Likelihood of engine failure

[steingar]

I could truthfully answer yes, I had carb ice form on a clear day and shut down the engine.  I was on final approach to an airport when when the carburetor heat happily brought it back to life.

Fairly obvious sampling error here, just about all the respondents to this "poll" have have engine loss.  That said Ron Wattanja does accident analysis for the EAA, inches dataset engine and mechanical failures are still a minority cause of crashes.  Miscontrol is still number 1, usually associated with weather.  Running out of gas is still up there too.

[Shadrach]

Obviously not a very reliable statistical method. That being said,  I know a lot of pilots that have had failures, it might even be close to half of the pilots I know.  The majority were able to dead stick at an airport.  If you do it long enough, something is bound to happen.  Flying well maintained and modern aircraft probably helps (In this case modern means post WWII).  I have seen strange issues with obscure machines maintained with whatever collection of NOS parts could be found to keep them going.  Things like crank flange separation in flight (bye bye prop) or a broken connecting rod.  One thing is for sure. A great deal of the failures that I am personally aware of never made it in to any sort of accident database.

[neilpilot]

31 minutes ago, chrixxer said:

Yeah, that too got me worried about it. And the fact that the A&Ps I've talked to all zeroed in on the servo being the likely culprit. Seems to be a "not unusual" point of failure. Is there anything we can do to guard against such issues? Have it checked every annual / 100 hour? Any pre-flight checks that can be done (unlikely to be anything feasible)?

I had intermittent issues for 3 years prior to the final event, usually months apart, maybe 2-3 incidents (typically unexplained rough engine operation) each year).  In all cases, while I found some minor issues (fouled plug, speck of dirt in an injector), I never really solved the problem. 

That's why I had my servo R&R.  I was averaging 80-100 hr/yr and 100-hr/annuals and pre-flight checks didn't reveal the problem.  Even during the post-accident servo bench test arranged by NTSB , fuel flow was initially in spec before intermittent flows were observed.  That led to an internal inspection and discovery of the blockage. 

I'm convinced, but have no proof, that that floating piece of o-ring is what caused very intermittent issues for years before I finally got unlucky.  In July 2012 the piece of rubber, which was likely floating around for years, moved into a position to essentially shut down fuel flow.

[David Lloyd]

Had one complete engine failure while flying.  70 Hours on a new engine and fuel pressure line from the distributor to a bulkhead fitting on the firewall came off.  Went from rough to nothing in maybe 15 seconds. The previous week was a trip to Pittstown Point.  Mechanic shot.  Twice.

Early morning, dark, I had a magneto that was not tight, slip causing some funky timing.  Did this 2-300 feet off the runway.  Must have either looked or sounded pretty intense as CLT tower cleared me to land any runway without my saying a word.  Engine just had magnetos timed.   Mechanic beaten.

Had a cylinder remove itself from the crankcase one night between the runway and hangar after a trip to and from MGM.  A few hundred hours before, the same cylinder was removed to replace a broken exhaust stud.  No other cylinder was ever off that engine.  Wonder if that was more than coincidence.  About 1950 hours on that engine.  

Had a valve stick and bend a pushrod bad enough the pushrod tube broke and was dumping oil.  Pushing the plane back after a trip home from LAL I looked down and wondered who spilled a quart of oil in front of my hangar.  Oh, it's me. It was down to 5 quarts from 12.  About 800 hours on that engine.  Sometimes stuff just happens.

So, I have become a little suspicious of work done on my airplanes.

[ArtVandelay]

1 hour ago, chrixxer said:

Yeah, that too got me worried about it. And the fact that the A&Ps I've talked to all zeroed in on the servo being the likely culprit. Seems to be a "not unusual" point of failure. Is there anything we can do to guard against such issues? Have it checked every annual / 100 hour? Any pre-flight checks that can be done (unlikely to be anything feasible)?

But if they have to disassemble it and risk causing problem in reassembly, you are better off leaving it alone.

[Hank]

2 hours ago, David Lloyd said:

Mechanic shot.  Twice.

Hmmm . . . . Seems to me that your Marksmanship instructor didn't beat you enough. Unless your mechanic was ducking, juking and weaving, two shots should not be required.

Glad you are still with us, and may this engine hold together a little better!  

Edited July 13, 2018 by Hank

[David Lloyd]

1 hour ago, Hank said:

Hmmm . . . . Seems to me that your Marksmanship instructor didn't beat you enough. Unless your mechanic was ducking, juking and weaving, two shots should not be required.

