A few thoughts on how the different gases in your cooling system should behave, which can help you understand what’s going on while observing different things:
air – may enter due to insufficient bleeding during initial filling.
It can be drawn back through the radiator cap as the system cools, due to an empty overflow tank or problems with the radiator cap valve or hose.
It will expand linearly due to absolute temperature increase (eg add 460*F to get absolute, so 240*F air will be (240 + 460) / (60 + 460) = 1.346 = 35% larger air bubble at constant pressure). The coolant overflow system is designed to accommodate the expansion of the coolant itself, which is much less than that. As the trapped air bubbles expand, the pressure will build, forcing the coolant out through the radiator cap after it exceeds the release pressure of perhaps 16 psig. With trapped air, some of that air will be released as the trapped pockets expand and escape, helping you eventually solve the problem, but it will come out too cold, raising the level of the overflow tank.
Then as the engine cools, those air pockets will shrink completely, the coolant itself will shrink a bit, and the coolant from the overflow tank will be drawn back in. As said, any air that came out should not be re-breathed, so an initial imperfect bleeding should be adjusted through this, which only requires a fluid refill after a day or two. So that’s why topping up coolant is usually pretty easy when you have the luxury of checking and topping up the next day (as DIYers do, but pros don’t).
water vapor – as the coolant heats up, if any part of the cooling system (eg, in the hottest part, in the cylinder heads, perhaps where there is not enough coolant flow for some reason) gets hot enough to cause a condition boiling, the water in the coolant-water mixture will boil, turning the water into water vapor, a gas, which has MUCH more volume than the water it came from; as a factor of 1000x instead of 30%. This increased volume will increase the pressure, which increases the boiling point, and if the liquid circulates well it is not a problem. But if a blockage is causing the fluid to not circulate properly, you may have a hot spot or two that increases the gas in your system too much, forcing the fluid past the radiator cap into your expansion tank, following the same method as air expansion.
But the BIG difference with water vapor versus air or combustion gases is that once cooled, the volume of the gas doesn’t just shrink by a percentage, it effectively goes to zero as the vapor condenses back into water.
combustion gases – (mainly CO2) unlike air and water vapour, combustion gases will steadily exit the cylinder into the cooling system, never escaping except through the radiator cap / expansion tank. If your cooling system gets a significant amount of combustion gases in it, they will not be corrected by condensing and any gas pockets will be constantly renewed with HG leaks, so the problem persists.
When the engine cools down, these gas pockets will contract just like the air, drawing some coolant back from the expansion tank. But unlike air, combustion gases are constantly renewed.
A “test block” kit uses color-changing fluid to detect the presence of CO2 gas in your cooling system, but it is possible to get a false negative (eg, if the combustion gases only pass the HG flow during conditions extremes of engine load, a simple block test result can come back negative) and false positives (LOL, but I speak from experience saying that if the test is not done properly, with the coolant level too high, it is possible to seep yellow coolant through the block.test fluid, turning it yellow [radiator hoses collapsing under vacuum can raise the coolant level unexpectedly during the test, as I have found]), so it is not 100% conclusive.
===
So your radiator cap and the overflow hose to the expansion tank are important parts of the system here to make sure they are working properly.
And if there is a partial blockage causing a localized boil-off problem, it should correct itself every drive cycle, as long as there is no leakage (from the top of the overflow tank) and the radiator cap/fill system is OK. That is, localized boiling will occur, the water vapor will increase the pressure, pushing the coolant into the expansion tank. But as long as it’s not bad enough to cause the reservoir to overflow, when things cool down, the coolant should be drawn back into the cooling system with no net change in level.
===
This is all general information to help you understand things. I still think it’s a HG problem, but you have to eventually figure this out, and also be prepared to find and fix any blockages that may be present.
