Cooling systems in specials without thermosiphon

[JonE]

Ivor posted a reply in another thread and I thought it merited a much wider discussion area on the basis of several insightful exchanges with Robert Leigh, Ian Williams and Greig Smith which started me thinking about thermodynamics.
Ivor mentioned "I ran a Super Sportsman for a number of years, which had an HC motor and twin 1 inch SUs, a cross flow radiator, which was very small and an aluminium header tank mounted just ahead of the bulkhead and had no issues with overheating despite not having a water pump...I would say go for it and watch the temperature..."

When one loses the tall radiator, the basis for flow of rising warm water from the engine is lost, but I can that with the above, there was still a higher point - the header - so was there still rising flow that was adequate enough for sustainable use, as long as the water is passing from head and back to the upper point? The header won't be an efficient shedder of heat but it is still less hot that the engine. What controls the strength of the thermosiphon effect - temperature difference presumably?

[Colin Wilks]

I am not going to pretend I have a clue about the science of thermodynamics beyond knowing heat rises and a hot liquid is less dense than a cold liquid and so the hot floats above the cold. In my simple way I think of gravity helping the colder water in the radiator to push the hot upwards through the engine block.

I suspect there is a critical point around where the rising hot water meets the apex of the system (i.e. where the top hose enters the radiator) and recall Nick Turley showing me a radiator from I think an Austin 20 which had been expensively rebuilt without a baffle in this area. This caused the car to overheat until this omission was identified and the radiator was taken apart again and the baffle reinstalled. Nick may have further and better particulars.

Ps. Nick has reminded me it was an Austin 12, and the purpose of the baffle was to ensure the turbulent hot water entering from the top hose did not cause a good portion of it to disappear down the overflow pipe, which begs the question why they didn't stick the overflow at the other end of the top chamber in the first place?