It is a very good point that when the oil consumption reaches the point where sufficient oil gets consumed during the fuel endurance of a “typical application aircraft” to take it from the max point to the min point (both of these are specified on the TCDS) then the engine is considered to need a repair.
AFAIK steel doesn’t have the fatigue problems which aluminium has. Steel parts, provided they are operated below some % of their elastic limit, have an indefinite life, but that is never the case with aluminium which will always fail via fatigue (unless, presumably, operated at some very low stress).
I recall a lot of press about some Chieftain crash many years ago (Deakin also wrote about it, pointing out a lot of nonsense in the report) and apparently that particular installation runs really hot, plus the pilots had no idea about engine management.
Engine certification is indeed done over a 150hr dyno run and the engine is stripped and measured before and after, but I wonder how they could possibly extrapolate 2000hrs from that, since there isn’t all that much wear in those 2000hrs. But maybe there is? Maybe the engine expected to be outside service limits, on some key parts? The exhaust valves, certainly, but those are usually outside service limits a lot sooner.
Peter wrote:
Engine certification is indeed done over a 150hr dyno run and the engine is stripped and measured before and after, but I wonder how they could possibly extrapolate 2000hrs from that, since there isn’t all that much wear in those 2000hrs. But maybe there is? Maybe the engine expected to be outside service limits, on some key parts? The exhaust valves, certainly, but those are usually outside service limits a lot sooner.
One way of doing it is to accelerate the engine wear by a torture test: maintaining takeoff power for more than the requisite 5 minutes, exceeding the red line altogether, etc.; even without a torture test, operating parameters during the dyno run are cycled more intensively than in normal operations. In addition to that, lots of engine parameters (temperatures, pressures, flow rates, vibrations, etc.) can be monitored during the run. After the run but before disassembly, one can check for friction in the bearings (for turbines, it’s usually the rotor rundown time after shutdown), ease of cold start, etc. Once the engine is stripped afterwards, it’s not only the measurable wear that is checked – you check for heat tints, various deposits, do an oil analysis, strength/hardness testing of critical parts, etc. Unlike normal inspections and overhauls, one is not limited to NDT, the parts may be sliced and diced and ripped to shreds.
For jet engines, manufacturers have approved methods to set inspection program and define the life cut off for a specific engine using past inspection and monitoring data, basically they calculate date of next inspection while varying plus/minus the TBO for a specific engine
This is mainly used in initial/continious certification to change inspection cycles/end life as they tend to be set very conservative in the early launch of a new engine rather than airlines production/maintenance
None of the airlines go that way (they could), they like to go for a well defined calendar and well defined engine life cut-off as it makes their maintenance & operation more predictible
The only interesting failures are those that happen in flight. An engine failing an inspection doesn’t count, nor does an engine that gives clear warning signs in flight without actually failing.
I agree. According to this, the critical failures are often infant failures, while the failures at the “end” of the bathtub are often less safety critical.
huv wrote:
while the failures at the “end” of the bathtub are often less safety critical
And also more easily detectable, even during routine maintenance (Busch lists: spalled cam, a cracked crankcase, worn or contaminated bearings, a worsening oil leak).
That’s actually a really good observation.
I reckon it may also have an impact on the overhaul cost – when one finally does an overhaul. Conrods never wear out (both bearing ends use bushings which are fixed relative to the conrod) and they get scrapped if out of spec on length (a conrod stretches or squashes – I can’t remember which – a tiny amount on each stroke) or if there is any trace of corrosion.
But crankshafts do wear, and crankcases can move about and wear on mating surfaces, and each of these two things can render the engine beyond-economical to overhaul (in Europe, especially).
I think worsened corrosion would be the greatest cause of additional cost, if it were to be the case for a deferred overhaul. For example, crankshaft corrosion would be more of an issue than crankshaft wear. Worn crankshaft bearing journals can be reground at least once, whereas corrosion pitting on crankshaft or gear teeth cannot be fixed beyond a certain depth.
Or simply an ingenious scheme setup by engine manufacturers to ensure your continuous business in return for steadily increasing engine safety for everyone, just not necessarily your specific engine.
I think the bathtub curve is one of the most misused concepts in anything technical. There is no reason to believe that a mass produced item has initial failures, unless the production facility is new and is undergoing “birth difficulties”, which is the original concept behind initial failures. Initial failures are tied to the production facility (which often is tied to production batches, and new to the market items), not the produced item specifically.
Overhaul is of course something that is supposed to bring the item back to “new” standard. The success of this is 100% dependent on the company doing the overhaul. There is no reason to believe that such a company is able to do that at all. The only way to be sure the item is in new condition, is to replace it with a factory produced new one.
It’s all depending on the factory or overhaul facility. Rotax has the production know how to get this right. Lycoming, probably not so much? It would be interesting to see the difference between the number of “initial failures” of a factory new Lycoming compared with a overhauled Lycoming.
@Lucius, this video is quite misleading in several aspects:
1. It assumes that among many pieces of the same part, some are inherently good and some are inherently bad and the reliability depends entirely on workmanship. The quality of individual pieces may indeed vary somewhat, but certainly not so dramatically as it says, and assuming that this quality will persist over several “generations” of rebuild is a total nonsense.
2. It remains silent about the difference between overhaul tolerances that apply to overhaul and new engine tolerances that apply to remanufacture.
3. It remains silent about the difference in service life of different parts.
4. It professes an, ahem, cavalier approach to statistics.
