Piston engine reliability (questions for high time pilots and mechanics)

[Capt. Geoffrey Thorpe]

Really? If you land in a field due to engine failure, there is no NTSB report?

Depends.

When this airplane (N1976W) was totaled following a landing on rail road tracks, it made the NTSB: http://www.ntsb.gov/aviationquery/brief.aspx?ev_id=20001211X12866&key=1

But I find no record of the same airplane landing in an onion field a few years earlier when a student ran out of gas on a cross country. ("I asked them to fill it, but I didn't check.") No damage, and the airplane was flown out of the field the same day.

 

[Henning]

The carb heat inlet has no filter or screen on a Piper Cherokee. This is not the normal induction air inlet. The same is true on at least some Cessnas. This is a design condition, not a maintenance condition.

They should be sealed shut when the system is not in use. If the carb heat is turned off, a wasp should not be able to enter the system. If it can, that is a maintenance issue. Take care of it or it will cost a couple thousand dollars for a crank on overhaul.

 

[Henning]

Really? If you land in a field due to engine failure, there is no NTSB report?

Yep, possibly anyway. Read the reporting requirements, they are on the PP written, you need to know them. Report back when you find them.

 

[Gmaper]

The number of GA deaths per million flight hours, is 22.43.

This means once you have 2000 flight hours, your odds of dying are. 4.5% (this incudes pilot error as well).

If we asume that 2/3 of those are pilot error, we are down to 3%

I am 43 years old. My guess is it will take me 20 years to get to 2,000 hours. So 3% is only 6 times higher then my odds of dying from natural causes in that time frame anyway.

I like those odds

I'm easily confused. How do you arrive at 4.5% after 2,000 hours. The way I read the numbers is that if we divide the 1,000,000 by 22.43 unlucky people then every 44,583 hours one person dies. I see that 2,000 hours is about 4.5% of the 44,583 but I don't understand how that applies to me, one of many thousands of pilots, hopefully not one of the 22.43. But I'm the last person that should ever be asked to explain statistics.

 

[DouglasBader]

Really? If you land in a field due to engine failure, there is no NTSB report?

Notification is required for sustained loss of power on two or more engines in flight: NTSB 830.5(a)(7)(ii).

Failure or damage to a single engine isn't even considered substantial damage, according to NTSB 830.2.

Landing off field? People land other places than airports all the time.

The carb heat inlet has no filter or screen on a Piper Cherokee. This is not the normal induction air inlet. The same is true on at least some Cessnas. This is a design condition, not a maintenance condition.

No, leaking of unfiltered air through your carb alternate air inlet is a maintenance problem, and it's robbing power in your engine, and enriching your mixture unnecessarily. It also allows contaminants past the normal induction and filter.

 

[Mafoo]

I'm easily confused. How do you arrive at 4.5% after 2,000 hours. The way I read the numbers is that if we divide the 1,000,000 by 22.43 unlucky people then every 44,583 hours one person dies. I see that 2,000 hours is about 4.5% of the 44,583 but I don't understand how that applies to me, one of many thousands of pilots, hopefully not one of the 22.43. But I'm the last person that should ever be asked to explain statistics.

Well, not sure my logic is correct, but how you described it, is how I arrived at the number.

The flaw in that number, is to asume it's a constant, and that everyone has an equal change of being the unlucky individual. We know that's not true.

We know for every 44,583 hours, someone dies. Let's turn it around, and make it a lottery. We are selling 44,583 tickets, and only one person wins.

If you only bought 1 ticket, you would have a 1/44,582 chance of winning. However if you bought 2,000 of them, you have a 4.5% chance of winning. This thought process is where I came to the number.

I realize it's flawed, but you have to come to a number somehow to quantify risk, so it's the one I chose

 

[Gmaper]

Well, not sure my logic is correct, but how you described it, is how I arrived at the number.

The flaw in that number, is to asume it's a constant, and that everyone has an equal change of being the unlucky individual. We know that's not true.

We know for every 44,583 hours, someone dies. Let's turn it around, and make it a lottery. We are selling 44,583 tickets, and only one person wins.

