Originally Posted by
DaveReidUK
While there aren't enough data points on the FR24 plot to determine exactly where the A320 became airborne, the last point shown with WoW is around 300 meters short of the upwind piano keys with a groundspeed of 67 kts (subject to usual FR24 caveats).
at 67kts GS, ~300mtrs to the end of TORA, the guys need to be accelerating at better than 0.35g to get to Vmu. However, they have another 1000' of runway beyond the displaced threshold, and a further 400' of clearway that is better than sand at least to run your wheels over.
Getting to 67Kts GS in around 1200' from the entry into the runway from the rolling on speed of say 15kts is a gentle acceleration, around 0.15g average, with a peak of only 0.2g. At some point the guys need to be adding some noise in order to not leave sand tracks, otherwise they don't get to Vmu on the concrete. If they are light, and add full thrust at the 1000' TR to the displaced threshold, then they can get to about Vmu with at the displaced threshold, and the aircraft has around 1400' of hard surface ahead. We would be hearing about the damage on the plane in that case, so that didn't occur. Rotating at around 110-115 kts instead would occur about 500' past the displaced threshold, and the wheels come off the ground around the beginning of the chevrons of the clearway, just east of the displaced threshold arrows.
That is the rough go case, and it needs adding of thrust promptly by the crew to make it work as it did.
The stop case is at 67GS 1000' TR, which gives 2400' to the sandpit, is stoppable. A couple of seconds later, it is not going to work, but the plane is going to be hitting frangible structures mainly, however, the aircraft on a soft surface will be impacting the structures at a lower height than they may have been designed for. Other than the frangible objects issue, that particular runway is about as good an outcome as you can expect outside of Edwards.
The acceleration curve is highly non linear, being roughly a 6th order polynomial on a normal takeoff from commencement of thrust addition or brakes off, through to maintaining V2 + XX kts. The curve from achieving thrust setting through to rotate is more or less a simple second order function, as drag from the airframe and the rolling friction act as a increasing force, with the thrust either being stable or slowly decaying (dependent on the engine characteristics, bypass, etc). Adding a thrust increase in the middle of that gives another 3 inflexion points... gets busy but can still be approximated, and doesn't need that anyway, it is recorded by the DFDR, and the QAR as well, so the info will arise eventually as to where the wheels were relative to the sand.
Takeaway is just having the right amount of runway in front of you when you set the thrust is a highly preferred state. Anything the system can do to increase the reliability of that simple situation is not a bad thing. The airlines "knee jerk" response is a good start to reinforce safety. Intersection departures increase operational risk through the potential of misidentification of the intersection used, miscalculation or no calculation of the performance change (long runway, light aircraft, no problem, don't bother checking), and potential for a loss of S.A./brain freeze leading to a wrong direction on the runway.
Beating up the crew in this instance doesn't stop the problem, they are now probably the least likely crew out there to mess up TO performance. They were also lucky, if ATC had cleared another aircraft onto the runway at full length (assuming that there was no WIP precluding that) then the obstacle would have been impossible to avoid in a go case, and the stop case would only occur if the crew recognised the problem early enough to react effectively. Would another aircraft follow ATC clearance to line up behind, and always note that the other aircraft is pointing in the wrong direction? maybe...