The $20/hour Cessna 172 experiment—Update

[Ted]

You mad about the GM v8 or the parts/assembly country of origin? There's nothing particularly complex about the electronics in an LS-engine, either. I've never had any electronics issues our of the 3 LS-engined vehicles I've owned. They've all burned excessive oil because they were AFM-variants, but it didn't affect engine operability in any way. If there was an automotive engine platform I'd trust as an aircraft engine, it'd probably be the LS-series. Ford's Coyote is a close 2nd, but it's a more complex animal with DOHC.

Any new V8 with variable valve timing and cylinder deactivation would have those removed, since those systems wouldn't add anything of value in an aircraft application. Less weight and less complexity/areas of failure. As I noted previously, there are some things you'd do differently for an ECU, which you'd want to be purpose-built for a certified application (and arguably for an experimental one, too).

But as far as the base engine goes, the LS is pretty darn bulletproof. To @wanttaja 's very valid point, auto engines have much higher failure rates in experimental aircraft than the purpose-built aircraft engines. But I think most of the time that has less to do with the core bottom-end of the engine and more with the support/control systems not being properly thought out or designed.

Ben Haas was probably one of the most successful auto derivative engines, with the Ford 347 (302 stroker) in his Zenith 801. He never publicly admitted it, but he did crack a piston on it, something he told me years ago. He attributed this to the very high EGTs he ran (I want to say north of 1600F) and after that he ran it a bit cooler, and had no issues. I was a huge fan of what Ben did, but he had the engine background and knowledge to do it well. Most homebuilders don't, so they're doing R&D without the necessary background for success.

 

[SoonerAviator]

Any new V8 with variable valve timing and cylinder deactivation would have those removed, since those systems wouldn't add anything of value in an aircraft application. Less weight and less complexity/areas of failure. As I noted previously, there are some things you'd do differently for an ECU, which you'd want to be purpose-built for a certified application (and arguably for an experimental one, too).

But as far as the base engine goes, the LS is pretty darn bulletproof. To @wanttaja 's very valid point, auto engines have much higher failure rates in experimental aircraft than the purpose-built aircraft engines. But I think most of the time that has less to do with the core bottom-end of the engine and more with the support/control systems not being properly thought out or designed.

Ben Haas was probably one of the most successful auto derivative engines, with the Ford 347 (302 stroker) in his Zenith 801. He never publicly admitted it, but he did crack a piston on it, something he told me years ago. He attributed this to the very high EGTs he ran (I want to say north of 1600F) and after that he ran it a bit cooler, and had no issues. I was a huge fan of what Ben did, but he had the engine background and knowledge to do it well. Most homebuilders don't, so they're doing R&D without the necessary background for success.

Agreed. Deleting the VVT/cylinder deactivation would be step 1. Heck, deleting the AFM on LS-engines is pretty common in auto applications as well, in order to solve the oil burning problem as well as not dealing with potential collapsed AFM lifters. ECU is a fairly easy problem to solve though, at least for the LS-engine as there are some aftermarket options with more compliant/secure harness connections. I'd think a lot of the old SOHC V8 engines would be a good fit if you could use a lightweight aluminum block/heads on them. However, in an ideal world, an all-aluminum diesel would be a good fit for most SE-piston aircraft as well as marine applications (doesn't help much in the cost-savings department over a traditional Lyco/Conti though).

 

[Ted]

ECU is a fairly easy problem to solve though, at least for the LS-engine as there are some aftermarket options with more compliant/secure harness connections.

Just to be clear, it's more than the security of the connectors. It's also having redundancy in multiple areas with multiple channels and some sort of voting scheme (basically two or more ECUs), it's multiple connectors oriented in such a way that one falling out won't result in a failure of the engine, and more. I'm not aware of any automotive application which has an ECU that meets the criteria, although the people on here who are more familiar might. If you do a Google search of FADEC, you'll see what I'm talking about. It's part of the reason why we have two magnetos in aircraft.

But then you get into a lot of the other requirements of both environmental and electrical which become far more complex than what we find in automotive. I can think of ways you could make some of the off the shelf units work, but not sure if you could make them meet the DO-160 requirements off the shelf.

There's a lot of stuff that goes on behind the scenes that make sure that your avionics work when you get too close to a thunderstorm at night and really need that ILS to correctly guide you to the runway. And then you take that one step further, and that's what your engine needs, too.

