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

Fly safe! home page

Decreasing your exposure to aerodynamic risk

5. Engine failure after take-off

Rev.5 — page content was last changed 14 June 2012

Google logo

Recreational light aircraft engine/propeller failures in flight are not rare and generally are not particularly stressful — if adequate route planning has preceded the flight — except if operating over water or when failure occurs in take-off or go-around modes.

The take-off and late go-around sequences in a very light aircraft are the most critical of all normal flight procedures; all the engine's available performance must be employed during the acceleration and initial climb, leaving no power in reserve. There is no potential energy of excess height or excess speed available so, during take-off, the pilot's options are very limited — even more so in high density altitude conditions.

All recreational light aircraft are low-inertia aircraft — the 'draggy' ones are low-inertia plus low-momentum aircraft — so are detrimentally affected by rough air at low levels.

If complete or partial loss of thrust occurs shortly after lift-off, the best and probably only option, is to promptly establish a safe descent attitude and speed, and land on the remaining airstrip, ground looping if necessary to avoid an obstruction. It is only if total loss of thrust occurs after achieving a 'decision height' that a safer option other than land more or less ahead MAY exist. A possibility of loss of control is introduced any time that total or partial loss of thrust is experienced in the climb or in a turn, the cardinal rule in those situations is to 'fly the aeroplane!'; i.e. maintain control of the situation.

In EFATO situations there is a strong compulsion to direct your thinking towards minimisation of airframe damage — very understandable if the aircraft has taken a long time to build or acquire — but the overriding priority has to be directed towards minimising risk to occupants, even if that means sacrificing part of the airframe.
What's meant by the term EFATO?
Once airborne, naturally any engine failure is a failure after take-off even if the aircraft is 100 nm from the take-off point. However, the EFATO term is usually accepted to mean a significant loss of thrust occurring while the aircraft and pilot are still in 'take-off or go-around mode'. For example, haven't yet set course, or raised take-off flap, or haven't yet reached 1000 feet agl if intending to operate above that height, or, if doing circuits have not yet completed the crosswind turn; i.e. a 'thrust deterioration at take-off' that occurs while climbing soon after lift-off or during a go-around when the aircraft has little energy to trade.

This module presumes the reader is familiar with the contents of the earlier 'Don't stall and spin in from a turn' and 'Don't land too fast in an emergency' modules, which are pertinent to this document.


5.1 What happens when the engine or propeller fails in the initial climb?

Pilot and aircraft reaction times
A recreational light aircraft established in the climb attitude at Vy (best rate of climb speed) has an aoa perhaps around 6–8°. At such angles there is significant induced drag so when thrust is lost, for any of a multitude of reasons, the aircraft may rapidly decelerate to stall speed — worse if the airframe also has much parasitic drag. The immediate action required is to convert the potential energy of height to a safer speed. When climbing at Vx (best angle or emergency climb speed) aoa could be around 8–12°, so deceleration following power loss is a greater hazard. Of course recreational light aircraft Pilot Certificate holders are aware of this and take immediate action to lower the nose to a position consistent with their estimate of the approach or glide attitude in pitch. Or do they? Material developed by the late Mike Valentine, the former RA-Aus Operations Manager and prestigious GFA stalwart, is included in this section. Mike conducted considerable research into pilot and aircraft reaction times following cable breaks in winched glider launches and engine failure after take-off in recreational light aircraft. Some research results — which were very similar for both aircraft categories — were published in the June 2004 issue of the RA-Aus journal, and are summarised as follows.
  • Following an engine/propeller failure in the climb, there is an initial delay while the pilot's brain adjusts to the shock of the event and then she/he pushes the control column forward. This reaction time appears to average around three to five seconds, much longer than might be imagined, but similar results are obtained in tests by other aviation bodies. Mental paralysis/disbelief, i.e. 'this can't be happening?', is the main contributor to that delayed reaction; meanwhile the aircraft is slowing at perhaps 2 to 4 knots per second. It can be exacerbated by slight panic if the power loss is accompanied by very unusual engine noises, smoke and/or violent shaking.

  • A quiet breakdown in the propeller speed reduction unit results in the unloaded engine's rpm increasing while the propeller is 'freewheeling' — producing no thrust — and it may take a little longer for the pilot to realise what has happened.

  • If the aircraft is equipped with an effective elevator trim system and the pilot has trimmed for the climb speed — which is generally similar to the best glide speed — then the aircraft will of course try to regain its trimmed speed when thrust decreases; however, this takes too long to stabilise, and the pilot must take firm control and push the stick forward.

  • As the pilot pushes over into the glide attitude the aircraft follows a curved flight path. During this manoeuvre, pitch attitude and wing loading are changing, and the aircraft still slows for two or three seconds before accelerating. When the desired attitude is eventually attained, the pushover is terminated and the aircraft is then apparently stable in its glide attitude.

