Engine Failure On Takeoff - 2 Cfi's At The Controls

[acaTerry]

Let's talk about retracting the landing gear after a normal takeoff.

I like your points, but there is one you miss, and it drives me crazy. I see it at DVT all the time by the academy aircraft and its only a matter of time...
They use the tower as a point to retract the gear (its 1/2 way down the runway). This is poor instruction. These instructors need to explain WHY they use the tower and then go on to explain that ALTITUDE GAINED AND SPEED are deciding factors as for when to retract gear for the decision to land or go.
I cannot believe how many times I see the Seminoles, after a go around, at 300 feet with gear DOWN until they pass the tower. Or follow through on a touch and go and bring the gear up at the tower when they are barely off the ground. The plane settles and nearly strikes the runway. I just love that diluted teaching!

 

[skiandsurf]

Its a good thing I am not into all of these books and number games. I cant imagine having a problem and trying to recite what chapter and paragraph I should paraphrase. Just do what it takes to stay off of CNN. If you are "flying" a twin, fly the airplane. I wouldnt even consider putting it back down, unless I had a loooooooong runway or a control problem. If the failure happens prior to takeoff, on a light twin and runway length is good, keep it on the ground. But to fly up to 75 feet, then try to descend, you are flying over a lot of unusable runway.

 

[UndauntedFlyer]

Verify with Throttle or Prop?

Let’s talk about technique for verifying the failed engine prior to shutdown and feather.

I use the throttle for verification in cruise and use the prop for verification if the engine failure was just after takeoff.

Let’s talk about the throttle first. In cruise, just use dead foot – dead engine, bring the suspect throttle all the way back to idle and then half way forward again for the trouble shooting process. Sometimes I see pilots on their tests trying to trouble shoot with the throttle all the way back, then announcing that the engine would not restart and feathering. This is totally unsatisfactory performance because when the throttle is all the way back the pilot is unable to see if the engine has actually re-started as a result of their own trouble shooting efforts.

Now let’s talk about using the prop only for verification. The prop is the best verification method for a time critical situation such as immediately after takeoff. Just use dead foot – dead engine, pull the suspect prop back about 1-inch and if there is no tone change it’s the right one, it’s verified so then just pull it all the way back to feather. If the incorrect prop is mistakenly pulled back just a little, that mistake will immediately be noticeable by the audio tone change. Of course, the best part of using the prop for verification is that very little power will be lost if the pilot makes a mistake and pulls the wrong prop for a moment. Then when the pilot realizes the mistake, he just moves his hand over to the correct prop and feathers. Also, it is important to consider that had the pilot pulled the wrong throttle back at 50-feet the loss of power would most certainly cause a crash of the airplane. So in this instance, the prop method of verification has advantages.

Another advantage of using the prop for verification in this time critical situation is that it’s so easy and simple to do while in the high demand situation of trying to to maintain control of the airplane following a total power loss just after takeoff. Just try using the throttle once and then having to find the correct prop. That is almost too much for this type of critical flight situation.

Questions or comments are welcome……

 

[acaTerry]

Sometimes I see pilots on their tests trying to trouble shoot with the throttle all the way back, then announcing that the engine would not restart and feathering. This is totally unsatisfactory performance because when the throttle is all the way back the pilot is unable to see if the engine has actually re-started as a result of their own trouble shooting efforts.

Well, not necessarily so.
If the restart technique / checklist is followed, it calls for 1/4 to 1/2 inch throttle open, which definitely would show signs of a successful restart. Don't forget that that there are signs of success besides a "louder" engine. Our mechanics have seen school planes come back with baffle damage due to restarts at high throttle settings (older PA series without the accumilator, which required starter assisted starts).
These same mechanics would also take issue with using the prop as a reference, especially in a turbocharged piston twin where 40" of MAP is there already at full RPM.
While I agree in theory with your use of the prop control (which IS a dead giveaway of the engine), it is no more useful than a carefully retarded throttle. If done correctly, no significant risk is taken by a smooth, careful pulling back of throttle.

 

[UndauntedFlyer]

While I agree in theory with your use of the prop control (which IS a dead giveaway of the engine), it is no more useful than a carefully retarded throttle. If done correctly, no significant risk is taken by a smooth, careful pulling back of throttle.

