Engine Failure After Liftoff (Simulated) (AMEL, AMES)

For AMEL or AMES the evaluator must select Task E or F; Task G; and at least one other Task. This is the heavyweight half of that first pair — Task E ends with both throttles closed on the runway; this one ends with a single-engine climb, a return to the departure airport, or an approach to the most suitable landing area available (AI.XII.F.S7).

It is also the Task with the most skill elements in Area XII — eleven of them — so budget your prep accordingly.

Most AFM/POH engine-failure-after-takeoff checklists direct the pilot to (AFH 13-31):

1. Assume VYSE.
2. Set takeoff power.
3. Retract the flaps and landing gear (on some airplanes, gear before flaps).
4. Identify, verify, and feather the failed engine.

Then the ACS layers on the instructor-level detail: establish VYSE, or if obstructions are present VXSE or VMC +5 knots, whichever is greater, until obstructions are cleared, then transition to VYSE (AI.XII.F.S2). Simulate feathering — after which the evaluator establishes zero thrust (S4). Simulate securing the inoperative engine (S8). Tolerances: heading ±10° and airspeed ±5 knots (S9).

Teach the memory items as confirmations, not blind actions: the purpose of the memory items is to either initiate the appropriate action or to confirm that a condition exists. Action on each item may not be required in all cases (AFH 13-32). The gear may already be up; you still call it.

• Identify — determine which engine failed. Confirmation on the engine gauges may or may not be possible, depending upon the failure mode. Identification should be primarily through the control inputs required to maintain straight flight, not the engine gauges (AFH 13-31). The memory aid: "dead foot — dead engine." Rudder pressure is exerted on the side of the operating engine, so the foot doing nothing is on the side of the failed one (AFH 13-33).
• Verify — retard the throttle of the engine thought to have failed. No change in performance when the suspected throttle is retarded is verification that the correct engine has been identified (AFH 13-31).
• Feather — bring the corresponding propeller control fully aft (AFH 13-31).

Where students go wrong: reading the gauges first (slow, and often uninformative), skipping verify (which is how a good engine gets feathered), and hesitating at the propeller control. That hesitation has a documented cause — see the windmilling-prop card below.

Because the airplane's entire climb margin is drag. Feathering stops engine rotation with the propeller blade streamlined with the airplane's relative wind; with the blade angle in the feathered position, parasite drag from the propeller is at a minimum, and in a typical multiengine airplane the drag from a single feathered propeller is a small part of the airplane's total drag (AFH 13-3). A windmilling propeller, by contrast, is a flat disc of drag — which is also why VMC is highest when the critical engine propeller is windmilling at the low pitch, high rpm blade angle (AFH 13-24).

The mechanism is worth knowing: on most multiengine airplanes the props are full feathering, counterweighted, oil-pressure-to-decrease-pitch designs, so the only thing that keeps them from feathering is a constant supply of high-pressure engine oil. Bringing the control fully aft dumps governor oil pressure and the counterweights, aided by a spring or high-pressure air in the propeller dome, drive the blades to feather — the entire process may take up to 10 seconds (AFH 13-5).

Gear and flaps are the rest of the drag budget: raising the landing gear as early as possible after liftoff drastically decreases the drag profile and significantly increases climb performance should an engine failure occur (AFH 13-16).

VSSE is the safe, intentional one-engine-inoperative speed — the minimum speed to intentionally render the critical engine inoperative (AFH 13-1).

It is the instructor's speed, not the pilot's. No engine failure should ever be introduced below VSSE. If no VSSE is published, use VYSE (AFH 13-18). And note how briefly real operations spend below it: other than training situations, the multiengine airplane is only operated below VSSE for mere seconds just after liftoff or during the last few dozen feet of altitude in preparation for landing (AFH 13-18).

So the three-speed hierarchy you teach:

• VMC (red) — the floor for control.
• VSSE — the floor for intentionally failing an engine.
• VYSE (blue) — the target for performance.

The ACS is explicit: on multiengine practical tests where failure of the most critical engine after liftoff is required, the evaluator must consider local atmospheric conditions, terrain, and aircraft type, and must not simulate failure of an engine until attaining an altitude of at least 400 feet AGL and at least VSSE, VXSE, or VYSE (ACS Appendix 2, Multiengine Considerations).

The AFH matches it and adds the teaching progression: initiation of a simulated engine-inoperative emergency at low altitude normally occurs at a minimum of 400 feet AGL and only after the learner has successfully mastered engine-inoperative procedures at higher altitudes. Initiating one at extremely low altitude, immediately after liftoff, or below VSSE creates a situation where there are non-existent safety margins (AFH 13-36).

