraft must slow, because of the increased induced drag at the higher aoa required to make the turn with no excess power available to counter it. Consequently the aircraft's pitch attitude in the turn must be reduced sufficiently to maintain safe airspeed.

Here is an extract from an RA-Aus serious injury incident report:

Weather conditions: wind calm, nil turbulence, 23° C. Witness report: "The Gemini took off from the 950 metre runway, and after initially climbing, appeared to slow with a nose high attitude approaching trees. The aircraft was then seen to bank to the left then rotated 180° whilst rapidly losing height before impacting the ground at an approximate 45° angle 100 metres from the runway ... the motor sounded to be operating normally".

Elevator trim stall
Most light aircraft are not particularly longitudinally stable at approach speeds. At each stage of the landing approach, a flap-equipped aircraft should be properly re-trimmed to maintain the desired airspeed at the current cg position and selected flap configuration and the elevator trim tabs exert quite a large control force at flight speeds. With full flap deflection on the approach some aircraft may need quite an amount of nose-up trim; under these conditions applying full power following a go-around decision may induce a very strong nose-up movement — exacerbated by the elevator trim setting — and this attitude change must be anticipated by the pilot. If the pilot is slow in applying forward stick pressure and adjusting the elevator trim, the pitch-up may result in a highly dangerous 'elevator trim' stall and particularly so if the aircraft is also turning while low in energy. In addition, a 'heavy' aircraft with an aft cg may require considerable forward stick pressure. And, of course, the pilot must make allowance for the normal go-around conditions such as engine torque effects and density altitude.

3.6 When a stall with wing drop is recognised, what's the best way to recover?

One standard recovery procedure is generally applicable to all stall events or attitude upsets in a three-axis aircraft, whether or not overbanked and/or overpitched — i.e. nose high/low — though this recovery procedure is not applicable to a fully developed spin, whether erect or inverted.

Stall recovery generally requires the following concomitant stages:
  1. Ease stick back-pressure to reduce aoa of the most stalled wing below critical — which immediately gets the aircraft flying and restores normal 3-axis control.

    For any aircraft type, the amount of elevator deflection required to unstall the most stalled wing depends on many variables and may range from just an easing of back-pressure to a firm but smooth push towards the neutral position. All aircraft have their own handling idiosyncrasies and pilots must be aware of them. The nose should be positioned sufficiently below the horizontal to achieve safe flying speed while still well clear of the terrain. It's a matter of balancing height loss and proximity to terrain against a quick return to a safe flight speed. If the forward stick movement is both excessive and abrupt, the result could be an aoa movement below the zero-lift aoa, in which case there will be a reversed lift force on the wings that hinders recovery. This may be particularly apparent with trikes. The negative g due to the bunt could adversely affect some engines at a critical time.

    In instances of extreme overbanking (past 60° or inverted) — where although the upset may be the result of a cross-controlled stall or perhaps wake turbulence — the inverted or near-inverted wing will not be stalled but the aircraft will be in an inverted descent. The forward control column movement is needed to reduce the angle of descent. However, there may be the possibility of an inverted stall if the control column is pushed into its extreme forward position.


    Warning: never pull BACK on the control column as the initial response to a perceived stall or an overbanked nose-low attitude.

  2. Halt downward wing movement with rudder or centre the slip ball.

  3. Increase power smoothly, possibly up to maximum. The slipstream will also increase rudder and elevator authority, and aircraft stability, through its effect on fin and horizontal stabiliser; though if the aircraft is near the wings-vertical position — or is inverted — the throttle must be closed. In the recovery from a stall in a climbing turn, full power should be maintained unless the nose is pitched too far down.

  4. Roll the wings level with aileron so that all the lift force will be directed away from the ground. If inverted, choose the roll direction that provides quicker return to a wings-level attitude and, of course, the right way up.


Following the preceding actions: adjust power as necessary; if flaps were fully lowered then adjust by stages to take-off position; hold attitude until speed has built up to Vy (perhaps Vx if there are terrain problems); then ease into a climb to a safe altitude, where you can assess what went wrong. Never attempt to continue a landing approach after such an event; go around, allowing plenty of time to assess the environment before re-approaching.

If the aircraft is properly balanced (i.e. cg is within the limits for that all-up weight), any cross-controlled stall condition is readily countered. Of course if the pilot doesn't wait for the airspeed to build to a safe speed before again applying control column back-pressure, there will be a high risk of a secondary stall which may be very hazardous, depending on the height loss from the first stall.


