Another update on the Corsair V8 project

[PeterNSteinmetz]

The only FAA involvement with the V8 kit would be for the owner to apply for the Special AWC under Experimental Exhibition and receive their certificate and operating limitations.

…
I believe so for rental, but do not follow this specific topic much. However, it is the same process used for E/AB or other Special AWC aircraft when rental, training, and other type ops are performed. For example, Corsair provided the V8 to a student to get their PPC to include the FAA check ride, but don’t know how they provided that aircraft to them. That event is also contained in their literature.

But no on standard commercial ops

Yes, except for obtaining a Special AWC and ops limits as mentioned above.

That does clarify, thanks. So I think it is clear that the regulations involved for the various operation types require additional work and paperwork, at a minimum. Thus they consume time, energy, and money. And therefore the regulations cause innovation to proceed more slowly than they would without them.

Now of course these regulations are intended to improve flying safety.

Let’s consider what sort of safety risks using the V8 in this manner present in aviation. What do you think those are? And let’s divide them into 2 categories: risks which are chosen by the person participating and risks to others. (Of course culpability is more likely a continuous aspect, but let’s divide along the lines the law would in terms of risks foreseeable by a “reasonably prudent man” and those which would not be.)

 

[Ron Wanttaja]

Let’s consider what sort of safety risks using the V8 in this manner present in aviation. What do you think those are? And let’s divide them into 2 categories: risks which are chosen by the person participating and risks to others. (Of course culpability is more likely a continuous aspect, but let’s divide along the lines the law would in terms of risks foreseeable by a “reasonably prudent man” and those which would not be.)

Auto engine conversions do not have a stellar record in homebuilt aviation. This graphic illustrates the percentage of accidents due to engine mechanical issues for about 22 years of homebuilt accidents. Note that the auto conversions actually come out worse than two-stroke engines.

Now, there's no question many of the affected conversions are done by persons who perhaps don't have the proper technical expertise, but commercially-available conversions have been available for quite some time and do not appear to be performing significantly better than the homemade conversions. However, the installed base is generally pretty low, so it's tough to quantitize the issue.

What this means, of course, is that the FAA is going to (quite rightly, in my opinion) insist on thorough qualification testing to PROVE the engines are safe. Wanna show your conversion is just as safe as an O-540? Then run the converted engine through the same testing that the O-540 went through. If it passes, there should be no question of getting an STC. But, again, because of the auto-engine conversion's poor showing in homebuilt aircraft, the FAA is likely skeptical.

The risks by the pilot vs. others depends on what kind of airplane the engine is installed in. Putting a Subaru conversion into a single-seat Fly Baby risks only the pilot (mostly). A V-8 conversion in a Cessna 206 can risk five passengers as well. Also, the larger the aircraft, the higher the impact speed after an engine failure, and the greater the potential for injuries or death. In addition, takeoff and initial climb is a high-probability period for engine failures, and more and more of our airports have urban buildings right outside the perimeter fence.

Ron Wanttaja

 

[rotorwrench]

And therefore the regulations cause innovation to proceed more slowly than they would without them.

Except the facts don’t support that conclusion. For one, in aviation, the innovation stage starts way before the regulatory stage with little or no regard for those regulations. And a prime example is the number of “innovative” eVTOL aircraft currently in the certification mix. If your conclusion were correct there would be no eVTOLs and especially within a 5-year time period.

In reality, what reduces or slows innovation in aviation is the one thing missing from your hypothesis: a viable market to sell those innovative aircraft. For example, look at the Toyota TAA-1 aircraft and the reasons why it never went to production.

Let’s consider what sort of safety risks using the V8 in this manner present in aviation. What do you think those are? And let’s divide them into 2 categories: risks which are chosen by the person participating and risks to others.

I think Ron’s chart above summarizes the risk point which illustrates a long-known fact that vehicle engines do not compare on many levels to purpose-built aircraft engines.