The method below 3,000 feet AGL: all in-flight simulated engine failures below 3,000 feet AGL should be introduced with a smooth reduction of the throttle, so the engine is kept running and available for instant use (AFH 13-35). At altitudes lower than 3,000 feet AGL, engine failure should be simulated by reducing throttle to idle and then establishing zero thrust (ACS Appendix 2).

When you simulate a failure, the learner responds with the memory items and retards the appropriate propeller control toward the FEATHER position. Assuming zero thrust will be set, the instructor promptly moves the propeller control forward and sets the appropriate manifold pressure and rpm (AFH 13-36).

Then you say it out loud. The AFH gives you the script: "I have the right engine; you have the left. I have set zero thrust and the right engine is simulated feathered." It is vital that the learner be kept informed of the instructor's intentions — any ambiguity as to who is operating what systems or controls increases the likelihood of an unintended outcome (AFH 13-36).

The ACS assigns the same division of labor: the applicant simulates feathering, and the evaluator should then establish zero thrust on the inoperative engine (AI.XII.F.S4).

Following a simulated engine failure, the instructor cares for the "failed" engine just as the learner cares for the operative engine (AFH 13-36). Concretely:

• If zero thrust is set to simulate a feathered propeller, the cowl flap is normally closed and the mixture leaned.
• An occasional clearing of the engine is desirable.
• If possible, avoid high power applications immediately following a prolonged cool-down at a zero-thrust power setting.

That last one is the go-around trap: you have spent five minutes shock-cooling a cylinder set, and now you want takeoff power right now. Plan the termination so you do not have to.

The learner's parallel duty is S6 — monitor the operating engine and aircraft systems and make adjustments as necessary. On the good engine that means temperatures, pressures, and cowl flaps, because it is now doing all the work.

Pick the decision point in advance, on the ground. If an engine fails before this point, the takeoff should be rejected, even if airborne, for a landing on whatever runway or surface lies essentially ahead. If an engine fails after this point, promptly execute the appropriate engine failure procedure and continue the climb, assuming the performance capability exists (AFH 13-12).

The gear is the usual marker: as a general recommendation, if the landing gear has not been selected up, the takeoff should be rejected, even if airborne (AFH 13-12), and once the gear is up, consider it a GO commitment if climb performance is available (AFH 13-16). But the gear should stay down as long as there is usable runway or overrun available to land on — this is not a license to retract on liftoff as a normal procedure (AFH 13-32).

And the thing to be avoided above all: attempting to continue flight when it is not within the airplane's performance capability to do so (AFH 13-11). If a climb is not possible at VYSE, the ACS wants you to maintain VYSE and return to the departure airport for landing, or initiate an approach to the most suitable landing area available (AI.XII.F.S7).

Use the AFH's own arithmetic, because it is more sobering than anything you can say. Under ideal circumstances the accelerate-go distance only brings the airplane to a point a mere 50 feet above the takeoff elevation — and to achieve even that, the pilot had to instantaneously recognize and react to an unanticipated engine failure, retract the landing gear, identify and feather the correct engine, all while maintaining precise airspeed control and bank angle as speed is nursed to VYSE. Assuming flawless airmanship, the airplane is now little more than one wingspan above the terrain, assuming it was absolutely level and unobstructed (AFH 13-12).

Then the climb: at a near 150 fpm rate of climb and a 90-knot VYSE, it takes about 3 minutes to climb the additional 450 feet to reach 500 feet AGL, during which the airplane travels 5 NM beyond the accelerate-go distance, at a climb gradient of about 1.6 percent. Any turn, such as to return to the airport, seriously degrades the already marginal climb performance (AFH 13-12).

Two caveats to add: not all multiengine airplanes publish accelerate-go distances and fewer still publish climb gradients, and where published, the figures were determined under ideal flight testing conditions — it is unlikely that this performance is duplicated in service conditions (AFH 13-12). For reference, the single-engine service ceiling is where the airplane can no longer maintain a 50 fpm climb OEI (AFH 13-11).

Exactly where you are least likely to be looking. The midair collision statistics describe this scenario: the vast majority of accidents occurred at or near nontowered airports and at altitudes below 1,000 feet, most in daylight with visibility greater than 3 miles, most involving aircraft not on any flight plan — and flight instructors were onboard in 37 percent of the accidents studied (AIH 9-11). A simulated engine failure at 400 feet AGL off the departure end puts you in the middle of that distribution with both pilots' eyes inside.

The mechanism the AIH names is teaching itself: an instructor conveying information while verifying the aircraft is flown safely may cause a decrease in attention to collision avoidance or loss of situational awareness, and heavy coaching may cause missed radio transmissions from ATC or aircraft in the pattern (AIH 10-9). During an OEI climb you are narrating identify-verify-feather, setting zero thrust, and watching airspeed and bank — that is the highest task-load moment in the syllabus.