 

[Bell206]

But I think most of the time that has less to do with the core bottom-end of the engine and more with the support/control systems not being properly thought out or designed.

Why do you suppose nobody has resurrected the Porche 3200 aircraft engine for use? As I recall Porche surrendered the TC so I imagine its for the taking.

 

[Bell206]

Iirc AlphaPilotFlyer had stated on this forum previously that the goal was never to get it certified or even stc'd. So I was surprised when the article mentioned those hurdles because I never thought it was the goal.

As they mentioned in a number of posts, their website, and other locales, they were in the process of pursuing an STC and had even received their initial G1 issue letter. This process was a topic of public distain for them on a number of occasions as in the article. Now whether they ever had a serious intent to complete the process is another matter which I dont believe there ever was.

 

[Ted]

Why do you suppose nobody has resurrected the Porche 3200 aircraft engine for use? As I recall Porche surrendered the TC so I imagine its for the taking.

As I recall the Mooney was slower and burned more fuel with the PFM than with the Lycoming, which incidentally is a common issue with an auto conversion.

Unless there's some loophole (or very understanding ACO) that I'm not aware of, I don't think someone could simply resurrect the PFM engine because the TC was surrendered. As I understand it, that means someone would have to start from scratch. Maybe if Porsche signed over the technical data used in certification that would be a lower burden, and you could potentially try to justify some previous test results. But as far as I'm aware, it would still be a recertification.

 

[SoonerAviator]

Just to be clear, it's more than the security of the connectors. It's also having redundancy in multiple areas with multiple channels and some sort of voting scheme (basically two or more ECUs), it's multiple connectors oriented in such a way that one falling out won't result in a failure of the engine, and more. I'm not aware of any automotive application which has an ECU that meets the criteria, although the people on here who are more familiar might. If you do a Google search of FADEC, you'll see what I'm talking about. It's part of the reason why we have two magnetos in aircraft.

But then you get into a lot of the other requirements of both environmental and electrical which become far more complex than what we find in automotive. I can think of ways you could make some of the off the shelf units work, but not sure if you could make them meet the DO-160 requirements off the shelf.

There's a lot of stuff that goes on behind the scenes that make sure that your avionics work when you get too close to a thunderstorm at night and really need that ILS to correctly guide you to the runway. And then you take that one step further, and that's what your engine needs, too.

right, I wasn't speaking necessarily of trying to have an FAA-certified option, I was just expounding on @Redneckpilot's assertion that the Mexico-assembled LS-engine w/cheap Chinese electronics would probably be pretty unreliable, despite the availability off-the-shelf ECUs which had better harness connections. You could probably just have two ECUs running simultaneously with an A/B selector switch to revert to in case of an ECU failure. Obviously that creates it's own complications, but it provides the redundancy needed. It's one of those things that probably has an incredibly low failure rate anyway.

 

[MauleSkinner]

As they mentioned in a number of posts, their website, and other locales, they were in the process of pursuing an STC and had even received their initial G1 issue letter. This process was a topic of public distain for them on a number of occasions as in the article. Now whether they ever had a serious intent to complete the process is another matter which I dont believe there ever was.

But the original intent was not certification…it was a conversion to make a certified airplane into an EAB.

 

[Sundancer]

Side note - it would be good (maybe great) for the FAA certification process to be audited by an independent operations research kind of outfit. Not necessarily to change technical specs and measures, but to examine the process, the layers of management oversight, look for redundant activities and paperwork, the workflow of approvals, opportunities for concurrent actions, etc. I remember a Lockheed engineer describing literally millions of pages of reports generated by DOD overseers, documents that no one read, that drove no action, yet delayed production and increased costs.

The FAA has some good folks, but they're working inside a dinosaur, one that any 1960s bureaucrat would recognize.

 

[Ted]

It's one of those things that probably has an incredibly low failure rate anyway.

Overall agreed. After all, how many times has an ECU quit on any of us while driving? I can only think of one ECU that I've had to replace, on my wife's Mercedes when it up and quit one day. Other issue I had was my Ford Excursion shutting down at random, which had to do with the security system that read the key - something you wouldn't have in a plane.