On the risk side, the regulations already separate the risk type between the owner/operator and the general public by allowing the V8 and other vehicle engines (except the Porche, Toyota, Austro variants) to be used only under Special AWCs and the experimental category. So I’m not following why you would need to divide the risk by category again?

 

[PeterNSteinmetz]

Except the facts don’t support that conclusion. For one, in aviation, the innovation stage starts way before the regulatory stage with little or no regard for those regulations. And a prime example is the number of “innovative” eVTOL aircraft currently in the certification mix. If your conclusion were correct there would be no eVTOLs and especially within a 5-year time period.

How could it even be possible that adding time and expense would not slow down innovation, at least to some degree?

It could easily be the case that potential innovators know very well the sort of obstacles they will encounter with regulation and often don’t even bother. So the timing does not preclude such an effect.

The last argument is almost a non-sequitur. Slowing things down does not imply that a thing will never happen, just that they are slower than they otherwise would be.

In reality, what reduces or slows innovation in aviation is the one thing missing from your hypothesis: a viable market to sell those innovative aircraft. For example, look at the Toyota TAA-1 aircraft and the reasons why it never went to production.

I suspect these two items are interrelated. If regulation increases costs and delays introduction of new products, that would decrease the viable market. I am not sure these can be completely decoupled.

I think Ron’s chart above summarizes the risk point which illustrates a long-known fact that vehicle engines do not compare on many levels to purpose-built aircraft engines.

On the risk side, the regulations already separate the risk type between the owner/operator and the general public by allowing the V8 and other vehicle engines (except the Porche, Toyota, Austro variants) to be used only under Special AWCs and the experimental category. So I’m not following why you would need to divide the risk by category again?

Because if one is going to evaluate the risk / benefit ratio of such regulations under most moral theories the potential risks which are freely chosen and those which are not are evaluated differently. I understand that under your legal - authoritarian moral theory that does not matter. It is my hope that you can see that in some other moral theories this does matter. That is why people may wish to understand the risks which fall into each category. Those who do not wish to factor that distinction into their moral decision making can simply lump the risks together.

Since I will not be explicitly discussing those tradeoffs with you given your non-falsifiable moral beliefs, I think it better to focus here on what those absolute risks are.

It seems that automotive engine conversions are about 3 times riskier than certified engines from Ron’s chart. So then to convert to absolute risk we could multiply by the absolute rates of such accidents. Maybe the Nall report?

 

[Ron Wanttaja]

It seems that automotive engine conversions are about 3 times riskier than certified engines from Ron’s chart. So then to convert to absolute risk we could multiply by the absolute rates of such accidents. Maybe the Nall report?

Again, though, my results are for Experimental Amateur-Built aircraft, for which there are NO standards that must be followed. If we're talking about an engine that will be used commercially, one would assume that, for liability reasons alone, the manufacturer would expend some effort to ensure reliability.

One thing about auto engines: The core engines themselves don't seem to be the issue. They don't seem to suffer mechanical failure at a rate much different from traditional engines. Where they DO get bit is in the ancillary equipment necessary to make the engine work on an airplane....the propeller speed reduction unit (PSRU), the liquid cooling system, the electronic ignition, etc.

This chart shows the relative frequency of accidents due to engine-related mechanical issues.

Note that the "Engine Internal" category (which is basically referring to the core engine) is only slightly higher for auto engine conversions. And look at the failures related to systems that traditional air-cooled aviation engines DON'T HAVE TO HAVE. If it's not there, it ain't gonna cause an accident by breaking.

Again, though, a commercial activity such as Corsair has the opportunity to bullet-proof these systems. The Rotax 912 has liquid cooling, a PSRU, and electronic ignition...and failures of those systems are few and far between. The engine is just about as reliable as the O-200 it's replacing.