If you only bought 1 ticket, you would have a 1/44,582 chance of winning. However if you bought 2,000 of them, you have a 4.5% chance of winning. This thought process is where I came to the number.

I realize it's flawed, but you have to come to a number somehow to quantify risk, so it's the one I chose

But there are thousands of pilots, not just 44,582. So I don't think the 4.5% works for me because I intend to be one of the many, many thousands that doesn't buy one of those 44,582 tickets.

 

[Mafoo]

But there are thousands of pilots, not just 44,582. So I don't think the 4.5% works for me because I intend to be one of the many, many thousands that doesn't buy one of those 44,582 tickets.

Every hour you fly, your buying another ticket.

Again, the logic is flawed, because a lottery ticket each has an equal chance of winning, and you can control the safety of each hour you fly, so they are not quite the same.

 

[Mafoo]

But there are thousands of pilots, not just 44,582. So I don't think the 4.5% works for me because I intend to be one of the many, many thousands that doesn't buy one of those 44,582 tickets.

As for the numbers... if we had 89,000 pilots, and each flew an hour, sadly, two of them will "win".

 

[LJS1993]

LJS,

insufficient data in your posting. Are you saying that in 250,000 miles, less than a TBO run in many aircraft/engine combos, you have never had the engine quit once, even momentarily?

Paul

In this motor I have not had to get the cylinder heads looked at, gaskets changed, timing belt swapped, fuel injection unit rebuilt, etc...... I've had the original fuel pump go out on me but that was entirely my fault since I ignored the warning signs of its impending doom. I had an alternator issue which I took care of and have had to replace one idler pulley. Overall from what I've learned around here about airplane engines is they are very maintenance heavy and seem to be rather delicate in terms of oil usage and changing. For their comparative usages I would say auto engines are doing a much better job.

 

[LJS1993]

It doesn't matter. An auto is an auto, and an airplane is an airplane. GA aircraft engines & auto engines are always discussed at length, within "experimental" aircraft forums, where actual results from using auto engine conversions are well known. More than not, the auto conversion never stands up to the reliability or performance of a purpose built GA engine. Some builders have success, but it's a lot of additional engineering & experimenting. In other words, if you think a modern auto engine is a well suited candidate for an aircraft, then you had better prepare to do a lot of research.

L.Adamson

I not once stated an auto engine would be suitable in an aircraft application. For their intended applications auto engines seem much more durable and much less maintenance heavy.

 

[olasek]

For their comparative usages I would say auto engines are doing a much better job.

Usage is a key word.
Automotive engines have rather an easy life compared to aircraft engines, majority of time they operate significantly below the red line, probably around 30% of power on the average, unlike aircraft engines that operate very close to red line majority of time. It is like comparing apples to oranges. Also the cooling system is very often the weakest part in automotive engines, how many times have I had problems with the cooling system, I would never want an automotive, liquid cooled engine in my aircraft.

 

[Henning]

Usage is a key word.
Automotive engines have rather an easy life compared to aircraft engines, majority of time they operate significantly below the red line, probably around 30% of power on the average, unlike aircraft engines that operate very close to red line majority of time. It is like comparing apples to oranges. Also the cooling system is very often the weakest part in automotive engines, how many times have I had problems with the cooling system, I would never want an automotive, liquid cooled engine in my aircraft.

Engines are not humans, they do not get tired or require rest, how long an engine runs is irrelevant, an engine is perfectly happy to operate at 100% power continuously until the bearings wear out, this is provided the inner cylinder pressures(ICP) and temperature are kept below the material values of the components. Comparing percent power of operation is also an incomplete comparison without considering the Power to Displacement ratios, compression ratios, and maximum ICPs.

The operating environment of an auto engine is far more abusive than an aircraft engine. Engines like steady states with lots of load and this is where aircraft engines operate. They also run with very low compression ratios comparatively in order to control the ICP. What causes the most expense and failure in aircraft engines themselves is that they get too much fuel. ROP operations are harder on your engine than anything. Not enough ROP and your ICPs skyrocket with your CHT. Use enough fuel to bring those under control and you start washing out the rings, even more and you glaze the cylinders (most people like their CHT far too cold IMO, anything under 390 when ROP is a ring problem waiting to happen)? The primary advantage auto engines have is that the operator can't control the mixture, the next is water cooling.