3.7 What are these illusory ground reference cues?

It is thought that some ground reference optical illusions may be a contributing factor in loss of control situations near the ground. Such illusions can cause no problem in the circuit if the pilot confines external scanning to the intended flight path and checks for conflicting aerial traffic, while maintaining the appropriate instrument scan and a minimum safe flying speed. The latter is 1.5 times Vs, or perhaps as low as 1.3 times Vs in the latter part of a stabilised final approach as long as 20° bank angle is never exceeded. Fixing the circuit pattern on particular ground reference points, rather than the landing strip (for example "turn downwind around the big tree"), may contribute to illusory ground reference cues.
Wind drift illusions
Ground track of 720° turnWhen wind speed is reasonably high relative to aircraft speed, then the aircraft's drift with reference to the ground is very apparent to the pilot operating at lower levels, and particularly at short, difficult airstrips. The diagram above represents the ground track of an aircraft conducting a level 720° coordinated turn with constant speed and constant bank angle, such that in the second 360° turn, the aircraft would be encountering its own wake from the first 360° turn — assuming that the wake didn't sink below the flight path. The movement of the air mass in which the aircraft is borne is toward the west (with an easterly wind) and the turn is clockwise when viewed from above. When in the region above the red line, ground speeds will be lower; when below the red line, ground speeds will be higher. The separation of the tracks for each 360° is exaggerated for clarity.

When entering the south-west quadrant of the first 360°, the ground speed is initially high but reducing. The drift away from a central ground reference would provide the illusion of skidding out of the turn. Passing through the north-west quadrant, the skidding illusion will disappear as ground speed reaches the minimum. Ground speed starts to increase slightly through the north-east quadrant. However, the increasing drift towards the reference point provides a very noticeable illusion of a slip into the turn. This reaches a maximum as the aircraft enters the south-east quadrant, where it abates as ground speed increases to the maximum.

So, in a 360° coordinated level with constant speed and constant bank, the aircraft (and its wake) drifts downwind relative to the ground at the wind speed rate. The cockpit instruments will of course show a constant airspeed, bank angle and a centred slip ball. However, the reference cues seen by a pilot looking at the ground during a low-level turn indicate increasing and decreasing airspeeds, alternating with decreasing and increasing slip into the turn.
The downwind turn illusion
An unaware pilot may get into a difficult situation in the low-level circuit when an aircraft is turning 90° from crosswind to downwind (as in the progress through the SE quadrant of the diagram above), when drift cues create an illusion of slipping into the turn. At the same time, the increasing ground speed might suggest increasing airspeed. The reaction of an unwary pilot is to increase bottom rudder pressure. This will increase the bank angle and lower the nose. The pilot's reaction may well be to apply opposite aileron to reduce the bank, while increasing control column back-pressure to bring the nose up and possibly reducing power to reduce airspeed. Thus the aircraft is cross-controlled and flying at an aoa with little margin in reserve. This is coupled with decreasing airspeed, reducing lift and the aircraft sinking with a consequent increase in effective aoa. Under such circumstances, there is a likelihood of the aircraft stalling and snapping over. The downwind turn illusion seems to have more potential for error if the aircraft is climbing in a downwind turn, as described previously.

        Note: sometimes you may read material which purports that an aircraft loses airspeed and might stall when turning from crosswind to downwind because the aircraft is changing direction relative to the wind direction, which of course is nonsense. However, airspeed must decrease in the turn if power is not increased to counter the extra induced drag. Although an aircraft can only stall if the critical angle of attack is reached, a combination of aircraft inertia and a wind shear or turbulence event encountered in the turn could result in a stall (particularly if it is still climbing) or, more likely, a loss of height. If turning very close to the ground to follow a particular ground path (close to trees when stock mustering, for example) the increasing drift into the turn must be allowed for.
Pivotal height and reversal height
Pivotal height or pivotal altitude is a term used by proponents of ground reference manoeuvres such as 'eights on pylons'. It is one particular height above ground at which, from the pilot's sight line, the extended lateral axis of an aircraft doing a 360° level turn (in nil wind conditions) would appear to be fixed to one ground point, and the aircraft's wingtip thus pivoting on that point.

Pivotal height Imagine an inverted cone with its apex sitting on the ground reference point and an aircraft flying around the periphery of its inverted base while maintaining a constant airspeed. The vertical distance from the reference point to the centre point of the inverted base is the pivotal height, and the distance from the edge to that centre point is the turn radius. The bank angle is formed between the outer wall of the cone and the radius line.

The pivotal height in nil wind conditions is readily calculated by squaring the TAS in knots and dividing by 11.3. So any aircraft circling at a speed of 80 knots would have a pivotal height (80 × 80 / 11.3) around 550 feet, no matter what the bank angle.

