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

[acaTerry]

The first thing as always is to maintain control of the aircraft and then secondly to slowly retard throttle to come down and land it.

Exactly! The plane is losing energy, and speed is lost as the aircraft pitifully attempts to continue where it was going. Think of this as an excercise in energy management.
A sudden chopping of the power (as recommended by the manual saying "immediately close both throttles") will only make bad turn to really bad.
Control the plane with available power and then reduce it appropriately to land.

 

[FL420]

Many thanks to UndauntedFlyer for starting this thread and to all who contributed. I am just starting to get back to flying piston twins after being away from them for about 15 years. Lots of really good info and insight here. I picked up some very good tips.

Thx again.

 

[UndauntedFlyer]

High Performance Twins & Blueline Considerations

Turbocharged High Performance Twins and Blueline Airspeed Considerations

One point in this thread is that a checkout in twin engine airplanes should include an engine failure after takeoff from a simulated airport hard-deck of 3000 feet AGL. This can be done safely, easily and without harming the engines in most any non-turbocharged light twin. However, if the airplane you’ll be flying is turbocharged or for any reason you are reluctant to abruptly fail and engine in this simulation, a performance test can still be accomplished by very gradually reducing power and then feathering the engine. After the engine is feathered the single engine climb performance can be validated. If it is sufficiently positive then it will only be that much better 3000 feet lower after a takeoff from an airport at that lower elevation.

Now let’s talk BLUELINE.

After the engine is feathered, the pilot should experiment with various airspeeds above and below “Blueline” (Vyse) as marked on the airspeed indicator.

When the engine is feathered on the twin as in the above recommendation, turbo-charged or not, the pilot should experiment with the performance capabilities of the airplane. As I have mentioned earlier in this thread, don’t use simulated “0-thrust” as that will give false performance results either positive or negative. With the engine feathered (real 0-thrust) it is now most important to determine the climb performance capability of the airplane at various airspeeds above and below blueline, such as blueline plus and minus 5 to 10 knots. Surprisingly, you will see that in effect, best rate of climb in feet per minute is practically the same for plus or minus 5 knots of blue line. Therefore, in a practical sense, the blueline should really be marked as 10 knots thick instead of being shown as only one exact airspeed. You will also see that when the airplane is at a light weight, such as with one person and half tanks of fuel, Vyse is, in effect, reduced by as much as 5 knots.

So as an example, if blueline (Vyse) is 100 knots at maximum gross weight in a typical high performance twin, best rate of climb in feet per minute will be immeasurably demonstrated to be the same at 95 to 105 knots. So in this case, if the airplane is very light, the best single engine rate of climb airspeed will actually be about 95 knots, or 90 to 100 knots considering the reduction in Vyse for the reduced gross weight.

What does this all mean? Simply this, if you were to lose an engine just after takeoff with no runway remaining and no option for a safe landing, the gear was retracted and you checked your airspeed as 5 knots below blueline, use what has been learned from the 3000 foot AGL test, hold that airspeed (Vyse-5k) because there is no performance value in pitching forward to obtain blueline, plus such a pitch down is certain to cost altitude when none can be given up. Furthermore, if the airplane is lightly loaded, and your tests at 3000 AGL show a best performance rate of climb airspeed to be blueline minus 5 knots, then as much as 10 knots below the marked blueline may be used to prevent a loss of altitude. Pitching forward to obtain the marked blueline which would be an increase of 10 knots in this case, will probably cost as much as 50 - 100 feet of precious altitude. So it would be best here to hold the current airspeed of blueline minus 10 knots (Vyse-5k) and climb to a safe altitude, then gradually increase airspeed for further climb. Of course, in any twin engine airplane Vmc is a critical airspeed that must be constantly considered for safety too, and in the above example of blueline minus 10 knots this may be dangerously close to Vmc in some airplanes. Also, when flying with and engine shut down/feathered it is always a good practice to be within gliding distance of an emergency use airport should the remaining engine fail for whatever reason.

And in conclusion on the topic of blueline, let’s talk about blue line on final approach.

In a practical sense, blueline airspeed is used on final approach in most every twin engine airplane for both single engine approaches and normal two engine approaches. Why is this? The answer here is that there is no good reason except that it pre-bugged, blue is a pretty color, and that speed is not too fast or too slow, it’s just about right.

Blue line is usually held on final approach until 300 to 500 feet AGL whereupon full flaps for landing are normally extended and airspeed reduced to the normal over the threshold Vref speed of 1.3 x Vso.

So should blueline be used on final approach in a typical twin? The answer is yes, but not for any real aerodynamic reason. It’s just a pretty color.

Your questions or comments on this topic are welcome……

 

[aerobaticspilot]

piston twins or single or anything piston just flat out scares me. 4 engine failures in piston aircraft i had enough. I will add this though at point where the engine fails you have about one second to decide to continue or land. personally i would continue the climb unless i know i was not going to clear an obsticale and if thats the case i would not be in that plane becuase that is flat out stupid. Another point is when you do a take off brief your briefing a lot of information and engine failure is one. You got to do what your brief states. please do not do the opposite becuase your not ready for it and the other pilot is not ready and you will kill yourself.

