Originally Posted by gathermewool
I don't fully understand how manually down-shifting the transmission to achieve >3k RPM will sufficiently raise valve temperatures by that much. More fuel is obviously needed, to account for the increased losses at higher RPM; however, the actual load is still the same (again, with the added load due to frictional and pumping losses).
Me either, mainly because I've previously read how Honda shift its auto transmission early when the engine is cold to increase the engine temperature faster, and how GM engine delay shifting into high gear when it detects lost of cooling to cool the engine down.
What it means, like you said, is the amount of fuel burnt per RPM is what heat up the engine. On the same vehicle load going from 1500 rpm to 3000 rpm does not increase the fuel burnt more than the air passing through the cylinder. This means the engine is actually cooling down instead of heating up.
One research paper I've seen in the past (from a Korean university) said that it is the engine load time, but having continuous load or not does not matter (i.e. 3 mins continuous high load vs 1 min high 1 min load cycled 3 times will have the same deposit reduction capability). What they found that you can do and that actually work, is group 4 oil changed frequently (maybe volatility and / or the amount of detergent used in group 4 form less deposit).