Glad you are still with us, and may this engine hold together a little better!  

I was still upset after the first shot.  And he deserved another.  Oh, and it was his second attempt on me.

Edited July 13, 2018 by David Lloyd

[jetdriven]

On 5/17/2018 at 11:33 AM, aviatoreb said:

Well - 3 weeks ago I would have answered no.

Now following my excitement about 2.5 weeks ago - a complete loss of power at 16,500 ft with a happy ending of a dead stick landing at an airport - now I will answer yes.

A (much more senior professor than me) mechanical engineer whom I know and respect, at Cornell, who owns and flies a C182, once said to me that he figures "something" happens roughly every 1000hrs of pilot time.

I must be in the minority as I've had a power reduction momentarily which caused me to find a new place to land twice in 3000 hours of piston time. Nothing more than that.. I've read the statistical liklehood of engine failure from a certified, conforming engine is 1 every 25K hours. Poorly maintained or operated airplanes, I suspect its much more.

[aviatoreb]

3 hours ago, jetdriven said:

I must be in the minority as I've had a power reduction momentarily which caused me to find a new place to land twice in 3000 hours of piston time. Nothing more than that.. I've read the statistical liklehood of engine failure from a certified, conforming engine is 1 every 25K hours. Poorly maintained or operated airplanes, I suspect its much more.

I wonder how that statistic was developed and how they define nonconforming?  In the case of my engine failure - the turbo failed.  Is that an engine failure according to that 25k statistic?  What if the fuel pump fails?  Would they have counted that as an engine failure?  What about fuel contamination?  Does that count in their statistics?  They may well all not be classified as engine failures in terms of counting stats for the engine itself but clearly the NTSB would count them as engine failures if a crash occurred. I think the real statistics as an engine failure must be devilishly difficult to accumulate real meaningful statistics, and I say that as a math professor in my day job.  Note that my engine failure did not result in a crash (thank goodness!) so while air traffic control was fully appraised of the situation as I declared an emergency, I doubt the incident was ever counted in any statistics.  And likewise the engineering approach to such statistics computation may end up excluding likely failure modes, like fuel pump problems, fuel contamination problems, and even turbo problems or the fuel pressure line from the distributor to a bulkhead fitting on the firewall coming off, etc.

P.S. The phrase "poorly maintained" may well refer to maintenance errors but by excluding the actual errors made by well meaning well trained professional mechanics from the statistics is almost like trying to make a statistic that says, "If we assume that nothing is wrong what is the probability that something is wrong?"  Mechanics are humans and they make errors at a given probability rate.  Presumably excellent mechanics make errors at a much less frequent rate that poor mechanics. (same with doctors, and any other professional).  Same with pilots - even great pilots make errors,  Hopefully most well trained pilots make errors at a minuscule rate but still it is just not true that humans exist that never make errors.  So what does the phrase "poorly operated" or "poorly maintained" really mean if trying to state statistics meaningfully?

Edited July 15, 2018 by aviatoreb

[N201MKTurbo]

Back in the 80s there was this nice lady Barbara. She was in her early 50s and decided to learn to fly. She had 4 engine failures that resulted in off airport landings before she got her private. This was in three different C150s. People were afraid to rent her an airplane!

All the engines had hard failures and all agreed that she had nothing to do with it. We made her a trophy.

[chrixxer]

On 7/13/2018 at 3:15 PM, Hank said:

two shots should not be required.

Rule 2.

[steingar]

The  analyses I have seen discussed loss of power.  All this said, it is very possible that lots and lots of pilots have experienced engine loss.  Not every loss of power results in a reportable accident.

[RobertGary1]

49 minutes ago, steingar said:

The  analyses I have seen discussed loss of power.  All this said, it is very possible that lots and lots of pilots have experienced engine loss.  Not every loss of power results in a reportable accident.

For sure. Non of my engine outages resulted in ntsb reporting even though one was a fsdo investigation When you come in deadstick opposite direction and they start holding jets for you the fsdo gets involved. Once they saw the engine failure wasn’t caused by a missing gear AD or unapproved cigarette lighter they lost interest  

-Robert 

[jetdriven]

On 7/13/2018 at 1:41 PM, neilpilot said:

The exact same comments would apply to N1310W, the 64E that was totaled in 2012.  The NTBS found a small piece of torn o-ring had blocked flow in my fuel servo.  The event occurred 130 flight hours and 25 months after the fuel servo R&R.
 

That’s what lawsuits are for. Totally justifiable too.