 

[SoonerAviator]

Overall agreed. After all, how many times has an ECU quit on any of us while driving? I can only think of one ECU that I've had to replace, on my wife's Mercedes when it up and quit one day. Other issue I had was my Ford Excursion shutting down at random, which had to do with the security system that read the key - something you wouldn't have in a plane.

Ford PATS for the win!

 

[Bell206]

I don't think someone could simply resurrect the PFM engine because the TC was surrendered.

There are available routes to take but Porche would have to sign off on it for it to work.

it would be good (maybe great) for the FAA certification process to be audited by an independent operations research kind of outfit.

FWIW: In general, the certification process is mainly handled by private 3rd party FAA ODA/DER/DAR designees. All of who have a direct FAA mentor to report to. When the process bogs down its usually due to a new technology (e.g. additive manufacturing, etc.) for which there are no existing standards, the need to change existing standards, or in some cases an applicant's decision not to follow the existing guidance. So the pace of the certification is generally dictated by the applicant resources more than the FAA resources in most cases. Using the existing guidance, it takes 5-9 years for a new type certification depending on complexity. And that timeframe is generally the same for any ICAO member country that produces and certifies aircraft under ICAO and other agreements. However, the FAA process for STCs tends to be much quicker and more flexible than those other countries which leads foreign companies to sometimes come to the US for STC certifications. So to add another "audit" layer to the process would only slow things down vs speed them up in my opinion. Unfortunately, since the MAX 737 issue, just about every FAA Designee/Mentor relationship is going through a reset which has caused a number of delays outside of the normal process.

 

[Ted]

There are available routes to take but Porche would have to sign off on it for it to work.

I was assuming you wouldn't have that blessing.

FWIW: In general, the certification process is mainly handled by private 3rd party FAA ODA/DER/DAR designees. All of who have a direct FAA mentor to report to. When the process bogs down its usually due to a new technology (e.g. additive manufacturing, etc.) for which there are no existing standards, the need to change existing standards, or in some cases an applicant's decision not to follow the existing guidance. So the pace of the certification is generally dictated by the applicant resources more than the FAA resources in most cases. Using the existing guidance, it takes 5-9 years for a new type certification depending on complexity. And that timeframe is generally the same for any ICAO member country that produces and certifies aircraft under ICAO and other agreements. However, the FAA process for STCs tends to be much quicker and more flexible than those other countries which leads foreign companies to sometimes come to the US for STC certifications. So to add another "audit" layer to the process would only slow things down vs speed them up in my opinion. Unfortunately, since the MAX 737 issue, just about every FAA Designee/Mentor relationship is going through a reset which has caused a number of delays outside of the normal process.

Overall, I agree with all of this. And if you have an cert applicant that really doesn't understand what they're trying to do from a cert perspective, that slows things up as well.

 

[Sundancer]

Thanks, good info - but I didn't mean to add a permanent audit layer - just get some (modern) folks in to review how they do business, identify opportunities for efficiency and process improvement. That other countries are worse I believe - but doesn't preclude improving what we have.

 

[Bell206]

just get some (modern) folks in

Ha. There are a number of "modern" and younger folks involved now with the process on both sides and review the process on a regular basis. As to improvements there have been a few, for example, they rewrote the entire Part 23 to streamline things by demoting about 2/3s of the original regulations down to Advisory Circulars based on consensus standards similar to an LSA. WHile I don't know if its possible to make government entity more efficient these days, I do know on the DAR/DER designee side things are efficient especially when it comes time to cut their checks and you see what you bought for those services. They can be fired unlike the other side. But all in all if you have your sheet together and the right people involved things move along at a decent clip.

 

[wanttaja]

Overall agreed. After all, how many times has an ECU quit on any of us while driving? I can only think of one ECU that I've had to replace, on my wife's Mercedes when it up and quit one day. Other issue I had was my Ford Excursion shutting down at random, which had to do with the security system that read the key - something you wouldn't have in a plane.

ECUs *are* one of the major bug-a-boos of auto engine conversions in aircraft. For instance, I've got 120 homebuilt accidents involving aircraft powered by O-200s. None of them were attributed to ignition systems, either conventional magnetos or electronic.

In contrast, I've got 201 accidents involving Subaru-powered homebuilts. Thirteen of them involved the ignition system.

