intro
Engine failures can be scary experiences, which is why it’s so critical to practice them frequently and have the process down cold so that you can respond correctly and ensure a safe outcome. One DPE I’ve worked with half-jokingly said that “the multiengine rating is really just more single-engine flying.” If, like many multi students, you’re doing your CMEL add-on to your CSEL, learning a new airplane and set of procedures will be a new challenge. The second engine means twice the maintenance and about twice the fuel burn, so the rental rates on twins like a Seminole are often about twice what you’d pay for a 172. That’s one reason why it’s crucial to have your procedures memorized ahead of time to make the most efficient use of your training in the actual aircraft. You should also start with a good understanding of VMCa, the gravity tax, and constant-speed props.
Memory items vs checklist
Different flight schools may list different amount of the OEI process as memory items versus checklists, and may also have different philosophies for checklist execution. Be sure to follow your school’s guidance on checklists so that you don’t fail any evaluation events. In my opinion, everything through feathering should be memorized stone cold, because in the real world you may not have time to go line by line if your engine fails at 500 AGL. If you’re an MEI, you should memorize everything, because you likely won’t have enough bandwidth to monitor flight path while cross-checking your student against your own checklist.
Conceptual framework
In the Seminole (and many other light twins), there’s a sequence of steps that you need to follow immediately upon power loss in order to keep the airplane flying. At a high level, these can be broken down into the following:
- Control the airplane: this may seem obvious, but it’s important to mention so that you don’t lose focus of it while executing the subsequent steps.
- Maximize thrust: make sure your powerplants are all producing as much thrust as they can, and check that you didn’t set anything up incorrectly.
- Minimize drag: clean up drag and verify that you didn’t inadvertently leave anything out that will slow you down.
- Identify the dead engine: figure out which one failed.
- Decide what to do: determine whether you’ll troubleshoot or immediately feather the engine. You should make this decision ahead of time so that you don’t waste time thinking in the moment. I usually avoid troubleshooting if a) I’m <1000 AGL, b) already joining or in the traffic pattern, c) <200 ft above FAF altitude on approach, or within 5 miles of the FAF. No rule fits all situations, so it’s important to brief adaptations to your plan.
- Troubleshoot: try to determine what caused the engine issue and try to fix it.
- Feather: eliminate drag from the failed engine.
- Secure the dead engine: shut off fuel, electrical, etc. to avoid fires and other issues. Depending on how much time you have (e.g. engine failure on the downwind), you should consider skipping this, because the costs of the distraction from flying the airplane outweigh the benefit. Again, brief the circumstances under which you’ll perform each action.
Step-by-step Execution
All of this comes straight out of the POH (3.5a). One limitation of the POH is that it defines an outcome, but not a lot about how to achieve it (“Fly the airplane right!”) That’s what this guide is for, along with some theoretical framework so that it makes more sense. In the event that I got something wrong, always follow the POH and send me a note so I can fix the article. Bold items are the memory aids.
Fly the airplane:
Fly the airplane:
- 88: maintain 88 knots or better. Blue line, AKA VYse, is your friend.
- Keep it straight: maintain directional control.
As your right foot advances, so does your right hand, keeping the string tight.
|
As the right pedal advances, so does the yoke in tandem with a roll input, keeping the string tight.
|
If you’re a visual person, I recommend always picking a visual point off the nose on the horizon before you throttle up on takeoff and set your heading bug to detect a change more quickly. You’ll likely see the yaw first, so get ready for that and make your compound input immediately. Looking at your visual spot will help you make minor corrections to keep the airplane pointing in the right direction.
In cruise, the procedure is a bit different: you typically have excess speed above blue line, which gives you more control authority, as well as a nose-level attitude, so the roll and yaw are not as drastic, and your pitch doesn’t need to go down as aggressively. Depending on how fast you are (e.g. >130 KIAS), it may make sense to swipe the trim up to near the takeoff setting to trade excess speed for altitude as you bring it back to blue line or just above (100ish should give you a good blend of speed buffer and drag). Keep in mind that you’ll need more rudder and aileron relative to the initial input as you decelerate and lose control authority.
