Question about NTSB accident review data

But there was direct forensic evidence of the seat position at impact, and an unsecured seat initially moving fully aft is consistent with the direct video evidence of the loss of control.

I’m not sure why you feel that your guesswork about the forces just prior to impact is sufficient to dispute this direct evidence. When the unsecured seat initially moved fully aft it may have jammed there; or friction from lateral g-forces may have stopped it moving forward again when the aircraft pitched nose-down.

I just can’t agree. The NTSB are skilled and experienced at forensic analysis of wreckage. There is no evidence for any other explanation for the loss of control, and we have a history of problems with this seat. All the evidence seems consistent here.

The NTSB don’t quantify their degree of confidence, but even if they did I don’t think any of us are qualified to judge that they should be saying they are only (say) 60% certain rather than 90%, which is what you seem to want here.

Professionally speaking …

Even those of us who are conscientious professionals make mistakes. And more often by omission than commission; that’s a feature of imperfect human cognition. And even with two of us, sometimes those mistakes are uncaught by both. Happens every day on many flights.

The vast majority of such fully uncaught mistakes are harmless. But when something bad has happened, it’s a near certainty one of those rare mistakes was part of the accident sequence.

It’s a form of Bayesian probability looked at from the other end of the process. Mistakes are kinda rare. Accidents are extremely massively rarer. But given an accident, what’s the probability of a significantly contributory mistake? Nearly 100%.

Switching from pro mode to amateur accident observer / explainer mode:

Something else to consider about a sliding seat is that the pilot’s weight sliding aft as his seat became unlatched would itself be unhelpful to an already precarious W&B situation. And that’s in addition to the monster control input as he inadvertently hauls back on the yoke.

I admit that I don’t place a great deal of faith in a lot of NTSB GA accident investigations. They often take a few pictures, talk to a few witnesses, publish a guess, and move on. I wish there was some more objective rating scale they applied to now rigorous the investigation and how confident they are about their conclusions. Them doing a lick-and-a-promise job on yet another VFR flight into IMC or light plane fuel exhaustion mishap is a completely reasonable prioritization of resources versus future accidents prevented. I just wish they were able to say up front in big print that that’s what they are doing. And therefore to apply appropriate caveats to what turns into the pilot’s aviation epitaph.


As to the seat stop witness marks, there may well be one set of marks where it slammed to the aft stop, and a different set of marks when the airplane hit the ground.

Digging into the photos in the NTSB docket, it’s pretty clear the seat was latched in the full aft position at impact. And failed forward from there. Both the pin(s) on the seat and the hole(s) on the track confirm that. So the question becomes: “How did the seat get full aft and locked when the pilot is too short to reach the controls with the seat there?” And the answer, which is surmise, but reasonable surmise, is “The seat wasn’t well-latched at takeoff and slammed aft when he first rotated, then over-rotated as the unlatched seat slid abruptly aft.”

From the vid as I interpret it, the airplane rotated so abruptly it probably drug the tail, or nearly so. Which is totally not the way to take off in an ICE or turboprop straight-winged airplane. “Rotation” is a swept-wing jet idea; that airplane should have more or less flown off in a level attitude with a smidgen of pitch up upon breaking ground.

The presence of this improper rotation suggests to me that the seat situation, or the trim setting, or the W&B situation was already going haywire while they were still on the runway. Maybe it was already unrecoverable by then, maybe it was not. But very quickly things got utterly out of hand.

I tend to discount a trim setting mistake as a full-up cause. Pitch trim in a low speed non-transport just doesn’t have enough authority to render the airplane unflyable all by itself. Unlike the all-moving stabs of big airplanes. Part of the reason big airplanes have elaborate systems to prevent stab trim runaways and elaborate backup systems to ensure failures don’t leave a stuck stab is precisely that they have enough leverage within their normal operating range to leave the airplane flat uncontrollable.

