In this context, what I have trouble with is the use of vertical acceleration as a yardstick, with a two decimal points resolution no less ....
I have no problem with using vertical speed, even at only one or two samples per second.... The whole event is pretty well defined by the vertical speed at the last few seconds (think inertia), and the geometry/characteristics of the landing gear.
Does somebody here have figures for the sample rates for vertical acceleration in a typical FDR or QAR?
Since at 10 fps the whole touchdown event is over in about a second, I would have thought you'd need something like 10 to 20 samples/sec to get an accuracy of even a few percent for peak G, which seems to be a waste of bandwidth and recording capacity.
Or am I missing something and is there a separate peak-reading device?