 

[brcase]

Eh not too sure about that one. My 88 Lincoln Mark 7 with a 5.0HO has about 250,000 miles on the original motor. Automobile engines have a lot longer lifespan and overall seem a LOT more dependable then the motors offered in your typical GA aircraft. If I am mistaken someone please explain to me my error in this statement.

I would guess your Lincoln burns an average for 3 gal per hour with and average HP setting of about 45hp, so maybe 20% power.

Seems like you have been driving about 10,000miles per year or about 166 hours per year.

If you run an aircraft engine like this I think it will likely last as long or longer.
I also think only a small percentage of 1988 cars last as long as yours has.

Brian

 

[Jaybird180]

Electric:
First, you are the weak link in the plane, not the engine...
Second, I'm not sure you are cut out for being a pilot - that is a whole lot of fear I'm smelling..

Aircraft piston engines are as reliable as we can make them, with a 100 years of experience status quo, detailed record keeping, boards of inquiry when they do fail, and directives from the FAA for correcting identified problems... Maintained by a competent mechanic they are vastly more reliable than the pilot...

FTFY

You mean pencil whip the maintenance, overfly AD's, and get away with anything possible? That kind of maintaining it like the airlines?

So that really happens - like my professor said???

 

[Jim Logajan]

Well, not sure my logic is correct, but how you described it, is how I arrived at the number.

The flaw in that number, is to asume it's a constant, and that everyone has an equal change of being the unlucky individual. We know that's not true.

We know for every 44,583 hours, someone dies. Let's turn it around, and make it a lottery. We are selling 44,583 tickets, and only one person wins.

If you only bought 1 ticket, you would have a 1/44,582 chance of winning. However if you bought 2,000 of them, you have a 4.5% chance of winning. This thought process is where I came to the number.

I realize it's flawed, but you have to come to a number somehow to quantify risk, so it's the one I chose

Your math is a reasonable approximation so long as the span of time you want to estimate for is small, but fails for larger spans. In your scenario buying all 44,582 tickets yields a 100% probability of winning. But you probably realize that you can fly 44,582 hours and not die.

The usual method (assuming equal probabilities for all hours, a generally invalid assumption) is to multiply together all the probabilities of the opposite result occurring, to get the net probability. So in one hour your probability of dying is 1/44,582 = 0.000022431 (0.0022431%). That means the probability of surviving that hour (the opposite result) is 1 - 0.000022431 = 0.999977569 (99.9977569%).

Your probability of surviving 2000 hours is 0.999977569^2000 = 0.9561 (95.61%) The opposite result of dying is 100 - 95.61 = 4.39%. Your result of 4.5% isn't far off.

But at 44,582 hours the probability of your surviving is 0.999977569^44,582 = 0.3679 (36.79%). The opposite result of dying is 100 - 36.79 = 63.21%. Your result of 100% dying is pretty far off, and the approximation method yields nonsense with any number over 44,582 hours.

Anyway, hope this helps for those whose recollection of statistics is rusty or wasn't something you ever had to learn.

 

[bobmrg]

Yes, they come to mind. Plus, numerous pilots at Kenmore Air Harbor. And, Dr. Callison, my friend and AME.

Mine as well. If he were still alive I would still have a first-class medical.

Within days of my going to work at Seattle Flight I was attending the funeral of an instructor who had gone in off Vashon in a Twinkie...we always thought that the student froze on the controls.

Bob

 

[Neal Howard]

Get a carb temp gauge, they aren't expensive and if you have an engine analyzer (which you should) most models accept an input for carb temp. This not only tells you when the conditions are hazardous for icing, it also let's you trim the carb heat and/or mixture to get you just clear without giving up more power or efficiency than you need to.

No engine analyzer... I've got an "old school" panel. Don't really have room for adding any more engine instruments either without cobbling up the looks of the panel. Some day, I'd like to upgrade to a glass panel, but I don't have the money for that any time soon.