 

[Espalda Mojado]

ENGINE FAILURE ON TAKEOFF – 2 CFI’S AT THE CONTROLS


OK, you’re rolling down the runway on a 5,000 foot runway in a Piper Seminole. There are two people on board and half-tanks of fuel. You rotate and an engine fails at 75 feet AGL. Your airspeed is about 83 knots (Vxse), or 5 knots below Vyse. The landing gear is still down because there is still runway/overrun left or because you were distracted by engine roughness.

What should you do in this situation? What does the FAA say you should do?

So what happed to these two instructors?

Well in short they crashed on to the runway, destroying the airplane. Fortunately, no one was hurt.

This accident was caused largely by the fact that no technique is being taught in any publication anywhere as to how to safely handle this emergency. In the FAA’s latest (2004) publication on this subject (FAA-H-8083-3A) the Airplane Flying Handbook, it simply says, “If the engine failure occurs prior to selecting the landing gear to the UP position, close both throttles and land on the remaining runway or overrun.” The problem here is that that is exactly what these two pilots did and that action destroyed the airplane. What’s wrong here? The problem is that the FAA’s publication and the PA-44 manual only say to land, but do not address technique. Unfortunately, poor technique oftentimes spells disaster.

COMMENTS/QUESTIONS are welcome.....

Yet another example of pilot error. These two should have known better.

 

[ReportCanoa]

This is all wrong.

Keeping the gear out in a piston twin WILL kill you.

I would bet you anything the Feds would have said nothing had these guys landed straight ahead with the gear up at a controlled rate of descent.

They crashed the plane BECAUSE the gear was hanging out. It is a TREMENDOUS amount of drag and compound that with an engine failure?
No wonder they pancaked it.

 

[xshuttlefa]

In a "learning experience" I had with a Beech Sierra one time, it is rather difficult to transition to landing again once airborne, even at low-altitude with runway remaining.

I had just rotated when the right door opened. This airplane did this all the time and since I had just rotated, I figured ( wrongly) that I'd just close the throttle and land on the remaining runway. Why waste time going around the pattern when I can just land again, right?

WRONG, VERY WRONG. I got the plane back down, not without porpoising it twices and having to really stand on the brakes to get it to stop. I was still at a fairly good pace as I turned off the runway at the end. Luckily I did not damage the plane, myself or anyone else. To save .2 Hobbs I could have ended up costing myself a lot more. I almost turned an inconvenience into a major emergency.

Just wanted to make the point that it is rather difficult to bring the plane back down once airborne. As for the situation in question, I guess factors like terrain, weather, length of runway remaining all play a role. As was mentioned, many light twins are negative performers when one engine quits.

Enough runway with a flat overrun, might as well put it back down. But with little runway and a bunch of trees, fences, roads, the plot thickens. As was also mentioned, you owe nothing to the plane if it fails and if it will take a controlled crash into a bunch of trees to save your life, so be it. But what else lies in those trees? What if it's dark or IFR or both?

I hope to begin my Multi soon and with that, an MEI. I'll find out then what I would really do, hopefully hypothetically.

 

[CMI]

NTSB Report
NTSB Identification: CHI06CA054

http://www.ntsb.gov/ntsb/brief.asp?ev_id=20060113X00074&key=1

 

[Luxor Two]

This is all wrong.

Keeping the gear out in a piston twin WILL kill you.

I would bet you anything the Feds would have said nothing had these guys landed straight ahead with the gear up at a controlled rate of descent.

They crashed the plane BECAUSE the gear was hanging out. It is a TREMENDOUS amount of drag and compound that with an engine failure?
No wonder they pancaked it.

Thanks for the thorough analysis, I can only guess your vast experience of raising and lowering landing gear in your sweet CRJ has given you the valuable insight to come to flightinfo and proclaim this to the rest of us that know no better.

Do you think they pancaked it in because the gear was out, or do you think it is possible the pitch attitude was not sufficient to provide a safe margin from stall speed?


On a side note, hopefully "UndauntedFlyer" will share more about what happened after the aircraft first stalled and how the instructor managed to fly a crippled airplane back around the pattern I know (I'm sure he has more knowledge of all the details than I've heard).

Besides the decision to perform the exercise in an actual aircraft off of a runway, which I disagree with....hard deck at altitude and sims {for procedure} are the place for this exercise...the MEI actually did a good job flying a twin with broken landing gear (broken and dangling), curled prop tips from rwy contact (yeah they hit pretty hard) and a wrinkled fuselage while getting no assistance from the "frozen in his seat" CFI.

This is an excellent thread and something that absolutely must be discussed and demonstrated many times not only in ME training but in rental checkouts, currency flights, and checkrides.