Now, my data above doesn't address the rate of failure COMPARED TO THE NUMBER OF ENGINES installed. Instead, it looks at the accidents involving aircraft with a given type of engine installed, and computes the percentage of those accidents that were due to a mechanical failure of the engine. It also includes only those cases where an NTSB report was generated. If the owner of a VW-powered airplane plops down in a field somewhere and trailers his plane home, it's not going to be listed.

Ideally, we'd be comparing the accident statistics to the number of engines installed. Unfortunately, the statistics just aren't out there. When you look at the FAA registry, 12% of the homebuilts are listed merely as having "AMA/EXPR" engines. That category, in fact, is the #3 most common installed engine on homebuilts.

Note that the #2 most common engine on homebuilts is a Rotax.. and half the Rotax listings don't specify the engine model. Which means we don't know how many of those ~4700 Rotax-powered homebuilts have two-stroke engines or four-stroke engine.

Note the small sample size for auto-engine conversions other than VW and Subaru. This plot cuts off at 100 aircraft. If the engine type isn't listed, that means there are fewer than a hundred on the FAA registry.

Now, there are undoubtedly MORE, since some of those "AMA/EXPR" engines are GM engines, or Mazda engines, or whatever. One noted converter claims there are thousands of their engines out there, but I only find a hundred or so actually in the FAA registry. Makes you lose faith in advertising.....

But we would like to know what are the annual odds of an engine failure are. I tackled a process to estimate how many of those AMA/EXPR engines were specific engine types...cross referencing the FAA registry cases with the NTSB accident database (the NTSB is better at identifying specific engine types). Came up with an estimated percentage of the AMA/EXPR engines that were Lycomings, Continentals, Hondas, etc. and developed an estimate as to the total number of engines installed in the fleet. THEN I compared it to the accident data....

Probably not a lot of surprise, here. I was actually surprised that the Rotax 912 came out as high as it did, but I was facing the same problems with the "Rotax" engine listings as the AMA/EXPR ones... there are thousands of homebuilts listed as merely having a "Rotax" engine without specifying whether it's a 582 or a 912.

So you can see why there's skepticism in the aviation industry as to the viability of converted auto engines in aircraft. The main question, I think, is whether the owner will be able to get insurance on an auto-engine-conversion Cessna 172....

Ron Wanttaja

 

[rotorwrench]

How could it even be possible that adding time and expense would not slow down innovation, at least to some degree?

Simple. In aviation, the innovative stage travels at its own pace, time frame, and cost. Its only when one has a viable, production-ready, flying aircraft that they even start the regulatory process.

It could easily be the case that potential innovators know very well the sort of obstacles they will encounter with regulation and often don’t even bother.

Not really. For example, with some of the latest aviation innovations, those innovators had no idea what regulatory obstacles they would face for the simple reason no pertinent regulations existed when they started the project. And in a few cases, no formal regulations even existed when they entered the certification stage.

So basically, it’s a moot point that regulation slows or adds cost at the innovative stage in aviation.

If regulation increases costs and delays introduction of new products, that would decrease the viable market.

In most aviation scenarios it’s the opposite: lack of market = no innovation. However, if you demonstrate a viable market to any OEM, they will innovate and bring a new product to that market.

Plenty of examples: Cessna Sky Courier - 50 firm orders/50 future orders; Beechcraft Denali – firm 250 aircraft market; Airbus H160 – firm 40 aircraft order; Hill HX50 – new design turbine-powered E/AB helicopter – 1000+ paid pre-orders. And so on.

Matter of fact, if regulations did increase costs and time, then the Part 23 rewrite should decrease costs and time for certification since it deleted 1000+ individual regulations.

But not all OEMs need a market, aircraft, or regulation structure to continually innovate. For example, the Bell Folding-blade propulsion system or the Airbus Racer are in-house technology demonstrators that may, or may not, lead to new and innovative aircraft in the future. So innovation has always been happening for many different reasons in aviation regardless the regulatory requirements.