In other than still air conditions the pivotal height varies with the ground speed. If the wind was northerly and the aircraft was turning anticlockwise (viewed from above), then ground speed would be lower on the eastern side of the turn and higher on the western side. When in the northern quadrant the aircraft would be drifting towards the centre point, while in the southern quadrant it would drift away. Drift would not be noticeable in the eastern and western quadrants but changed ground speeds would.

At 70 knots ground speed, the pivotal height is reduced to 450 feet, at 90 knots it is about 750 feet.

(Thus an exercise requiring a continuous 360° balanced turn at constant speed around a ground reference point, whilst holding pivotal height, involves continually changing the height above ground so that the line of pivot around each point is held constantly — rather than maintaining a constant distance from the 'pylon'. The bank angle must also be changed constantly as the wind drifts the aircraft towards or away from the pivot point. It is not an easy exercise to do well, and requires an ability to manoeuvre accurately whilst including the ground reference point in the normal scan pattern. Usually two ground reference points, about five seconds apart, are included for a figure eight pattern — otherwise known as 'eights on pylons'.)

Height reversal diagram Now imagine two cones — the upper one is the inverted cone with the aircraft flying around the edge of its inverted base and below that is a second cone with its base on the ground and its apex connecting with the apex of the upper cone. The vertical distance from the ground through the cone intersection to the centre point of the inverted base is the aircraft height.

So when an aircraft is turning at pivotal height in nil wind conditions, the wingtip appears to be fixed to a single point in the landscape. But when at any height other than the pivotal height, the wing tip will appear to move across the landscape. When an aircraft is turning at a height greater than the pivotal height, which is the normal situation in flight, the wingtip appears to move backwards over the landscape — path A in the diagram. However, when an aircraft is turning at a height less than the pivotal height (thus close to the ground), the wingtip appears to move forward over the landscape — path B in the diagram.

Thus, when a turning and descending aircraft descends below pivotal height there is an apparent reversal of the wingtip movement from backward to forward, which is the reason why pivotal height is sometimes termed reversal height. There is some thought that the reversal illusion may cause problems to unaware pilots during the final turn on approach to landing, because the turn may well pass through reversal height — at 50 knots ground speed, the reversal height is about 200 feet, at 60 knots it is about 300 feet and at 70 knots it is about 450 feet.

If the aircraft is in a banked turn below reversal height, and if the pilot looks down over the wingtip, she/he may get the impression that the aircraft is not turning and may then add additional bottom rudder so that the wingtip then appears to move backwards in the turn — the normal movement. This will cause a yaw and the aircraft's nose will move down, the aircraft may then appear to be nose-low and the pilot's reaction is to increase back-pressure on the control column. Low speed, excessive bottom rudder and an increasing control column back-pressure are the prerequisites for the aircraft to stall and roll toward the lower wing — an entry to incipient spin. All pilots should be aware of this illusion and that wind drift will exacerbate it — the base to final approach turn is probably the most important ground reference manoeuvre that recreational pilots regularly perform.

3.8 So how can low-level stall/spin events be avoided?

The rules to avoid low-level stall/spin situations are:
  1. Always expect the unexpected. (A rather trite statement but, hopefully, it conveys the message.)

  2. Good energy management: if potential energy of height is very low then kinetic energy should be reasonably high; i.e. maintain a safe speed (1.5 × Vs1) near the ground.

  3. Avoid distraction: maintain a scan appropriate to the situation. Never concentrate on one task, or a ground object/surface scenery, or a search for another aircraft thought to be in the circuit.

  4. Envisage the wing aoa while manoeuvring, keep the wing flying and don't exceed 30° bank angle. A good pilot is able to feel the onset of a stall — before any wing-drop — and catch it with just a slight forward stick input.

  5. Fly accurately. Keep the slip ball centred and never apply an excess of rudder in an attempt to tighten any turn or change direction.

  6. Never turn while still climbing at Vx. Restrict climbing turns at Vy to less than 15° bank angle.


The next article in this series, titled 'Don't land too fast in an emergency', discusses minimising ground impact following in-flight engine failure.

'Decreasing your exposure to risk' articles

| Introduction and contents | Recent RA-Aus accident history | Don't fly real fast | Don't stall and spin in from a turn |

| Don't land too fast in an emergency | Engine failure after take-off | The turn back: possible or impossible — or just unwise? |

| Wind shear and turbulence |


Copyright © 2007-2012 John Brandon     [contact information]