After this, all of the actions we take are geared toward maximizing performance, not necessarily controllability. Maximizing thrust and cleaning up the airplane, for example, are good for performance, but bad for Vmc. That’s why “88, keep it straight” comes first and needs constant attention (think “Blue line means life, red line means death”).
Maximize Thrust
Reds: full mixture: march your mixtures all the way forward. This will help you make maximum power or close to it. On a hot/high day you may be too rich and lose a few percent, but that’s usually better than missing a fuel delivery problem. The excess fuel will help cool your good engine and reduce the likelihood of it overheating.
Blues: full prop: maximize the RPMs.
Blacks: fully advance your torque levers. It’s important to do RPM before torque to avoid going past the oversquare limits.
On the upwind, this is mostly an idiot-proofing check and shouldn’t change performance. If it does, keep flying the plane. At partial power and RPM (e.g. 21/21), the blues and blacks will create a secondary and tertiary yaw and roll moment. Advancing the RPMs will create more drag on the windmilling side and more power on the good side. Advancing the torque levers will create more power on the good side. Plan for these and check your flight path references so you can quickly adjust your roll and rudder inputs. This is especially important to keep in mind on OEI approaches.
In cruise, the procedure is a bit different: you typically have excess speed above blue line, which gives you more control authority, as well as a nose-level attitude, so the roll and yaw are not as drastic, and your pitch doesn’t need to go down as aggressively. Depending on how fast you are (e.g. >130 KIAS), it may make sense to swipe the trim up to near the takeoff setting to trade excess speed for altitude as you bring it back to blue line or just above (100ish should give you a good blend of speed buffer and drag). Keep in mind that you’ll need more rudder and aileron relative to the initial input as you decelerate and lose control authority.
After this, all of the actions we take are geared toward maximizing performance, not necessarily controllability. Maximizing thrust and cleaning up the airplane, for example, are good for performance, but bad for Vmc. That’s why “88, keep it straight” comes first and needs constant attention (think “Blue line means life, red line means death”).
Maximize Thrust
Reds: full mixture: march your mixtures all the way forward. This will help you make maximum power or close to it. On a hot/high day you may be too rich and lose a few percent, but that’s usually better than missing a fuel delivery problem. The excess fuel will help cool your good engine and reduce the likelihood of it overheating.
Blues: full prop: maximize the RPMs.
Blacks: fully advance your torque levers. It’s important to do RPM before torque to avoid going past the oversquare limits.
On the upwind, this is mostly an idiot-proofing check and shouldn’t change performance. If it does, keep flying the plane. At partial power and RPM (e.g. 21/21), the blues and blacks will create a secondary and tertiary yaw and roll moment. Advancing the RPMs will create more drag on the windmilling side and more power on the good side. Advancing the torque levers will create more power on the good side. Plan for these and check your flight path references so you can quickly adjust your roll and rudder inputs. This is especially important to keep in mind on OEI approaches.
Minimize drag
Flaps away: make sure flaps are retracted. They’re mostly just a drag device, so get rid of them if you’re already at blue line. The Seminole has the handbrake lever for the flaps, so you can eliminate this drag quickly (~1 second), which is probably why it comes early in the process. For planning purposes, you’re usually better off planning no-flap takeoffs.
Gear in the bay: retract the gear: the gear takes a while to retract, which is probably why it’s the second in the list.
Leg that’s slack, check that black: your slack leg corresponds to your slack engine (also “dead foot, dead engine”). Pull some throttle on that side and check the engine gauges for any changes. If you have partial power, you may want to keep it running until it quits (e.g. to get out of a valley or clear a close-in obstacle). If nothing changes, it’s dead.
