Engine Failure after V1 but before V2 what do you do?

[B767Driver]

Originally Posted by jtrain609

Erm, you don't have special engine out procedures that include obstacle avoidance? I'm pretty sure ours are just for us, but don't other people have that stuff?

Most airlines will subsribe to an engineering product that analyzes obstruction clearance out to 30 miles or so. Mostly, the obstruction clearance guarantee with these products will be ensured with the runway analysis...since obstruction clearance will largely depend upon which runway you depart.

The Part 25 profile is not runway dependent. Since you are simply meeting a climb gradient profile...the runway is not an issue.

Takeoff and climb performance data will have two numbers. A runway limited weight and a climb limited weight. You will notice the climb limited weight will always be the same regardless of which runway you takeoff from. The runway limited weight may be different even for runways of the same length. This would be the data accounting for obstacles in the departure path.

[B767Driver]

Originally Posted by higney85

Ok without getting beyond knee deep think about the basics.

What is V1- what does it ensure?

What is V2- what does it ensure?


[/b]

Getting back to these questions:

V1 is the speed at which the critical engine fails PLUS the speed gained during the time interval during which the pilot recognizes and reacts to the engine failure. Typically Vef plus 1 second.

V2 must be at least 20% faster than the stall speed and 10% faster than Vmc. It also must provide for the minimum climb gradient specified in Part 25.

VR must not be less than V1, must be 5% over Vmc, and using normal rates of rotation will ensure that V2 is attained by 35' and that the tail will not strike the runway surface.


These numbers ensure that you have enough runway to stop, enough runway to continue and to meet Part 25 OEI climb gradients on departure.

The pilots must ensure that these gradients will not penetrate obstacles on departure. Most airlines have an independant analysis of every airport they fly from and will have worked the performance data accordingly to reduce crew workload.

[Dugie8]

Originally Posted by jtrain609

Erm, you don't have special engine out procedures that include obstacle avoidance? I'm pretty sure ours are just for us, but don't other people have that stuff?

That "stuff" you have is company and airplane specific, I can't use your OEI procedure in a Citation for say KSLT. The data you have gives you a temperature and weight combination for a given runway. You are limited by either climb performance or runway lenght to meet either accelerate stop or accelerate go. That procedure is usually VASTLY different that any obstacle DP for that airport.

[jtrain609]

Originally Posted by Dugie8

That "stuff" you have is company and airplane specific, I can't use your OEI procedure in a Citation for say KSLT. The data you have gives you a temperature and weight combination for a given runway. You are limited by either climb performance or runway lenght to meet either accelerate stop or accelerate go. That procedure is usually VASTLY different that any obstacle DP for that airport.

Obviously, but do you have the same stuff for your airframe or is your company not investing in it?

[higney85]

Each company has specified single engine climb out instructions for performance critical airports. At Pinnacle we now have the LIDO charts which gives you pretty colors and somewhat of an idea of where NOT to go. I have only flown for 1 part 121 carrier so I cannot talk for all or any other carrier. What I can tell you is that in places such as Helena Montana we have a different departure to fly in the event of an engine loss on takeoff or during the first segments of climb. I have no idea how it compares to the other carriers that operate in/out of HLN. What I can guarantee is that it must be something close to ours because the mountains are in the same place whether you are in a CRJ or ERJ.


Moral of this topic- At/after V1 you are going flying, before V1 the idea is to stop. If there are issues such as contaminated runway or a performance limiting item your V1 will be adjusted accordingly to be lower and in theory allow you to stop. I cannot speak for other A/C but the CRJ climbs like a pig but has some amazing brakes!

[Dugie8]

Originally Posted by jtrain609

Obviously, but do you have the same stuff for your airframe or is your company not investing in it?

Nope, 135 operator we have to use the published DPs and then see if we can make it using the charts (HUGE PITA!). ATI and Piedmont had it however.

[Doug Taylor]

Originally Posted by jtrain609

Erm, you don't have special engine out procedures that include obstacle avoidance? I'm pretty sure ours are just for us, but don't other people have that stuff?

We've got 'em for the DOH, B767Driver beat me to it.