 

[Henning]

No engine analyzer... I've got an "old school" panel. Don't really have room for adding any more engine instruments either without cobbling up the looks of the panel. Some day, I'd like to upgrade to a glass panel, but I don't have the money for that any time soon.

That won't happen living around Wichita Falls. You at Kickapoo or the Valley?
Al Martin still alive?

 

[Neal Howard]

That won't happen living around Wichita Falls. You at Kickapoo or the Valley?
Al Martin still alive?

I'm at the Valley. Al Martin passed away a couple months ago. http://www.timesrecordnews.com/news/2012/jun/01/al-martin/

 

[DouglasBader]

So that really happens - like my professor said???

If your professor said so, he or she was correct.

 

[Henning]

I'm at the Valley. Al Martin passed away a couple months ago. http://www.timesrecordnews.com/news/2012/jun/01/al-martin/

Bummer, he was a good guy to work for, drilled and drove a hell of a lot of rivets with him. I remember when we moved everything from Piggly Wiggly Aircraft in Nacona out to the building in Electra. What's happenning with all that stuff, you know?

 

[Neal Howard]

What's happenning with all that stuff, you know?

Don't really know. Last I'd heard was that it was all up in Oklahoma... Grandfield and some other small town airport. I reckon that his business partner probably ended up with all of it.

 

[Henning]

Don't really know. Last I'd heard was that it was all up in Oklahoma... Grandfield and some other small town airport. I reckon that his business partner probably ended up with all of it.

Never knew him to have a partner except his cousin, I forget her name, and she just did upholstery. Oh well, there were a few treasure parts in those piles of metal.

 

[Tom-D]

It seems like everyone that has more than 2000 hours in piston aircraft had at least one engine failure. Is that really the case?

no, most of the engine failures I have experienced were shut down and feathered due to a chips light, or a fire warning light, and were shut down to prevent a casitrophic
material break down.

I have only had one engine quit on me, that was a C-85 equipped with a old Tillitson carb that had no accelerator pump installed on a Fisher celebrity when I stabbed the throttle in close to the numbers it quit, resulting in a hard landing. My bad, adding the throttle too quickly made the engine quit.

 

[BurtM]

It seems like everyone that has more than 2000 hours in piston aircraft had at least one engine failure. Is that really the case?

No.

I'm a fairly new pilot (170 hours) and it seems rather discouraging to know that if I plan on flying piston singles for a long time I should expect at least one deadstick landing in the future. Especially, considering that 1 out of 10 NTSB reports due to engine failure involves a fatality.

I think you are misinterpreting accident stats. But if that's what you want to hang your hat on, and the odds seem bad, then you should probably stop flying.

Would an engine that is well cared for vs the engine on a rental aircraft be significantly more reliable?

What's the difference? Realistically, you're probably less likely to have an engine failure with a well-used rental than anything that sits in a hangar all week long.

How well can you manage the risk...

The best way to manage the risk is to realize that 8 times out of 10, you're going to be the problem and not the engine, or anything else mechanical on the plane.

 

[peppy]

But for some reason FADEC seems very slow in making its way to piston aircraft. Is there any factory new Cessna or Piper piston with FADEC, or a Cirrus, I don't think so.

There is one Cirrus with a FADEC retrofit but no factory new installations. The Continental system is expensive and adds weight but seems to work well. here have been several planes with FADEC standard including Diamond diesels and Liberty XL2.

 

[Velocity173]

There is one Cirrus with a FADEC retrofit but no factory new installations. The Continental system is expensive and adds weight but seems to work well. here have been several planes with FADEC standard including Diamond diesels and Liberty XL2.

Yeah good luck in finding someone that can work on FADEC. The Liberty I was renting finally sold because no one in the area was familiar with it.

 

[Jaybird180]

There is one Cirrus with a FADEC retrofit but no factory new installations. The Continental system is expensive and adds weight but seems to work well. here have been several planes with FADEC standard including Diamond diesels and Liberty XL2.