Technique tip: I like to use the index and middle finger on my inside hand for this to remove ambiguity. Let’s say I’m seated left side and the left engine is out. I’ll tap my left leg with my right hand’s index finger (the left of the index/middle pairing, then keep wiggling that finger as I go to the throttle quadrant. The wiggling finger in the pair should then correspond to the correct throttle. Confirm on the gauges before pulling it back.
Troubleshooting
Room to play? Determine if you have time to troubleshoot the engine problem. You should already have made this decision ahead of time based on planned workload, so just evaluate your situation against the briefed criteria.
Yea: save the day: try to troubleshoot the engine (GAS-O)
G: gas: how does our fuel delivery look? Confirm selector is on, boost pumps are on, check fuel pressure and gauge volume. (Instructor tip: the selector is a great way to fail the engine in flight and see if your student runs the full troubleshooting procedure. Make sure to guard it with your hand until they get there.)
A: Air: see if there’s an airflow problem. Pull the carb heat to see if it’s iced up, and look for other indications like abnormal MP.
S: Spark: check the mags to make sure they’re both on. If the engine is running rough, cycle them to see if that fixes it.
O: oil: check temp and pressure to look for signs of loss.
Nay: feather right away: proceed with feathering.
Feathering
Leg that’s slack, pull that black: bring back the dead throttle (Tap in a simulated failure).
Pull same blue: pull the dead prop handle all the way into the feather stop (Tap in a simulated failure, instructor should then give you 11 MP to simulate feathering). As the feathering drag goes away, you will get a yaw back toward the good side. Anticipate this change and back out some of your rudder input while looking at either your HSI needle or your outside reference to keep it smooth and track a straight line.
Pull same red: pull the mixture all the way to ICO (Tap in a simulated failure).
Technique tip: I like to use the wiggling finger method for all three of these as a kinesthetic backup of what I’m doing. In a simulated failure, tapping the appropriate lever, instead of pulling it, while announcing each step (e.g. “Simulating left prop to feather detent”), will signal to your instructor that you’re following the correct procedure. Brief your callouts with your instructor or student ahead of time to avoid turning a simulated emergency into a real one.
Split the ball, and raise the dead: you should have about half a ball (or the brick on the G1000) of slip indication towards your good engine, and 2-3 degrees of dead side up. It’s good to start dialing this in at the start, but worth verifying once you’re done.
Securing the engine: CAMPSEX
C: cowl flaps: open on the good engine for better cooling, closed on the dead one for reduced drag. On a warm day, you’ll want to get to this quickly before the good engine overheats, especially in a climb.
A: alternator: switch off on the dead engine.
M: mags off dead engine
P: boost pump off dead engine
S: fuel selector off dead engine
E: electrical: load shed if required
X: X-feed fuel if required. Keep in mind that you should land drawing fuel from the onside tank, so if you’ll need the fuel in the other side, don’t wait until your good side is almost empty.
Technique tip: the “finger wiggle” works well for the securing process as well.
If you want it all in one consumable format, here’s the short version:
Flaps away: make sure flaps are retracted. They’re mostly just a drag device, so get rid of them if you’re already at blue line. The Seminole has the handbrake lever for the flaps, so you can eliminate this drag quickly (~1 second), which is probably why it comes early in the process. For planning purposes, you’re usually better off planning no-flap takeoffs.
Gear in the bay: retract the gear: the gear takes a while to retract, which is probably why it’s the second in the list.
Leg that’s slack, check that black: your slack leg corresponds to your slack engine (also “dead foot, dead engine”). Pull some throttle on that side and check the engine gauges for any changes. If you have partial power, you may want to keep it running until it quits (e.g. to get out of a valley or clear a close-in obstacle). If nothing changes, it’s dead.
Technique tip: I like to use the index and middle finger on my inside hand for this to remove ambiguity. Let’s say I’m seated left side and the left engine is out. I’ll tap my left leg with my right hand’s index finger (the left of the index/middle pairing, then keep wiggling that finger as I go to the throttle quadrant. The wiggling finger in the pair should then correspond to the correct throttle. Confirm on the gauges before pulling it back.
