There does not seem to be anybody here who has flown an aircraft with a Lyc 320 or 360 with a 4×CHT engine monitor with downloadable temp data.
Hence the guessing going on.
The issue, which actually runs through a lot of GA “surveys”, is whether something is being reported.
A lot of stuff isn’t, for all kinds of straightforward reasons. In this case, there are
Then you get non-reporting due to perhaps less than fully honourable reasons 
I have heard of school planes getting cracks, but it does seem that they get a lot fewer considering the power-off abuse they get almost daily during training.
Raven wrote:
You have aluminium heads on the steel barrels.
Interesting point – but actually adding more questions than answers:
- If we use the simple model “Al-Case over Steel Barrel” the static stress due to different thermal expansion coefficients is order of magnitude higher than the dynamic stress while cooling
- The steel barrel in any case is quite thin and therefore should have almost the same temperature as the Al directly adjacent. Don’t see a big shift in temp gradient between the two
- Why is it a bigger problem with hotter cylinders? With your explanation, the problem should actually become smaller when hotter as the Al is softer and softer material is less prone to cracking in mechanical stress (might be deformed more easily but doesn’t crack as easy). So for cracking a very cold engine would be the problem.
Overall: If we look at flight school engines which are often abused by students, fly simulated engine failures, power off stall trainings, etc. (in brief: typically have to digest much more shock cooling than any privately operated plane) there is no significantly shorter cylinder life than in other planes (more the opposite).
Therefore it is imho safe to say, that even if shock cooling was a problem, it’s order of magnitude smaller than the problem of flying the engine infrequently and/or less than 200hrs/yr.
Malibuflyer wrote:
If we use the simple model “Al-Case over Steel Barrel” the static stress due to different thermal expansion coefficients is order of magnitude higher than the dynamic stress while cooling
Reality is not a simple model which you are using. Aluminium head is screwed ONTO a barrel (barrel has outside thread and head inner).
So expansion lowers the stress but contraction increases.
Engine manufacturer recommends some max. cooling rate. Are they wrong?
Malibuflyer wrote:
- Why is it a bigger problem with hotter cylinders? With your explanation, the problem should actually become smaller when hotter as the Al is softer and softer material is less prone to cracking in mechanical stress (might be deformed more easily but doesn’t crack as easy). So for cracking a very cold engine would be the problem.
The heads are still far away from being soft however weaker to sustain cracks. It’s just how it is. It’s not as you would think with your “simple model”.
The problem is higher with turbocharged engines (which normally operate higher CHT). For example Senecas operated in flightschool tend to change some cylinders every 300-400 hours just because of frequent engine shutdowns/feathers for training.
C150/C172 normally aspirated heat up only to 300-330F on level flight and therefore may be much much more resistant to fast cooling when simulating engine out.
Malibuflyer wrote:
Therefore it is imho safe to say, that even if shock cooling was a problem, it’s order of magnitude smaller than the problem of flying the engine infrequently and/or less than 200hrs/yr.
Here I fully agree with you. Statistically number one reason of destroyed cylinders is corrosion – long periods between flights – bad preservation, high humidity.
This discussion is really extremely enlightening for me, poor (in the monetary sense) operator of a horrendously expensive TSIOL-550. L as LIQUID COOLED.
The horrendous cost is artificial and caused by the fact that Conti discontinued spare parts sales. So we pay a fortune for a spare cylinder… But this apart:
The TSIOL-550 is marginally heavier than a 350 HP air cooled engine.
It has massively lower CHTs (max 270 F, typical 240 in cruise)
As a result it can be operated leaner at high power settings resulting in better fuel efficiency
There is no possibility of shock cooling as the operating temperature is already cool enough
The cooling system – which is part of the airframe – can be designed in a way that is very efficient in ALL operation regimes (using a thermostat and maybe an thermostat operated cowl flap). This would provide hot day performance increase while increasing efficiency through minimum drag in all flight regimes.
In my view, the TSIOL 550 is an engineering gem that fell victim of a deadly combination of manufacturing issues and poor knowledge in the field, which caused a cascade of problems, as follows:
- cylinder cracks
- cylinder replacement – bad workmanship
- more cracks…
Considering the above debate and the informally accepted fact that a turbocharged Cirrus will need a top overhaul at 1000 hours or less, even if operated cautiously, here a simple question:
Doe we have a business case for a liquid cooled, turbocharged IO-550 derated to 315 HP as an alternative to a factory overhaul?