When an engine fails at 100 feet or so, especially if flying on instruments at night that is a most demanding emergency in a time critical situation. The most important priority in that case is to maintain control of the airplane and second priority is to reduce drag by feathering the failed engine assuming the gear is already up. While focusing 90% of your scan on the flight instruments only 10% can be used to find the engine & prop controls by sight and feel. It is no harder to find the throttles than the prop controls and then no harder to find the correct throttle or the correct prop control, but why do it twice? Plus by using the throttles you always have that possibility of verifying and pulling the wrong control back causing a certain crash as a result of power loss in the mistaken verification. But with the prop control none of those problems exist plus the whole process takes only seconds to complete: Verify by pulling prop control back one inch and then feather by pulling back to the feather position.

 

[acaTerry]

I could not agree more. But the point you make is for an ACTUAL emergency, and the point I was making was for training purposes. Unfortunately, we can not always train for real emergencies because examiners / PTS too often apply standards & procedures which would not always be practical in an emergency, but work well for the training environment.
Also, you make a good point that too often gets looked over, and that's aircraft control...too many times pilots go to their death performing actions of procedure while letting the airplane run away from them.

 

[UndauntedFlyer]

Also, you make a good point that too often gets looked over, and that's aircraft control...too many times pilots go to their death performing actions of procedure while letting the airplane run away from them.

This is very true and the point of my "Deer in the Headlights" post.

“Deer in the headlights” syndrome, (freezing-up) that’s the problem.

I see this many times when I am testing a multiengine applicant who is only prepared for what he or she thinks is the usual simulated engine failure after takeoff. What I mean is that some applicants are only prepared for the instructor/examiner to simulate an engine failure by bringing a throttle back at about 600 - 800 feet AGL after takeoff when the landing gear would already have been retracted and the power reduced to climb power. After the examiner brings back a throttle, the applicant mumbles a few things such as, “Mixture, props, throttles, gear-up, flaps-up, verify and feather.” Then the examiner moves the throttle and prop and announces that “zero thrust” is set. In my opinion this whole scenario has very little realism to an actual engine failure just after takeoff at 50, 75 or 100 feet with no runway ahead and no where to land.

Instead of testing with the above scenario I prefer a more realistic simulation of a real engine failure after takeoff. As I have mentioned earlier in this thread, I prefer to climb to 3000 feet AGL for the safety of a simulated hard-deck takeoff with the gear down, full power and airspeed between Vxse and Vyse or, Vsse if such as speed is published. With a mixture cut at 50 feet above the simulated hard deck, all that’s required is to apply rudder to maintain directional control, level the wings while simultaneously pitching forward to hold airspeed, then retracting the landing gear and feathering (for real) the propeller. This simulation is realistic. Every instructor that knows me and how I test this task prepares their students for this simulation. Almost all applicants pass this task by clearly knowing the correct action following an engine failure after takeoff. They have a clear plan of action for this emergency. And as for performance, all modern training twins such as the Seminole and the Duchess climb out with no difficulty.

On the other hand, if I happen to have an applicant for testing that is recommend by a CFI outside my area who might be unfamiliar with the 3000 foot AGL hard deck method, those applicants perform similar to a “Deer in the Headlights” following the engine failure after takeoff. They are startled and do nothing for a short while and then they start mumbling something about mixtures, props, throttles, checklists, carb heat, boost pumps, flaps and so on (even though all of these were set for takeoff and don’t need change). This brings an element of confusion and second guessing until so much time is lost along with airspeed. Of course, the next thing we know the airspeed has deteriorated to very near Vmc causing examiner intervention for safety of flight. That check ride is over.

So you see, what is necessary to prevent the “Deer in the Headlights” roll over accident is a clear plan of action for an engine failure right after takeoff. If the runway is less than about 5000 feet it would be very difficult to abort a takeoff once airborne and above 50 feet. But if you know the airplane is light enough and the performance will be satisfactory, then be prepared to do exactly what is necessary and don’t perform actions that are unnecessary for that takeoff that day. For example don’t talk about or try to retract the flaps when they aren’t being used for takeoff. Have a clear plan of action and review that plan in your mind before takeoff:

• Maintain directional control while pitching forward to maintain airspeed
• Retract landing gear
• Verify and feather prop
• Maintain at least Vxse, or Vyse if that speed has been obtained

Having a clear plan of action will prevent the “Deer in the Headlights” syndrome which is the cause of Vmc roll-over (upset) accidents.

Comments/questions?

 

[nosehair]

Glad to see this is being discussed. I suspect most multi pilots are not aware of the technique/practice/skill needed to execute what they brief about "landing on the remaining runway".

Why? BECAUSE WE DON'T PRACTICE IT!!!

See how many respondents have declared how stupid and unsafe it is to practice this maneuver?