Yet, electronic ignitions ARE compatible with installation on aircraft. I've got 391 accidents involving Rotax 912-powered homebuilts, and just two involved the ignition.

The trouble is largely design- and implementation-based. The ECUs themselves aren't often at fault, but failure to keeping power supplied to the ECU often is.

The meaning of "redundancy" needs to be pounded into a lot of homebuilder's heads. It does NOT mean running a separate wire and switch from the same battery to the same ECU.

Ron Wanttaja

 

[Ted]

The trouble is largely design- and implementation-based. The ECUs themselves aren't often at fault, but failure to keeping power supplied to the ECU often is.

The meaning of "redundancy" needs to be pounded into a lot of homebuilder's heads. It does NOT mean running a separate wire and switch from the same battery to the same ECU.

And therein lies (part) of the problem I was getting at. Good data, Ron.

 

[Bell206]

The meaning of "redundancy" needs to be pounded into a lot of homebuilder's heads. It does NOT mean running a separate wire and switch from the same battery to the same ECU.

And maybe with a little murphy-proofing to boot. Just look at the DA62 thread where it had 5 independent/interconnected power sources that failed to keep at least one ECU powered due to "improper wiring."

 

[Ted]

And maybe with a little murphy-proofing to boot. Just look at the DA62 thread where it had 5 independent/interconnected power sources that failed to keep at least one ECU powered due to "improper wiring."

Or how the early diesel DA42 that lost both engines because the battery was low following a jump start (which granted was prohibited) and they pulled the electric gear up.

 

[Bell206]

the battery was low following a jump start (which granted was prohibited)

Whats funny is with the ECU back-up batteries installed they now allow jump-starting for Day VFR only. Better hope all your IPads are charged before you leave.

 

[Redneckpilot]

You mad about the GM v8 or the parts/assembly country of origin? There's nothing particularly complex about the electronics in an LS-engine, either. I've never had any electronics issues our of the 3 LS-engined vehicles I've owned. They've all burned excessive oil because they were AFM-variants, but it didn't affect engine operability in any way. If there was an automotive engine platform I'd trust as an aircraft engine, it'd probably be the LS-series. Ford's Coyote is a close 2nd, but it's a more complex animal with DOHC.

Both.

The complexity is relative.
It is extremely complicated when compared with a Continental or Lycoming aircraft engine with magnetos. I simply do not see the value in aircraft applications.

My issues with my GM have been mostly mechanical in nature, though I’ve had plenty of electronic problems as well.
I don’t want to get into the whole GM vs Ford deal here but I will say the sole GM product I own far exceeds my maintenance costs of the three Ford products I own.
That said, I don’t want either in an airplane.

If I have to resort to pouring a bunch of oil in my airplane engine it’s gonna be a big, loud, round one!

 

[Ted]

The complexity is relative.
It is extremely complicated when compared with a Continental or Lycoming aircraft engine with magnetos.

Not really. You've got slightly fewer moving parts with an LS. Still a pushrod setup, you can set one up with solid lifters, no VVT, and even a carburetor if you're so inclined. Could create a very simple/dumb solid state coil driver setup.

As I've said already, there are complexities with design and certification. But fundamentally, it's not really more complex. Most of what you see that adds complexity to these things in a modern automotive application are driven by emissions and CAFE requirements, not the core engine itself.

I've had lots more magneto failures than coil pack/coil-on-plug failures.

 

[jsstevens]

Based on my analyses of homebuilt accidents, auto engines have a much higher failure rate. About 22% of all accidents involving auto-engined-powered homebuilts were due to mechanical failures of the engine (either spontaneously or due to builder/maintainer errors, vs. about 7% of all accidents where traditional certified engines were installed.

With that said, the problems with auto engines were NOT in the basic engines. Both had roughly the same failure rates due to bearings, pushrods, cylinders, etc. The numbers get even closer when you ignore those due to builder error.

Ron Wanttaja

Does this mean that the engine failures in automotive conversions are mainly in external components like gearboxes, belt drives, etc?

 

[wanttaja]

Does this mean that the engine failures in automotive conversions are mainly in external components like gearboxes, belt drives, etc?

Pretty much.