  • Apparently? Yes because, although the aircraft is in the required nose low attitude, it has just been through an energy-changing manoeuvre without the benefit of thrust to sustain it. Its inertia, aided and abetted by its drag, prevents it from immediately attaining the airspeed appropriate to the glide attitude; some seconds must be allowed for the aircraft to build to that speed. Gravity alone can't instantly accelerate an aircraft to a safe speed through a 10 or 15 degree pitch attitude change.

  • The real EFATO event will be noticeably different from that experienced in a simulated EFATO because there is no residual thrust from an idling engine. If the propeller is windmilling there will be additional drag and thus a bit steeper descent path. Also the lack of a cohesive propeller slipstream over the tailplane will make the elevators feel different — and less effective.

  • Any attempt to start manoeuvring the aircraft without allowing sufficient time for the indicated airspeed to build to, and stabilise at, a safe speed will risk loss of control — and don't think there is a discernible lag in the ASI; there isn't if it is in good condition and the pitot-static system is unobstructed. If the pilot lowers the nose to the glide attitude and immediately performs just a moderate 'bank and yank' manoeuvre, the aircraft may stall and spin. At least five seconds will elapse from the moment the pilot pushes the stick forward to the time the airspeed margin over stall is safe enough to carry out a gentle manoeuvre. The diagram below represents the result of Mike's tests in a simulated (and placid) EFATO when climbing at 55 knots (about 1.3 Vs of 42 knots). Similar results were found in other tests. The diagram doesn't show the 3–5 seconds reaction time for the average pilot, as the pilot for the test series was conditioned to an expectation of the throttle being pulled by the observer. EFATO inertia
  • During the pushover, the control column was pushed forward smartly enough and far enough to unload the wings to perhaps 0.5g or less, so that the aircraft is still totally controllable even if the airspeed reduces below the normal Vs1 of 42 knots. At 0.5g the airspeed will build relatively quickly because the lift will be nearer zero and thus induced drag is reduced to nearer zero.
Unloading the wings is a good practice to practise
As mentioned in the flight envelope section of the 'Don't fly real fast' article a light aircraft can be safely held at sub-Vs speeds for several seconds by unloading the wings so that the aircraft is operating in the reduced-g band between zero g and +1g, but not in negative g. The stall speed between +1g and 0g is still proportional to the square root of the wing loading g ratio, as indicated in Table 4.1.

Table 4.1: stall speeds at positive loads below +1g
Load factorSquare rootStall speed
+1g 1 42
+0.75g 0.86 36
+0.5g 0.70 30
+0.25g 0.5 21
0g 0 0

Note: when the wings are unloaded, ailerons and rudder can be used in ways that would be regarded as excessive at 1g loads. This unloading technique also has value as a stall recovery exercise (at a safe height) for pilots to really comprehend what is going on. It involves unloading the wings to perhaps 0.25g by pushing sufficiently forward on the control column so that you feel very light in the seat but not yet constrained in the harness as you would be if imposing negative g — or if dirt and dust start floating up from the floor. When unloaded — which takes an instant — roll the wings level (holding near zero g of course) using full aileron and whatever rudder is necessary (often quite a lot), and centre the aileron and rudder as soon as the wings are level. As drag at that minimum aoa is much reduced, speed will build more quickly and thus dive recovery is started earlier. With practice, the total height loss by taking such decisive action may be less than in a gentle reaction, and the speed will stay well within the allowable envelope in most recreational light aircraft. There will not be any fuel system problems as long as negative g is not applied. However, if forward pressure is slightly relaxed and the aircraft allowed to return to its normal 1g state while airspeed is below Vs1, the wing will promptly stall.

5.2 Practice good energy management in the take-off!

Planned energy management during the initial climb
Following engine failure in the climb, the total energy available is the sum of kinetic energy and potential energy of height. As shown above, a lot of that kinetic energy is lost to drag in the 6–8 seconds following loss of power. The potential gravitational energy must be converted to kinetic energy so that the total energy level of the aircraft is maintained, albeit at a lower level than that immediately prior to the power loss.

There may not be enough time available to regain enough speed within the remaining height to have sufficient energy to arrest the rate of sink (i.e. flare) for a normal landing. A heavy or very heavy landing is then almost inevitable. For example, the low-momentum Thrusters and Drifters have thick high-lift wings that give their best climb rate [Vy] at about 50 knots. They probably need about 150 feet to build enough airspeed to enable the aircraft to be flared for a normal landing; obviously, more slippery aircraft need less height.

