Super 80 on takeoff. It’s sometimes better to continue rather than abort.
Before we begin this month’s story about the decision to abort or continue a takeoff, and before explaining our eccentric subtitle, we should thoroughly understand these takeoff-distance definitions:
Accelerate-go distance is the total distance to accelerate from a stop, lose an engine (or other reject criteria) just before V1, recognize the failure as you reach V1, and continue the takeoff to cross 35 feet AGL at the end of the runway and at your takeoff safety speed (V2). Note: after liftoff, if your airspeed is slower than V2, pitch to accelerate to and maintain V2. If your airspeed is faster than V2, pitch to maintain an airspeed no faster than V2 + 10. Some flight guidance systems will recognize an engine failure and display this profile through the flight director’s V-bars.
Accelerate-stop distance includes the total distance to accelerate from a stop, lose an engine (or other reject criteria) just before V1, recognize the need to reject as you reach V1 (officially, a three-second delay is acknowledged, but the reject must begin at or before V1) and stop the airplane using idle thrust, wheel brakes and speed brakes/spoilers before reaching the end of the runway.
Note: most aircraft performance data for an abort (reject) do not include using thrust reverse, so if used, a single thrust reverser (single engine, remember?) is a bonus in reducing stopping distance, but directional control may become an issue. Also, the commonly accepted order (after throttles idle) for the use of deceleration devices is: 1. max braking and 2. simultaneously deploy speed brakes & spoilers, and then 3. thrust reverse.
Balanced field length is where the accelerate-go and accelerate-stop distances are identical.
Now that we’ve had our refresher, on to the story.
The most difficult thing is the decision to act; the rest is merely tenacity.
-Amelia Earhart
There I Was
I was trying to analyze, rationalize, and explain away the offending engine parameter as being ‘close enough’ so that we could continue. Continue how we have for the last eight or nine thousand takeoffs, comparing left to right and assessing limitations. Why won’t it just quit—or give me a fire light, an overtemp, or at least something out of limits?
I know this airplane better than I’ve known any other machine in my life, better than the back of my hand. I know what close enough looks like. Maybe this is close enough. Nope, this is not close enough—in fact, it’s getting worse, further away from ‘close enough.’ It’s acting too different from the previous nine-thousand times. The right engine N1 was slowly decreasing, and the EGT was increasing. In three seconds, it will be too late to stop. In another eight, it will quit; I know it will.
Reject! Throttles idle and max manual brakes. Verify the auto-spoilers deployed and throw out the buckets. My FO tells the tower we’re aborting as we’re pushed slightly forward into our shoulder harnesses. The instant I made the decision, I was mad. Not mad at myself for making the wrong decision. Ironically, I was mad at the engine for not failing more definitively—more deliberately, more exuberantly. A dramatic failure that the passengers could see and be grateful that we stopped.
Dopamine Singularities
Everything you have read so far transpired in three seconds as we accelerated from 110 kts to 120 kts: 25 kts below V1. Seven seconds later, we were below 50 kts and had more than one-third of the runway remaining. I moved my hand from the yolk to the tiller and steered onto the high-speed taxiway at 20 kts, made a PA to the folks, and waited for ARFF to check us for hot brakes. Total elapsed time from brake release to taxi speed: 40 seconds. Perceived time: five minutes. I’ve had eight engine failures: two in the Duke and the rest in turbines—and two of the turbine events were during takeoff. One of them happened at gear retraction (see T &T September 2010), and the above event was my only engine failure, a high-speed abort. And it’s always the time compression that amazes me. I’ve had a handful of events, like this one, that were intense enough to cause the linear time anomaly. If you’ve never experienced it, you will. Can it be explained away as simply an adrenaline and dopamine-induced change in perception?
Black and white is easy; gray is much more difficult. Fires, engine failures or flight control problems are easy; they’re all on a short list of reasons to stop. A sickly engine that is dying a slow death, a funny noise or vibration, or a strong smell as you approach V1 means you must use judgment and decide; is it safer to fly or stop? (more in a bit about botching the abort) You’re the one that must decide, often in a hurry. The decision is made using the knowledge, training, current conditions and experience you can muster in the few seconds you have before V1.
