Auto Engines Aren't Designed to Take Full Power for More Than a Few Minutes...

[Hephaestus]

It fits the scenario that by the time the investigator gets around to opening the cowl and checking things out, the ice has melted. So, another "unexplained" loss of power.

It bugs me I really wish they would would tweak those categories to include a "probably carb ice" if it looks like a duck and quacks like a duck... Unexpected power loss, wx favorable to icing - no other visible cause... Call it probable icing.

The fact that the traditional aero engines are double the icing rate of the others says it all. We see the attitude here regularly "I live in the south it won't happen, I don't need carb heat"

 

[cheapracer]

In my opinion, it's certainly possible to make an automobile conversion that's just as reliable as traditional engines. But what statistics tells us is that the average homebuilder can't achieve that goal.

I don't think there's any surprises there, the quality and knowledge of what they are doing varies considerably, and there is no authority to check their work.

I was going to ask you about the cores, thanks for the answer, but I would like to see VWs separated from the rest of the auto conversions, with the presumption that the bulk of the others are somewhat more modern engines, and water cooled. Is that available to you?

 

[cheapracer]

"Can't?" or "Didn't?" It seems to me that it is possible, with a lot of attention to detail and (especially) careful attention to what has worked for others in practice.

I started life out repairing cars, motorcycles, lawnmowers, "you name it" vehicle types.

The amount of 'home fixes' I have seen simply tell me that Ron's statistics on auto conversions are no where near as bad as I thought they would be purely based on the varying levels of workmanship I have seen over the years.

 

[Wanttaja]

Funny unknown is so high and carb ice is so low. That doesn't seem to fit with the GA narrative...

The sound you hear is me grinding my teeth about "The GA Narrative". 1/2

There's a lot of myths about accident causes. One of these days, I'll probably write an article about them.

However, back to your point: "Undetermined" is so high because homebuilts are not a priority for the NTSB. The reason they investigate accidents is to attempt to identify trends that may lead to changes that reduce the probability of the same type of accident happening again.

Homebuilts are, technically, one-of-a-kind aircraft. The NTSB doesn't see a lot of benefit in spending a lot of time on every homebuilt accident. So about 10% of all homebuilt accidents are shown as "Undetermined Power Failure" vs. about 4% of Cessna 172s.

Ron Wanttaja

 

[BoKu]

Gotta watch out for that "undetermined." It's a killer.

 

[Wanttaja]

Funny unknown is so high and carb ice is so low. That doesn't seem to fit with the GA narrative...

It fits the scenario that by the time the investigator gets around to opening the cowl and checking things out, the ice has melted. So, another "unexplained" loss of power.

And production-type aircraft have coolers under the cowl that keep the carb ice from melting....?

"Can't?" or "Didn't?" It seems to me that it is possible, with a lot of attention to detail and (especially) careful attention to what has worked for others in practice.

Like I said, they *can* be as reliable. Can.

The problem is, if you want a, say, Subaru conversion to be as reliable as, say, a Continental C-85, you have to spend the same money it would take to buy a low-time Continental.

In my opinion, the problem lies in the unending search for a "cheap" aircraft engine. Nothing more expensive than an unreliable airplane powerplant.

As for "Can't" vs. "Didn't"? Last time I did a full analysis was about ten years ago. Since then, the auto-engine statistics have gotten worse....

Interesting that the "other" four strokes (non aircraft, non auto) have about 1/2 the internal failure rate of certified engines. I'm not sure what those engines are, but maybe the little lawnmower engines are as tough in the air as they are on the turf.

What I call the "Non-Certified Four Strokes" are primarily the Rotax 912 and similar. The're about 2/3rds of the engines in the category in my accident database.

Ron Wanttaja

 

[Wanttaja]

Gotta watch out for that "undetermined." It's a killer.

And all too often, they restart the engine after the accident and it runs normally.....so it ain't just mechanical issues.

Ron Wanttaja

 

[Hephaestus]

I just have an issue there, personal issues I guess

When the Lycoming/continental show double the rate of icing compared to others... But the undetermined is so **** high...

If there's engine power loss, risk of carb ice is high, call it "likely carb icing" Put an actual number to it...

 

[Wanttaja]

I was going to ask you about the cores, thanks for the answer, but I would like to see VWs separated from the rest of the auto conversions, with the presumption that the bulk of the others are somewhat more modern engines, and water cooled. Is that available to you?

Here's a look at it in tabular form.

Failure

Undetermined

Engine Internal

Fuel - Engine

Ignition

Drive system

Oil System

Carb Mech

Carb Ice

Cooling System

[TD1]Traditional Engines[/TD1][TD1]All Auto Engines[/TD1][TD1]VW Engines[/TD1] [TD1]39.8%[/TD1][TD1]35.7%[/TD1][TD1]37.8%[/TD1] [TD1]12.4%[/TD1][TD1]12.7%[/TD1][TD1]13.3%[/TD1] [TD1]9.0%[/TD1][TD1]5.4%[/TD1][TD1]3.3%[/TD1] [TD1]5.3%[/TD1][TD1]14.0%[/TD1][TD1]8.9%[/TD1] [TD1]0.8%[/TD1][TD1]7.2%[/TD1][TD1]3.3%[/TD1] [TD1]7.4%[/TD1][TD1]4.1%[/TD1][TD1]4.4%[/TD1] [TD1]8.4%[/TD1][TD1]9.0%[/TD1][TD1]20.0%[/TD1] [TD1]16.6%[/TD1][TD1]6.3%[/TD1][TD1]8.9%[/TD1] [TD1]0.3%[/TD1][TD1]5.4%[/TD1][TD1]0.0%[/TD1]

Engine innards seem pretty comparable. Looks like the biggest bugaboo with the VWs is the carburetor and carb heat.