So divide the labor explicitly and brief it:

• Say who owns the scan. The instructor should emphasize that both are responsible for maintaining a lookout to see and avoid other air traffic (AIH 10-7), but during the failure the student's eyes are on blue line and the ball. Take the outside scan yourself and say so: "you fly it, I own the traffic."
• Own the radio. Announce the simulated failure and your intentions on CTAF. Traffic behind you cannot see that your climb rate just collapsed.
• Do not let scanning lapse become a habit. Any observed tendency of a learner to enter flight maneuvers without first making a careful check for other air traffic needs to be corrected immediately (AIH 9-11) — including when the maneuver is an emergency drill.
• Pick the airport and the time. A busy nontowered field on a Saturday afternoon is the statistical worst case for this exercise.

• Feathering the wrong engine — skipping verify. Correction: drill identify-verify-feather as three separate spoken steps.
• Failure to maintain VYSE — pitching for altitude instead of speed. Correction: VYSE is maintained with pitch control (AFH 13-32); the bank and the power are separate levers.
• Wings level, or banking the wrong way. Correction: "raise the dead" — the inoperative engine is raised with a very slight bank toward the operating engine (AFH 13-33).
• Excessive bank or excessive rudder — chasing the ball instead of zero sideslip.
• Failure to reduce drag — leaving gear or flaps out, or leaving the prop windmilling.
• Rushing the securing checklist and mis-actuating a control. Correction: after the memory items, the remaining items are done deliberately and without undue haste unless a fire is suspected — there is a distinct possibility of actuating an incorrect switch or control if the procedure is rushed, and other than closing the failed engine's cowl flap, none of these items, if left undone, adversely affect climb performance (AFH 13-32).
• Turning back too soon. Climb straight ahead or with shallow turns to avoid obstacles to an altitude of at least 400 feet AGL before attempting a return to the airport (AFH 13-32).

That it treats a certification result as a physical constant. VMC is a fixed airspeed only for the very specific set of circumstances under which it was determined during aircraft certification. In reality, VMC varies with a variety of factors — the VMC noted in practice, demonstration, or actual OEI operation could be less or even greater than the published value, depending on conditions and technique (AFH 13-23).

The historical certification definition: the sea level calibrated airspeed at which, when the critical engine is suddenly made inoperative, it is possible to maintain control with that engine inoperative and thereafter maintain straight flight at the same speed with a bank angle of not more than 5° (AFH 13-23). The dynamic determination is done by highly experienced test pilots during certification and is unsafe to be attempted outside of those circumstances; there is also a static determination, and if the two differ, the higher is published (AFH 13-23).

From the historical 14 CFR 23.149 conditions (AFH 13-24):

• Power on the operating engine — VMC increases as power increases. With normally aspirated engines VMC is highest at takeoff power and sea level and decreases with altitude; with turbocharged engines it stays constant up to the critical altitude and then decreases.
• Inoperative propeller drag — VMC increases with increased drag, so it is highest with the propeller windmilling at the low pitch, high rpm angle.
• CG position — VMC increases as CG moves aft, because the rudder's moment arm and therefore its effectivity are reduced. For a typical light twin the aft-most CG is the most unfavorable.
• Weight — VMC increases as weight is reduced.
• Landing gear — VMC increases when the gear is retracted; extended gear aids directional stability.
• Flaps — determined with flaps in the takeoff position (for most twins, 0°).
• Bank angle — VMC increases as bank angle decreases.

VMC may increase more than 3 knots for each degree of bank reduction between 5° and wings-level. Since VMC was determined with up to 5° of bank, loss of directional control may be experienced at speeds almost 20 knots above published VMC when the wings are held level (AFH 13-25).

Explain the mechanism, not just the number: with bank toward the operating engine, the horizontal component of lift generated by the bank balances the side force from the rudder, rather than using sideslip to do so. Sideslip requires more rudder deflection, which in turn increases VMC (AFH 13-24). The 5° allowance works in the manufacturer's favor — but the AFH is candid that the method may result in unsafe flight from both the large sideslip and the need to increase angle of attack to maintain the vertical component of lift (AFH 13-24).

Also worth stating so nobody conflates the two: the 5° bank does not inherently establish zero sideslip or best single-engine climb performance. Zero sideslip, and therefore best single-engine climb performance, may occur at bank angles less than 5° (AFH 13-25).

Zero sideslip is the airplane presenting its smallest possible profile to the relative wind — if a yaw string were mounted on the windshield it would align itself straight up the center (AFH 13-27, 13-28). It is achieved by a particular combination of aileron and rudder, used together; used individually, neither is correct. Used together in the proper combination, zero sideslip and best climb performance are achieved (AFH 13-27).

