Walter Atkinson:
I have taken a cursory look at the issues on the ink above and found several interesting issues.
1) The "assumption" by the researchers that the observed increase in EGT with respect to RPM is flawed. The measured change in combustion temperature with increasing RPM is actually lower due to the delayed thetaPP during the higher RPMs, therefore so is actual EGT. The effects are actually very small but the reading changes are significant. Why, then are they (and we in instrumented aircraft) seeing higher EGT READINGS with increases in RPM? The reason this is the case is that the probe has less refractory time to cool between pulses of hot air.--it will read the same (or even a lower) temperature as higher. This has only recently become appreciated (the last 20 years or so).
You have overlooked where they specifically state that the EGT is
"not the true exhaust gas temperature but merely indicated the temperature that may be attained by an engine part subjected to the flow of exhaust gases at the same position"; which is more relevant than instantaneous gas temperature. Nonetheless it stands to reason that EGT would increase with RPM, like it does with retarded timing.
"Mass-averaged exhaust temperatures are higher than time-averaged or thermocouple determined temperatures. All these temperatures increase with increasing speed, load, and spark retard, with speed being the variable with the greatest impact". Heywood JB, Internal Combustion Engine Fundamentals, chap 6, p234, 1984.
What you describe as "it will read the same (or even a lower) temperature as higher" is actually thermocouple equilibrium temperature approaching the temperature of the "pulses of hot air" as RPM increases.
The measured change in combustion temperature with increasing RPM is actually lower due to the delayed thetaPP during the higher RPMs, therefore so is actual EGT.
It does not follow that EGT is lower. In the test they
maintained imep as they varied RPM. In the diagram below the red pecked line represents the pressure trace with retarded PP. For the EGT to be cooler the red pecked line would have to drop into the brown area. That cannot happen whilst
maintaining imep.
2) If one looks at the change in exhaust valve temperature as CHT is altered while holding everything else consent, one quickly sees the most direct correlation of CHT-EVT as compared the other factors.
They did not do a test to alter CHT and EGT while holding everything else constant. I would be very interested to know how you think one would go about such a test...
3) The unusually high EGTs in this study are due to the low compression (6.5:1) of the engine. As expected.
The CR was 6.65 which was normal for that engine and seemingly typical for blown engines of that time, eg the Wright TCs (CR 6.7). Therefore the EGTs in the test were representative under the conditions stated. The mixture temp was 200°F which would be modest based on NACA estimates:
In the test of variable CR the difference between CR of 6.65 and 7.5 (typical GA turbo) was 55°F:
...not huge and the point of the test was to investigate how all the temps respond, not establish absolute figures for any specific CR.
4) Research and analysis of hundreds of events of pre-ignition have debunked the "assumption" during this test that pre-ignition is caused by high exhaust valve temperature. This is simply not supported by reams of data from pre-ignition events where engine monitors have been present to record the events
A proper reading of the report reveals that the valve temperature was actually
lower when they experienced preignition, due to the insulating effect of the valve deposits. It is those deposits that is the problem they identified rather than the valve temperature per se;
"The foregoing results indicate that the scale and deposits probably reached a temperature sufficiently high to cause preignition whereas the temperature of the valve steel as measured by thermocouple remained substantially below this temperature". All backed up by the engineering literature:
"The source of preignition is usually the exhaust valve covered with mineral deposits coming from fuel and lubricant additives" Guibet JC, Duval A, New Aspects of Preignition in European Automotive Engines, SAE Technical Paper, 1972.
"Preignition is caused by the mixture igniting as a result of contact with a hot surface such as an exhaust valve." Stone R, Introduction to Internal Combustion Engines, chap 3, p74, 1992.
"The parts which can cause preignition are those least well cooled and where deposits build up and provide additional thermal insulation: primary examples are spark plugs, exhaust valves....Under normal conditions using suitable heat range spark plugs, preignition is usually initiated by an exhaust valve covered in deposits". Heywood JB, Internal Combustion Engine Fundamentals, chap 9, p453, 1984.
...but perhaps no longer the same problem as with the fuels of 1947.
5) Some of the "assumptions" from the 1930s and 40s are taking a long time to dispel (recently a CMI representative giving a talk to 300 LAMEs in the US assigned effects of pre-ignition to having been, incorrectly, detonation--and that's not an uncommon error)
Interesting anecdote but it has nothing to do with this test. On the two occasions when they encountered preignition, had that actually been detonation
they would have recognised it due to audible knock. Your opinion that these experienced NACA researchers mistook preignition for detonation is not supported by studying the series of tests undertaken at the Cleveland lab to explore those very issues.