[BobDDuck]

The single engine climb procedures are NOT Jepp (or LIDO) data. They are third party (ours is Aerodata) publications that take into account the minimum performance your aircraft type can fly when single engine and then give you a departure corridor that you SHOULD be able to maintain assuming you follow the weight/performance restriction in the rest of the data. Some of them are pretty straight forward (turn to heading XXX). Some are more complex (at X DME track heading XXX to join the XXX VOR radial XXX). Either way, our weight restrictions are predicated on being able to meet the required climb on both the 2 engine normal departure and the single engine special departure.

What Jepp (and NOS and I think LIDO) does have is a normal "Obstacle Departure Procedure". This just gives you the climb gradient that you must hold to clear terrain. That data is generic and applies to any aircraft departing the runway. If you don't have third party data, it's up to you to figure you if you can meet it with one or two engines.

[Doug Taylor]

Ours at Southernjets are AUAeropilot publications!

[kellwolf]

Originally Posted by BobDDuck

The single engine climb procedures are NOT Jepp (or LIDO) data.

I think what Ken was referring to was the specific airline procedures (LIDO) or the 10-7 pages in Jepp. I don't think he meant that LIDO actually developed the data. They do have pretty colors, though.

Side note: had to fly with a deferred FMS from IND to BOS yesterday, and I discovered I actually like the LIDO enroutes better than Jepps. They're not as cluttered, so it's easier to find stuff like airway names, VOR freqs and mileage data.

[staplegun]

Originally Posted by Dugie8

You are going to have to show me an airport that has a published (for everyone) one engine inoperative (OEI) procedure for obstacles???

Rio De Janeiro is one, there are several others throughout the world...

pg0060im.gif


(I could only find a picture of SBRJ's, but SBGL's is virtually identical...)



Kevin

[germb747]

Every airplane/airline might do it slightly differently, so check the applicable manuals, but generally you'll want to accelerate and climb at V2 until a specified height (800ft for us, or clear of obstacles), then accelerate in level flight and clean up.

Our procedure is that if an engine failure occurs above V2 but below V2+10, climb at the airspeed that exists at the time of the failure. If the engine failure occurs above V2+10, climb at V2+10.

The purpose of calculating a V1 speed is to give yourself a decision point to continue the takeoff or reject. Below 80kts, we reject for all kind of things. Above 80kts and below V1 we reject only for engine failure/fire/safety of flight. Does that mean that if you wait one knot over V1 before initiating the abort, you're going to run off the end of the runway? Maybe or maybe not. Depends on where you are and the conditions. If you're taking off from the dry lake bed at Edwards AFB or the 15,000 ft runway at NASA's shuttle landing facility on a dry day, you'll probably have more than enough room to stop if rejecting after V1. (It's still incorrect procedure to do that, by the way). I'm just making the point that V1 can be calculated a number of ways--it can be "stop" oriented or "go" oriented, provided it fits between a few set values. Biases can be put into the computation to, for example, provide a certain stop margin. In the military, we remember it like this:

Vmcg, Vcef < V1 < Vrotate, Vrefusal, VBMax

I'll try to define some of these terms, because I know the civilian world may use slightly different terminology:

Vmcg - Minimum Control Airspeed/Ground. With the critical power unit failed, runway centerline can be maintained within 25feet using rudder force.

Vcef - Critical Engine Failure Speed. The speed one can accelerate to, lose an engine, and either accelerate and takeoff or recognize the failure and stop in the same distance (critical field length). You may know this as a "balanced field"

Vrefusal - The no-kidding speed you have to abort by to stop in the available runway (which may be very high if we're talking about Edwards AFB, for example).

VBMax - Speed at which maximum brake energy absorption capability occurs if brakes are applied.

So you can see a V1 speed can be safely computed as long as it is more than the highest of Vmcg and Vcef but less than the lowest of VRotate, VRefusal, and VBMax. Normally, for us, V1 will be based on VRotate, so delaying the abort past rotation speed would be illogical since you're already flying at that point. If we're to have "split markers"; i.e. a V1 less than VRotate, you'd be looking at big trouble if you tried to abort beyond V1 because you're probably looking at a low refusal speed and would actually go off the end of the runway.

So, to keep it simple, never abort beyond V1 unless you know you're going to crash otherwise. Just commit yourself to go flying after V1; it's the correct procedure and it's the safest thing to do.