Lightplanes solve the problem the other way, by limiting the trim authority (by range of motion and size of tab) such that a pilot of normal strength can outmuscle even worst case trim stuck at one or the other stop. To be sure badly mis-set trim can be a real startle as you’re accelerating down the runway and the pitch situation starts developing a mind of its own. But if nothing else goes wrong that “should” be handle-able by the reasonably prepared & skilled private pilot.

We know, or strongly suspect that “nothing else” wasn’t the case that day.

Or it may have just chosen that moment to lock properly. If the pin was initially not fully engaged, then a good shove backwards might have caused it to glide over the intervening holes–until it hit the aft stop, at which point the pin dropped into place and fully locked. Simple bad luck could have exacerbated this, say if the aft hole was slightly larger than the rest, or was better centered relative to the pin.

I’m just an interested reader here with not a lot of aviation knowledge, and I’m genuinely curious but I suspect it’s a typo. What does this sentence in your post mean?

“It has about 3x the moment arm of the main tanks.”

No typo. See here for an explanation of how the center of gravity is calculated to include the variable weights and positions of fuel and payload, including a simple example:

Center of gravity of an aircraft - Wikipedia.

Follow the link I posted above to the docket to find the weight and balance calculation the NTSB included with this accident report. You will see that the arm of the transfer tank is 138" aft of the datum, while the arm of the main tanks is 43" aft of the datum.

Ok, thank you! I can see I’m in way over my head here. I take it that the poster’s meaning is that the transfer tanks would have contributed to the CofG being off and that fuel being used from there would help the situation.

Yes, what the NTSB report calls a “transfer tank” was an auxiliary range-extending fuel tank that had been installed in the aft baggage compartment. The fact that this was full (when the extra fuel was not required for the planned trip) was the greatest contributor to the aircraft being overweight and unbalanced.

Yes, this was I was getting at, so I’m glad to hear it confirmed by a professional. I don’t expect that they got it wrong per se. An expert probably could have just eyeballed the incident and figured it in a few hours, given what they saw, and given what normally happens in that situation. The report is probably correct and it serves its intended purpose. I was just hoping the dots would be better connected for me to understand it, but the final report just gives whiffs of boilerplate to me. Maybe I was expecting too much, maybe this really incident really was nothing more remarkable than the boilerplate event it sounds to be.

Yes, as I mentioned before this was the first turboprop he owned. I don’t know how many hours he had in turboprops, if any. The preliminary report noted the engine only had 2 hours on it. Enough for a couple of checkout flights with 1 instructor and no baggage.

Thanks for addressing that point. I had no strong opinion on it, but in reading other reports I understood this to be a possible contributor to past issues with the aircraft, and maybe something this pilot hadn’t much dealt with before. So I appreciate your treatment of that topic.

I appreciate everyone’s insights previous to now, and thanks @lslguy for chiming in with additional professional perspective.

Riemann has it. An airplane is just a see-saw, and the moment arm is just the distance from the fulcrum. If there’s some mass 3x further out, it contributes 3x as much to moving the CoG around, pound-for-pound. Likewise, removing mass from a place 3x further out moves the CoG in the other direction 3x as much.

The “moment” (not moment arm, just moment) column in the NTSB weight-and-balance chart shows the overall contribution: it’s just weight times (moment) arm (divided by 1000 for convenience). The transfer tank had a moment of 24.8, which is almost as high as the main tanks. Nothing else on the aircraft contributed nearly as much.

Nitpicking myself a tad: the “fulcrum” (called the datum) is essentially arbitrary, and the math works out the same no matter where it’s put. Usually, it’s put in a location such that all the distances come out positive (for calculation convenience). One could move the datum to the aft tank, though, and it would then appear to have zero contribution. Nevertheless, mass that’s far away from the center of lift (not the same as the datum point) is going to contribute more to a bad CoG than mass closer to the center of lift.

Thank you. The see-saw/lever analogy helps he a lot!