What is Infant Mortality TBF for Lycoming IO-360?

Jaybird180

Final Approach
Joined
Dec 28, 2010
Messages
9,034
Location
Near DC
Display Name

Display name:
Jaybird180
Anyone know what the acceptable range is when a new engine is considered low risk? I haven't been able to find any good data, though it is generally agreed that infant mortality is a real thing and a real risk that should be respected.

I have however, seen a curved chart that risk is highest at low time and high time for an engine, but no numbers included in the chart, so I dismissed it as anecdotal.

Thanks.
 
I’ve heard 200 hours to prove an engine. Dave Weber, a machinist/ pilot, posted an article (aopa maybe?) where he states with 200 hours most stress fractures from manufacturing will show up. After that you should be good results.
 
Anyone know what the acceptable range is when a new engine is considered low risk?
Too many variables to give a general range. As mentioned above who did the work, plus how was it broke in, who maintained it, etc. all figure into the infant mortality rates. There are high level reports and analysis out there like the FAA's FMEA reports and the Australian ATSB reliability reports that give you stats on OEMs and specific failure rates. Regardless by simply following all the OEM recommendations gets you through those initial hours.
 
Ran a quick comparison on my EAB database, comparing the aircraft time at time of accident. Not necessarily the same as engine hours, of course, but does seem to show the IO-360 is pretty close to the O-320. Note that this is *all* accidents, not just those engine-related, but if the IO-360 were significantly worse, I'd expect it to show.
io360 and o320.jpg
Sample size was 199 for the IO-360, 373 for the O-320.

Ron Wanttaja
 
What can we ascertain from the graph? Perhaps a refutation of infant mortality? I'm hesitant.
I suspect that considering the fleet-wide use of the IO-360 this data set (not using engine hours) isn't extensive enough to see if the accident rate plateaus or goes down as a function of time.

So, I have an idea of what it doesn't tell us, but just not sure of what it does tell us.
 
I've always heard that if the connecting rods are not torqued properly that catastrophic failure will manifest itself within 200 hours.
 
What can we ascertain from the graph?
Maybe it’s telling us that engine failures don’t necessarily result in accidents.

I’ve had half a dozen engine failures in 45 years of flying… no accidents yet!

That said, I’m very cautious during the first 30-50 hours of engine operation… two oil change intervals (first change at ten hours…) No night, no IMC, no mountain flying, only selected passengers…

Paul
 
What can we ascertain from the graph? Perhaps a refutation of infant mortality? I'm hesitant.
I suspect that considering the fleet-wide use of the IO-360 this data set (not using engine hours) isn't extensive enough to see if the accident rate plateaus or goes down as a function of time.

So, I have an idea of what it doesn't tell us, but just not sure of what it does tell us.
My thought was that it showed that, whether the IO-360 has an elevated rate of infant mortality, it seems to be no worse or better than the common O-320 engine.

In comparison, here's the same plot, with the Volkswagen engine conversions added:
io360 and o320 and VW.jpg
You can see how the VW-powered aircraft's accident rate is much higher earlier in the aircraft time. This is an indication of higher infant mortality.

As Paul says, though, this is ALL accidents, not just those related to engine failures. Luck and pilot skills will determine if a typical engine failure might end in a reportable accident. IO-360s are more likely to be installed on higher-performance homebuilts, though, and the likelihood of a successful forced landing is somewhat less.

My homebuilt accident database shows that about 6.1% of all O-320-powered homebuilt accidents are due to mechanical failure issues with the engine. The rate is 8.1% for the IO-360; a bit higher, but the actual numbers of cases are pretty low (less than 30) for each engine. The VW conversions are at 22%.

Ron Wanttaja
 
I don’t know if I’m interpreting the graph correctly, but if so, I’d say you’re still well within the infant mortality zone for those lycoming engines even at 400 hours.
More than half of all accidents are below 400 hours? :eek:
 
I don’t know if I’m interpreting the graph correctly, but if so, I’d say you’re still well within the infant mortality zone for those lycoming engines even at 400 hours.
More than half of all accidents are below 400 hours? :eek:
This isn't accidents just due to engine issues; they're due to all causes. And it isn't driven by the installed base, there are hundreds, if not thousands, of aircraft that don't have accidents.

Ron Wanttaja
 
I don’t know if I’m interpreting the graph correctly, but if so, I’d say you’re still well within the infant mortality zone for those lycoming engines even at 400 hours.
More than half of all accidents are below 400 hours? :eek:

Well, now you must take other data into the decision matrix. How many of those hours are early hour student pilots? Etc. -Skip
 
I dunno. From 200 to 400 hours the increase was nearly 20%. And the sheer fact that half of engines crash before they make it to 400 hours whether mechanical failure or not is insane to me.
 
It's rare that the engine itself quits. It's failed magnetos, dirty/watery fuel, busted engine controls. Owners, especially, are famous for running components to failure.

In E-AB, a builder can overhaul his own engine. If the overhaul manuals aren't closely followed, failure is much more likely. Con rod bolts, for example, should not be reused. RTV has no business being used as a sealant for fuel or oil stuff. If there is a propstrike history, just about everything in there needs NDT.
 
I dunno. From 200 to 400 hours the increase was nearly 20%. And the sheer fact that half of engines crash before they make it to 400 hours whether mechanical failure or not is insane to me.

Again, that's half of the airplanes *that crashed*. Not half of all homebuilts with that engine.

Here's a plot I recently posted to another forum. It shows the number of Lycoming engines installed in the homebuilt fleet as of January 2023, both from active registrations and from de-registered airplanes:
Lycoming Engines.jpg
Almost 6,000 homebuilts have had one of the O-360 series installed in them. This does not include the ~3,300 homebuilts that don't list a specific engine type ("AMA/EXPR"). About a quarter of those are production-type engines. I think we can safely round it up to 6,000 airplanes.

Out of those 6,000 homebuilts, there have been 38 cases of accidents caused by an O-360-series engine mechanical failure...over a 22-year period. Less than two per year.

As far as "crashing before 400 hours," remember, that MOST homebuilts don't fly as much as production airplanes. The FAA General Aviation Survey estimates homebuilts fly about 47 hours a year. My own estimate based on accident report data is about 50 hours per year. Using the same process on my ~3300 Cessna 172 and 210 accidents, I get 250 hours a year. The AVERAGE Cessna 172/210 in that database has 5,200 hours, compared to 355 hours for the homebuilts.

A typical GA airplane with a thousand hours is considered almost new. A homebuilt with a thousand hours on it is admitted to a special parking area at Oshkosh, and no doubt has the aircraft equivalent of walkers available. Shoot, my Fly Baby was completed in 1980, and, other than a few repair periods, has been in continuous use since. It's got ~800 hours, of which ~700 were flown by me.

So *most* of the homebuilt accidents you see are going to be low-time aircraft, at least by the standards of production-type GA aircraft.

Ron Wanttaja
 
Yeah, I understand that I’m not understanding the data. But I’m still not understanding the data. Lol
 
Back
Top