It is my hope that you can see that in some other moral theories this does matter. That is why people may wish to understand the risks which fall into each category.

Either I’m missing something in your narrative or you’re over thinking this. Today, right now, people have that ability with aircraft. No moral theories required.

To illustrate this, I propose the following:

Two identical 172s are at a fly-in giving free rides. One is powered by a non-certified V8 and the other is powered by its certified O-320. A group of people wait in line and are allowed to personally pick which 172 they fly in.

How does your moral theory-based system make it any easier for people to assess risk and “freely choose” which 172 to fly in over the current FAA regulatory-based system?

 

[PeterNSteinmetz]

Matter of fact, if regulations did increase costs and time, then the Part 23 rewrite should decrease costs and time for certification since it deleted 1000+ individual regulations.

That would be an excellent test, wouldn’t it? Does the FAA or some other part of the government have methods in place to measure this?

All of the other arguments above are essentially theoretical with no real empirical backing (some anecdotes but those don’t really demonstrate anything, though may suggest further investigation.) Personally that is not going to persuade me that aviation is somehow an exception to the large literature on the effects of regulation on markets. IIRC, none has ever been adduced to support this hypothesis over several years of discussion now. I realize you are very convinced this is true, but I’m from Missouri on this one, show me a real proper study, then we have something meaningful to discuss.

Either I’m missing something in your narrative or you’re over thinking this. Today, right now, people have that ability with aircraft. No moral theories required.

To illustrate this, I propose the following:

Two identical 172s are at a fly-in giving free rides. One is powered by a non-certified V8 and the other is powered by its certified O-320. A group of people wait in line and are allowed to personally pick which 172 they fly in.

How does your moral theory-based system make it any easier for people to assess risk and “freely choose” which 172 to fly in over the current FAA regulatory-based system?

I think that a free market approach without the government trying to make the decisions would result in both more data being available to consumers to judge the risk of one versus the other and to them being aware that they should weigh their choice against the costs. What if the V8 flight was 20% cheaper and there was a charge for both?

That sort of weighing of cost versus benefit is precisely the sort of weighing that is made by regulators when implementing and enforcing a regulatory scheme. Weighing those is making value choices, and how to perform such weighing is the essence of a moral theory.

Another deleterious effect of coercive regulatory schemes by the government is that they decrease information available to and attended to by consumers. Most consumers just figure that if the government says it is ok or licensable, it must be fine. This then focuses providers on providing the lowest common denominator that meets the requirements because consumers will not pay much more for anything above that. There is really also no incentive to focus consumer attention on safety, other than “we are fully certified by the FAA”.

I think this also illustrates an assumption being made in many of the arguments above. That is that the cost of regulation is zero. Sort of like the free rides being given. And in that case, of course add more regulation, provided it does not decrease safety. Why not “if it even saves just one life” is a perfectly logical conclusion.

But if the cost is not literally zero, then a weighing is required. And that requires a moral theory. And in the real world, nothing is ever truly “free” with zero cost. That includes FAA regulation of aviation.

 

[Ron Wanttaja]

There's a major factor that has to be considered, here: The insurance companies.

You can whip the FAA into simplifying the certification process, but insurance rates won't be fooled. The insurance company will set rates based on actual accidents; if a particular type of aircraft has more accidents, the rate is higher. However, the industry is also very conservative. LACK of data is not going to produce low rates.

Take a look at an insurance executive setting the rates for two 172s. One has a Lycoming certified under the traditional processes. The other has an automotive engine "certified" under new, simpler processes whose primary goal was to foster innovation...not safety. The rates will be higher, and the company may refuse to cover it at all.

Not, perhaps, that big of an issue for privately-owned aircraft...many owners operate "bare". But a major factor for any commercial enterprise.

So "fixing' the FAA may not solve the ultimate problem....

Ron Wanttaja

 

[PeterNSteinmetz]

There's a major factor that has to be considered, here: The insurance companies.