Reasons to Stop: Fire, failure, stall, controllability, wind shear
The weight and balance and takeoff performance software used by my part 121 carrier provided minimum runway lengths and V-speeds to define the accelerate-stop/go and balanced field length parameters. The program I use in the Citation instead provides a maximum takeoff weight and V-speeds for a user-specified runway to meet the balanced field length definition—any weight below that calculated weight gives you runway to spare. Also, an engine failure is not the only reason to reject a takeoff. The commonly accepted reasons to abort include fire, failure, stall, controllability and wind shear. That is ANY fire (cockpit, cabin, cargo, or engine), engine failure, wing or compressor stall, aircraft controllability concerns and wind shear or micro-burst.
Other Bad Things
Up to a predetermined speed early in the takeoff, we are trained to watch for any reasons to stop, not for reasons to go. In high-performance twins and most jets, the first pilot-elected ‘decision point’ is around 60-80 kts—while our kinetic energy is still relatively low. A typical list of reasons to stop before 60-80 kts would begin with the big things: fire, failure, stall, controllability and wind shear. But, at that first slow-speed decision point, we can also add other potentially bad things: slow acceleration, tire failures, unusual noise or vibration, a sudden, strong smell, bird strikes, and possibly a call from your cabin crew or a scream from the passengers. In other words, just about anything.
[A pilot] must have good and quick judgment and decision, and a cool, calm courage that no peril can shake.
-Mark Twain
Once we exceed that predetermined, slow-speed decision point, we will continue accelerating to V1. In the time from our elected decision point to our official, performance-based decision speed at V1, we will then only abort for the big ones: fire (cockpit, cabin, cargo, engine), failure (engine and maybe total electrical), stall (wings or engines), controllability (flight control issue: runaway trim, failure to set the flaps, a deployed thrust reverser, rudder hard-over), and wind shear or microburst. I hear you asking: how can we have a wing stall during takeoff? I don’t know—wing ice or possibly AOA failure. But do you really want to leave the ground with the stall warning blaring and/or the stick shaker going off? Once we reach V1, we are taking the vehicle into the air, no matter what, lest we attempt an abort above V1, exiting the paved surface at a kinetically energetic (½ mv2) and likely deadly velocity. Now, about botching that high-speed abort.
We Totally Botch them
A low-speed, low-altitude, heavy-weight engine failure is dangerous and difficult to manage. So why even have this discussion about whether to go or stop? Why not simply decide right now to abort for just about anything? Why is the decision not black and white? Because of that (½ mv2) thing, my dear Watson. And our propensity to totally botch the abort—despite practicing them in the sim. We perform aborted takeoffs least often of all the maneuvers. I bet you have never practiced a high-speed abort in an actual aircraft; at least, I sure hope not. We botch them for a variety of reasons: no experience, low ability, late decision, bad decision, bad abort techniques, bad runway conditions, bad aircraft equipment, bad runway components and sometimes, just bad luck.
We are encouraged to continue the takeoff in jets and turbines because we have plenty of power to fly all day long on one motor—if done properly. If the airplane is capable of flight and not on fire, statistically, we will have a better outcome if we continue the takeoff and work the problem in flight. The reason we’re encouraged to continue the takeoff between those slower ‘criteria’ speeds and V1 is because, as a group, we have shown that, more often than not, we will totally and completely botch a rejected takeoff and depart the paved surface anyway—either off the sides or off the end. And we will catch the tires, wheels and brakes on fire and hurt a bunch of folks during an evacuation. Sometimes we catch the whole blasted airplane on fire.
It’s better to be on the ground wishing you were in the air, than in the air wishing you were on the ground
It’s better to be on the ground wishing you were in the air, than in the air wishing you were on the ground If you ‘go’ when you should have stayed, you will likely regret it, maybe a lot. If you ‘stay’ when you could have gone, you may also regret it, but probably much less, unless that is, you totally botch the abort. Remember: each runway, your takeoff weight, the wind & weather, your level of tiredness and time-of-day will affect your plan. Once you appreciate that (½ mv2) thing and have a personal list of low-speed and high-speed decision events, and you have a plan for when one of them happens, stick to it. The confidence you gain from simply having a plan will serve you well. And whether you stay or go, when one of your ‘events’ does happen, the changes in your body’s adrenaline and dopamine levels will almost certainly alter your perceptions and give you a moment of ‘Time in a Bottle.’
Author’s note:
Grateful acknowledgment to Jim Croce for his classic 1973 folk-rock tale, “Time in a bottle”.