Ron Wanttaja

 

[Redneckpilot]

Not really. You've got slightly fewer moving parts with an LS. Still a pushrod setup, you can set one up with solid lifters, no VVT, and even a carburetor if you're so inclined. Could create a very simple/dumb solid state coil driver setup.

As I've said already, there are complexities with design and certification. But fundamentally, it's not really more complex. Most of what you see that adds complexity to these things in a modern automotive application are driven by emissions and CAFE requirements, not the core engine itself.

I've had lots more magneto failures than coil pack/coil-on-plug failures.

Can’t argue complexity much when comparing base engine vs base engine.
However the installation in the article isn’t simply a base engine. This is a huge part of my problem with it.

Regarding magnetos, on a traditional a/c engine if I lose a magneto I still have one. If something happens to a plug or a plug wire I still have another set.
If it eats a valve chances are I can change a cylinder vs wrecking the engine.

Something I haven’t seen mentioned is the fact automotive V-8 engines are liquid cooled.
There’s a whole other field of opportunities for issues. Yes, I am aware that P-40s and P-51s were as well and there were problems there also.

I just don’t see automotive engines as a viable option to traditional piston aircraft engines.
If costs weren’t so ridiculous for parts I doubt this would even be a topic.

 

[jsstevens]

Pretty much.
View attachment 112516


Ron Wanttaja

Nice graph! This raises the question (in my mind) what kind of cooling system failures are there with a certified engine? It's oil/air cooled and a oil related failure would be engin internal, yes? Unless this includes Rotax as certified.

 

[SoonerAviator]

Can’t argue complexity much when comparing base engine vs base engine.
However the installation in the article isn’t simply a base engine. This is a huge part of my problem with it.

Regarding magnetos, on a traditional a/c engine if I lose a magneto I still have one. If something happens to a plug or a plug wire I still have another set.
If it eats a valve chances are I can change a cylinder vs wrecking the engine.

Something I haven’t seen mentioned is the fact automotive V-8 engines are liquid cooled.
There’s a whole other field of opportunities for issues. Yes, I am aware that P-40s and P-51s were as well and there were problems there also.

I just don’t see automotive engines as a viable option to traditional piston aircraft engines.
If costs weren’t so ridiculous for parts I doubt this would even be a topic.

Bolded #1: again, with an engine like the LS or even the Coyote from Ford, you have 8 individual "magnetos" in the coil-on-plug system. Obviously they aren't self-contained like a magneto and require 12V power from a battery to run. So redundancy with ECU and battery power is a good idea. If a COP goes out, it only affects one cylinder and is easily diagnosed and replaced (every auto parts store in the country will have a COP on-hand). LS-engines don't "eat valves" with any regularity, but even it it did it rarely involves a full engine teardown. Unless it ate the valve and punched a hole in the piston, you could have it back up and running in a weekend, probably for far cheaper than the cost of a new cylinder on a Lycoming.

Bolded #2: that's mostly the point. Using an engine that itself is 1/5 of the cost of anything certified, and then having off the shelf parts that cost very little and are abundant. Furthermore, no more messing with hot start/cold start procedures, setting/adjusting mixtures, run ups, spark plug fouling. Better fuel economy as well as the lower cost of Mogas.

 

[wanttaja]

Nice graph! This raises the question (in my mind) what kind of cooling system failures are there with a certified engine? It's oil/air cooled and a oil related failure would be engin internal, yes? Unless this includes Rotax as certified.

The teeny-tiny blue block there represents a single case (0.3% of the engine failures): CHI07CA313. Probable Cause, "The loss of engine power due to the blocked cooling and induction systems because of the pilot in command's failure to remove the cowling covers before the flight." Lycoming O-320.

In my database, I don't use "Probable Cause." Instead, I designate a single Initiator for each accident, and however many Secondary Initiators are appropriate. This particular accident's Initiator is listed as "Inadequate Preflight," but the result was an overheated engine. None of the 13 auto-engine cooling-related accidents were due to blocked intakes.

Ron Wanttaja

 

[jesse]

I’ve owned more than a few trucks and I can tell you if you’re turning 3000-4000 RPM for any extended period in any truck built in the last 20 years you’re either running 90 MPH or pulling something very heavy everywhere you go.