A little OT, but I saw a print ad for a Cirrus STC that adds the blue knob

 

[kmead]

In the last 11 months I have gone on the road and installed engines in 3 downed aircraft. All three lost their oil suddenly in flight. Normally a loss of oil in the flight levels results in a total loss of the engine. Almost every time it will be a defective turbo or a broken oil control ring on a cylinder on my type of ships. There has been some broken crankshafts and departing cylinders too. One just lost it's turbo today! The training level on our aircraft is very good now and most have landed on an airport. This makes my job easier.

In way to many cases the pilot or mechanic had signs that something was really wrong on the last flight before the big event. Like- 1) I had to add 9 qts of oil before I left. 2) I heard a whine in flight that changed with the mixture. 3) The oil analysis showed no problem. 4) I thought it was normal for fuel to drain from that tube 5) My favorite- We never look at the filter, I didn't know it had a suction screen.

It comes down to looking at trends, something a renter doesn't have the option of. My rule is- if it's changed and is different from the last flight, don't fly and have it checked out. The engine monitor does help at times, but many times it doesn't present until after the failure. I have had aircraft arrive with dead cylinders and nothing changed on the engine monitor.

With reciprocating and turbine engines, they will only be as reliable as the level of maintenance attention as they get. Also the pilot should always be looking for changes in operation and performance. Always think about loss and be proficient with engine out procedures.

I sure like to rant. I operate two reciprocating aircraft. Neither has had an engine failure (yet). My wife has good confidence in them, mainly because she has seen how some of the planes come in with parts missing or broken with no operational issues. I hope I can keep cool if it ever happens to me. I had an owner that the engine seized because of bearing failure. He broke out near minimums and made the runway.

 

[tiger]

Guys I would take another approach.
The fatalatiy rate per 100,000 miles is ten times that of the automobile. However:
1/3 is ran it out of gas
1/3 is weather that you were not equipped nor qualified for.
~1/4 is cocamamie stuff you can eliminate on preflight.

Do the math. This can be as safe as an automobile, if you make it so.

True, Dr Bruce, and the number of those types of accidents is depressing. But if you want to knock out the 'dumb' flying accidents you could also knock out the 'dumb' driving accidents. The stats above really just mean that the 'careful pilot' can be as safe as the 'average driver'.

However, you can 'de-risk' the driving too - don't drive after consuming alcohol, avoid cellphone use, drive appropriately to the conditions, don't be young... In other words, if you're a careful operator of both types of vehicle (which I expect most people here attempt to be) I think you'll still be running more risk per mile in the airplane.

 

[Henning]

Originally Posted by bbchien
Guys I would take another approach.
The fatalatiy rate per 100,000 miles is ten times that of the automobile. However:
1/3 is ran it out of gas
1/3 is weather that you were not equipped nor qualified for.
~1/4 is cocamamie stuff you can eliminate on preflight.

Do the math. This can be as safe as an automobile, if you make it so.

True, Dr Bruce, and the number of those types of accidents is depressing. But if you want to knock out the 'dumb' flying accidents you could also knock out the 'dumb' driving accidents. The stats above really just mean that the 'careful pilot' can be as safe as the 'average driver'.

However, you can 'de-risk' the driving too - don't drive after consuming alcohol, avoid cellphone use, drive appropriately to the conditions, don't be young... In other words, if you're a careful operator of both types of vehicle (which I expect most people here attempt to be) I think you'll still be running more risk per mile in the airplane.

If if was a skiff we're all be sailing... you just can't fix stupid.

 

[bbchien]

True, Dr Bruce, and the number of those types of accidents is depressing. But if you want to knock out the 'dumb' flying accidents you could also knock out the 'dumb' driving accidents. The stats above really just mean that the 'careful pilot' can be as safe as the 'average driver'.

However, you can 'de-risk' the driving too - don't drive after consuming alcohol, avoid cellphone use, drive appropriately to the conditions, don't be young... In other words, if you're a careful operator of both types of vehicle (which I expect most people here attempt to be) I think you'll still be running more risk per mile in the airplane.

Maybe. But the unmitigatable risk is the OTHER guy doing drugs, texting, drinking....and driving.

