VVN,
While I agree with your friends basic formula, there is more to this than just flow being proportional to RPM. This totally disregards the effect of head or frictional loss. While it is true that flow in a centrifigual pump goes up in direct proportion to the RPM, it is also a fact that frictional loss goes up by the square of the flow. This means that the frictional loss (head) will quadruple if you double the flow. Given the very open impeller design on a stock pump, I would guess that the impeller has a very steep loss of flow as head increases. If the pump flows a certain flow at 4400 RPM, it will flow way more than half that at 2200 RPM! Consequently, it will probably not flow much more than that no matter how fast you spin it due to the dramatic increase in head due to increased flow. Of course you also have to factor in the ability of the impeller's ability to increase head as RPM increases, but this increase is not even quite proportional to flow, so the frictional losses will still eat up flow at an alarming rate. The only way to know for sure is to obtain a pump curve for the pump that plots out RPM,flow, and head. The NPSH or net positive suction head also has to be taken into account. If an engine has more restriction in the suction side of the pump than the discharge side, the pump can run below the minimum NPSH for a pump, which can cause the pump to cavitate and lose flow dramatically. The most common cause of this is running no thermostat, which decreases head on the discharge of thepump while not decreasing it on the suction, causing the pump to cavitate. The resulting loss of cooling is usually blamed on running the water through the block too quick, when in reality, it is exactly the opposite. A pump needs to run against the head it was designed for, and deviating from that design to much can cause either cavitaiton or excessive horsepower consumption.
By the way, I would agree with sticking with the mechanical pump!
Regards, John McGraw