I don't even see enough practice engine failures on take-off in a single engine. Which is probably one of the underlying reasons for this lack of airmanship.

Anyway, I have seen the system go like this: "Hey! We're getting too many accidents in this area, let's quit training there." That's what they said about spins.

I don't know about you, but to me, when the accident rate goes up in a pilot skill area, THERE SHOULD BE MORE TRAINING.

It makes my ulcers fire up when I here/see our people in authority (FAA, Instructors, Examiners) say we shouldn't do engine cuts on take-off right after rotation because they are dangerous. THAT'S WHY WE NEED TO PRACTICE THEM!!!

I cut my baby teeth on multi-engine instructing by instructing initial multi-engine in a Baron for the U.S. Army during VietNam.

We did engine cuts right before, during, and immediately after lift-off, 50-75 feet after lift-off, everywhere during that critical phase of flight that you could do to insure the student could handle a cut anywhere, anywhere during take-off. It was just the normal thing to do. The student was definitely going to get an engine cut somewhere during that critical phase on his stage check, and the test was to see that he could manage the airplane at all times in a variety of possible events. That he could, in fact, either land straight-ahead or continue and clean-up.

We never had accidents as a result of any of this. We routinely did engine cuts on the single engine training on Vx climbs. The students knew how you have to push the nose over rapidly and then almost as rapidly pull back into the flare. It just takes actual hands-on training with competently trained instructors.

There just are no trained instructors for this.

I was appalled when the Army gig was up and I discoverd civilian pilots don't do this. They are brain-washed to believe it is unsafe. It's only insafe because we haven't practiced it.

Which is why so many light twins roll over in an engine out on T.O.
They haven't practiced it. That's the cause of the deer-in-headlights syndrome.

 

[jws717]

Ok then, the question is: Can this takeoff be made safely following an engine failure at 75 feet?

Comments please.....

according to gidance givin in the new seminole manual, continued T/O below 75 kias is impossible.

The new manual also gives the performace hits for things like gear retaction, flaps out and prop windmilling, Doing some simple math for a seminole in average conditions.

(s/e roc) - (performace hits)= negative ROC untill the gear is up and the prop feathered. at the negative rate of climb aprox 300fpms you will be underground before the gear has time to retact!


BOTTOM LINE, GEAR DOWN LAND! GEAR UP GO!
This is why retacting the gear at the appropriate time durring T/O is critical. When you dont want to land anymore pull it up otherwise leave it down.

by the way to see how much runway it would use to reject at 50ft add the obastacle t/o to the obsatcle landing, 5000ft is plenty.

last note; most accidents in a twin are stalls/vmc associated with an engine fail or just pilot error in a good airplane. Would you rather VMC and die or overrun the runway and walk away from a totaled and insured airplane that isnt yours anyhow...

 

[UndauntedFlyer]

according to gidance givin in the new seminole manual, continued T/O below 75 kias is impossible.

The new manual also gives the performace hits for things like gear retaction, flaps out and prop windmilling, Doing some simple math for a seminole in average conditions.

(s/e roc) - (performace hits)= negative ROC untill the gear is up and the prop feathered. at the negative rate of climb aprox 300fpms you will be underground before the gear has time to retact!


BOTTOM LINE, GEAR DOWN LAND! GEAR UP GO!
This is why retacting the gear at the appropriate time durring T/O is critical. When you dont want to land anymore pull it up otherwise leave it down.

by the way to see how much runway it would use to reject at 50ft add the obastacle t/o to the obsatcle landing, 5000ft is plenty.

last note; most accidents in a twin are stalls/vmc associated with an engine fail or just pilot error in a good airplane. Would you rather VMC and die or overrun the runway and walk away from a totaled and insured airplane that isnt yours anyhow...

Thank you for your comments but as I read your post I would ask the following:

1. Did you read the entire thread on the points you question? I ask this because most all your points are addressed and answered in the thread.

2. Have you ever tried the 3000 foot hard deck simulation of the engine failure on takeoff at 75 feet in a Seminole with 2 people and half tanks of fuel?

I have done this test many times and I find that the gear really makes very little difference if it is retracted right away. And if the prop is feathered without delay the climb out continues without any altitude loss. That has been my experience.

I am only presenting what the results of the 3000 foot hard deck test have been for me in the Seminoles I have flown. If anyone should happen to go out and do this test and find the results to be the same or different please post them here.

So bottom line, the decision to "Clean up and go," or to "Chop and drop" should be made on expected performance......if it is known to you.

Your comments or questions are welcome....