Just had my coffee break and was reading accident reports in AS Magazine. This is a full list of NTSB accidents. Note how many are 'power related'. Whilst these aren't necessarily the engine suffering physical damage and stopping, the interesting fact is, they're not LOC, Terrain, etc that are according for the accidents.
Here is a summary
Beech P35 - wrong fuel tank selected
PA-28-140 - substantial water recovered
Cessna 182H - engine failed
Stinson 108-3 - engine failed
PA-28-235 - random crash
Cirrus SR22 - random
Flight Design - random
PA22-125 - substantial water recovered
PA34-200 - random
C162 - random
Beech 23 - power loss, intake blocked
Cirrus SR22 - engine failure
C172 - random
RV6A - random
and I thought engines accounted for 10% or so of accidents...
[List reformatted. Use 2 spaces at the end of a line, to force a line break without a blank line. See Posting Tips]
Well 3 engine failures out of 14 gives you a 95% confidence interval of 5% - 51% of crashes being caused by engine failures so the 10% statistic could still be true.
7/14 gives a 95% confidence interval of 23% to 77% of accidents being caused by power loss.
A binomial confidence interval calculator:
Well 3 engine failures out of 14 gives ...
I would also count the "wrong fuel tank selected", "substantial water recovered" and "power loss, intake blocked" as engine failures - the only difference to the other three unspecific failures is that in these cases the cause for the failure is known. Which brings the percentage up to 50.
But I think when people talk about engine failure they mean actual mechanical failure not pilot-induced engine failure ie running out of fuel. Actual mechanical failure is a smaller percentage.
We all tend to divide accidents into things we have control over and things we don't. If I hire a plane and the engine explodes I won't feel to blame, though I will if I run out of fuel or have a CFIT. I have some influence over water in tanks too.
My point is - the prop stopping to turn, from one reason or another is accounting for half of accidents in this period. Not blaming the engine but there are simple things manufactures could do to reduce accidents (like designing mechanism to stop the pilot screwing up)
there are simple things manufactures could do to reduce accidents>
I think it's the training regime, not the manufacturers, that could (SHOULD) do something. There are three recent Youtube videos where there was little or no injury. In one, an instructor lands in a field. In one, the pilot puts the plane into the biggest available tree, ignoring a road and a field. The third one is an interview with a student, who, engine still running, stalled at slow speed, from low altitude, uphill into scree rocks. Teaching keeping control,while getting the speed off at a low height, rather than insisting on flying a pattern in a PFL would be better. A friend who had failed to get his hours for renewal had to sit a GFT. He was given an engine failure near a microlight field. While the instructor agreed he would have got in, he failed him for not flying a pattern. Some one on another blog commented that the instructor who did the forced landing didn't fly a pattern, apparently thinking that mattered.
Frankly speaking, history repeats itself - humans do forget to change tank, do forget carb heat, don't open cowl flaps, run too rich. Why expect the next generation of pilots to suddenly provide different statistics? I suspect a little micro controller could solve 20% of accidents...
The bigger problem is that pilots won't accept that we as humans are incapable of erroring. I get pleasure from flying but I don't want a choke, mixture control and left and right manual valve tank in my car!
I used to be a real skeptic re human factors. Read the Naked Pilot and you get a better appreciation of how these things happen. Really good pilots make mistakes. But it is important not to confuse pilot mistakes with mechanical failure. Even in pistons mechanical failure is very rare.
I suspect a little micro controller could solve 20% of accidents...
Or to take things to their logical conclusion as per Cirrus and Airbus, a really big microcontroller ought to be able to prevent 100% of accidents.
(I half agree with you, though when I've flown Rotax engined aircraft I miss the extra levers)