GM has all kinds of V8s in moving trucks, and, you aren't going to cruise down the interstate in one at anything less than 3 grand. My boat has a GM V8 and its normal cruise setting is 3,000 RPM. It has been doing this happily for over a thousand hours and twenty plus years. I've also ran it at 5,000 RPM for hours and nothing went kabang. All kinds campers being pushed around by GM V8s as well, and all of those do it at 3000 RPM or better.

To be clear, there are all sorts of challenges with mounting a GM V8 into an airplane and expecting it to go well. None of those challenges have to do with the engine spinning too fast for its own good. Instead, they have to do with the engine spinning too fast for the propellor's good...cooling details...fire protection...ECU redundancy......etc.etc....but nowhere on the problem list is "a GM v8 is not reliable at 3000 RPM"

My opinion is based upon my experience. YMMV.

Your experience seems rather limited. Perhaps you should base your opinion on something more meaningful?

 

[Initial Fix]

One of these days an up a coming company will realize that good engineering and not hope will weigh out.

 

[flyingcheesehead]

Or how the early diesel DA42 that lost both engines because the battery was low following a jump start (which granted was prohibited) and they pulled the electric gear up.

Whats funny is with the ECU back-up batteries installed they now allow jump-starting for Day VFR only. Better hope all your IPads are charged before you leave.
View attachment 112497

That's because there are more power sources for the ECUs now, thanks to an AD precipitated by that accident.

 

[Redneckpilot]

GM has all kinds of V8s in moving trucks, and, you aren't going to cruise down the interstate in one at anything less than 3 grand. My boat has a GM V8 and its normal cruise setting is 3,000 RPM. It has been doing this happily for over a thousand hours and twenty plus years. I've also ran it at 5,000 RPM for hours and nothing went kabang. All kinds campers being pushed around by GM V8s as well, and all of those do it at 3000 RPM or better.

To be clear, there are all sorts of challenges with mounting a GM V8 into an airplane and expecting it to go well. None of those challenges have to do with the engine spinning too fast for its own good. Instead, they have to do with the engine spinning too fast for the propellor's good...cooling details...fire protection...ECU redundancy......etc.etc....but nowhere on the problem list is "a GM v8 is not reliable at 3000 RPM"

Your experience seems rather limited. Perhaps you should base your opinion on something more meaningful?

Darn…you got me.

I never realized how little I have seen in my half century on this spinning rock until now.

You seem to be full of yourself. Perhaps you shouldn’t assume. It usually doesn’t work out.

 

[Sundancer]

Ha. There are a number of "modern" and younger folks involved now with the process on both sides and review the process on a regular basis. As to improvements there have been a few, for example, they rewrote the entire Part 23 to streamline things by demoting about 2/3s of the original regulations down to Advisory Circulars based on consensus standards similar to an LSA. WHile I don't know if its possible to make government entity more efficient these days, I do know on the DAR/DER designee side things are efficient especially when it comes time to cut their checks and you see what you bought for those services. They can be fired unlike the other side. But all in all if you have your sheet together and the right people involved things move along at a decent clip.

Thanks for the education - I defer to your insider knowledge, though I'd sure like to hear that some operations folks with no dog in the fight (Independent entities) had taken a pass at it. I have seen an FAA org chart - it looks like it's from the 1960s, and as far from "flat" as it gets. But that doesn't mean they haven't streamlined the real work.

 

[Albany Tom]

I'm surprised to see this thread still going on. Seemed simple to me. Somebody has an idea to try to cash in on a predicted death of 100LL by getting an automotive engine certified, with a hope that somehow this would make the approval process easier. 100LL didn't go away, and the process didn't, and shouldn't have, gotten easier.

The engine not getting certified seemed pretty certain to me, unless for some goofy reason GM decided to support it. Why would GM support this? It would probably take creating a custom engine/parts division for 1/1000th share of their overall production. Or...they have to run every tiny change of that engine through the FAA? AND...they'd have to deal with the liability associated with producing aircraft engines. Or in other words, I can't see it being certifiable without the manufacturer on board. And I can't see any sane reason why, in 2022, GM would do such a thing.

The hybrid? Seems like either a) these guys aren't good at math or aren't familiar with the laws of thermodynamics or b) they're just having fun getting venture capital to justify building toys. Perhaps I'm being harsh.

Ok, rant over. Just seems way to close to the ads at the back of Popular Mechanics that would turn water into gasoline or whatever.