fdr,
As to standing up for a point .. regardless of whether it be right or wrong ... be prepared to be shot down if the aim is to reduce payload beyond that permitted by regulation. That view may be cynical but it is realistic. An individual operator has to make a commercial/risk decision regarding whether, and to what extent, any additional conservatism might be appropriate.
There is no simple answer .. no matter how you look at it .. and the world, unfortunately, is not an idealised place ... and, no matter how much we may not like it .. the bean counters rule the world these days ...
AFM data is
(a) performance model predicted data, which is
(b) validated during certification testing,
and has not a lot to do with real world observations unless both (aircraft/environment) conditions and (pilot) techniques are very similar to those which went into the numbers in the first place. It would be inappropriately idealistic to presume that the manufacturer might contemplate building into these data any conservatism not required by the rulebook ....
If you saw (as we all do routinely) any particular takeoff which had AEO liftoff very close to the end of the runway (or low achieved screen), this could be due to any of a host of reasons .. some of which would include
(a) engine(s) down on predicted thrust, either due to the engine(s) or, more likely, crew operation
(b) line up distance squandered prior to spinup
(c) spinup not in accordance with AFM techniques
(d) dragging brakes
(e) substantial error in the loading sums versus the loading ... a common tale .. errors in empty weight, errors associated with the use of standard weights, loading scale errors in freight operations .. etc., etc....
(f) late and/or slow rotation resulting in longer flare distance and speeds higher than AFM-predicted ... generally the case.
(g) for 4-motor birds, the AEO performance is going to be less spectacular than for 3-motors .. which is less than for 2-motors .. as the OEI case for continued takeoff generally sets the level of the bar ...
The latter points constitute the most likely story. Keep in mind that flight test techniques may tend to be somewhat more aggressive than what one sees routinely on the line.
If you feel the need to work on the presumption that the routine AEO case should see 35ft with a 15 percent pad, then do ensure that this relates to the TODA and not the TORA ... which could be considerably less.
The fact that we don't see routine AEO overruns or terrain collisions suggests that the risk levels associated with the AEO problem you identify do not present an unacceptable situation .... ?
Alex,
You might have put me on the spot if the aim is to find a reference on which I can hang my particular words ... although I suspect that we are saying the same thing ? ... the worst case is going to be the aft cg wet runway .. so that is what has to be addressed ... ? in any case, at aft cg, the steering is not going to be particularly effective even on a dry runway if the yawing moment results in lateral skidding due to the reduced vertical load on the tyre. This discussion, of course, could go around in circles for a long time ..
Useful references -
(a) FAR 25.149(e) "VMCG, the minimum control speed on the ground, is the calibrated airspeed during the takeoff run at which, when the critical engine is suddenly made inoperative, it is possible to maintain control of the airplane using the rudder control alone (without the use of nosewheel steering) ....."
(b) AC 25-7A 23(3)
(iv) Control of the airplane should be accomplished by use of the rudder only.....
(vi) VMCG testing should be conducted at aft c.g. and with the nose wheel free to caster, to minimize
the stabilizing effect of the nose gear. If the nose wheel does not caster freely, the test may be conducted with
enough nose up elevator applied to lift the nose wheel off the runway.
(vii) For airplanes with certification bases prior to Amendment 25-42, VMCG values may be
demonstrated with nose wheel rudder pedal steering operative for dispatch on wet runways. The test should be
conducted on an actual wet runway. The test(s) should include engine failure at or near a minimum VEF associated
with minimum VR to demonstrate adequate controllability during rotation, liftoff, and the initial climbout. The
VMCG values obtained by this method are applicable for wet or dry runways only, not for icy runways.
.. or have I just fogged it up a bit more ?