can you apply what you've learned? I didn't.

[Keith Smith]

Had an eye-opening moment during a sim session last week when I was visiting FlyThisSim in San Luis Obispo, CA.

I'm not a rated twin pilot (I hold an ASEL certificate with instrument, that's it), but I have tinkered with twin operations for a few years, done some reading, etc. I've encountered maybe 100 simulated engine-out operations in my simulated twin adventures, almost entirely self inflicted. Those that were not self inflicted were normally at tradeshows with Precision Flight Controls up at pattern altitude with someone working the instructor console, making my life difficult without warning. In either case, though, I've either known which engine just failed (because I failed it), or I've had ample time to work out which engine had failed (the critical "IDENTIFY" step of the engine out procedure in a twin).

I was exposed to an interesting situation during the filming of a scenario where an engine was failed during an unexpected go-around. This happened at very low altitude, so there wasn't much time to deal with it. Here's where I discovered a serious deficiency in my skill set for flying the twin. I knew how to identify an engine, but not to a level where I could do it within 2 seconds. Given more time, I absolutely would've identified the correct engine, but in a pinch, I picked the wrong one, going by instinct based on rudder input. Sadly, it was the wrong instinct. All this time, I've been breezing through the 'identify' step because I've either known what just failed, or I've had tons of time to work it out.

I had no idea that I was actually horrible at identifying the dead engine. It's easy to brush this off and say, "who cares? You're don't hold a twin rating, if you did, you'd have the training and everything would be fine."

Here's what bothers me about this...I have a feeling I could get through a twin engine program and STILL come out of it with the same deficiency UNLESS they put me in a situation where there was a requirement to identify the dead engine with lightning speed and take appropriate action. Imagine the training is taking place entirely in an airplane with no sim available during the process. Are they actually going to fail the engine at 200-300ft on takeoff with no warning? I'd be surprised. Not sure there are many instructors that would want to be put in that position with a random client in the left seat working the controls.

Now, I could be way off base, and perhaps all twin programs do a perfect job of teaching you to identify the engine with lightning speed. That would be great, and I'd be relieved to hear it. It was still an eye opener for me, though. It's one thing to kinda-sorta know how to handle something when it comes up....it's another to have complete mastery of a subject area where you'll know exactly what to do with minimal time spent before action is taken.

The trick, then, and my take away from this, is to identify what sorts of scenarios and failures would require this level of mastery...and then acquire that level of knowledge. Examples...engine out for a single...requires immediate action if it happens at low level, I'm good there. Alternator failure? Who cares....you've got time. Engine fire, GPS outage, com failure, Vacuum failure, glideslope failure...the list goes on. I'm going to rexamine what could go wrong and make sure that I identify cases where not knowing the right thing to do within the first few seconds could kill you, versus those where you have time to fix it before it can kill you.

It's been a while since I learned something significant with fixed wing flying...this was a good one. Chalk up another win for sims.

[Keith Smith]

Another thought about the training process is that you're often put in a mindset where you're going to practice failures. You're primed for it, you've briefed the procedure, you're good to go. I imagine this is what it's like for recurrent training in a Level D FFS. The V1 cut comes along, and *gasp*, you handle it like a champ.

If only the real world gave you such a great heads up. In this case, I was caught completely off guard and then had to rely on a skill set that, as it turns out, was not very well developed. I could've practiced engine outs by myself another 100 times in the sim and still had this same problem. The trick, I guess, is to be go through scenarios which have the ability to expose this weakness. That either means you need a competent sim instructor, or some decent software to manage the scenario and truly surprise you.

[Ryan B]

Great story. Sims are important... I'm not sure how grand it would be to practice that in real life hehe.

On a side note I landed without a clearance today in SNA!

Filed a NASA form argh! Stupid military pilots! I actually feel bad and it was just on the sim!!

[Peter Grey]

As a rated MEI I feel like I should chime in on this a bit.

My read of the situation actually doesn't show a deficiency in identify (assuming you know how rudder input = failed engine). What it showed was a failure in the verify step.

After identifying the dead engine you should verify it by taking the dead engine throttle to idle. If correct nothing happens, if incorrect your yaw changes and you know 1 of 2 things. You've identified the wrong engine (yaw goes away), or it's only a partial failure (yaw gets worse). If you are actually feathering the wrong engine that means either a. you had a momentary "which engine was it" moment (not that case here). Or b. you didn't verify (which I think happened here).

So let's talk altitude. I used to fly a PA-44 from a field at 5000' with a DA upwards of 9000'. In our airplane feathering the engine was worth around 300 ft/min. Looking at some performance charts at 5500' during a summerish day (30 C) at MGTW we were at -100 ft/min.

It's hard to find a twin that performs worse then a PA-44. So at 200' you have 2 minutes to get it figured out. Of course this assumes a bunch of stuff. But even so 30 sec to 1 minute is reasonable. CPDIVE (the abbreviation we used for stuff to do prior to feathering an engine [control, power, drag, identify, verify, evaluate {which may result in a feather}]) takes 15-20 seconds for the average person. So don't rush it, if you really have poor enough performance and are low enough for time to be an issue, then plan on an off field landing. These decision points are day to day specific and need to be determined ahead of time.

Are they actually going to fail the engine at 200-300ft on takeoff with no warning?