The solution to this potential problem is planned energy management during the initial climb. Don't use the recommended speed for best rate of climb, use a climb speed perhaps 10–20% higher until at 200–250 feet, then steepen the climb a little to maintain Vy. The loss of initial climb performance won't be particularly significant but the additional speed in hand will make a difference if you lose thrust at a critical height. Of course, you may prefer to maintain the higher speed as a cruise-climb speed, particularly if there is a reasonable headwind or a tendency to overheat.
What about using the best angle of climb speed for initial climb?
Vx should not be used in normal operations — it should be regarded as an emergency climb speed. The high pitch attitude, high aoa and low speed provide a very limited safety margin if power is lost. If an airstrip is so marginal that you consider you must use Vx to clear obstructions at the end of the strip — or worse, out-climb rising terrain — then you should not be using that airstrip. If you absolutely have to use Vx for obstacle clearance then lower the nose to a safer climb speed as soon as possible.

5.3 Always be ready to implement plan B!

Have a mental 'what if?" action plan
Pilots must always be prepared for the possibility that the engine/propeller will lose thrust during the take-off and climb out (or at any other time during flight), and have simple pre-formulated mental action plans for the particular airfield/strip/runway conditions and various failure modes — remembering that, depending on height if the engine fails, there may be little time to do much else but keep your eyes outside the office, select the landing run and fly the aeroplane. One thing though — it is important to close the throttle early enough to avoid the engine suddenly regaining full power at an inopportune time; e.g. just as you are about to flare, thus driving the aircraft into the ground — which has happened on occasion.

If there is any thought that something is not quite right during taxying, run-up or the take-off ground roll, the flight should be abandoned immediately. A surprising number of pilots disregard indications/warnings that something is not as it should be and press on to an inevitably expensive reminder that engine/fuel/propeller problems cannot fix themselves. It'll be okay? Not likely!
On-field landing
If the aircraft is very low when the engine fails the only option is to keep the wings reasonably level, the slip ball centred and land more or less straight ahead. So the minimum action plan would be:
  1. If loss of thrust or other problem is evident, immediately push over into the approach attitude while keeping the slip ball centred.

  2. If loss of thrust is accompanied by extreme vibration or massive shaking of the aircraft (possibly due to a propeller blade failure), it is important to immediately shut down the engine to avoid it departing from its mountings.

  3. Do nothing else while waiting the few seconds for the aircraft to stabilise at a safe speed — except hold that attitude, keep your eyes outside and decide the landing run; probably there will be little time or opportunity to conduct any cockpit or radio drills prior to touchdown.

  4. Ensure the throttle is closed, lower full flap or sideslip if height permits, then land the aircraft. Be careful to avoid wheelbarrowing.

  5. Brake hard and/or ground loop if necessary to avoid collision. The groundloop is induced by booting in full rudder (and brake) on the side to which you want to swing and will probably result in some wing tip, undercarriage and propeller damage, unless you impact something other than the ground. Running it into long grass will help slow the aircraft.

There have been occasions, even at small airfields, where a recreational light aircraft losing power at 200 feet or less had sufficient height to safely turn 60–90° and land on, or parallel with, an intersecting strip. Of course, the pilot in those reported cases has been quite familiar with the aircraft's capabilities and had commenced take-off with little or no runway behind.
Off-field landing
If some height has been gained but there is no possibility of landing on the airfield, then an off-field landing is mandatory. Look for somewhere to put it down but don't immediately fix on the first likely landing site spotted straight ahead of you — there may be a more suitable site closer. However, you have to rapidly assess your height and airspeed (i.e. your energy level), and the turn possibilities available; i.e. can you safely turn through 30° or 45°, perhaps 60°, and still make it to that much better looking site? Will the wind assist or hinder? How much height will be lost in the turn? It has to be a quick decision because at best you have just a few seconds available to plan the approach. If any doubt go for 'into wind' and remember you can't stretch the glide!

Do not choose the site at marginal distance, even if it's perfect. Close by is better because the height in hand can be used for manoeuvring the aircraft into the best approach position. Because you have no power available you must always have an adequate height margin to allow for distractions, misjudgements, additional loss of height in turns, adverse wind shifts, sinking air, turbulence and other unforeseen events — and you can dump excess height quickly using full flap or sideslipping. Remember that the rate of sink whilst sideslipping is high and the slip must be arrested before the flare.

Some major factors affecting the outcome of a forced landing are highlighted in the previous module 'Don't land too fast in an emergency' and it is not my intention to list all factors that might be assessed in the decision making process following EFATO. Suffice to say, it is impossible to assess everything in the few seconds available, hence the need for prior knowledge of the airfield environs, plus a pre-established plan B and intuitive procedures for any situation that may occur before you are established at a safer height.

Apart from being clearly within range the choice of landing site is affected by:
  • wind strength and direction

  • ground run availability and direction; a short into-wind site may be preferable to a longer but crosswind/downwind site for an aircraft with a low stall speed; the reverse applies for an aircraft with a high stall speed. It all relates to kinetic energy and stopping distance

  • approach obstructions; final approach may require some diversion around/over trees, under/over power-lines plus avoidance of other obstructions. Can the near-ground turns be handled safely? Is there sufficient margin for misjudgement and/or wind gusts?

