What causes cylinder head separations? Are they commonplace on our engines? Should I expect one to occur, sooner or later? How can I prevent one from taking place ‘on my watch’? Assuming no damage from a resulting forced landing, and no prop strike, will a complete engine teardown be required? Won’t a lot of oil be lost from the broken cylinder’s pushrod tubes and oil drain-back tubes?
Cylinder head separations very seldom occur on a 320 or 360 Lycoming. It literally never happens on a new Lycoming steel jug, or on a second-run Lycoming steel jug that has been bored oversize for the first overhaul. When it does happen, it is virtually always on a used-exchange cylinder, a used-repaired exchange cylinder, an overhauled-exchange cylinder, or a chromed-exchange cylinder. In other words, a cylinder on which there is absolutely no documentation on how many hours are on the barrel and head. The FAA has never established any requirement to serialize, and track hours on cylinders, despite their obvious risks. This was probably because there are literally so many hundreds of thousands of cylinders out there; and because during the heyday of aviation (with the big radials), cylinder turnover was a massive business. The advent of extremely reliable turbine engines in virtually all commercial passenger service made engine failures a relatively moot point, in light of the labor and paperwork that would be needed to track individual cylinders. The moral is, once you make it a practice to accept an exchange, as opposed to buying new or having your first-run cylinders bored for oversize pistons at overhaul, you can probably expect to have a cylinder head separation somewhere down the road.
By far the best choice on cylinders that have worn out on their first run since new, is to have them bored oversized for new pistons. The break-in will be immediate, they will have a long service life, and will have no surprises. This assumes that the seats, valves, guides, etc. are properly repaired or replaced. The guides should always be replaced. The exhaust valve should always be replaced. The seats and intake valve are sometimes repairable. The exhaust port studs should always be replaced. Bored cylinders will get you through one more TBO; after that you will need to buy new.
If your existing cylinders have an unknown history due to being exchange units, are known to be on their second or subsequent run, or are of unknown history, the only reliable option is replacement with new Lycoming cylinders. Never believe anyone who tells you that any other cylinders are just as good an option, when it comes to Lycoming engines. There are some aftermarket cylinders that are better than factory cylinders for Continentals; but not for Lycomings.
The core problem is that aluminum, like all metals, has a finite fatigue life. Heat-treated aluminum alloys are more susceptible than some other metals; that’s a key reason why planes certificated under FAR23 must have a life-limit established for things like the wings. Examples are the Skipper, the Duchess, and even the composite Cirrus. Every time a cylinder heats up then cools off, and every time the cylinder ‘pulses’ from the combustion event, the aluminum is stretched and relaxed. Eventually it cracks. Ideally the initial cracks will show up during proper maintenance. For example, just doing a compression test is insufficient, during the Annual Inspection. A bright light and mirror must be used to examine the top and bottom of each cylinder head. You are looking for any unusual stains, oil leakage, blown-gases markings, lost paint in a suspicious pattern, etc. You must also listen for any unusual noise during the compression test. It is very rare for a cylinder crack to progress to the point of separation, without having provided some sign beforehand, if someone is watching for the early warning signs.
If you literally lost all oil pressure while the engine was still producing power, you must split the case. If the engine was at idle power before the oil pressure went below 5 PSI or so, there will be no damage from lack of oil pressure. Any oil leak would be from the separated drain-back tube; or more accurately, from the exposed pushrods and pushrod tubes. That is a relatively slow loss rate as compared to a leak on the pressure side of the system. So the need for a tear-down really depends on how low the pressure got, and the engine power circumstances at the time. An idling engine needs next to no oil pressure. There should be no internal engine contamination, since the piston kept the cylinder sealed up.