 

[Dan Thomas]

Yeah good luck in finding someone that can work on FADEC. The Liberty I was renting finally sold because no one in the area was familiar with it.

I bet the owners of Model Ts in 1910 had the same problem.

But will GA survive and grow like the auto industry did back then?

Dan

 

[tiger]

Maybe. But the unmitigatable risk is the OTHER guy doing drugs, texting, drinking....and driving.

Agreed. Pilots are very rarely taken out by another pilot.

I was horrified when I saw the number of DUI arrests in my small, suburban city every day - assuming that only a tiny fraction are caught, there must be an awful lot of impaired drivers out there.

 

[Graueradler]

They should be sealed shut when the system is not in use. If the carb heat is turned off, a wasp should not be able to enter the system. If it can, that is a maintenance issue. Take care of it or it will cost a couple thousand dollars for a crank on overhaul.

The Piper Cherokee carb heat valve is a horizontal flapper valve on the bottom of the carb. The top of the flapper is always directed toward the selected source. Select carb heat on run up. Nest breaks loose, slides down scat tube and ends up on top of flapper valve. Now de-select carb heat. mud dauber nest is still on top of flapper but is now on the normal induction side. No maintenance deficiency in the system, just a design deficiency. Now apply full power for take off and velocity of air flow carries nest into carb throat. When carb heat is turned off, the system is filtered and a wasp or wasp nest cannot enter the system.

 

[Graueradler]

No, leaking of unfiltered air through your carb alternate air inlet is a maintenance problem, and it's robbing power in your engine, and enriching your mixture unnecessarily. It also allows contaminants past the normal induction and filter.

I'm not talking about leaking from the carb heat side while the normal induction side is selected. I'm talking about migration to the top of the flapper valve by sliding down the scat tube while carb heat is selected during run up. When the flapper valve is returned to normal induction, what is on top of it migrates to the normal induction side.

 

[Electric]

IMO, anything under 390 when ROP is a ring problem waiting to happen)?

Interesting, I always thought you should keep the CHT below 380, even in climb. What should be the top of the CHT range? Also, what about LOP?

3) The oil analysis showed no problem. 4) I thought it was normal for fuel to drain from that tube 5) My favorite- We never look at the filter, I didn't know it had a suction screen.

Could you elaborate on those. Which tube are you referring to? What's a suction screen? And what's wrong with the oil analysis showing no problems?

 

[DouglasBader]

I'm not talking about leaking from the carb heat side while the normal induction side is selected.

I know what you're talking about. It doesn't change my comments in the least.

 

[Henning]

Interesting, I always thought you should keep the CHT below 380, even in climb. What should be the top of the CHT range? Also, what about LOP?



Could you elaborate on those. Which tube are you referring to? What's a suction screen? And what's wrong with the oil analysis showing no problems?

Because you start washing out the rings at the bottom of the cylinder which leads to premature failure. LOP eliminates the problem because it eliminates the excess fuel. I've been an engine guy since I was a kid, and when I got to airplanes at first I couldn't understand why I was being told to operate engines in such an inefficient destructive manner especially when I had a control to fine tune the mixture at my hand. A few hours in I helped an old guy hang a 520 on his Navion and went flying with him and saw him operate in what I thought was a proper and efficient manner so I asked him about it. He gave me the history of it from the war, the whole Lindbergh Doctrine and how the whole thing changed during the war and said he was also confused how people had lost the lessons learned there. Since most all the "LOP will fry your engine" came from mechanics and that most pilots were mechanically naive, we surmised that it was due to the rising costs in college tuitions and the mechanics need to bring in work to pay it. That made him laugh so hard he used to take me on any flight he made and I was available for.

I'm not Ken but I'll make a guess on the rest. The tube is the one that is the tube that drains the fuel overboard from a plural space in the injector pump should the seal fail, it's purpose is to prevent fires in the instance of seal failure. The suction screen is on the fuel inlet, if you don't clean it during inspections, it may clog. Not all fuel starvation problems are running out of fuel. If the path from the tank to the engine gets clogged you can starve the engine with full tanks. Lastly oil analysis is but one tool to monitor engine health, but some rely on it to do more than it can.