We have ECi cylinders on our parachuting aircraft (C185 : IO-520D) and are happy to state that in just under 700 hours in service with this engine, we have not had to have any work done to them what so ever. The oil samples are coming back with grade "A" reports at every 50 hours (with just one grade "B" some while ago after the aircraft sat idle for the best part of sie to seven months). Plugs alway have very little to no lead at each 50 hours check and thus, valves and guides have needed no attention in this 700 hour period!

We also have a JPI EDM930 fitted which enables us to acurately keep the CHT's to no more than 380 degrees C and fuel flow turned-up an extra 1.5 USG/hr for take-off etc. Agressive 'leaning' is used for all ground manouvres, with the mixture control basically left where it is at the top of climb for the whole descent! EGT's are kept to no more than 1,330 degrees C in the climb at between 80knts, reducing to 75knts at altitude.

Oh yes! ..... leak down tests on last check were 3x 80's, 2x 79's and 1x 78 ... which is fairly consistent on all checks. Incidently, on the previous engine with factory cylinders fitted, three were removed for valve/guide work within the first 400 hours and five had work to them by the end of it's 1,800 hour life. (In fairness, the JPI EDM930 was not fitted at the time).

Steve Holder New Zealand

The lesson might be to make sure that digital engine management readouts, giving readings for each cylinder, be installed when ever new heads or cylinders are needed so pilots can avoid overheating (and save fuel). Of course that would be possible only if their price was not so inflated due to FAA certification requirements.

The company I used to work for operated 10 twin Cessna's - I believe we had 2 ECi head separations - otherwise no complaints. My impression, the ECi's may be slightly more prone to head cracking, but is less apt to have valve problems, so at the end of the week I'd buy them. I have ECi Titans on my Continental powered 172 (I know different engine application) and love them - at 800 hours, zero issues and compressions 79/80.

ECI had issues with their Titan cylinders for the O-540 a couple years back. There was an AD, and I think that one was far more reasonable, requiring inspections of the cylinders every 50 hours an hour limit on a subset that were determined to be the most likely to fail, and mandatory retirement at TBO (I hope I got that right). I had put a set of those cylinders on my Cherokee Six, and I specifically selected the Titans for the superior corrosion resistance. Those cylinders are still on the airplane, running quite fine and the frequent inspection is a relatively minor cost for some peace of mind. Why the FAA can't allow frequent inspections, perhaps borescope inspections for cracks as well, and maybe even mandate an engine monitor, all of those would be far cheaper than a wholesale replacement that may precipitate other failures.

Of the three aircraft using TSIO 520 NB engines on our field using ECI cylinders, one has had 100% of it's cylinders replaced prematurely, one 60% and the other 1 cylinder. Non-were by the barrel/head but rather by the fuel injection and leaking fuel. The replacement cylinders were of a different casting design to allow improved feed metal during the casting process.

Good analysis Paul. For something this major, why doesn't FAA have an AMOC like inspections? This has been commonly done in the past for other ADs. If this type of broad but weakly-identified AD is a new tack by the FAA I'm afraid we could all be grounded soon. Let's hope not.

By the way, the 0.0012 figure is the fraction of failures, not the percentage (0.0012 = 36/30,000). The percentage would be 0.12%. Just words, but important ones.

I am an A&P / IA, and your article fails in several respects: a) allowing owners to choose and decide? Owners are generally unable to technically decide and will typically base a choice on cost - and if they have the choice they'll say "They've been fine, leave them alone" and do nothing; b) the failures are catastrophic. Means sudden, without warning. They come apart at the identified zone of weakness. c) the past incidence is indicative, and will continue. The fatigue failure incidence is past and the future can expect the same rate or higher as the fleet TIS grows; d) Your premise that the failure of a cylinder may not lead to total engine failure is a presumption and appears reckless. Any element of failure, especially a cylinder failure, is a major failure.

The FAA is a regulator and has laboratories that can conduct metallurgical testing, and can evaluate risk. I trust their analysis over that of an aircraft owner. As a regulator they are charged with the responsibility of exactly that. Their only tool is to modify data, but in this case there is no modification of the existing cylinders available, so it's a choice of replacement or persevering with the existing latent problem.

I suggest that you reflect these responsibilities in your otherwise excellent article.

"Owners are generally unable to technically decide and will typically base a choice on cost."

To which I would reply, what's wrong with that? Aircraft ownership is a constant balancing act between what you spend and what you get. Suppose the FAA came along and said you must install active traffic equipment because it will improve safety. Would you then say we have to follow the regulators directive just because they are regulators?