[AUAeropilot]

Originally Posted by Doug Taylor

Ours at Southernjets are AUAeropilot publications!

Yes sir, yes sir. By the way, Bogota is about to go up stairs for some head scratching and sim flying here shortly. We got rid of the bowtie and got out to the west to hold, away from everything coming inbound via BOG.

If anyone has any questions about this stuff let me know. I'm still relatively new to it, but there is a guy in our department who did the engine out stuff for Pan Am, in fact he was like 1 of a few engineers they had in flight operations engineering. We call him "the professor" for many good reasons. If I don't know the answer to something I can definitely go to a few extremely reputable sources. I'll divulge as mush information as they'll let me regarding some of our challenges, but I can definitely answer or clear up the more general stuff.

We do all our engine out procedures in house, courtesy of yours truly. Most of our existing airports need little to no modifications on the procedures expect for "sharpening our pencils" with regards to splay width (obstacle accountability area) and actual aircraft performance; it's the new markets that'll suck a lot of time into the development. That is unless our marketing department is hitting the bong again and is bound and determined to get an equipment type into somewhere it just cannot go, from a performance standpoint. That happens way too many times.

In regards to the questions that were asked, let me go home and try and put together a response, hopefully not too technical or lengthy, that try and answers the questions thus far. It's Friday and our little guy, almost 2 now, has some daddy time with Thomas the Tank Engine planned.

[AUAeropilot]

Originally Posted by Foxcow

...At the airspeed suggested, Vmcg is not a factor. ...

The topic of Vmcg actually approaching V1 speeds was covered in the Boeing conference I went to this year. It was with regard to the 777-200LR, which Delta is the US launch customer for [/shameless plug], and the higher thrust engines being placed on the aircraft with no real appreciable enlargement of the vertical tail or rudder size. It really only comes into play when using a derated takeoff setting and then firewalling the throttles upon engine loss at V1.

An excerpt from the article is below:
Fixed derate takeoff limit thrust in takeoff calculations (VMC speeds are set by derated thrust level), yielding lower takeoff speeds and shorter field lengths. Pushing thrust higher upon engine failure violates procedures and reduces or eliminates margin to Vmcg and Vmca, and reduces margin to loss of directional control speed.

Ok, nerd moment over with. Back to the show.

[Dugie8]

Originally Posted by staplegun

Rio De Janeiro is one, there are several others throughout the world...

Attachment 4627


(I could only find a picture of SBRJ's, but SBGL's is virtually identical...)



Kevin

Is that for obstacle(s) or for dumping fuel to return to the field?

[BobDDuck]

It's got a climb gradient and states "a provision for dumping fuel" so I'm guessing it's an obstacle predicated departure with the fuel dumping thrown in for good measure.

[staplegun]

Originally Posted by Dugie8

Is that for obstacle(s) or for dumping fuel to return to the field?

It's definitely for obstacles; the fuel dump is just thrown in and is not an integral part of the procedure...


Kevin

[Doug Taylor]

Thanks AUA! You're a tremendous asset to the site.

Hey! BTW, I shared a beer or four with the guy responsible for designing ATL's southernmost runway. Poor guy was shocked when I went off on him about how the overpass resembles a high speed taxiway.

Fantastic guy though. Works in Kristie's new office.

[AUAeropilot]

Originally Posted by Doug Taylor

Thanks AUA! You're a tremendous asset to the site.

Hey! BTW, I shared a beer or four with the guy responsible for designing ATL's southernmost runway. Poor guy was shocked when I went off on him about how the overpass resembles a high speed taxiway.

Fantastic guy though. Works in Kristie's new office.

Do me a favor and never let him step foot in the general office complex, particularly the 4th floor, 3rd floor FLOE offices, or the training center. Ok, just kidding...a little. You didn't hear this from me, but there is a distinct possibility that that runway will be extended in the next 10 years or so. I haven't heard that is set in stone, but it has been thrown around before. I haven't had the opportunity to get the 20 minute taxi from out there to the gates...yet.

I'm thinking I'm gonna do some type of write up about performance determination from our side of things. It'll be general information simplified as best I can. Hopefully it'll answer some questions and clear some things up. Of course it could open up a can of worms completely and make everything more confusing. It'll basically be how performance numbers are generated from the "chase me" charts in the AFM.