  • ground surface and obstructions, including livestock, during the ground roll. Can you steer to avoid them? Are livestock or kangaroos likely to take fright and run into your path?

  • the energy absorbing properties of the vegetation

  • ground slope: the possibilities of landing downslope may range from difficult to impossible; moderate upslope is good if the pre-touchdown flare is well judged. There is a much greater change in the flight path during the flare; for example, if the upslope has a one in six gradient (about 15°) and the aircraft's glide slope is 10° then the flight path has to be altered by 25° so that the aircraft is flying parallel to the upslope surface before final touchdown. A higher approach speed is needed because the increased wing loading during the flare (a turn in the vertical plane) increases stall speed. If the wind is upslope then a crosswind landing may be feasible

  • if a rural road is chosen can you avoid traffic, larger trees, drainage ditches, wires and poles, particularly in a crosswind situation?

  • a final approach into a low sun should be avoided so that vision is not obscured.
All of this is impossible to assess in the few seconds available, hence the need for prior knowledge of the airfield environs and a pre-established emergency procedure for any situation that may occur before you are established at a safe height.

As height increases, the options increase for turning towards and reaching more suitable landing areas, making a short distress call and doing some quick trouble shooting.
When trouble-shooting full or partial power loss remember the first edict — constantly 'fly the aeroplane!'. If the engine is running very roughly or died quietly (i.e. without obviously discordant sounds associated with mechanical failure) and time is available, then apart from the engine gauges, the obvious things to check or do are:
  • Fuel supply: switch tanks (making sure you haven't inadvertently switched to the 'fuel off' position), fuel booster pump on, check engine primer closed.
  • Air supply/mixture: throttle position and friction nut, throttle linkage connection and mixture control position. Apply and maintain carburettor heat (while engine is still warm), setting the throttle opening at the normal starting position. Apply carburettor heat or select alternate air to bypass the air intake filter — which could be blocked by grass seeds or a bird strike.
  • Ignition: position of ignition switches — and try alternating switches in case one magneto is operating out of synchronisation.
  • Or: reverse the last thing you did before the engine packed up.
  • And then: try a restart. There is no point in continuing with a forced landing if the engine is really okay.
Cockpit check prior to touchdown
  • Pilot and passenger harnesses must be tight and maybe remove eyeglasses. Seats should be slid back and re-locked in place (if that is possible without adding to the risk) but be aware of the cg movement.
  • Advise the passenger of intentions, warn to brace for impact and advise evacuation actions after coming to a halt.
  • Unlatch the doors so that they will not jam shut on impact. If the aircraft has a canopy or hatch take similar safety action, if that is possible without the canopy affecting controllability or detaching and damaging the empennage.
  • If equipped with a retractable undercarriage, leave the wheels down unless surface conditions indicate otherwise.
  • To minimise fire risk turn the ignition, fuel and electrics off.

It is important to research and develop your own safety plan, including the cockpit and radio drills, so that it is more deeply ingrained and appropriate to your capabilities and the aircraft being flown. Don't just adopt a plan published by someone else. Before moving onto the runway for take-off, do a mental rehearsal of plan B; such rehearsal is a powerful safety aid.

As height achieved before engine failure increases, the options increase for trouble-shooting, turning towards and reaching more suitable landing areas; making a distress call on a selected frequency; properly securing the fuel, ignition and electrical systems; and for an adequate cockpit check prior to touchdown — but all in accordance with the plan.
Partial thrust loss
If the engine/propeller does not fail completely but is producing sufficient thrust to enable level flight at a safe speed then, if you can't determine the fault, it may be possible to return to the airfield. Make only moderate turns, maintaining height if possible without the airspeed decaying, and choose a route that provides potential landing sites in case the engine loses further power. It's a judgement call whether you should take advantage of a possible landing site along the way because the off-field landing may damage the aircraft and perhaps injure the occupants. But that must be weighed against the chance of further power loss producing a more hazardous situation; it is usually considered best to put the aircraft down at the first reasonable site. If there is insufficient power to maintain height, then of course you must set up an off-field landing.
Intermittent power
If the engine is producing intermittent power, and you can't determine the cause using your Plan B trouble-shooting schedule, it is probably best to use that intermittent availability to get to a position where a glide approach can be made to a reasonable off-field site. Intermittent power negates the ability to conduct a controlled approach and could get you into a dangerous situation. So having achieved a position where you can start a final approach, then secure the engine by shutting down the fuel, ignition and electrical systems. Securing the engine early means it will be colder at touchdown, reducing fire risk, but it mainly gets that job out of the way so you can concentrate on flying.

The next article in this series discusses 'The turn-back: possible or impossible — or just unwise?'.

'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]