The catch that trips up single-engine pilots: in a multiengine airplane with an inoperative engine, the centered ball is no longer the indicator of zero sideslip due to asymmetric thrust. In fact, there is no flight deck instrument that directly indicates conditions for zero sideslip (AFH 13-27).

So teach the substitute values. Without specific manufacturer guidance for zero sideslip, use a bank of 2° and one-third to one-half ball deflection on the slip/skid indicator toward the operative engine (AFH 13-32). The actual bank angle for zero sideslip varies among airplanes from one and one-half to two and one-half degrees (AFH 13-29), and the precise condition varies slightly by model and available power. Note also that the zero sideslip ball position for straight flight is also the zero sideslip position for turning flight (AFH 13-29), and that these recommendations apply to reciprocating multiengine airplanes flown at VYSE with the inoperative engine feathered (AFH 13-29).

The stall-spin in a twin, from asymmetric thrust at low speed. In order to spin any airplane, a stalled condition needs to exist. At the stall, the presence or introduction of a yawing moment can initiate spin entry. In a multiengine airplane, the yawing moment may be generated by rudder input or asymmetrical thrust. Spin awareness should therefore be greatest during VMC demonstrations, stall practice, slow flight, or any condition of high asymmetrical thrust, particularly at low speed and high AOA (AFH 13-18).

The stakes: no multiengine airplane is approved for spins, and their spin recovery characteristics are generally very poor (AFH 13-18). Very few twins have ever been spin-tested — none are required to be — so recommended recovery techniques are based only on the best information available, and the departure may be abrupt and possibly disorienting (AFH 13-18). Recovery, per most manufacturers, is these actions taken as near simultaneously as possible, held until rotation stops:

• Retard both throttles to idle.
• Full rudder opposite the rotation.
• Full forward elevator with ailerons neutral.

It will take considerable altitude (AFH 13-18).

Your prevention: pay strict attention to the maintenance of proper airspeed and bank angle as the learner executes the procedure (AFH 13-18). Airspeed and bank. Those are the two you guard, and the two you take the controls over.

Because a turning propeller looks like a working engine. A windmilling propeller, in many cases, has given the improperly trained multiengine pilot the mistaken perception that the engine is still developing useful thrust, resulting in a psychological reluctance to feather, as feathering results in cessation of propeller rotation (AFH 13-36).

The AFH prescribes the cure directly: the flight instructor should spend ample time demonstrating the difference in the performance capabilities of the airplane with a simulated feathered propeller (zero thrust) as opposed to a windmilling propeller (AFH 13-36). Fly both, at altitude, and let the student read the VSI. That comparison does more than any lecture.

When you do feather for real in training, do it at altitudes and positions where safe landings on established airports may be readily accomplished if the propeller will not unfeather, and plan unfeathering and restart to be completed no lower than 3,000 feet AGL (AFH 13-36). Be aware that at some elevations, in many popular training twins, that may be above the single-engine service ceiling and level flight will not be possible (AFH 13-36). And repeated feathering and unfeathering is hard on the engine and airframe — do it only as necessary to ensure adequate training (AFH 13-36).

Usually not, and this is a nuance students miss because training always simulates a total failure. Not all engine failures result in complete power loss. If there is a performance loss when the throttle of the affected engine is retarded, some power is still available. In that case the pilot may consider allowing the engine to run until the airplane reaches a safe altitude and airspeed for single-engine flight — and while shutting down a malfunctioning engine may prevent additional damage in certain circumstances, shutting down an engine that can still produce partial power may increase risk for an accident (AFH 13-33).

The counterweight: catastrophic failure accompanied by heavy vibration, smoke, blistering paint, or large trails of oil indicates a critical situation. Then:

1. Feather it.
2. Complete the securing checklist.
3. Divert to the nearest suitable airport.
4. Declare an emergency with ATC for priority handling (AFH 13-34).

Because it compresses every margin at once. Engine failure on takeoff, particularly with obstructions, is compounded by the low airspeeds and steep climb attitudes used in short-field takeoffs. VX and VXSE are often perilously close to VMC, leaving scant margin for error as VXSE is assumed. If flaps were used for takeoff, the engine failure situation becomes even more critical due to the additional drag incurred (AFH 13-17).

The AFH's mitigation is a planning decision, not a piloting one: if VX is less than 5 knots higher than VMC, give strong consideration to reducing useful load or using another runway so that a short-field technique is not required (AFH 13-17). Teach students to make that call at the fuel pump.

Related: the use of wing flaps for takeoff virtually eliminates the likelihood of a single-engine climb until the flaps are retracted (AFH 13-32). That is why the ACS puts flap retraction in the skill elements (AI.XII.F.S3) rather than leaving it to the checklist.