Our floor was 500' AGL and yes no warning. The FAA PTS limit is 400 ft.

Not sure there are many instructors that would want to be put in that position with a random client in the left seat working the controls.

You underestimate our bag of tricks, and in the end if it's all going wrong you just "unfail" the engine and problem solved. I've had interesting experiences on this one. Imagine wrong rudder and trying to use aileron to fix it...

[Keith Smith]

Thx Peter. There is a full video of the event that's being edited, so I'll have to look closely to see what the altitude was when it failed on the go-around and how much time actually passed.

The plane was bleeding energy badly when the failure happened and the props were full forward, so I felt extremely rushed to identify/verify. That's where my process failed. At higher altitudes, I do use the throttle techniques you described to verify which one has failed. For some reason (either reality, or just perception), it didn't seem as though I had time to do that and I had to make a call. I hurried and made the wrong call. I quickly realized I'd nuked the good engine and then landed just off the runway, eventually rolling back onto the center line, stopping by the end of the runway.

It'll be very interesting to watch the video to see what the actual altitude and altitude/energy loss actually was with everything full forward.

I'll try to arrange for this to happen randomly in some of my sim flights and will time my performance against the numbers you described.

[wmburns]

I mainly fly the X-Plane default Barron BE58. In my X-plane set up the MTBF has been set slightly lower than default. Sometimes I can go weeks without any failure. Then a couple of weeks ago two engine failures on two separate flights. The failures did not occur at a critical flight phase so I had time.

I found significant rudder was needed to control yaw as expected. I was also surprised how fatiguing it was to hold the constant rudder.

This caused me to do some offline flying. At a level cruise shut off one engine. I attempted to trim out the dead engine with rudder trim. What I found was that the default Barron BE58 didn't seem to have enough rudder trim to counteract the dead motor (shut off and feathered).

First question. In the real world, is the rudder trim used to counteract the dead engine?

I played around for perhaps an hour flying on one engine trying to get a feel for the airplane. What happened next surprised me. When I attempted to restart the shut off engine I could not. I tried everything I could think of including a windmill start. No joy.

Question. Did I allow the engine to cool so much at altitude that a restart was no longer possible? I did check to see there were not any failures active at the time. If this is true, it's amazing that level of detail is simulated.

Thanks for sharing this story Keith.

[Keith Smith]

Doubtful that there would be enough authority with rudder trim for hands off. Peter will have more information on that, though.

The stock Baron is also fairly under powered in X-Plane. Single engine flight is a real challenge with that airplane.

Not sure why the engine isn't restarting, will tinker with that and get back to you. It's been a while since I've played with the stock Baron.

[Peter Grey]

In terms of using the rudder trim for OEI (one engine inop) the answer is "it depends".

In the PA44 we generally never used rudder trim as the control forces weren't that bad. In the DA42 it's basically mandatory unless you want to have your leg go numb. In the DA42 once trimmed you could fly "feet off" relative to being OEI.

A warning here. If you use rudder trim make sure you are aware that when you pull power to land you will have a LARGE yaw force. Be ready for this (or trim out to neutral on final).

In terms of there being enough rudder trim authority the answer is "it depends".

Here's what the book has to say about it (FAR part 23.143):

Values in pounds force applied to the relevant control
Yaw
For temporary application: 150
For prolonged application: 20

and 23.161

General. Each airplane must meet the trim requirements of this section after being trimmed and without further pressure upon, or movement of, the primary controls or their corresponding trim controls by the pilot or the automatic pilot. In addition, it must be possible, in other conditions of loading, configuration, speed and power to ensure that the pilot will not be unduly fatigued or distracted by the need to apply residual control forces exceeding those for prolonged application of §23.143(c). This applies in normal operation of the airplane and, if applicable, to those conditions associated with the failure of one engine for which performance characteristics are established.

So an aircraft does not have to be trim-able to a feet off position, just have to be able to get required force under 20 lbs.

In terms of being able to restart the answer is.....wait for it.....it depends:

The PA44 could fly around all day OEI and restart in an instant.

The DA42 could not. It had a limitation of 2 minutes OEI below 10000' and restart "immediately" above 10000'. Reason for this was rather complex and beyond the purpose of this thread. Note the DA42 version I flew was a diesel so YMMV in other DA42 variants.

To help with this one I looked up a BE58 POH and here is what it said on the topic:

Nothing, there is vague statement about using power to "avoid the difficulties of restarting an engine and preserves the availability
of power to counter potential hazards. "

with regards to simulating OEI by actual shutdown, but that's it.

[ChrisS]

There are so many good points in here, I don't even know where to start. Okay, actually I do.

1) Using a sim as a training tool is effective provided it can accurately simulate the failure / result. You have to condition yourself to those events and reactions in the sim through practice, and then it will translate to the real world to an extent proportional to the realism of the sim. (Thanks, super expensive psychology degree!)

2) Here's where the miracle of shared cockpit comes in. So long as you're training yourself to perform the correct action under the correct circumstances, you can have someone fly with you and 'fail' components as you would in real life (or through a controller on PE if you have that pre-arranged). That way, you're not really expecting it.

3) In a CRM environment, the fuel control levels are an ICM switch (Indicate, Confirm, Move), so if you have to shut down an engine, you would say 'Number one engine failure' and the FO would have to confirm the same with you before you can shut it down permanently.