The TC of the TSIOL-550 is here and Conti say they no longer make or support the engine.
Business opportunity?
Stanley wrote:
There does not seem to be anybody here who has flown an aircraft with a Lyc 320 or 360 with a 4×CHT engine monitor with downloadable temp data.
Hence the guessing going on.
I do (Cessna 172S with G1000) and I have some files laying around. I’ll dig into it and be back — but probably not for a few days.
Stanley wrote:
There does not seem to be anybody here who has flown an aircraft with a Lyc 320 or 360 with a 4×CHT engine monitor with downloadable temp data.
Even if you have load of real-time data from thermal/mechanical loads sensors, it is not obvious how you would calculate the lifetime of the cylinders?
Watching erratic curves may not tell you anything on the health of your engine, unless it matches data for one benchmark that you have opened….
I used to work on similar topic for jet engines (CFM56 & M88), we had load of sensor data but few broken engines to get decent regressions or descriptive conclusions
Even full simulations were worthless without data from regular inspections (whole thing torn apart and measured to micrometer), it is hard to predict when engines blow up just using sensor data, I am not sure if this applies to piston engines reliability but I am more than happy to fly on CFM56 flown by Rafale pilots whatever spikes I see in the live sensor data
Raven wrote:
Reality is not a simple model which you are using. Aluminium head is screwed ONTO a barrel (barrel has outside thread and head inner). So expansion lowers the stress but contraction increases
This might be an opportunity to introduce the way some Lycoming cylinders (e.g. higher compression 8.5:1 engines) have ‘choked’ or tapered bores, meaning the top end of the cylinder is bored (cold) to a very slightly smaller diameter than the bottom end. When the engine is warmed to normal operating temperature, the top end is warmer and expands more, resulting in closer to constant operating piston clearance, lower oil consumption etc. In similar fashion, high performance engine builders of some air cooled motorcycle engines heat the cylinders to operating temperature before they are machined to final size.
I think the issues with air cooling aircraft cylinders are (1) given variation in cylinder operating temperature it takes careful design to minimize oil consumption and eliminate cracking (particularly on large cylinders) but (2) you don’t have to fuss with glycol, head gaskets, thermostats, outrageously priced hoses, coolant pumps and so on, and (3) individually replaceable air cooled cylinders make ownership much easier – you don’t have to remove and ship the engine for compete disassembly and overhaul every time there is a minor issue on one cylinder – which is often the case in the real world. I think personally I’d rather maintain two 160 HP O-320s than one liquid cooled monoblock aircraft engine, never mind two of them, and it’s also got me thinking that maybe a Twin Comanche with two small IO-320 Lycomings makes more sense to own and maintain than a 300 HP aircraft with one IO-520 or IO-550 Continental.
To supply some actual figures for a Piper 28-181.
A recent flight (2nd flight of the day) 2 POB (75kgx2), full fuel, OAT 13C
(I have owned the a/c 11 years and it is in good condition. I am the only pilot. The engine monitor records 2sec intervals)
Prior to rolling on the runway Cyl 3 (hottest) 349F
At t/o C3 391F, 2464 RPM full rich (climbing Vy+10 asap)
At 980 feet C3 415F at this ALT power reduced using 2410prm (normally I would reduce power earlier but got distracted)
(this stopped C3 rising further)
To depart downwind at 1000 agl, 2300 rpm set and this produced a cooling spike of 40F/min.
No instructor I had ever advised to reduce power in the climb, so if departing cross country it would have been full power
to cruise altitude with no consideration of engine temps.) Plus in the circuit, power would be reduced in an instant from
full to 2100rpm with a heavy cooling spike corresponding.
These aircraft are also used in the summer in four-up flights fully loaded and that will be fun in more ways than one.
Shock cooling is a myth
https://www.avweb.com/ownership/shock-cooling-time-to-kill-the-myth/
Cylinder cracks via shock cooling are absolutely real but happen only if the CHT is high enough to start with.
Any further reading/sources? I’m very curious (trying to learn as much as possible about piston aircraft engines currently).