Keep in mind that my "VW" column includes all VW-derived engines.

Ron Wanttaja

 

[Wanttaja]

I just have an issue there, personal issues I guess

When the Lycoming/continental show double the rate of icing compared to others... But the undetermined is so **** high...

If there's engine power loss, risk of carb ice is high, call it "likely carb icing" Put an actual number to it...

When I do my assessment of each accident, I read the narrative, not just the probable cause.

Remember, the investigator does not make the "probable cause" decision. He or she writes the factual report, and the board itself makes the final determination.

I've seen a number of accidents where the investigator says, "OAT was XXX, humidity was YYY, and the table shows a high probability of carburetor icing," and the Board actually states the accident was due to engine failure for undetermined causes.

When that happens, *my* database lists it as carburetor icing. I did a quick check, I've got 15 cases like this, out of about 100 cases where carb icing was a factor.

Also, keep in mind that not every builder installs carburetor heat, or installs *adequate* heat.

Ron Wanttaja

 

[BJC]

I've seen a number of accidents where the investigator says, "OAT was XXX, humidity was YYY, and the table shows a high probability of carburetor icing," and the Board actually states the accident was due to engine failure for undetermined causes.

The only surprise there is that the NTSB hasn’t recommended that the FAA issue a rule requiring that the laws of physics be changed to eliminate the possibility of carburetor ice.


BJC

 

[cheapracer]

Here's a look at it in tabular form.

Thanks for replying, appreciated.

 

[Daleandee]

It bugs me I really wish they would would tweak those categories to include a "probably carb ice" if it looks like a duck and quacks like a duck... Unexpected power loss, wx favorable to icing - no other visible cause... Call it probable icing.

It appears that they figured it out at least once ...

https://app.ntsb.gov/pdfgenerator/ReportGeneratorFile.ashx?EventID=20170422X03928&AKey=1&RType=Final&IType=LA

This was a very well built airplane and nearly an exact clone of the one I have. The MA3-SPA needs carb heat. I have it and I ain't afraid to use it!

Dale
N319WF

 

[Vigilant1]

The fact that the traditional aero engines are double the icing rate of the others says it all.

I don't think Ron's graph tells us that traditional aero engines have double the icing rate of the others. For each type of engine, the graph shows us the relative frequency of the major causes of power loss. But we can't use the chart to determine if icing (or anything else) is more common for one type of engine than another. For example, if a "traditional engine" quit running, then 16% of the time it was due to carb ice. If a non-traditional, non automotive engine stopped running, 8% of the time it was due to carb ice. But, if traditional engines only experience power loss (all causes) half as often as nontraditional 4-strokes, then they both experienced the same rate of power loss due to carb icing.

 

[BBerson]

I think engines need to be compared on the failure rate per 1000 hours. I doubt that data is available for homebuilts.

 

[Wanttaja]

I think engines need to be compared on the failure rate per 1000 hours. I doubt that data is available for homebuilts.

That would certainly give you hard data.

However, most folks are interested in comparing reliability between two types. We can assume, for instance, that a Glasair owner flies the same number of hours per year as a Lancair owner. Similarly, I would generally assume that a person with a Lycoming-powered RV-9 would operate it the same way as the person would if it had a Subaru engine.

However, my data does show that auto-engine failures occur a LOT earlier in the life of the aircraft.... Almost 20% of the accidents involving auto-engine homebuilts happen in the first ten hours.

I'm generally pretty calm when I read about dumb decisions in accident narratives, but it still gripes my wagger when a guy's having trouble getting the engine to run right on the GROUND, and decides that it would improve if he was in the air....

Ron Wanttaja

 

[BJC]

I know of an Eggenfeller Subaru equipped HBA that had a new engine fail (the core engine, not the PSRU) within the first three hours of flight, followed by another new engine, not the failed one rebuilt, that failed within its first ten hours of flight. Neither resulted IN damage to the airframe (both occurred over airports) and neither was reported to the FAA.

The airplane has had many hours of safe flight with the factory new Lycoming IO-360 that the owner installed after his very expensive and time consuming attempt to use a “modern technology” engine.


BJC

 

[cheapracer]

The airplane has had many hours of safe flight with the factory new Lycoming IO-360 that the owner installed

So it wasn't one of these 1,100 engines then?

https://www.aopa.org/advocacy/advocacy-briefs/regulatory-brief-lycoming-crankshaft-replacement

 

[BJC]

So it wasn't one of these 1,100 engines then?

https://www.aopa.org/advocacy/advocacy-briefs/regulatory-brief-lycoming-crankshaft-replacement

No, not one of those. But there were two conversion failures that are not in the FAA records, and, therefore, not in Ron W’s statistics.

The point is that there have been, and continue to be, lots of undocumented engine failures / forced landings, with and without damage, that the FAA has no records of. I have a friend who rebuilt his last airplane, a J-3, while in his early 90’s. He had five engine failures, all over the airport with no damage, before he finally got it right.


BJC

 

[TFF]

If the FAA is not sent out for a statement,no one knows. There was a guy near here that flew his Cessna 170 off his farm for a couple of decades. No license, no annuals. Only found out during Prebuy when someone had to look at the plane. In LA or NY, you can’t get away with much, too expansive. Anywhere else you can drive 30 min to an hour and be in the middle of nowhere. Of course if you reverse it, you might be somewhere. Most stories don’t get told until the storyteller is out of the game, and if you do know, it’s a story among friends.