 

[UndauntedFlyer]

NTSB Narrative

NTSB Identification: CHI06CA054.
The docket is stored in the Docket Management System (DMS). Please contact Records Management Division
14 CFR Part 91: General Aviation
Accident occurred Thursday, December 08, 2005 in Rockford, IL
Probable Cause Approval Date: 3/28/2006
Aircraft: Piper PA-44-180, registration: N2196B
Injuries: 2 Uninjured.
The airplane sustained substantial damage during a hard landing. The flight instructor and dual student were engaged in an instructional flight at the time of the accident. The flight instructor held a multi-engine airplane instructor rating and was providing training to the other pilot at the time of the accident. The dual student held a commercial pilot certificate with single and multi-engine airplane ratings, and a flight instructor certificate with a single-engine airplane rating. He was obtaining training in preparation for the multi-engine flight instructor rating at the time of the accident. The flight instructor stated that he intended for the student to execute a simulated forced landing on the remaining runway immediately after liftoff. He reported that he instructed the dual student to back taxi the full length of the runway. The tower subsequently cleared the flight for takeoff followed by an immediate landing. He noted that the dual student held the brakes and applied full engine power. Upon reaching full power, the student released the brakes and began the takeoff roll. After liftoff the student established the airplane in a climb at 88 [knots]. The instructor stated that upon reaching an altitude of 200 feet above ground level (agl), he instructed the student to "simultaneously reduce both throttles to idle while pitching for a landing attitude that maintain[ed] 88 [knots]. He reported that the resulting descent rate "did not appear favorable" so about 100 feet agl he "commanded" the dual student to execute a go-around. He noted that full engine power was applied, however, the descent continued until the airplane contacted the runway in a "flat attitude." It subsequently bounced back into the air. He recalled that the stall warning horn sounded shortly before runway contact. The flight instructor stated that he assumed control of the airplane at this point and decided to continue the go-around since the engines seemed to be producing full power. He noted that during the flight around the traffic pattern, the rudder and stabilator were "less responsive than usual." He subsequently executed a no-flap landing on the departure runway. A post accident inspection revealed that the fuselage skin on the left and right sides, forward of the windshield was buckled. The left main landing gear strut was bent aft approximately 90 degrees. Both engine propellers were in the feathered position. The blade tips were curled and twisted consistent with runway contact. The National Transportation Safety Board determines the probable cause(s) of this accident as follows:
Failure of the commercial pilot (dual student) to maintain a proper descent rate and a safe airspeed during the simulated forced landing. Additional causes were the failure of the flight instructor and the dual student to initiate a go-around in sufficient time to prevent the hard landing and a failure by the flight instructor to provide timely remedial action, which allowed an unsafe condition to develop. A contributing factor was the inadvertent stall encountered prior to runway contact.

 

[Northern Lights]

AOPA INSTRUCTOReport

I must say that this has been a very interesting and informative thread. When I was instructing, I would put the student in the Frasca FTD, and fail one engine right after liftoff. If they brought both throttles to idle, they would crash -- every time! I know that the FTD is different from the actual plane, but the lesson was that it was necessary to keep quite a bit of power on the good engine if the intention was to actually land.

On another note, has anyone read the third quarter AOPA Instructor report? The front page article is called "The multiengine dilema -- Chop and drop, or clean up and go?" Me thinks that the author is UndauntedFlyer
---------------------------------------
Here is an accident from when I was a student. the CFI choped the power and pancaked it in causing some major damage to the plane.

NTSB Identification: CHI96LA202 .
The docket is stored in the Docket Management System (DMS). Please contact Records Management Division
14 CFR Part 91: General Aviation
Accident occurred Friday, June 14, 1996 in GRAND FORKS, ND
Probable Cause Approval Date: 7/25/1996
Aircraft: Piper PA-44, registration: N853ND
Injuries: 2 Uninjured.

The instructor reported that he was attempting to demonstrate a rejected takeoff. Instead of initiating the maneuver during takeoff roll he attempted to initiate the maneuver on final approach to the runway. The instructor entered the maneuver from 150 feet AGL at 70 to 80 KIAS. As the airplane crossed the runway threshold, he reduced power on both engines to idle and applied forward elevator, establishing a descent for landing. He started to 'round the airplane out about 30 to 50 feet above the runway.' The descent rate was too high, so he continued to increase the pitch. He reported the stall horn sounded 'only milliseconds' before the airplane impacted the runway. The instructor reported the demonstration he performed of a rejected takeoff was not a recognized or an approved maneuver. He reported that the he had never had this maneuver demonstrated to him.

The National Transportation Safety Board determines the probable cause(s) of this accident as follows:
the instructor pilot used an improper procedure when he demonstrated a rejected takeoff and he misjudged the flare.

 

[UndauntedFlyer]

And here is the link about the multi engine "chop and drop" accident.

http://www.aopa.org/asf/publications/inst_reports2.cfm?article=5661

And here is the one about the "Deer in the Headlights" syndrome.

http://flighttraining.aopa.org/cfi_tools/publications/inst_reports2.cfm?article=5897


Questions/Comments!

 

[Holding Short]

[Me thinks that the author is UndauntedFlyer
---------------------------------------
Me too. I think I met him one time. If it's the same guy that was giving checkrides at KPWK.

HS