In this case, the FAA has not demonstrated a high degree of risk, only a pattern of failures that appears to be somewhat worse than competitors. As for the increase in failures, ECI insists the reverse is true: this population has shown an attenuation in incidents as fleet hours increase.

I keep coming back to the relative risk, which appears trivial. We the regulated should have a say in how the regulators make decisions which impact cost of ownership. The FAA hasn't demonstrated a clear and present danger to flight safety or the general public from airplanes falling out of the sky.

They may get there, but they haven't yet.

"Owners are generally unable to technically decide and will typically base a choice on cost."

To which I would reply, what's wrong with that?"

What's wrong with that is the FAA is entrusted with the task of protecting the general public. Owner decisions can be all over the map. I might decide that carrying my grandchildren in my twin with ECI cylinders is not acceptable, and have them replaced. Guys across the field running charters into ski country with 414s might try to eek out another marginal season with what might be a ticking time bomb, or not. Should they make the decision? What are their qualifications for doing so? Who should really decide how to protect the public?

Does lean of peak operation and not monitoring CHTs a factor here?

Edd,

That is an easily solvable issue, and it can be handled the same way the FAA deals with all things related to the difference between part 91 and part 135/121.

Make inspections and/or replacement mandatory only for commercial operations (even if part 91). Add a carve-out for flight training and private rental.

Now, anyone carrying passengers or cargo for hire needs to have the extra caution that you're describing, thus protecting the public. Owners and flight schools are still free to do as they see fit.

Further, doesn't the FAA have to prove there's clear and present diminution of flight safety here, whether applied to for-hire or Part 91 operations? And if so, has it?

I'm far from convinced.

A regulator's strongest tool remains bad publicity.

I haven't found anywhere what ECI is planning to do to ensure these cracks don't occur going forward.

Paul you state: Maddeningly, the government's own NPRM process precludes the FAA from supplying specific information on its methodology or the underpinnings of its conclusion to demand removal of these cylinders.

Couldn't figure out what exactly additional info you were referring to. The FAA regulatory process has a pretty good track record and is pretty robust. However, what I couldn't find in my short look thru the docket is the cost-benefit analysis that is typically required. At 80+ million, 5.7 million per life (DOT policy last time I looked) they must be estimating around 15 lives. But that analysis should be part of the docket. That's not a hard requirement and sound judgment can override the data but again that logic should be part of the docket.

"Couldn't figure out what exactly additional info you were referring to."

I can be specific.

As they are required to do, the regulators are asking for comments on the NPRM, which presume they are seeking opinions. One needs data and facts to form same, plus answers to questions. My chief question is this:

We think the FAA used two statistical models to predict head-barrel separations in the named cylinder population, one called Weibull, the other Crow-AMSAA. Between these two projection methods and the two cylinder populations, the total head/barrel separation events were supposed to be about 50 between 2011 and 2013. But none occurred, according to ECI. None have occurred in cylinders manufactured since 2008, says ECI.

So my question is: did these models underpin the FAA's decision to broaden this AD? And if they did, what input assumptions were made? And why is the prediction so divorced from reality and shouldn't this be a consideration in asking for such a broad remedy? As you know, government NPRM rules preclude these discussions. They should not.

As for 15 lives, there have been no accidents, fatalities or injuries related to these cylinder problems. Maybe I just dense, but I couldn't find this in the docket and neither does ECI say it's there.

It just seems that this is a lot of money, effort and disruption to perhaps correct a small problem. And let's not forget the potential for introducing problems when opening up 5000 engines and replacing cylinders.

To me, the numbers just add up in an action where the FAA admits it doesn't understand the mechanism for these failures.

The lack of the statistical analysis of the failures doesn't particularly surprise me - AD's are issued (proposed) when the FAA determines an unsafe condition exists, not the public, and sometimes includes proprietary data. Though I would of thought ECI had visibility to them. Personally, I don't believe their data to be overly conservative - too many operators of light airplanes don't report problems they just fix them and how does one account for that in ones analysis.