You can whip the FAA into simplifying the certification process, but insurance rates won't be fooled. The insurance company will set rates based on actual accidents; if a particular type of aircraft has more accidents, the rate is higher. However, the industry is also very conservative. LACK of data is not going to produce low rates.

Take a look at an insurance executive setting the rates for two 172s. One has a Lycoming certified under the traditional processes. The other has an automotive engine "certified" under new, simpler processes whose primary goal was to foster innovation...not safety. The rates will be higher, and the company may refuse to cover it at all.

Not, perhaps, that big of an issue for privately-owned aircraft...many owners operate "bare". But a major factor for any commercial enterprise.

So "fixing' the FAA may not solve the ultimate problem....

Ron Wanttaja

A very good point Ron. The insurers actually have to bet their bottom line on the risks and benefits. Especially for commercial operations involving billions of dollars. If they get it wrong, they pay. Thus they are very incentivized to make sure to be as accurate as possible.

This is a very different set of incentives than those acting on government regulators. They usually have some personal motivation to try and achieve safety but they suffer literally no penalty for being overly cautious relative to the real risk. In fact they are rewarded for it by having larger numbers of employees to supervise more regulations more strictly, higher salaries and better pensions. They are also incentivized to minimize their view and evaluation of the cost of regulation. Indeed they often simply ignore those costs.

Which in my view points to the “solution”, if there is one. Minimize the role of the FAA in the aviation safety business. Make them just a policeman for clear aggressions against others - like deliberately flying your plane in a manner which clearly and imminently threatens the safety of others. Not policing that there might be some slight increase in risk of injury well below the risks of daily living. We just normally ignore those and prosecute them after the fact as needed and appropriate.

Let the companies and their insurers worry about these things. They can develop appropriate certifications for aircraft and components. Heck, they could even initially adopt the FAA ones and then the market will evolve.

Yes, I know that will blow many minds here. But novelty does not necessarily imply something is false. And there are small incremental steps we can take in this direction. For starters, eliminate 3rd class medicals entirely. Let private pilots fly on driver’s licenses, like LSAs. In terms of maintenance, let owners decide on what they want to do, even on “certified” aircraft. In fact, perhaps eliminate that distinction entirely and let insurers decide what standards they want to impose. Then also strictly enforce any torts dealing with aviation. Another step. Repeal the liability limits for commercial carriers. Make them pay for the risks of their profit making activities.

 

[rotorwrench]

That would be an excellent test, wouldn’t it? Does the FAA or some other part of the government have methods in place to measure this?

Yes x2. But since the applicant controls all the costs, I believe there is a voluntary program to report the rewrite effects to the FAA.

All of the other arguments above are essentially theoretical with no real empirical backing

Not really. All the examples and descriptions I posted on this topic are based on experience, historical record, actual events with real aircraft, and supported by factual data which doesn’t get more empirical than that.

I think the only way for us to have a meaningful discussion on this topic is if you can also provide factual data or historical precedent that your moral theory/free market has actually worked in the real world and not merely existed only on paper. Remember, I’m a realist, not an idealist.

I think that a free market approach without the government trying to make the decisions would result in both more data being available to consumers to judge the risk of one versus the other and to them being aware that they should weigh their choice against the costs. What if the V8 flight was 20% cheaper and there was a charge for both?

Interesting. So using your version of the original 172 example above, same question… how does your “free market” approach make it any easier for people to assess risk and value, then freely choose which 172 to fly in over the current FAA regulatory-based system?

That sort of weighing of cost versus benefit is precisely the sort of weighing that is made by regulators when implementing and enforcing a regulatory scheme.

But isn’t that the same “cost vs benefit” premise you question above with the V8 172 being 20% cheaper?

Another deleterious effect of coercive regulatory schemes by the government is that they decrease information available to and attended to by consumers.

In what manner do the FARs decrease information available to the consumer?