Troubleshooting
Room to play? Determine if you have time to troubleshoot the engine problem. You should already have made this decision ahead of time based on planned workload, so just evaluate your situation against the briefed criteria.
Yea: save the day: try to troubleshoot the engine (GAS-O)
G: gas: how does our fuel delivery look? Confirm selector is on, boost pumps are on, check fuel pressure and gauge volume. (Instructor tip: the selector is a great way to fail the engine in flight and see if your student runs the full troubleshooting procedure. Make sure to guard it with your hand until they get there.)
A: Air: see if there’s an airflow problem. Pull the carb heat to see if it’s iced up, and look for other indications like abnormal MP.
S: Spark: check the mags to make sure they’re both on. If the engine is running rough, cycle them to see if that fixes it.
O: oil: check temp and pressure to look for signs of loss.
Nay: feather right away: proceed with feathering.
Feathering
Leg that’s slack, pull that black: bring back the dead throttle (Tap in a simulated failure).
Pull same blue: pull the dead prop handle all the way into the feather stop (Tap in a simulated failure, instructor should then give you 11 MP to simulate feathering). As the feathering drag goes away, you will get a yaw back toward the good side. Anticipate this change and back out some of your rudder input while looking at either your HSI needle or your outside reference to keep it smooth and track a straight line.
Pull same red: pull the mixture all the way to ICO (Tap in a simulated failure).
Technique tip: I like to use the wiggling finger method for all three of these as a kinesthetic backup of what I’m doing. In a simulated failure, tapping the appropriate lever, instead of pulling it, while announcing each step (e.g. “Simulating left prop to feather detent”), will signal to your instructor that you’re following the correct procedure. Brief your callouts with your instructor or student ahead of time to avoid turning a simulated emergency into a real one.
Split the ball, and raise the dead: you should have about half a ball (or the brick on the G1000) of slip indication towards your good engine, and 2-3 degrees of dead side up. It’s good to start dialing this in at the start, but worth verifying once you’re done.
Securing the engine: CAMPSEX
C: cowl flaps: open on the good engine for better cooling, closed on the dead one for reduced drag. On a warm day, you’ll want to get to this quickly before the good engine overheats, especially in a climb.
A: alternator: switch off on the dead engine.
M: mags off dead engine
P: boost pump off dead engine
S: fuel selector off dead engine
E: electrical: load shed if required
X: X-feed fuel if required. Keep in mind that you should land drawing fuel from the onside tank, so if you’ll need the fuel in the other side, don’t wait until your good side is almost empty.
Technique tip: the “finger wiggle” works well for the securing process as well.
If you want it all in one consumable format, here’s the short version:
Advice for instructors
One of my colleagues drove the charge for an improved system of simulated versus actual failure callouts to better distinguish between real and simulated engine problems and adapt callouts to improve instructor/evaluator/student understanding. In the sim, we would practice both a “real” OEI situation (sim instructor tells the sim to have an engine quit) and the student pulls all the levers as in a real emergency. We also practiced the simulated failures where an instructor would pull a torque lever (or selector/mix at altitude to test troubleshoot process) and the student would go through their process until they got to whichever switch or lever you pulled, at which point the instructor would say “My left throttle, simulated flow” and the student would tap the remaining switches in the process while calling out “Simulated feather left engine,” “Simulated mix left engine” and so forth. This made our students proficient with both sets of actions in the sim before they got to the airplane and minimized the risk of applying them incorrectly in an environment where it could be dangerous. Wiggling the onside finger and calling out the correct lever was a further technique to ensure the correct one was selected, thus engaging visual, auditory, and kinesthetic learning centers in the brain.
I always reserved the callout “Not a drill” for a real engine problem. Fortunately, I only had to say that once for a minor power loss that we quickly attributed to carb ice and resolved.
I always reserved the callout “Not a drill” for a real engine problem. Fortunately, I only had to say that once for a minor power loss that we quickly attributed to carb ice and resolved.