So while they probably welcome any comment, they specifically want comments on the regulatory, economic, environmental, and energy aspects of the proposal seems like an aviation consumer magazine might post comments on behalf of the consumers to question the FAA's regulatory evaluation e.g: -How did the the FAA determine that this proposed change is not a significant regulatory action under Executive Order 12866? That the cost of this change is significant to this segment of the industry reaching across a large number of aircraft. So a cost benefit analysis should be conducted taking into account DOT treatment of the value of preventing fatalities and injuries in preparing economic analysis and OMB policy and the spirit of the Presidents Order. -How did the FAA determine that only 6,000 engines would be impacted when clearly not everyone replaces every cylinder so the impact to the number of consumers is far greater and the FAA's analysis should take into account. -The 82+million estimate doesn't account for lost revenue/time on the products, particularly since a portion of these engines are used for commercial operations. This should take into account however many weeks the airplane will be down for inspection, removal, replacement of any cylinders and the amount of revenue these airplanes won't generate. I could easily see another 20 to 30 million in costs to the industry but that is something those who use these engines need to pony up; -How did the FAA determine that there would be no impact on interstate commerce in Alaska when clearly these engines are installed in aircraft used for commercial use where aviation is the only means of transportation in Alaska (examples warranted). -Also it is hard to believe the AD won't impact any number of businesses that qualify as small and entities -That when the above is provided by the FAA the public should be afforded 90 days to review and comment prior to proceeding with issuance of any AD.

19 (or 30) head/barrel separations? Wow. I guess I have really bad luck. The twin Cessna I was flying from 2001-2010 had 3 in flight head/barrel separation, AFTER having all 12 cylinders replaced under warranty due to a proposed AD that later went into effect. We also caught 3 more head/barrel separations on the ground with inspections we started doing every 50 hours. Each of these cylinders was replaced under warranty, so I would assume they became part of the data pool. After 6 failures, we lobbied the engine installer for an entire new set of non-eci cylinders, which we got. We had no further problems. After the first 2 in flight failures, we installed an engine monitor. While operating the engine under the same parameters as before the failures, we found cylinder head temps did not exceed 410 in any phase of flight, yet we opted to run even more fuel and with cowl flaps open slightly to keep cylinder head temps below 400 (alarms were set on the engine monitor to help ensure this). Yet we still had 4 more failures. And after the third full set of cylinders were installed (ECI overhauled Continental cylinders), we had no more problems with any cylinder cracking. These cylinders had about 500 hours on them if I remember correctly. In flight failures cause a lot of vibration. While the engine may still produce some power, it might also damage something with the vibration if operated very long. I opted to shut it down each time. Thankfully I had two engines. If I had been in a P-210, this would have been a major emergency. This experience led me to believe I no longer want to fly in turbocharged single engine planes. In one of the in flight failures, the separation was so severe, it tore the baffling and bulged the cowling. During this time, another operator on the field had a separation where the head went through the cowling and departed the airplane, however I never heard if his was an ECI cylinder. We spent a lot of time with the engine shop trying to figure out why this kept happening. They had no issue with the manner in which we operated the engine. The cylinders apparently bore all of the blame. That being said, if they have fixed the problem and there have not been more failures, why the AD? On the other hand, I find it hard to believe I represent 20-40% of all of the failures, depending on which data you look at. It seems there may be a good bit of under reporting going on.

I was flying an A36, and had just gone wheels up, when the cylinder head on the #3 (ECI) cylinder separated from the barrel. It caused a lot of vibration and reduction in power. Outside of that, no other issues. I was able to maintain altitude at pattern altitude so I decided to go ahead and lower the gear and continue through the pattern with constant eye outside to find a place to put it down if the need arose. I was able to fly a full pattern and land without incident. Now to address the total number of reported head/barrel separations. In reading through these (at the time) 16 comments, I am counting a total of 15 (including my one) issues with ECI cylinders. With a sample size 16 accounting for 78.9% of ECIs claim of 19 cylinder issues, I suspect that the FAAs count is a bit more accurate, if not conservative itself.

A club I flew with in the mid-90s had four ECI cylinders have head/barrel separation on three different aircraft resulting in three forced landings, one of which was off-airport. The forth failure was detected during run-up.

Do other manufacturers have issues with head/barrel separations? It seems that the shrink fit heads inherently puts high stress on the head/barrel.

Historical evidence suggests that plated cylinders have a greater propensity to suffer head-to-barrel failures in the thread-engagement area than non-plated cylinders-regardless of manufacture. Acid etching of the barrel prior to plating has been shown to generate corrosion in areas made difficult to clean-specifically, the threaded portion of the head. This corrosion compromises a highly stressed area resulting in head separation. If failure analysis identifies the alloy as the root cause, justification for the AD exists. If, however, the plating process is to blame, the remedy is simply procedural.