Perhaps we can use the 172 example to illustrate your point: how does your “free-market” system provide more data to those 172 consumers than the current regulation-based system?

That is that the cost of regulation is zero. Sort of like the free rides being given.

Except that is not what I stated. We were discussing the cost and time effects of regulation on innovation. My statements were only in that context.

Let the companies and their insurers worry about these things.

Curious. Will your free-market system mandate companies to carry insurance?

 

[PeterNSteinmetz]

So you can see why there's skepticism in the aviation industry as to the viability of converted auto engines in aircraft. The main question, I think, is whether the owner will be able to get insurance on an auto-engine-conversion Cessna 172....

Nice data and analysis Ron. So going back to my earlier question, I would like to know what sort of absolute risk rate risk is potentially created if all engines were to be auto conversions in the entire GA fleet of 4 seat single engines.

I was just looking at the Nall report and it seems that the total GA non-commercial fixed wing accident accident rate is about 6 /100k hours of flight. And about 20% of those are mechanical of some type. So that means that at most 1.2/100k could be engine failures. Only about 8% of those are lethal, so the fatality rate due to engine failure is at most 0.1/100k.

Now my understanding is that less than 10% of fatalities in GA flying are people not in the plane. So that means that the risk to people who have not chosen to fly in a small plane is about 0.01 / 100k flight hours for a fatality. And even if one tripled this rate because of using auto conversions throughout, the absolute rate would be only 0.03/100,000. This is 33 times lower than the 1/100k total risk of fatality for piloting the plane. And that amounts on a per mile basis to about the same as riding a motorcycle on dry pavement.

Thus the risk to people who do not choose to fly in a small plane from converting the entire fleet to auto conversions seems to much lower than the normal risks of everyday living. For example. for a 20 year old male all cause mortality is 0.17/100k hours.

 

[PeterNSteinmetz]

Not really. All the examples and descriptions I posted on this topic are based on experience, historical record, actual events with real aircraft, and supported by factual data which doesn’t get more empirical than that.

But those don’t tell us anything about the question of whether the regulations slow down and increase the cost of innovation. Perhaps the issue with this sort of argument is best illustrated more schematically.

We have argument X: more A decreases B
Then argument Y: but there is B, therefore X is false.

Perhaps it is clearer in this form that the mere existence of B says nothing about the rate of change of B due to changes in A.

What has been presented so far in the above are anecdotes and current observations of innovations (B). One has to examine the rate of change of innovation as a function of regulation (A) to be able to make a statement about the effects of regulation on innovation.

So far as I have seen there is no real such comparison in aviation that we can refer to. This may be due to the fact that many of the regulations we are concerned about were put in place when they didn’t instrument them properly.

Thus I have repeatedly noted that the data in non-aviation fields indicates that regulation decreases innovation. That has been rejected above with the claim that aviation is unique and so those data don’t apply.

I think the best thing to do at this juncture is to ask if there is any evidence, which is at least possible, which would persuade you that regulation decreases innovation in aviation? I don’t think anything that would be persuasive has been suggested and I don’t think having observed the way these arguments have been presented that there is likely to be.

I know you really believe this hypothesis is false so that may be an unusual way to think about it. Essentially it is the question a scientist asks about a hypothesis that appears false - what would prove the hypothesis correct?

Perhaps we can use the 172 example to illustrate your point: how does your “free-market” system provide more data to those 172 consumers than the current regulation-based system?

This was answered in my prior post #27. Please see and feel free to ask additional more specific questions.

Curious. Will your free-market system mandate companies to carry insurance?

That would in my view be an enormous improvement. I really don’t wish to discuss the specific moral considerations however given your non-falsifiable belief in a legal authoritarian moral perspective.

How about discussing the absolute risk levels of converting the entire 4 seat single engine fleet to auto conversions? I think Ron has shown some very informative data and analysis above. And see post #31.