[svxistentialist]

Tony, I think in fairness to Trevor's reservations about the system and circuit being non standard when tested at atmospheric it should be remembered that the reason for pressurising the system is because water [and in this case coolant] has a different boiling point when under pressure.

As your test rig was running normally [as I understand it] at atmospheric and without a head then said rig could not realistically emulate high temperature conditions in the engine under normal load and operating conditions. Local conditions near the pump could cause localised boiling when running at a lower than normal pressure.

The changes you have made to your test system have allowed improvements to the flow path without positively proving that this will also work the same when under radiator tank pressure.

I for one accept that your trial and error method works, and I would be following the same path if I was planning to race or boost an EG33 to serious horsepower.

Using a larger outlet from the lower radiator tank is a given improvement as I see it, but I'm less convinced the outlet needs to be in the centre. Also I would expect your modification [assuming it works, as I do] to improve when operating under a head, compared to circulating at atmospheric.

And for that reason as you are now saying you did pressurise at one point and it made no difference, I am now suggesting that Trevor should accept your modification as being proven.

Trevor?

If you feel more needs to be proven here, please provide a detailed methodology in line with current thinking so that proof can be tested and accepted.

Joe